OCTOBER-DECEMBER 2020 Vol.26, Number 4, 321-425

ISSN 1451 - 9372(Print) ISSN 2217 - 7434(Online) OCTOBER-DECEMBER 2020 Vol.26, Number 4, 321-425

www.ache.org.rs/ciceq Journal of the Association of Chemical Engineers of Serbia, Belgrade, Serbia

EDITOR-In-Chief Vlada B. Veljković Faculty of Technology, University of Niš, Leskovac, Serbia E-mail: [email protected]

ASSOCIATE EDITORS Jonjaua Ranogajec Srđan Pejanović Milan Jakšić Faculty of Technology, University of Department of Chemical Engineering, ICEHT/FORTH, University of Patras, Novi Sad, Novi Sad, Serbia Faculty of Technology and Metallurgy, Patras, Greece University of Belgrade, Belgrade, Serbia

EDITORIAL BOARD (Serbia) Đorđe Janaćković, Sanja Podunavac-Kuzmanović, Viktor Nedović, Sandra Konstantinović, Ivanka Popović Siniša Dodić, Zoran Todorović, Olivera Stamenković, Marija Tasić, Jelena Avramović, Goran Nikolić, Dunja Sokolović

ADVISORY BOARD (International) Dragomir Bukur Ljubisa Radovic Texas A&M University, Pen State University, College Station, TX, USA PA, USA Milorad Dudukovic Peter Raspor Washington University, University of Ljubljana, St. Luis, MO, USA Ljubljana, Slovenia Jiri Hanika Constantinos Vayenas Institute of Chemical Process Fundamentals, Academy of Sciences University of Patras, of the Czech Republic, Prague, Czech Republic Patras, Greece Maria Jose Cocero Xenophon Verykios University of Valladolid, University of Patras, Valladolid, Spain Patras, Greece Tajalli Keshavarz Ronnie Willaert University of Westminster, Vrije Universiteit, London, UK Brussel, Belgium Zeljko Knez Gordana Vunjak Novakovic University of Maribor, Columbia University, Maribor, Slovenia New York, USA Igor Lacik Dimitrios P. Tassios Polymer Institute of the Slovak Academy of Sciences, National Technical University of Athens, Bratislava, Slovakia Athens, Greece Denis Poncelet Hui Liu ENITIAA, Nantes, France China University of Geosciences, Wuhan, China

FORMER EDITOR (2005-2007) Professor Dejan Skala University of Belgrade, Faculty of Technology and Metallurgy, Belgrade, Serbia Journal of the Association of Chemical Engineers of Serbia, Belgrade, Serbia

Vol. 26 Belgrade, October-December 2020 No. 4

Chemical Industry & Chemical Engineering CONTENTS Quarterly (ISSN 1451-9372) is published quarterly by the Association of Chemical Pongayi Ponnusamy Selvi, Rajoo Baskar, CO2 absorption in Engineers of Serbia, Kneza Miloša 9/I, nanofluid with magnetic field ...... 321 11000 Belgrade, Serbia D. Revathi, Dr.K. Saravanan, Experimental studies on Editor: hydrodynamic aspects for mixing of non-Newtonian Vlada B. Veljković fluids in a Komax static mixer ...... 329 [email protected] Sanja Radonjić, Helena Prosen, Vesna Maraš, Lea Demšar, Tatjana Košmerl, Incidence of volatile phenols in Editorial Office: Montenegrin red wines: Vranac, Kratošija and Kneza Miloša 9/I, 11000 Belgrade, Serbia Cabernet Sauvignon ...... 337 Phone/Fax: +381 (0)11 3240 018 E-mail: [email protected] P.C. Nnaji, B.I. Okolo, O.D. Onukwuli, Luffa cylindrica seed: www.ache.org.rs Biomass for wastewater treament, sludge generation study at optimum conditions ...... 349 For publisher: Kristina Jančaitienė, Rasa Šlinkšienė, Influence of cellulose Ivana T. Drvenica additive on the granulation process of potassium dihydrogen phosphate ...... 359 Secretary of the Editorial Office: Slavica Desnica Bruna Muriel F. Costa, Carolina Monteiro Coelho, Cássia Ribeiro Souza, Gabriela Drumond A. Duarte, Maria Marketing and advertising: Bernadete Pinto, Vítor J. O. Silva, Marcelo S. Batista, AChE Marketing Office Study of the drying kinetics of the malt bagasse in a Kneza Miloša 9/I, 11000 Belgrade, Serbia pneumatic transporter ...... 369 Phone/Fax: +381 (0)11 3240 018 Marija Bodroža-Solarov, Nevenka Rajić, Lato Pezo, Jovana Publication of this Journal is supported by the Kojić, Jelena Krulj, Bojana Filipčev, Rada Jevtić Muči- Ministry of Education, Science and babić, The rheological properties of wheat dough Technological Development of the Republic of containing zeolite residue ...... 377 Serbia Rim Jouraiphy, Islaam Anouar, Hamid Mazouz, Badr El Kor- Subscription and advertisements make payable tobi, Rachid Boulif, Ahmed Mahrou, Amina Ibnlfassi, to the account of the Association of Chemical Solid-liquid separation techniques for efficient recovery Engineers of Serbia, Belgrade, No. 205-2172- of P2O5 from phosphoric sludge ...... 385 71, Komercijalna banka a.d., Beograd Jesús M. Casas, Josué Lagos, Drowning-out crystallization of sodium molybdate in aqueous–ethanol solutions ...... 395 Computer typeface and paging: Vladimir Panić Maja Čolnik, Maša Knez Hrnčič, Mojca Škerget, Željko Knez, Biodegradable polymers, current trends of research Printed by: and their applications, a review ...... 401 Faculty of Technology and Metallurgy, Corrigendum to Enhanced sorption of Cu(II) ions from Research and Development Centre of Printing aqueous ions from aqueous solution by ionic liquid Technology, Karnegijeva 4, P.O. Box 3503, impregnated nano-silic and nano-alumina particles 11120 Belgrade, Serbia (Chem. Ind. Chem. Eng. Q. 23(2) (2017) 207-216) ...... 419

Abstracting/Indexing: Contents: Vol. 26, Issues 1–4, 2020 ...... 421 Articles published in this Journal are indexed in Author Index, Vol. 25, 2019 ...... 423 Thompson Reuters products: Science Citation Index - ExpandedTM - access via Web of Science®, part of ISI Web of KnowledgeSM

Activities of the Association of Chemical Engineers of Serbia are supported by: - Ministry of Education, Science and Technological Development, Republic of Serbia - Hemofarm Koncern AD, Vršac, Serbia - Faculty of Technology and Metallurgy, University of Belgrade, Belgrade, Serbia - Faculty of Technology, University of Novi Sad, Novi Sad, Serbia - Faculty of Technology, University of Niš, Leskovac, Serbia - Institute of Chemistry, Technology and Metallurgy, University of Belgrade, Belgrade, Serbia Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 321−328 (2020) CI&CEQ

PONGAYI PONNUSAMY CO2 ABSORPTION IN NANOFLUID WITH SELVI1 MAGNETIC FIELD RAJOO BASKAR2 1Department of Chemical Article Highlights Engineering, Kongu Engineering • Experiments carried with/without the magnetic field of the structured packed absorpt- College, Erode, India ion column 2 • Nanofluids were prepared by dispersing Al O nanofluids in water with/without mag- Department of Food Technology, 2 3 netic field Kongu Engineering College, Erode, • The mass transfer performance of the structured packed absorption column was India investigated

SCIENTIFIC PAPER Abstract

UDC 504.055:66.081:54 Acidic gases like CO2, SO2, NO2, H2S, etc., are to be removed as these are polluting the atmosphere in one way or another by inducing temperature rise which further results in undesirable climatic change. Among all these gases,

CO2 is the most responsible for the environmental issues and its capture becomes one of prime importance. The objective of this work is the enhance-

ment of the CO2 absorption by employing nanofluids in the presence of magnetic field. The nanofluid used in this work is Al2O3/water in the concentration of 0.0015%. The maximum flux obtained is 0.014 mol/(m2s) (without magnetic 2 field) and 0.015 mol/(m s) (with magnetic field) for lower CO2 flow rate of 30 L/h. The nanofluids along with the magnetic field show a positive performance

towards the absorption of CO2.

Keywords: CO2, flux, flow rate, global warming.

Due to human activities and since the industrial acid gas. Among all the chemical solvents [9], amine- th revolution from the 17 century, 289 ppm of CO2 -based solvents are used to capture CO2. But this increased to 406 ppm in 2017 resulting in a 2 °C raise method has several advantages and certain disad- in atmospheric temperature. To avoid dangerous eco- vantages like high power consumption, low gas load- logical and environmental issues, CO2 sequestration ing, high equipment corrosion, amine degradation etc. has to be carried out [1,2]. The major sources of Hence, a novel solvent with low cost, higher capacity greenhouse gases [3] come from the burning of fossil and efficiency with lesser decomposition and corros- fuels, fermentation, deforestation, chlorofluorocar- ion rate is the aqueous ammonia [10,11]. After the bons, agricultural activities [4,5], etc. CO2 is the major solvent selection, the next absorption is to be carried contributor to global warming which results in the loss out in the wetted wall column or packed column. in biodiversity and extreme changes in agricultural Among these two, the wetted wall column is most activities [6,7]. CO2 has been captured by various preferable with a film promoter. The film promoter technologies [8] like adsorption, membrane separ- improves the formation of a stable liquid film and ation, cryogenic method, biological fixation and wet liquid distribution inside the column. Apart from the scrubbing or absorption. Wet scrubbing is the cost conventional solvents, nanofluids [12–14] are used to effective and widely accepted method to capture the enrich the absorption rate. With Al2O3 nanofluids [15– –17], the Brownian movement and grazing effect have been enhanced by the application of magnetic field. Correspondence: P.P. Selvi, Department of Chemical Engineer- ing, Kongu Engineering College, Erode-638 060, India. CO2 Sequestration methods E-mail: [email protected] Paper received: 25 December, 2018 There are various technologies available for acid Paper revised: 18 October, 2019 gas removal. They are adsorption, membrane separ- Paper accepted: 28 February, 2020 ation, cryogenic method and absorption process. https://doi.org/10.2298/CICEQ181225008S

321 P.P. SELVI, R. BASKAR: CO2 ABSORPTION IN NANOFLUID… Chem. Ind. Chem. Eng. Q. 26 (4) 321−328 (2020)

Absorption means adding limestone or carbon compounds to Absorption or wet scrubbing is the transfer of separate CO2 from the gas stream. Calcination one or more species from the gas phase to the liquid means heating to very high temperatures in the pre- phase. The species transferred to the liquid phase are sence of oxygen. Biosorption is the process of rem- known as solutes or absorbate. Absorption occurs by oving CO2 from the gas stream by the usage of mic- physical and chemical means. Physical absorption roorganisms. This process has been carried out by occurs when there is no chemical reaction between employing specific microorganisms suitable for CO2 the solute and the solvent. Absorption with chemical absorption. So, a thorough knowledge on the applic- absorption involves the removal of impurities from the ation and availability of microorganisms should be gas phase and dissolving them into liquid. identified. Membrane separation Reaction mechanism The membrane permeates the chemical species The reaction between aqueous ammonia and more easily than the other conventional methods. The carbon dioxide mainly occurs in the liquid phase of porous or semi-porous structured membranes like the gas-liquid interface, the reactions in the liquid polymers, zeolite or inorganic membranes like palla- phase of the CO2-ammonia system are given as fol- dium are used. These gas separation membranes will lows. permeate one of the components more quickly than The total reaction involved in the process is the other. The separation capacity is increased by described as follows: performing multiple stage operation but it leads to CO2(g) + NH3(aq) + H2O  NH4HCO3(aq) (1) high energy consumption, operational cost and dif- The actual process is more complicated and can ficulty of operation. be described in the step-by-step process:

Cryogenic method - + CO2(g) + 2NH3(aq)  NH2COO (aq) + NH4 (aq) (2) The cryogenic method involves gas compression and cooling it to the minimum temperature At the same time the complex balance of solute thereby the separation takes place by distillation up to forms ionic reactions in the solution and the equations

90%. This method is used for high CO2 concentrat- are as follows: ions, high pressure gases and high energy is needed + - NH3(aq) + H2O(l)  NH4 (aq) + OH (aq) (3) for refrigeration. The unwanted component like water NH HCO (aq)  NH +(aq) + HCO -(aq) (4) is to be removed before cooling the gas stream to 4 3 4 3 + 2- avoid blockages. The light olefins and paraffins are (NH4)2CO3(aq)  2NH4 (aq) + CO3 (aq) (5) studied by cryogenic distillation. Alternate to distil- - - 2- OH (aq) + HCO3 (aq) CO3 (aq) + H2O (6) lation, absorption was carried out with mass transport CO 2-(aq) + CO (g) + H O(l)  2HCO -(aq) (7) but distillation was carried out with energy transport. 3 2 2 3 The absorption reduces the capital cost and energy Reaction (2) is very fast and irreversible and requirements of aromatic separations. Copper- and instantaneous. Reaction (4) is too slow to influence silver-based solutions were used to separate olefins the rate of the absorption directly. Therefore, the and paraffins. reaction between aqueous ammonia and carbon Adsorption dioxide is mainly controlled by reaction (2). It is a second-order reaction with first-order for CO and The flue gas stream is passed through the solid 2 NH , respectively. On the other hand, reactions (1) adsorbent such as zeolite, activated carbon and CO 3 2 and (2) are reversible, with ammonium carbonate or is held on the surface particles of the adsorbent. bicarbonate as the products. The forward reactions These adsorbents are heated to desorb the CO . 2 are dominant at room temperature. The backward Desorption is carried out either reducing the pressure reactions occur at temperatures of around 38-600 °C. or by increasing the temperature. This method is not carried out for large scale separation of CO2 from the Enhancement of CO2 absorption flue gas and hence this method is used along with Enhancement using nanofluids other methods like pressure or electric swing ads- Nanoparticles are of very small in size and dis- orption. solved in the solvent. These particles have very large Other methods surface areas compared to those of the normal par-

The other methods used for CO2 capture are ticles. Hence the increased surface area increases carbonation, calcination, biosorption, etc. Carbonation the number of gas and liquid molecules to react sim-

322 P.P. SELVI, R. BASKAR: CO2 ABSORPTION IN NANOFLUID… Chem. Ind. Chem. Eng. Q. 26 (4) 321−328 (2020) ultaneously, increasing the mass transfer rate. These based sorbents, sodium based sorbents, other metal- fluids have novel properties useful in many applicat- -based sorbents, pressure and temperature swing ions, like refrigerators, chillers, heat exchangers, adsorption, mixed technology, acid catalyst, carbon pharma industries, fuel cells, microelectronics, etc. nano tubes, microchannel reactors, etc. Solid absor- Nanofluid synthesis include metallic particles, oxide bents like calcium absorbents from limestone are low particles, carbon nanotubes, graphene and ceramic price, non-toxic in nature, and CO2 capture is easily particles. The thermophysical properties of nanofluids employed in a fluidized bed reactor at a very high include viscosity and thermal conductivities (Farghali temperature of around 500 °C. The sorbent can be et al. 2013) [3]. reused in cement production and hydrogen product- Also, the Brownian movement of the nanopar- ion. Carbon nanotubes are employed for low tempe- ticles in the solvent increases the mass transfer rate. rature and regeneration of the solvent is easy. The effect known as the grazing effect plays a major role in explaining the role of nanoparticles in the sol- EXPERIMENTAL SETUP vent. The nanoparticles adsorb the gas molecules from the bulk gas and after few seconds release the The experimental setup consists of a wetted- gas molecules into the bulk liquid. Here, the nanopar- -wall column, solvent container, liquid distributor, CO2 ticles are assumed that they only act as catalyst and vessel, pump and flow meters (Figure 1). Height and do not take part in the chemical reaction of the internal diameter of wetted wall column are 0.5 and process (Park et al. 2006, 2007, 2008) [8]. The major 0.012 m, respectively. The liquid flows inside the wet- disadvantage in this case is that the nanoparticles ted-wall column to cover the internal surface of the have to be in suspension medium in the solvent. If the wetted-wall column as a very thin layer. The liquid nanoparticle concentration increases beyond the opti- flow rate varies in the range of 0.7–1.1 L/h and the gas mized level, agglomeration of the particles and hind- flow rate is adjusted at 30 and 90 L/h. rance to the absorption of gas molecules into the bulk liquid take place. The nanofluids are employed in gas absorption studies and furthermore, external magnetic field is induced in the absorption column for the enhancement of gas absorption. Enhancement using magnetic field

CO2 absorption was enhanced with the introduction of smaller-sized nanoparticles. The nanoparticles used in CO2 absorption studies are Al2O3, TiO2,

SiO2, NiO, Fe3O4, Fe2O3, graphene oxide, etc. The iron oxide magnetic particles when dissolved in the base solvent and exposed to external magnetic field resulting in the segregation of nanoparticles in the nano fluid. As a result, the interfacial area is inc- Figure 1. Schematic diagram of wetted-wall column. reased. In the shuttle mechanism, at the gas–liquid interface nanoparticles adsorb the carbon dioxide in To generate a homogeneous magnetic field, a the film layer and after some time desorb them in the coil of 1.1 mm thick copper wire with 10000 windings solvent and this increases the rate of absorption. In is wrapped around the external surface of the abs- the Brownian mechanism the nanoparticle disturbs orption column. The coil is winded throughout the the flow and enhances the mixing of the liquid leading entire length of the inner tube of the column (0.5 m) to increase in residence time and good gas-liquid and is placed in the middle of the column. A maxi- contact within the absorption column. This accounts mum electrical current of 2 A is applied to the coil with to increase in rate of CO2 absorption by nanofluids. the aid of transformer. The nanofluid consists of a The presence of the magnetic field increases the mixture of Al2O3 and water. Average particle size of randomness of the nanoparticles in the fluid account- Al2O3 used is 53 nm. The CO2 (gas) and nanofluid ing to increased mass transfer rate. (liquid) is passed into a wetted-wall column in a counter-current manner. The CO is passed at the Other methods for absorption enhancement 2 bottom of the wetted-wall column while the nanofluid Other methods for CO2 absorption enhancement liquid is allowed to enter at the top of the column. The includes the usage of solid absorbents like calcium nanofluid starts to absorb CO2 gas. A copper coil is

323 P.P. SELVI, R. BASKAR: CO2 ABSORPTION IN NANOFLUID… Chem. Ind. Chem. Eng. Q. 26 (4) 321−328 (2020) wrapped around the wetted-wall column to generate the magnetic field. The heat generated during the process is cooled by circulating cold water through a coil which is placed between the magnetic field coil and the outer column. The magnetic field is applied to enhance higher mass flux, residence time and mass transfer coefficient. The absorbed CO2 is collected at the bottom of the wetted-wall column. The absorbed

CO2 is titrated against sodium hydroxide by adding Figure 2. SEM image of Al2O3. phenolphthalein as an indicator. By varying the flow rates and voltage, the readings would be taken with nanofluid, it was observed that the change in molar and without magnetic field. Care is taken to ensure flux with respect to solvent flow rates follows a poly- that a uniform thin layer of nanofluid is established on nomial trend. The relationship between molar flow the inner surface of the column. rate and solvent flow rate can be better described by Eq. (1): Nanofluid preparation (λ-Al2O3) = 2 Aluminum sulfate pentahydrate, Al2(SO4)3·18H2O FaS (8) (4 g) and polyethylene glycol (5 g, PEG2000) are dis- where F is molar flux and S is the solvent flow rate solved in 100 mL distilled water. Aqueous ammonia is and by neglecting the higher order terms. added to the above solution, giving rise to milky precipitates at pH 9. The mixture is then stirred for 1 h at Effect of solvent flow rate on molar flux without room temperature and subsequently placed in a Parr magnetic field and nanofluid Teflon-lined stainless steel vessel [18]. The vessel is The effect of solvent flow rate on molar flux are completely sealed and heated at 130 °C for 48 h. The carried out for 30, 60 and 90 L/h of CO2 flow rate. The mixture is gradually cooled until room temperature is observations are shown in Figure 3. reached. The resulting precipitate is filtered and It is observed that at 60 L/h of CO2 flow rate the washed three to four times using double distilled change in molar flux with respect to the change in water and ethanol to remove polyethylene glycol and solvent flow rate follows a polynomial trend. It has other impurities. The synthesized Al2O3 powder been shown in Figure 4 that the maximum flux (Figure 2) is dried at 100 °C for 1 h and is calcined at obtained at 60 L/h is 0.0058 mol/(m2s). At this point 400 °C for 2 h [19,20]. the solvent flow rate is 1.1 L/h without the aid of magnetic field. RESULTS AND DISCUSSION It was observed that as the flow rate of the solvent increases the molar flux also increases. This was From the experiments conducted in a wetted- due to the mechanism that as solvent flow rate inc- -wall column for the absorption of CO using Al O 2 2 3 reases the number of solvent molecules to absorb the

Figure 3. Molar flux vs. solvent flow rate for different CO2 flow rates: 30, 60, and 90 L/h (without magnetic field).

324 P.P. SELVI, R. BASKAR: CO2 ABSORPTION IN NANOFLUID… Chem. Ind. Chem. Eng. Q. 26 (4) 321−328 (2020)

CO2 increases. As a result of more solvent molecules, the viscosity effect shoots up.

Figure 4. Molar flux vs. solvent flow rate for CO2 flow rates of 60 L/h(without magnetic field).

Effect of solvent flow rate on molar flux with magnetic field without nanofluid Figure 5. Molar flux vs. solvent flow rate for different CO2 flow rates: 30, 60, and 90 L/h (with magnetic field). Figure 5 illustrates the influence of solvent flow rate on molar flux in the presence of the magnetic movement of solvent molecules would be promoted field, for different CO2 flow rates. by the magnetic field produced in the wetted-wall From Figure 6 it can be interpreted that when column. the magnetic field is used, greater molar flux can be Influence of nanofluid (0.0015%) on change of flux with obtained for the same molar flow rates. Molar flux of respect to solvent flow rate (without magnetic field) >0.002 mol/(m2s) can be achieved at a lower flow rate of solvent (1.1 L/h) and CO2 (30 L/h) when the mag- Figure 7 shows that there is an increase in the netic field is used. At higher CO2 flow rate of 90 L/h, rate of change of the molar flux (positive slope). The high molar flux was observed even at lower solvent higher values of molar flux are at a low CO2 flow rate flow rate. Figure 6 shows greater change in molar flux of 30 L/h. The change in flux at this flow rate is better for increasing solvent flow rates and flux as high as described in Figure 8. 2 >0.009 mol/(m2s) was observed at 1.1 L/h of solvent A maximum molar flux of 0.0014 mol/(m s) was flow rate. observed for 30 L/h of CO2 flow rate. The presence of With the presence of the magnetic field, the flux nanoparticles increases the mixing effect which was increases even in the higher solvent flow rates. The due to the Brownian motion. The increase in the

Figure 6. Molar flux vs. solvent flow rate for CO2 flow rate of 90 L/h (with magnetic field).

325 P.P. SELVI, R. BASKAR: CO2 ABSORPTION IN NANOFLUID… Chem. Ind. Chem. Eng. Q. 26 (4) 321−328 (2020)

Figure 7. Molar flux vs. solvent flow rate for different CO2 flow rates: 30, 60, and 90 L/h, using nanofluid (0.0015%) without magnetic field.

Brownian motion was due to the fact that it increases Influence of nanofluid (0.0015%) on change of flux up to a critical value and beyond this value the inter- with respect to solvent flow rate (with magnetic field) particle interaction hinders its motion. This mech- The Brownian movement and microconvection anism is known as shuttle or grazing effect. These of nanoparticles is responsible for the achievement of nanoparticles may increase the interfacial area by high mass transfer. Brownian movement of nanopar- covering the gas bubble surface and preventing the ticles is the random movement of smaller-sized par- coalescence of these bubbles [21,22]. ticles in the fluid. The mass diffusion occurs by the diffusive transport of species due to concentration difference. When the magnetic field is applied, the absorption increases better than without magnetic field by 66.10%. Nanoparticles with magnetic field promotes the Brownian movement and the grazing effect, thereby the removal efficiency increases with a high mass transfer rate [23]. Figure 9 depicts the change

of molar flux with solvent flow rates for different CO2 flow rates. The graph reveals that higher flux can be

best achieved at a lower CO2 flow rate of 30 L/h. This higher molar flux for different solvent flow rates even

at low CO2 flow rate was achieved because of the use of nanofluids (0.015 mol/(m2s)). It was also observed Figure 8. Molar flux vs. solvent flow rate for CO2 flow rate of 60 L/h, using nanofluid (0.0015%) without magnetic field. that when nanofluids were used in the presence of the

Figure 9. a) Molar flux vs. solvent flow rate for different CO2 flow rates: 30, 60, and 90 L/h, using nanofluid (0.0015%) with magnetic field;

b) molar flux vs. solvent flow rate for CO2 flow rate of 30 L/h, using nanofluid (0.0015%) with magnetic field.

326 P.P. SELVI, R. BASKAR: CO2 ABSORPTION IN NANOFLUID… Chem. Ind. Chem. Eng. Q. 26 (4) 321−328 (2020)

magnetic field, higher CO2 flow rate resulted in lower and its molecular structure, since ammonia had molar fluxes for different solvent flow rates. higher reactivity than amine solutions.

Hence, a lower CO2 flow rate was considered sufficient for obtaining better molar fluxes when REFERENCES nanofluids were used in the presence of a magnetic field. [1] H.K. Lee, M.S. Jeong, J.W. Park, H.S. Park, J.H. Cho, Kor. J. Chem. Eng. 7 (1990) 13-17 CONCLUSION [2] M. Finkenrath, Chem. Eng. Technol. 35 (2012) 482-488 [3] W. Yu, H. Xie, X. Wang, X. Wang, Phys. Lett., A 375 (2011) 1323-1328 The CO2 gas absorption studies have been carried out in a wetted-wall column. The selection of [4] M. Afkhamipour, M. Mofarahi, RSC Adv. 7 (2017) 17857- suitable solvent, comparing the performance of ran- –17872 dom and structured packing material, enhancement of [5] H. Chen, Y. Ding, C. Tan, New J. Phys. 9 (2007) 367-382 absorption by adding nanomaterials in solvent and [6] E. Alper, B. Wichtendahl, W.D. Deckwer, Chem. Eng. Sci. the induction of a magnetic field were studied in 35 (1980) 217-222 [7] Y. Xuan, W. Roetzel, Int. J. Heat Mass Transfer 43 (2000) detail. An effort has been made to enhance CO2 3701-3705 sequestration with the nanofluid addition and in the [8] R. Krupiczka, A. Rotkegel, Z. Ziobrowski, Renew. presence of a magnetic field. The mass transfer flux Sustain. Energy Rev. 96 (2018) 502-525 increases with the magnetic field as opposed to with- [9] L.S. Tan, A.M Shariff, K.K. Lau,M.A.Bustam, J. Ind. Eng. out the magnetic field. When nanofluid (0.0015%) is Chem. 18 (2012) 1874-1883 used along with the magnetic field, the flux increases. [10] C. Chen, J. Kim, W. Ahn, Kor. J. Chem. Eng. 31 (2014) Further study can be carried out for a higher ratio of 1919-1934 nanofluids (>0.0015%). For the nanofluids, as the [11] S. Srivastava, G. Preet, Phys. Chem. Liq. 53 (2015) 174- volumetric ratio increases the absorption also inc- –186 reases. Hence, the acid gas absorption has been car- [12] S.S. Ashrafmansouri, M.N. Esfahany, Int. J. Therm. Sci. ried out in a wetted-wall column with higher mass 82 (2014) 84-89 transfer rate, and lower solvent flow rate and CO2 [13] J.K. Kim, J.Y. Jung, Y.T Kang, Int. J. Refrig. 29 (2006) gas flow rate. The following conclusions were drawn 22-29 out of the study: [14] H. Zhu, B.H. Shanks, T.J. Heindel, Ind. Eng. Chem. Res. 1. Among different solvents considered in this 47 (2008) 7881-7887 study, NH3 was found to be better for CO2 gas abs- [15] J. Jiang, B. Zhao, M.Cao, S. Wang, Energy Prod. 37 orption. It was observed that the structured packing (2013) 518-524 resulted in higher efficiency as opposed to random [16] C. Pang, J.W. Lee, Y.T. Kang, J. Mech. Sci. Technol. 28 packing. (2014) 2925-2936

2. It was also observed that the % of CO2 rem- [17] Y.T. Kang, H. J. Kim, K. Lee, Int. J. Refrig. 31 (2008) 850- –856 oval was found to increase with increase in NH3 concentration. [18] S. Krishnamurthy, P. Bhattachara, P.E. Phelan, Nano 3. With the introduction of nanofluids the per- Lett. 6 (2006) 419-423 [19] J. Salimi, M. Haghshenasfard, A. Moheb, Chem. Eng. centage of CO2 removal increased. Increase of Technol. 37 (2014) 462-470 ammonia concentration and decrease of temperature have a positive effect on the density and viscosity. [20] J. Salimi, M. Haghshenasfard, A. Moheb, Heat Mass Transfer 51 (2015) 621-629 Aqueous ammonia also has a non-volatile state and a [21] Z. Zhang, Y. Li, W. Zhang, J. Wang, M. R. Soltanian, A. high-surface tension, which could be useful for pro- G. Olabi, Rew. Sus. Energy Rev. 98 (2018) 179-188 viding a better CO absorption performance. 2 [22] N. Zhang, Z. Pan, L. Zhang, Z. Zhang, Fuel 242 (2019) 4. pK is a major physicochemical parameter a 470-478 which was used to analyse the reaction mechanism [23] Z. Zhang, J. Cai, F. Chen, H. Li, W. Zhang, W. Qi, Rew. Energy 118 (2018) 527-535.

327 P.P. SELVI, R. BASKAR: CO2 ABSORPTION IN NANOFLUID… Chem. Ind. Chem. Eng. Q. 26 (4) 321−328 (2020)

1 PONGAYI PONNUSAMY SELVI APSORPCIJA CO2 NANOFLUIDOM U MAGNETNOM RAJOO BASKAR2 POLJU 1Department of Chemical Engineering,

Kongu Engineering College, Erode, Kiseli gasovi, kao što su CO2, SO2, NO2, H2S i drugi, treba ukloniti, jer zagađuju atmo- India sferu na ovaj ili onaj način izazivanjem porasta temperature, što dalje ima za rezultat 2 Department of Food Technology, neželjene klimatske promene. Među svim ovim gasovima, CO2 ima najveći uticaj na Kongu Engineering College, Erode, životnu sredinu, pa je njegovo “hvatanje” od primarne važnosti. Cilj ovog rada je pobolj- India šanje apsorpcije CO2 primenom nanofluida u prisustvu magnetnog polja. Nanofluid koja

se koristi u ovom radu je Al2O3/voda u koncentraciji 0,0015%. Dobijeni maksimalni fluks NAUČNI RAD je 0,014 mol/(m2s) (bez magnetnog polja) i 0,015 mol/(m2s) (sa magnetnim poljem) za

protok CO2 manji od 30 L/h. Otuda, nanofluidi, zajedno sa magnetnim poljem, pokazuju

pozitivan uticaj na apsorpciju CO2.

Ključne reči: CO2, fluks, protok, globalno zagrevanje.

328 Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 329−335 (2020) CI&CEQ

D. REVATHI EXPERIMENTAL STUDIES ON Dr.K. SARAVANAN HYDRODYNAMIC ASPECTS FOR MIXING Department of Chemical OF NON-NEWTONIAN FLUIDS IN A Engineering, Kongu Engineering KOMAX STATIC MIXER College, Perundurai, India

SCIENTIFIC PAPER Article Highlights • Performance of a Komax static mixer was determined for non-Newtonian ﬂuid streams UDC 66.063:544:53 • Pressure drop studies were carried out for different flow rates with different concentrations • The results of the experimental runs for the designed elements were presented • Complete mixing is achieved with pressure losses of 75% for the designed element • Pressure losses for the designed element are less than that of conventional static mixers

Abstract Mixing is the degree of homogeneity of two or more phases and it plays a vital role in the quality of the final product. It is conventionally carried out by mechanical agitators or by static mixers. Static mixers are a series of geometric mixing elements fixed within a pipe, which use the energy of the flow stream to create mixing between two or more fluids or to inject metered liquid into a continuous process. The objective of this work is to predict hydrodynamic aspects of the static mixer designed. The mixing performance of a Komax static mixer has been determined for the blending of non-Newtonian ﬂuid streams with identical or different rheology by using experimental study. The energy needed for mixing comes from the force created by the liquid due to turbulence and the geometry of the static mixer. Pressure drop in the static mixer depends strongly on the geometric arrangement of the inserts, properties of fluids to be mixed and flow conditions. Hence, pressure drop studies are carried out for different flow rates of fluids with different concentrations of two non-Newtonian fluids. Starch and xanthan gum solutions are used as working fluids. It is observed from the experimental results that the pressure drop per unit length increases as the fluid flow rate increases, and the nature of fluid flow varies with the velocity of the fluids. Keywords: pressure drop studies, mixing performance, non-Newtonian fluid blending, fluid rheology, static mixer.

Mixing is a vital unit operation that ensures the of the product and process [1]. Most of the industrial homogeneity of phases by converting a heterogene- processes employ mixing in one form or another. ous mixture to a homogeneous mass by dissolution, However, those industries which employ processing dispersion, and diffusion. It enables the transfer of of jams, inks, polymers and other viscous fluids result heat and mass across one or more streams or phases in mixing operations under laminar flow conditions in and influences the operating time, safety and quality order to homogenize the product. In laminar flow conditions, the level of mixing is low which results in non- homogeneities in composition [1,2]. Mixing in food Correspondence: D. Revathi, Faculty of Chemical Engineering, industries encounters many challenges in maintaining Kongu Engineering College, Erode, Tamil Nadu, India. the homogeneity, consistency and enhancement of E-mail: [email protected] Paper received: 17 October, 2019 the texture of final products as they are processed Paper revised: 27 January, 2020 under laminar conditions. Paper accepted: 28 February, 2020 https://doi.org/10.2298/CICEQ191017009R

329 D. REVATHI, DR.K. SARAVANAN: EXPERIMENTAL STUDIES ON HYDRODYNAMIC… Chem. Ind. Chem. Eng. Q. 26 (4) 329−335 (2020)

A static mixer is a device designed incorporating fluids are fewer. Some such studies have shown that motionless elements in a tubular arrangement which these fluids have better mixing quality than Newtonian may enable the flow of fluids in various directions fluids [12]. Shear-thinning fluids get deformed readily, based on the design in order to provide continuous as the shear-thinning viscosity enhances mixing. For mixing of fluids. The metallic elements cause inter- low elasticity fluids, mixing is enhanced till shear- material contact due to fluid splitting and recombining, thinning effects appear [13]. The first study on the and thus enhance mixing in systems ranging from hydrodynamic aspects of non-Newtonian fluids has simple blending operations to chemical reactions and been done recently at low primary flow rates for cost- heat transfer [3]. Further, in static mixers, the flow is effective mixing [14]. So, this paper focuses on the split into layers which are rearranged and forced to experimental study of hydrodynamic aspects of mix- flow in radial direction. The first ever static mixer ing non-Newtonian fluids for a Komax static mixer. designed was a single element mixer [4]. Static mixers have numerous advantages com- MATERIALS AND METHODS pared to those of the conventional mixers. They are cost-effective and consume less power. The absence Experimental setup of moving parts makes it easier for maintaining them. The experimental study on the mixing of non- The compact design makes a static mixer easy to Newtonian fluids is carried out in a continuous flow install and it occupies less floor space. The mixing setup, as depicted in Figure 1. The experimental elements are either self-cleaning or disposable, thus setup consists of two gear pumps each of 5 L/min preventing erosion or scale formation [5]. There is a capacity to pump the solutions. Starch and xanthan wide range of static mixers like Komax, Kenics, SMX, gum are used as working fluids. Electrical resistance etc. that are commercially available. Very few res- tomography depicts the effect of xanthan gum sol- earches have highlighted the performance of static ution rheology on the quality of mixing in the chaotic mixers [6,7]. Recent studies have compared the per- SMX static mixer, showing that the xanthan gum sol- formance of all the commercially available static ution at higher concentration results in a more homo- mixers with pressure drop and length as criteria [8]. genous mixing, as given in literature [15]. Kenics and Multiflux mixers were initially deve- Six Komax static elements with opposite twists loped to mix fluids under laminar conditions [9]. Static are interlocked with each other in series and are mixers have evolved over the years offering a wide placed in the pipe of a specified dimension. Fluid flow variety of applications in almost all chemical process is determined using flow sensors and pressure is industries that involve mixing of chemical reactants, measured at the inlet and outlet of the static element heating and cooling, scrubbing, stripping and water setup using pressure transducers. Pressure across treatment. Inline mixers provide quality processing of the pipe section is displayed in the Arduino board, edible products that are viscous in nature. They can which converts and records the signals from the be used in industries that process jams, oils, juices, pressure transducers into analog signals. A mesh is chocolates, beverages and dairy products [10]. Rec- placed at a distance of 8 cm from the inlet of the pipe ent application of static mixers was as catalyst sup- to prevent the formation of a dead zone. The spe- port for heterogeneous photocatalysis [11]. The stu- cification of the static mixer setup is provided in Table dies on static mixer applications for non-Newtonian 1. The static mixer designed is shown in Figure 2.

Figure 1. Schematic representation of the experimental setup.

330 D. REVATHI, DR.K. SARAVANAN: EXPERIMENTAL STUDIES ON HYDRODYNAMIC… Chem. Ind. Chem. Eng. Q. 26 (4) 329−335 (2020)

Table 1. Specifications of the static mixer mixing is determined by collecting samples from several ports located along the length of the pipe. The Specification Dimension mixed fluid, thus obtained, drains to the outlet tank Inner diameter of the pipe 25 mm through the outlet nozzle. The experiments are rep- Outer diameter of the pipe 32 mm eated by varying the flow rates and the concentration Thickness of the pipe 3.5 mm of process fluids in order to study the mixing perform- Length of the pipe 290 mm ance of static mixer. Length of the mixing element 210 mm Development Zone 80 mm Table 3. Specifications of the flow sensor Number of elements 6 Details of flow sensor Technical specification Diameter of the static element 24 mm Manufacturer Sea Thickness of static element 1.2 mm Model YF-5301 Number of twists 4 o Working range 0.3-10 L/min Angle of twist 30 Working pressure < 0.5 MPa Sensor material Plastic

Table 4. Specification of the pressure transducer

Details of pressure transducer Technical specification Manufacturer Generic Working voltage DC, 5.0 V Output voltage DC, 0.5-4.5 V Sensor material Carbon steel alloy Working pressure range 0-1.2 MPa

Power requirement The energy consumed by a static mixer can be Figure 2. Designed static element and static mixer: a) static element before twist; b) static element after twist; c) arrange- determined using the following equation: ment of six static elements in series; d) designed static mixer. Power =ΔQP (1)

Experimental procedure The energy is supplied by the pump used to create the flow of the fluid through the mixer for The starch and xanthan solutions of specific homogenization of two or more liquids. Static mixers concentrations are pumped to a static mixer setup reduce standard deviation or coefficient of variance. using rotary gear pumps powered by motors. The The reduction of variance is the product of shear rate specification of the rotary gear pump is provided in and time, and variance is equal to L/D . Early inves- Table 2. The flow sensor detects the flow rate of t tigations into the performance of static mixers have process fluids and its specifications are provided in focused on the pressure drop across the mixer with Table 3. The process fluid enters the mixer pipe various fluids. It is found that at low Reynolds num- through the inlet nozzle. The fluids split and are bers, there is an increase in pressure drop as fluid recombined inside the pipe by mixing elements. Fur- elasticity increases [16,17]. Depending on the flow ther, the inlet and outlet pressure are determined rate and liquid viscosity, the pressure drop varies. In using the pressure transducers P1 and P2, respect- laminar flow conditions, the performance varies widely ively. The specifications of the pressure transducers for different mixers and mixing efficiency does not are provided in Table 4. Likewise, the intensity of the necessarily follow Sauter mean drop diameter, but pressure drop needs to be considered. For a Kenics Table 2. Specifications of rotary gear pumps mixer, ∆Pmixer is about 6 times that of an empty pipe in Details of pump Technical specification laminar flow, and for a Sulzer SMX mixer it is 64 times Manufacturer Fluid Tech Systems higher. The pressure drop in a static mixer depends Model FT-025 strongly on the geometric arrangement of the inserts. Suction and delivery size ¼ ”X ¼ ” It is simply defined in relation to the pressure drop ∆P Capacity at 1440 rpm 5 L/min in an empty tube given by Darcy’s equation: Pressure 10 kg/cm2

331 D. REVATHI, DR.K. SARAVANAN: EXPERIMENTAL STUDIES ON HYDRODYNAMIC… Chem. Ind. Chem. Eng. Q. 26 (4) 329−335 (2020)

ρ 2  Δ= LV ε = L Pf (2) Z  (8) Dt 2 Dt

ΔP It is suggested that the product of the pressure K = mixer (3) Δ Pempty pipe drop ratio (Z) and the value of L/Dt would provide a useful quantitative estimate of mixer efficiency given K depends on the mixer type and its value can be in Eq. (8). This is less ambitious than the ratio values obtained from the manufacturer’s literature. used in previous comparative tests, and provides a Energy and efficiency of mixing basis for the comparison of data from various sources.

Mixing efficiency is accomplished by controlled RESULTS AND DISCUSSION vortex structures generated by the patented low- pro- file tab geometry. This provides uniform blending Rheological properties while limiting the mixer length to less than half of the The rheological behavior of the process fluids pipe diameter. Complete mixing is achieved with like xanthan gum and starch is determined by plotting pressure losses of 75% which are less than that of shear stress as a function of shear rate for varying conventional static mixers [18]. An understanding of compositions. The starch and xanthan solutions are the energy requirement of a static mixer is necessary prepared by first dissolving the appropriate amount of with respect both to the establishing of installed pres- starch and xanthan into deionized water warmed to sure drop and flow rate requirement. The empirical about 40 °C to make a solution with a concentration of relation for friction factor is as follows [19]: varying weight percentages. Rheological measure- 2DPΔ ments are carried out using a Brookfield DV-III ultra f = t (4) ρVL2 programmable rheometer. Both starch and xanthan- -based fluids show a slight shear-thinning for low ρ = VDt shear rates. Figure 3 shows the consolidated graph NRe (5) μ depicting the rheological behavior of process fluids at the concentrations of: i) 2 and 3% starch solution, and The two dimensionless groups are based on the ii) 0.1 and 0.2% xanthan gum solution. empty pipe diameter (D ), including the value of V, t The shear-thinning behavior of process fluids which is a superficial velocity. In principle, it is has been depicted by plotting the viscosity as a func- required only for the value of ∆P to be measured at tion of shear rate. Viscosity is reported as function of one value of V to define the product fN for any static Re shear rate for starch and xanthan solutions at dif- mixer, but a range of measurements can give useful ferent concentrations. Figure 4 shows the decrease in information regarding the slight volubility of fN and Re viscosity as the shear rate increases. the upper limit of the laminar flow regime in terms of

NRe. In more recent studies, a simple method has been used where the pressure drop characteristics are described as a ratio of mixer pressure drop and empty pipe pressure drop for the same flow rate and diameter (Harnby). The mixer pressure drop ratio is commonly measured by the design factor K, (Z-factor) which is defined as follows [20]: f K = mixer (6) fempty pipe

ΔP ()fN Z ==mixerRe mixer (7) ΔPfN() empty pipe Re empty pipe

As both fNRe and NRe are constant for any mixer or pipe in the laminar regime, it follows that Z is also constant. Further, the relationship among various Figure 3. Shear stress vs. shear rate. parameters is very simple to determine:

332 D. REVATHI, DR.K. SARAVANAN: EXPERIMENTAL STUDIES ON HYDRODYNAMIC… Chem. Ind. Chem. Eng. Q. 26 (4) 329−335 (2020)

Effect of friction factor and Reynolds number on mixing efficiency Mixing efficiency is determined by plotting the Z(K) factor as a function of the Reynolds number. The friction factor is calculated using Eq. (4) while Z-factor is calculated using Eq. (7). Figure 6 depicts the variation of the friction factor as a function of the Rey- nolds number. As fluid flows through the mixer tube, pressure drop across the tube occurs with frictional forces on the fluid. The friction factor, f for static mixers, is high at laminar regime and decreases gradually, almost to constant value with the increase in the Reynolds number, which ensures turbulence. The friction factor in Komax mixers varies with twist angle, Figure 4. Viscosity vs. shear rate. thickness, and length of the static element. At turbulent flow conditions, the number of mixing elements Pressure drop observations decreases with flow velocity approaching constant The pressure drop is found using pressure values for high Reynolds numbers. The mixing effici- transducers placed at the inlet and outlet of the static ency is high under laminar flow conditions as well, but mixer pipe. Pressure drop for different flow rates of gets reduced as the flow begins to approach turbulent starch and xanthan at varying concentrations across conditions. It is obvious that under a turbulent regime, the empty tube and mixer tube with elements are as velocity increases, the pressure increases as this investigated and the readings are noted. The pres- is proportional to the square of velocity. sure drop is calculated using Eqs. (2) and (3). Plotting pressure drop against set flow rate and measured using a flow sensor shows the variation of the former. An increase in pressure drop is observed with the increase in flow rate. The results remain the same when the experiments are carried out by varying flow rates. For a turbulent flow system, the greater value of pressure drop shows better mixing in a static mixer system. The variation of pressure drop with respect to flow rate is shown in Figure 5.

Figure 6. Friction factor vs. NRe.

Effect of pressure drop ratios, Z- factor, on mixing efficiency Pressure drop ratios, Z, for a static mixer depend on the design of the mixer and the Reynolds number. Figure 7 shows the variation of the Z-factor with respect to the Reynolds number. The development zone provided in the static mixer ensures the negligible backmixing. The mixing efficiency of a sta- Figure 5. Pressure drop vs. flow rate. tic mixer is the product of pressure drop ratios and the mixing length required to achieve a specified CoV of 0.05. In order to assess the accuracy of the experiment, the Z-factor results are compared with the lite-

333 D. REVATHI, DR.K. SARAVANAN: EXPERIMENTAL STUDIES ON HYDRODYNAMIC… Chem. Ind. Chem. Eng. Q. 26 (4) 329−335 (2020) rature. From literature, measured pressure drop ratio, ρ Density of the fluid (kg/m3)

Z for Komax mixer, is found to be 25 for a L/Dt ratio in V Velocity of the fluid (m/s) the range of 29–38 and as per the investigation it is µ Viscosity of the fluid (kg/m s) found to be around 23 for a L/Dt ratio of 8.4. The Z Pressure drop ratios for static mixer 2 available experimental data shows variability due to ∆Pmixer Pressure drop in static mixer (N/m ) 2 slight difference in measurement locations. Mixer sel- ∆Pempty pipe Pressure drop in empty pipe (N/m ) ection is best based on mixing efficiency, design with fmixer Darcy’s friction factor for fluid in static mixer low pressure drop and shorter length. The investi- fempty pipe Darcy’s friction factor for fluid in empty pipe gation shows that this designed mixer with a compar- NRe Reynolds number atively less L/Dt ratio is a better design for the sel- e Mixing efficiency of the static mixer ected fluid system. Dimensionless numbers ρVD Reynolds Number, N = t Re μ

REFERENCES

[1] A. Ghanem, T. Lemenand, D. Della Valle, H. Peerhos- saini, Chem. Eng. Res. Des. 92 (2013) 205-228 [2] J. Ramsay, M.J.H. Simmons, A. Ingram, E.H. Stitt, Chem. Eng. Sci. 139 (2016) 125–141 [3] E.L. Paul, V.A. Atiemo-Obeng, S.M. Kresta, Handbook of Industrial Mixing Science and Practice, John Wiley & Sons, Hoboken, NJ, 2004, p. 1017 [4] W. S. Sutherland, (Sutherland), UK, Patent No. 206,642 (1874) [5] R. K. Thakur, Ch. Vial, K. D. P. Nigam, E. B. Nauman and G. Djelveh, Trans. IChemE 81 (2003) 787–826 Figure 7. Z-Factor vs. NRe. [6] M. H. Pahl, E. Muschelknautz, Chem. Ing. Tech. 52 (1980) 285–291 CONCLUSION [7] D. Rauline, JML Le Blevec, J. Bousquet, P.A. Tanguy, Trans. IChemE 78 (2000) 389–396 In this present study, a static mixer (Komax) is [8] H.E.H. Meijer, M.K. Singh, P.D. Anderson, Prog. Polym. designed, fabricated and tested in a pipeline for inline Sci. 37 (2012) 1333–1349 mixing. From experimental result, it is found that the [9] H.P. Grace, Chem. Eng. Commun. 14 (1982) 225–277 pressure drop per unit length increases with the inc- [10] A. Cybulski, K. Werner, Int. Chem. Eng. 26 (1986) 171– rease in flow rate, and for a turbulent flow system, –180 better mixing is achieved for higher values of the [11] A.M. Díez, F.C. Moreira, B.A. Marinho, J.C.A. Espíndola, pressure drop. Rheological behavior of various fluids L.O. Paulista, M.A. Sanromán, M. Pazos, R.A.R. Boa- is studied for the absolute viscosity of the process ventura, V.J.P. Vilar, Chem. Eng. J. 343 (2018) 597-606 fluid. The increase in velocity is twofold for turbulent [12] S. Liu, A.N. Hrymark, P.E. Wood, Chem. Eng. Sci. 61 flow and it is directly proportional to velocity for lami- (2006) 1753–1759 nar flow. The friction factor decreases with the inc- [13] P.E. Arratia, G.A. Voth, J.P. Gollub, Phys. Fluids 17 rease in the Reynolds number, and a sharp decrease (2005) 1-10 is observed for the laminar region. In further study, [14] S. Jegatheeswaran, F. Ein-Mozaffari, J. Wu, Chem. Eng. CFD modeling for the flow profile of static mixing sys- Process. 124 (2018) 1–10 tem has to be done and compared with the experi- [15] S. Jegatheeswaran, F. Ein-Mozaffari, J. Wu, Chem. Eng. mental run. J. 317 (2017) 215–231 [16] N.F Shah, D.D Kale, Chem. Eng. Sci. 46 (1991) 2159– Nomenclature –2161 3 Q Flow rate (m/s) [17] K.G. Chandra, D.D. Kale, Chem. Eng. Sci. 47 (1992) ΔP Pressure drop (N/m2) 2097–2100 f Friction factor [18] L.F. Albright, Albright’s Chemical Engineering Handbook,

L Length of the static mixing element (m) CRC Press Taylor & Francis Group, New York, 2009, pp. 683-694 Dt Inner diameter of the static mixer tube (m)

334 D. REVATHI, DR.K. SARAVANAN: EXPERIMENTAL STUDIES ON HYDRODYNAMIC… Chem. Ind. Chem. Eng. Q. 26 (4) 329−335 (2020)

[19] W.L. Wilkinson, M.J. Cliff, An Investigation into the [20] M. Laporte, C. Loisel, D. Della Valle, A. Riaublanc, A. Performance of a Static In-Line Mixer, in Proceedings of Montillet, J. Food Eng. 128 (2014) 119–126. Second European Conference on Mixing, Cambridge, England, 1977, pp. A2-15–A2-29

D. REVATHI EKSPERIMENTALNO ISTRAŽIVANJE DR.K. SARAVANAN HIDRODINAMIKE MEŠANJA NENJUTNOVSKIH Department of Chemical Engineering, TEČNOSTI U STATIČKOM MEŠAČU KOMAX Kongu Engineering College, Perundurai, India Mešanje utiče na stepen homogenosti dve ili više faza i igra vitalnu ulogu u kvalitetu konačnog proizvoda. Konvencionalno se vrši mehaničkim mešalicama ili statičkim NAUČNI RAD mešačima. Statički mešači su niz geometrijskih elemenata za mešanje pričvršćeni u cevi, koji koriste energiju strujanja za mešanje dve ili više tečnosti ili ubrizgavanje odme- rene količine tečnosti u kontinualni proces. Cilj ovog rada je da predvidi hidrodinamičke parametre projektovanog statičkog miksera. Istražene su performanse statičkog mešača Komax za mešanje nenjutnovskih tečnosti sa identičnom ili drugačijom reologijom. Energija potrebna za mešanje potiče od sile koju tečnost stvara zbog turbulencije i geo- metrije statičkog mešača. Pad pritiska u statičkom mešalču jako zavisi od geometrijskog rasporeda umetaka, svojstava fluida koji se mešaju i uslova protoka. Zbog toga je istra- žen pad pritiska za različite protoka tečnosti sa različitim udelima dve nenjutnovske teč- nosti. Kao radne tečnosti, koršćeni su rastvori skroba i ksantana. Iz eksperimentalnih rezultata se zapaža da se pad pritiska po jedinici dužine povećava sa povećanjem protoka fluida, pri čemu priroda strujanja fluida varira u zavisnosti od njegove brzine.

Ključne reči: pad pritiska, efekat mešanja, mešanje nenjutnovskih tečnosti, reo- logija fluida, statički mešač.

335

Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 337−347 (2020) CI&CEQ

SANJA RADONJIĆ1 INCIDENCE OF VOLATILE PHENOLS IN HELENA PROSEN2 MONTENEGRIN RED WINES: VRANAC, VESNA MARAŠ1 3 KRATOŠIJA AND CABERNET SAUVIGNON LEA DEMŠAR 3 TATJANA KOŠMERL Article Highlights 1Research and Development • VP were found mostly in Kratošija wines, followed by Vranac • Sector, “13. Jul Plantaže“ a.d., Higher 4-EP concentrations were formed in Vranac wines with BM4x4 yeast and without LAB addition Podgorica, Montenegro 2 • Higher 4-EP content was observed in Kratošija wines with BDX yeast and without Faculty for Chemistry and LAB addition Chemical Technology, University of • Higher 4-EG content was observed in Kratošija wines with BM4x4 yeast and without Ljubljana, Ljubljana, Slovenia LAB addition 3 Biotechnical Faculty, University of • VP can appear in aged wine with low pH, high alcohol content and free SO2 (∼35 mg/L) at Ljubljana, Ljubljana, Slovenia bottling

SCIENTIFIC PAPER Abstract The main objective of this study was to determine the impact of three different UDC 663.2(497.16):66 commercial yeasts and two lactic acid bacteria, as well as the use of oak (chips and staves) on the volatile phenol (VP) formation in Montenegrin aged red wines (Vranac, Kratošija and Cabernet Sauvignon) of the 2013 vintage. Standard chemical parameters of the wines were determined in order to establish relationship with VP. After three years of bottle aging, chemical para-

meters of wines (alcohol content, residual sugars (RS), pH, free and total SO2, anthocyanins) and VP (4-ethylphenol and 4-ethylguaiacol) were determined. Volatile phenols were found in 18 samples of Kratošija and 5 samples of Vra- nac. Volatile phenols were not detected in Cabernet Sauvignon wines. Krato- šija as a grape variety from the Montenegrin terroir during 2013 vintage had the highest potential for the VP formation, while the highest content of 4-ethylphenol (2.84 mg/L) and 4-ethylguaiacol (0.52 mg/L) was determined in Vranac wine. It was demonstrated that VP would be formed in wines even with high

alcohol content (>12.50 vol.%), low pH (<3.30) and with free SO2 concentrations at bottling of 30-35 mg/L. Keywords: red wine, yeast, lactic acid bacteria, oak alternatives, volatile phenols.

Volatile phenols (VP), i.e., 4-ethylphenol (4-EP) pounds in wine are commonly described as phenolic, and 4-ethylguaiacol (4-EG) in wine are the products horse sweat, leather, stable or medicinal [2]. Hyd- of the metabolism of Brettanomyces spp. from phe- roxycinnamic acids are the precursors of VP and are nolic acids that are derived in grape and in wine [1]. natural compounds of grapes and wine. Metabolism Their production leads to a loss of freshness and fruit- of Brettanomyces spp. consists of two enzymes with iness [1], and the off-flavour related to these com- consecutive activities. First, the phenolic acids are cleaved by cinnamate decarboxylase resulting in the formation of the corresponding vinylphenol and vinyl- Correspondence: S. Radonjić, Production Sector, “13. Jul Plan- guaiacol (p-coumaric acid is cleaved to 4-vinylphenol taže“ a.d., Put Radomira Ivanovića, 81000 Podgorica, Monte- and ferulic acid is cleaved to 4-vinylguaiacol), fol- negro. E-mail: [email protected] lowed by a conversion of 4-vinylphenol into 4-EP and Paper received: 13 August, 2019 4-vinylguaiacol into 4-EG using vinylphenol reductase Paper revised: 24 October, 2019 [ ] Paper accepted: 5 March, 2020 3 . It is known that Brettanomyces yeasts can grow during bottle storage over long periods and they can https://doi.org/10.2298/CICEQ190813010R

337 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020) produce 4-EP and 4-EG, that exceeds critical olfac- to positive or negative sensory properties of wine tory thresholds during the first months of storage [4]. [17]. Lactic acid bacteria can detoxify phenolic acids, Other yeast species, such as Saccharomyces cere- or use them as alternative energy sources [18]; thus, visiae have only cinnamate decarboxylase activity, their cinnamoyl esterase activity has an important role resulting in the 4-vinylphenol and 4-vinylguaiacol [19]. Several studies reported the presence of phen- formation. B. bruxellensis can utilize the free 4-vinyl- olic acid decarboxylase, which is one of the enzymes phenol produced by other organisms present in the involved in the formation of VP, in O. oeni, as well as medium to form 4-EP [5]. The concentrations of differ- in L. plantarum [20]. Cappello [21] highlighted the use ent VP may vary depending on the grape variety, of starter cultures for induction of malolactic ferment- yeast strains, viniculture conditions and winemaking ation (MLF) to improve wine quality. Cappello [21] practices [1]. In addition to resistance to high ethanol also reported that LAB play an important role in the content, SO2, as well as oxygen and sugar depletion, development of wine aroma. Standard chemical para- Brettanomyces bruxellensis is one of the yeast spe- meters of Montenegrin red wines (Vranac, Kratošija cies that is well habitual in dry red wine medium [6,7]. and Cabernet Sauvignon) were done in order to est- Variation of the pH, temperature and nutrient content, ablish relationship with VP to improve the overall as well as the concentration of total anthocyanins can quality of wine. Microbiological and technological affect the SO2 activity. It was shown that this species parameters that affect VP production were studied in is tolerant to the high ethanol concentrations in wine the wines of autochthonous Montenegrin grapevine (14.5-15 vol.%) [6]. It was also reported that even if varieties Kratošija and Vranac and international grape the autolysed S. cerevisiae is present and if the con- variety Cabernet Sauvignon. Based on literature data, centrations of sugar are below 150 mg/L, Brettano- grape variety Kratošija has been present earlier and myces still have the ability to grow and produce ethyl- was cultivated before Vranac variety [22]. Work on phenols [8]. A possible approach toward reduction of autochthonous grapevine varieties is of interest con- the precursors of ethylphenols is the use of hydroxy- sidering that Vranac and Kratošija are related (parent- cinnamate decarboxylase activity positive S. cerevi- offspring) and it is proposed that Montenegro is the siae strains that are able to carry out the alcoholic best variety for the origin and spreading centre of fermentation [9]. The formed vinylphenols can react Kratošija and its group of synonyms (Primitivo, Zin- with grape anthocyanins, resulting in the formation of fandel and Crljenak Kaštelanski) [23]. The main aim vinylphenolic pyranoanthocyanins. The presence of of the investigation was to determine the impact of these molecules that are stable under oenological different commercial yeasts and lactic acid bacteria, conditions can reduce the concentration of free ethyl- as well as the use of oak alternatives on VP formation phenol precursors [10]. in Montenegrin aged red wines. Wine quality degradation caused by the presence of VP can present an economic problem [11], MATHERIAL AND METHODS particularly with respect to top quality wines that mature for a long period in oak vessels [12]. In addition, Grapes and winemaking Dekkera intermedia and Brettanomyces custersii deg- The trial was carried out during 2013. The har- rade a polysaccharide which is a structural wood vest of the three grape varieties (Vranac, Kratošija compound, by metabolizing cellobiose [12]. Matur- and Cabernet Sauvignon) was performed manually. ation in oak vessels is still the number one choice in The grapes that were used for this study from Krato- the production of premium quality wine. Cost effective šija and Vranac vineyards are at the same micro- alternatives are available, but it has been related to locality (Dinoš) at Ćemovsko field (2310 ha), while the spoilage caused by the Brettanomyces presence Cabernet Sauvignon is planted at the micro-locality in wine [13]. These alternatives include different sizes (Pista) at Ćemovsko filed. PVC barrels (500 L) were and shapes of oak shavings, chips and pieces [14]. used for alcoholic fermentation within all trials and It was also shown that beside B. bruxellensis traditional methods were applied. Potassium metabi- there are other microorganisms that produce 4-EP sulphite (Agroterm KFT, Hungary) was added (8 g/ and 4-EG, such as some lactic acid bacteria and non- /100 kg of grapes). Enzymes, yeast, lactic acid bac- Saccharomyces yeasts [10]. Some Lactobacillus spe- teria and nutrient were purchased from Lallemand, cies have the ability to form these compounds, France. For alcoholic fermentation for all examined although in lower quantities [15,16]. Several studies varieties, three commercially available yeasts that are have also reported on the VP formation due to the used for red wine production were chosen, i.e., Lalvin LAB activity using hydroxycinnamic acid, which lead ICVD21, Lalvin BM4x4 and Enoferm BDX. These

338 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020)

yeasts were added in quantity of 30 g/hL. Beside the SO2. For the determination of wine chemical para- trials with addition of different yeasts, the control wine meters, methods of [24] the European Union, Com- without the addition of commercial yeast was also mission regulation (EEC) No. 2676/90 were used. For made. All chosen yeasts were Saccharomyces cere- determination of the total anthocyanins, pH differential visiae with some differences in microbial and oeno- method was applied [25]. logical properties. In the beginning of alcoholic fer- Determination of 4-ethylphenol and 4-ethylguaiacol mentation, a stimulant and protectant for yeast (Go- content Ferm Protect) was added (30 g/hL) and the enzyme for maceration Lalvin EX-V (2 g/100 kg). Yeast nut- Volatile phenols (VP) were determined after rient, Fermaid E (25 g/hL) was added after 1/3 of three years of bottle aging, in three different bottles alcoholic fermentation passed. When alcoholic fer- per each sample. using solid phase micro extraction mentation was finished, racking was done and wines, (SPME) with GC-MS method, which was modified which were inoculated with commercial yeast, were and previously described by [26]. Compounds were separated in three equal quantities (lots) for malo- sampled for 45 min at 50 °C and fibre desorption for 5 lactic fermentation. Two wine lots were induced by min at 250 °C. commercial LAB and the third underwent spontane- The temperature program on the GC column ous MLF. For inducing MLF, Lalvin VP41 and Eno- Rtx-20 was 50 °C (2 min) - 12 °C/min - 180 °C - 50 ferm ALPHA were used (1 g/hL). Commercial LAB °C/min - 240 °C (20 min). Carrier gas was helium at a selected for MLF inoculation were Oenococcus oeni constant pressure. Temperature of the injector was strains that possess features aimed particularly for 250 °C, temperature of the GC-MS interface was 250 winemaking conditions in term of tolerance to high °C. The linear range for the determination of 4-EP alcohol content and low pH. Additionally, LAB are cin- and 4-EG was from 0.02 to 0.75 mg/L. LOQ for both namoyl esterase negative, which means they cannot compounds was 0.02 mg/L, while LOD was 0.006 form ethylphenol precursors. Ten samples were mg/L. obtained within each grape variety after MLF, i.e., Statistical data analysis control wines (without addition of commercial yeast), The data were assessed statistically using the BDX-ALPHA, BDX-CTRL, BDX-VP41, BM 4X4- SAS (SAS/STAT, 1999). The basic statistical para- -ALPHA, BM 4X4-CTRL, BM4X4-VP41, ICVD21- meters were calculated using the MEANS procedure. -ALPHA, ICVD21-CTRL and ICVD21-VP41. The determined data (wine chemical parameters and Wines were racked after MLF and adjusted with volatile phenols) were tested for normal distributions, potassium metabisulphite, according to the levels of and analysed according to a general linear model free and total SO2 determined in wines at that point of (GLM) procedure. The statistical models included the the experiment, therefore the dose of potassium main effects of the commercial yeast, LAB and oak metabisulphite was different and ranged from 5 to 8 addition (OA; chips, staves, and control) as well as g/hL, and added in order to achieve 35 mg/L of free their interaction: yijk = µ + CY(L)i + OAj + CY(L)OAij + SO2 in all wines. After MLF, all wines inoculated with e . The least square mean values for the experi- commercial yeast with and without LAB were divided ijk mental groups were obtained using the LSM test, and into three equal volumes and allow to mature for three compared at the 5% error probability level. months in contact with oak alternatives, i.e., control wines without addition of oak, and with addition of RESULTS AND DISCUSSION staves and chips (4 g/L), which were medium plus toasted French oak, purchased from Pronektar, Ton- Since reducing SO2 levels in winemaking is a nellerie Radoux, France. After three months of matur- global trend, biological alternatives in the form of ation, wines were bottled and stored for bottle aging. starter cultures tailored to control spoilage microorg- Wine samples were bottled using natural cork cap anisms are studied. In this study, the effect of different and wines aged in a room with the temperature of ca. commercial yeasts and LAB on different chemical ° 15 C. Bottles were placed in a horizontal position. parameters and volatile phenol production was exam- Determination of chemical parameters ined. Chemical parameters of wines after aging are presented in Tables 1-6. The alcohol content and pH After three years of bottle aging, chemical para- values (Tables 1 and 2) were not statistically signific- meters were determined in three different bottles per antly affected by commercial yeasts and LAB, nor by each sample. These analyses included determination addition of oak alternatives in all varietal wines. Mon- of alcohol content, reducing sugars, pH, total and free tenegrin wines had a high alcohol level (>12.50 vol.%)

339 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020)

Table 1. Alcohol content in examined wines (vol.%); values presented as least square means; SE – standard error of mean; PCYL – statistical probability of commercial yeast and LAB addition; POA – statistical probability of oak addition

Grape variety Yeast BDX Yeast BM4x4 Yeast ICVD21 Yeast Control SE PCYL LAB ALPHA Control VP41 ALPHA Control VP41 ALPHA Control VP41 Oak type Cabernet Chips 13.02 13.21 13.09 13.10 13.14 12.99 13.10 13.07 13.08 0.65 1 Sauvignon Control 13.10 12.98 12.89 13.11 13.03 13.05 12.98 13.04 13.04 12.80 0.65 0.9999 Staves 13.05 13.12 12.91 13.08 13.00 13.01 13.10 13.00 13.00 0.65 1 SE 0.65 0.66 0.65 0.65 0.65 0.65 0.65 0.65 0.65 0.64

POA 0.9886 0.9117 0.9181 0.9984 0.9627 0.9934 0.9670 0.9913 0.9888 Kratošija Chips 12.97 13.01 12.98 13.10 13.18 13.07 12.93 12.69 13.02 0.65 0.9967 Control 13.02 12.91 12.88 13.04 12.95 12.84 12.88 12.73 12.96 12.81 0.65 0.9999 Staves 12.85 12.88 12.80 13.02 13.06 12.98 12.96 12.79 13.08 0.65 0.9993 SE 0.65 0.65 0.64 0.65 0.65 0.65 0.65 0.64 0.65 0.64

POA 0.9473 0.9675 0.9435 0.9879 0.9124 0.9099 0.9884 0.9815 0.9749 Vranac Chips 13.46 13.49 13.61 13.78 13.72 13.78 13.89 13.98 13.74 0.69 0.9893 Control 13.50 13.33 13.30 13.74 13.84 13.85 13.72 13.84 13.70 12.94 0.68 0.7853 Staves 13.54 13.37 13.45 13.73 13.72 13.73 13.80 13.87 13.74 0.68 0.9905 SE 0.68 0.67 0.67 0.69 0.69 0.69 0.69 0.69 0.69 0.65

POA 0.9895 0.955 0.8562 0.9956 0.9702 0.9774 0.9558 0.967 0.9966

Table 2. pH value in examined wines; values presented as least square means; SE – standard error of mean; PCYL – statistical probability of commercial yeast and LAB addition; POA – statistical probability of oak addition

Yeast BDX Yeast BM4x4 Yeast ICVD21 Yeast

Grape variety LAB Control SE PCYL ALPHA Control VP41 ALPHA Control VP41 ALPHA Control VP41 Oak type Cabernet Chips 3.75 3.70 3.73 3.70 3.64 3.66 3.74 3.64 3.62 0.18 0.986 Sauvignon Control 3.75 3.70 3.72 3.65 3.63 3.67 3.75 3.66 3.62 3.64 0.18 0.9904 Staves 3.74 3.70 3.72 3.68 3.62 3.65 3.71 3.65 3.62 0.18 0.9923 SE 0.19 0.19 0.19 0.18 0.18 0.18 0.19 0.18 0.18 0.18

POA 0.9972 1 0.9971 0.9458 0.9909 0.9911 0.9636 0.991 1 Kratošija Chips 3.46 3.71 3.65 3.62 3.49 3.62 3.70 3.59 3.69 0.18 0.666 Control 3.47 3.72 3.65 3.61 3.49 3.62 3.72 3.61 3.71 3.67 0.18 0.6894 Staves 3.46 3.70 3.64 3.62 3.47 3.61 3.70 3.58 3.69 0.18 0.636 SE 0.17 0.19 0.18 0.18 0.17 0.18 0.19 0.18 0.18 0.18

POA 0.9967 0.9913 0.997 0.9969 0.9869 0.9969 0.9884 0.9786 0.9884 Vranac Chips 3.33 3.33 3.31 3.28 3.24 3.25 3.36 3.35 3.34 0.17 0.9854 Control 3.28 3.26 3.26 3.28 3.24 3.30 3.33 3.35 3.32 3.44 0.17 0.9344 Staves 3.32 3.32 3.31 3.28 3.24 3.26 3.36 3.35 3.35 0.17 0.9882 SE 0.17 0.17 0.16 0.16 0.16 0.16 0.17 0.17 0.17 0.17

POA 0.9271 0.8576 0.9132 1 1 0.9254 0.9686 1 0.9753

that can be unfavourable for the growth of certain sugar, free and total SO2 and total anthocyanins con- strains of Brettanomyces spp. yeast [27], while other tent in all three wines. strains tolerate higher ethanol concentrations [28]. Volatile phenol concentrations can also be From this research, it can be concluded that high affected by the presence of anthocyanins, since vinyl- ethanol concentrations (13.98 vol.%) was not a limit- phenols formed by the decarboxylation of hydroxycin- ing factor in volatile phenol formation. It is shown that namic acids, which can condense with anthocyanins there is a statistically significant effect of applied com- and form stabile pyranoanthocyanin adducts [29]. mercial yeasts and LAB on the content of reducing Therefore, differences in VP concentrations can also

340 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020)

Table 3. Reducing sugar content in examined wines (g/L); values presented as least square means; SE – standard error of mean; PCYL – statistical probability of commercial yeast and LAB addition; POA – statistical probability of oak addition; mean values with a different letter (A, B, C) within the column are statistically significantly different (P ≤ 0.05, significance of differences between the oak types); mean values with a different letter (a, b, c, d) within a row are significantly different (P ≤ 0.05, significance of differences between the used commercial yeast and LAB)

Yeast BDX Yeast BM4x4 Yeast ICVD21 Yeast

Grape variety LAB Control SE PCYL ALPHA Control VP41 ALPHA Control VP41 ALPHA Control VP41 Oak type Cabernet Sauvignon Chips 2.90a 2.90a 2.67bacA 2.47c 2.60bc 2.67bac 2.58bc 2.81ba 2.60bc 0.13 0.0097 Control 2.81bac 2.88ba 3.01aA 2.78bdac 2.60dc 2.71bdc 2.64bdc 2.64bdc 2.52d 2.80bac 0.14 0.0109 Staves 2.98a 2.92ba 2.54dcA 2.60dc 2.72bc 2.75bac 2.45d 2.70bdc 2.68bdc 0.14 0.0032 SE 0.14 0.15 0.14 0.13 0.13 0.14 0.13 0.14 0.13 0.14

POA 0.4113 0.945 0.0143 0.0712 0.4821 0.7783 0.2551 0.3624 0.3818 Kratošija Chips 2.40dc 2.35d 2.39dc 2.90a 2.38dc 2.80ba 2.60bc 2.60bc 2.54dc 0.13 0.0003 Control 2.39c 2.38c 2.40c 2.84a 2.38c 2.72a 2.65ba 2.65ba 2.43bc 2.30c 0.13 0.0081 Staves 2.40b 2.37b 2.40b 2.75a 2.35b 2.72a 2.54ba 2.58ba 2.47b 0.13 0.0052 SE 0.12 0.12 0.12 0.14 0.12 0.14 0.13 0.13 0.12 0.12

POA 0.9931 0.9516 0.9931 0.4718 0.9386 0.7251 0.6088 0.802 0.5765 Vranac Chips 1.72cC 1.72cC 1.64cC 2.30a 2.50a 2.40aB 2.30a 2.03b 2.42a 0.11 <0.0001 Control 2.40cbdB 2.97aA 2.97aA 2.24ed 2.30ced 2.59bA 2.30ced 2.12e 2.50cb 1.34f 0.12 <0.0001 Staves 1.52eC 1.64eC 1.72eB 2.50cb 2.59b 2.84aA 2.35c 2.12d 2.40cb 0.11 <0.0001 SE 0.10 0.11 0.11 0.12 0.12 0.13 0.12 0.10 0.12 0.07

POA <0.0001 <0.0001 <0.0001 0.0777 0.0681 0.0176 0.8346 0.5155 0.5963

Table 4. Free SO2 content in examined wines (mg/L); values presented as least square means; SE – standard error of mean; PCYL – statistical probability of commercial yeast and LAB addition; POA – statistical probability of oak addition; mean values with a different letter (A, B, C) within the column are statistically significantly different (P ≤ 0.05, significance of differences between the oak types); mean values with a different letter (a, b, c, d) within a row are significantly different (P ≤ 0.05, significance of differences between the used commercial yeast and LAB)

Yeast BDX Yeast BM4x4 Yeast ICVD21 Yeast Grape LAB Control SE PCYL variety ALPHA Control VP41 ALPHA Control VP41 ALPHA Control VP41 Oak type Cabernet Chips 28.16ba 24.32dcB 25.60c 25.60c 26.08bc 23.04dA 24.32dcA 25.60cA 29.44aA 1.29 0.0003 Sauvignon Control 25.60c 30.72aA 23.04de 25.60c 24.32dc 29.44baA 28.16bA 21.76eA 24.32dcA 28.16bA 1.31 <0.0001 Staves 25.60dc 29.44aA 25.60dc 25.60dc 24.32d 25.60dcA 24.32dA 28.16baA 26.88bcA 1.31 0.0020 SE 1.32 1.41 1.24 1.28 1.25 1.31 1.28 1.27 1.35 1.41

POA 0.0882 0.0033 0.0703 1 0.2166 0.0028 0.0158 0.0024 0.0102 Kratošija Chips 23.04bB 30.72aA 29.44aA 15.36cB 16.64cB 15.36cB 21.76bB 16.69cB 16.64cB 1.07 <0.0001 Control 16.64eC 32.00aA 23.04cA 14.08fC 19.20dB 17.92edB 19.20dB 19.20dB 14.08fC 26.88bA 1.05 <0.0001 Staves 20.48bB 23.04aB 23.04aB 16.64dC 17.92cdC 19.20cbC 24.32aA 23.04aB 16.64dB 1.03 <0.0001 SE 1.01 1.44 1.27 0.77 0.90 0.88 1.09 0.99 0.79 1.34

POA 0.0007 0.0005 0.0012 0.0187 0.0358 0.0047 0.0037 0.0007 0.0111 Vranac Chips 17.92e 20.48d 17.92e 25.60bA 28.12aA 23.04cA 14.08fC 12.80fC 16.64eB 1.01 <0.0001 Control 19.20cd 19.20cd 19.20cd 21.76bB 20.48cbB 17.92edB 16.64eC 20.48cbBA 21.76bB 28.16aA 1.03 <0.0001 Staves 19.20bc 19.20bc 17.92c 19.20bcB 20.48baB 21.76aB 11.52eB 12.80eC 15.36dB 0.89 <0.0001 SE 0.94 0.98 0.92 1.12 1.17 1.05 0.71 0.79 0.91 1.41

POA 0.2356 0.2601 0.2233 0.0012 0.0003 0.0024 0.0004 <0.0001 0.0003

arise from different anthocyanin concentrations and ethylphenols. In this study, Kratošija wines had lower their reactions with vinylphenols, eliminating the avail- total anthocyanins (171.0–282.0 mg/L), compared to ability of these precursors for further conversion into Vranac (522.0–786.0 mg/L), where higher concentrat-

341 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020)

Table 5. Total SO2 content in examined wines (mg/L); values presented as least square means; SE – standard error of mean; PCYL – statistical probability of commercial yeast and LAB addition; POA – statistical probability of oak addition; mean values with a different letter (A, B, C) within the column are statistically significantly different (P ≤ 0.05, significance of differences between the oak types); mean values with a different letter (a, b, c, d) within a row are significantly different (P ≤ 0.05, significance of differences between the used commercial yeast and LAB)

Yeast BDX Yeast BM4x4 Yeast ICVD21 Yeast Grape LAB Control SE PCYL variety ALPHA Control VP41 ALPHA Control VP41 ALPHA Control VP41 Oak type Cabernet Chips 88.32aA 75.52bA 76.80b 85.76a 87.04aA 83.20ba 75.52b 83.20ba 85.76a 4.12 0.0036 Sauvignon Control 69.12cA 88.32aA 76.80b 76.80b 78.08bB 79.36b 81.92ba 83.20ba 84.48ba 87.04a 4.03 0.0004 Staves 81.92baA 75.52bcA 75.52bc 78.08bac 74.24cA 75.52bc 83.20a 83.20a 76.80bac 3.91 0.0492 SE 4.01 4.00 3.82 4.02 4.00 3.97 4.01 4.16 4.12 4.35

POA 0.003 0.0116 0.8956 0.0675 0.0198 0.138 0.116 1 0.074 Kratošija Chips 69.12d 79.36cA 84.48bcA 60.16e 79.36cB 83.20bc 89.60ba 67.84dB 94.72aA 3.97 <0.0001 Control 65.28e 81.92dA 106.24aA 60.16e 97.28bcA 84.48d 92.16c 64.00eB 67.04eB 102.40ba 4.18 <0.0001 Staves 67.84dc 60.16eB 84.48aA 62.72de 78.08baA 81.92a 81.90a 74.24bcB 72.96bcA 3.71 <0.0001 SE 3.37 3.72 4.62 3.05 4.27 4.16 4.40 3.44 3.96 5.12

POA 0.4191 0.0007 0.0017 0.5313 0.0025 0.7624 0.0661 0.0289 0.0003 Vranac Chips 49.92cd 52.48cbB 55.04bC 47.36edC 42.24fC 66.56aB 43.52efC 33.28gC 56.32b 2.52 <0.0001 Control 51.20dc 69.12bB 69.12bB 34.56fC 42.24eC 55.04cB 47.36dC 51.20dcC 53.76c 81.92a 2.86 <0.0001 Staves 51.20cd 48.64edC 56.32bC 38.40fC 35.84fB 69.72aB 38.40fB 44.80eC 53.12cb 2.48 <0.0001 SE 2.54 2.87 3.02 2.02 2.01 3.20 2.16 2.19 2.72 4.10

POA 0.7835 0.0003 0.0023 0.0007 0.0119 0.0032 0.0066 0.0002 0.3745

ions of VP were detected, and a higher level of antho- -CTRL and in BDX-CTRL-CHIPS wines, respectively, cyanins content which is a varietal characteristic of while in other Kratošija wines, 4-EP varied from 0.01- Vranac [30]. Contrary to our expectations, the results –0.89 mg/L. A significantly high content of 4-EG in indicated that there is no relation between total antho- Kratošija wines was found in all samples where alco- cyanins and VP formation. A statistically significant holic fermentation was carried out with BM4x4 yeast, difference in anthocyanin content was found between without LAB addition, regardless of the addition of the different yeasts used for Kratošija and Cabernet oak alternatives. Statistically significant differences Sauvignon wines, while for the majority of these wines were also found in 4-EP/4-EG ratio between wines no statistical difference was found for oak addition. with added commercial yeast and LAB as well as Statistically significant differences in anthocyanin between wines where different oak alternatives were content were found for Vranac wines fermented with used. The ratio 4-EP/4-EG (Table 7) was calculated commercial BDX and ICV D21 yeasts (Table 6). and notable variations were found from 2.27 to 7.27 in After 3 years of bottle aging, 4-EP and 4-EG Kratošija wines and from 4.97 to 7.69 in Vranac were determined. Results are shown in Figure 1. In wines. These findings corroborated the results of pre- varietal wines, Kratošija wines had the greatest num- vious studies, where notable variations in 4-EP/4-EG ber of samples with VP detected (18 samples). In Vra- ratio were found within the same grape variety [32], nac wines, only five samples contained VPs, while no as well as between different grape varieties [33-35]. VPs were detected in Cabernet Sauvignon wines. Pollnitz [34] analysed different varietal Australian red Statistically significant differences in VP content were wines that are commercially available and found that found between wine samples with added commercial an average ratio of 4-EP/4-EG was approximately yeast and LAB as well as between wines where differ- 10:1 for Cabernet Sauvignon, 8:1 for Merlot, 3.5:1 for ent oak alternatives were used. The high concen- Pinot Noir and 9:1 for Shiraz. This is in contrast to tration of VP in Kratošija wines was expected, since results obtained in this study, where VP were not det- these wines contained the highest concentrations of ected in Cabernet Sauvignon wines. The variations of phenolic acids, especially p-coumaric acid, resulting 4-EP/4-EG ratio could be related to different concen- in higher concentrations of 4-EP, compared to 4-EG trations of the VP precursors in wines as well as to [31]. In Kratošija wines, a significantly high content of metabolism variability of Brettanomyces spp. strains 4-EP (1.51 and 1.56 mg/L) was found in BDX-CTRL- [32]. One of the potential sources of VP, whose con-

342 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020)

Table 6. Total anthocyanins content in examined wines (mg/L); values presented as least square means; SE – standard error of mean;

PCYL – statistical probability of commercial yeast and LAB addition; POA – statistical probability of oak addition; mean values with a different letter (A, B, C) within the column are statistically significantly different (P ≤ 0.05, significance of differences between the oak types); mean values with a different letter (a, b, c, d) within a row are significantly different (P ≤ 0.05, significance of differences between the used commercial yeast and LAB)

Yeast BDX Yeast BM4x4 Yeast ICVD21 Yeast Grape LAB Control SE PCYL variety ALPHA Control VP41 ALPHA Control VP41 ALPHA Control VP41 Oak type Cabernet Chips 218.0cC 249.0b 270.0ba 224.0c 267.0ba 248.0b 281.0a 250.0b 280.0aB 12.75 <0.0001 Sauvignon Control 221.5cC 260.0ba 270.0a 249.0ba 270.0a 263.0ba 270.0a 239.0bc 264.5aB 256.0baB 12.84 0.0023 Staves 250.0bdcB 238.0dc 249.0bdc 231.0d 280.0a 263.0ba 258.5bac 262.0bac 240.0bdcB 12.64 0.0050 SE 11.51 12.46 13.16 11.75 13.62 12.90 13.50 12.53 13.10 12.80

POA 0.0272 0.1774 0.1582 0.0932 0.5123 0.3278 0.2055 0.1596 0.0300 Kratošija Chips 280.5a 255.0b 252.0cb 177.0e 255.0b 198.0d 198.5d 232.5cB 203.0dC 11.52 <0.0001 Control 265.5a 262.5a 252.0a 171.0e 258.0a 193.5cd 204.5cb 186.0cdeC 179.0deC 222.0bB 11.11 <0.0001 Staves 282.0a 262.5ba 270.0a 184.5e 243.0bc 201.0ed 190.0e 238.5cB 214.0dB 11.71 <0.0001 SE 13.81 13.00 12.91 8.88 12.60 9.88 9.89 11.01 9.96 11.10

POA 0.3371 0.7294 0.2233 0.2534 0.3671 0.6643 0.2722 0.0021 0.0100 Vranac Chips 729.0A 682.5A 733.5A 691.5 772.5 711.0A 729.0A 682.5A 703.5A 35.78 0.1046 Control 597.0cA 574.5cdA 574.5cdA 735.0a 730.5a 670.5bA 621.0cbA 522.0dA 583.5cA 522.0dA 30.87 <0.0001 Staves 786.0A 747.0A 777.0A 720.0 735.0 765.0A 786.0A 747.0A 700.5A 37.60 0.1252 SE 35.42 33.59 35.02 35.79 37.31 35.83 35.76 32.87 33.24 26.10

POA 0.0016 0.0020 0.0010 0.3800 0.3800 0.0500 0.0040 0.0004 0.0100

Figure 1. Volatile phenols content present in examined wines. The first five samples from the left to the right refer to Vranac wines, and the other 18 samples represent Kratošija wines. centration is also dependent of the degree of toasting, Within Kratošija wines in which VP were det- is oak and oak alternatives used in wine production ected, it can be concluded that there were statistically [36]. In this study, statistically significant effects of dif- significant effects of different commercial yeast, LAB ferent oak alternatives on VP production were found and oak addition to VP formation. Regarding Vranac (Table 8). wines in which VP were detected, it was noticed that significantly high concentrations of 4-EP were formed

343 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020)

Table 7. 4-EP/4EG ratio in examined wines; values presented as least square means; SE – standard error of mean; PCYL – statistical probability of commercial yeast and LAB addition; POA – statistical probability of oak addition; mean values with a different letter (A, B, C) within the column are statistically significantly different (P ≤ 0.05, significance of differences between the oak types); mean values with a different letter (a, b, c, d) within a row are significantly different (P ≤ 0.05, significance of differences between the used commercial yeast and LAB)

Yeast BDX Yeast BM4x4 Yeast ICVD21 Yeast Grape LAB Control SE PCYL variety ALPHA Control VP41 ALPHA Control VP41 ALPHA Control VP41 Oak type Kratošija Chips <0.01e 6.63aB 5.24bA 0eC 2.72dB <0.01eB <0.01e <0.01eB 3.75cB <0.01 <0.0001 Control <0.01h 7.27aA 4.76bB 3.49eA 2.27gC 2.38fA <0.01h 4.41cA 3.80dA <0.01hA <0.01 <0.0001 Staves <0.01e <0.01eC <0.01eC 3.08dB 3.09cA <0.01eB <0.01e 5.66aC 4.15bA <0.01 <0.0001 SE <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01

POA - <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 - <0.0001 <0.0001 - Vranac Chips <0.01c <0.01c 7.69aA <0.01c 4.97bC <0.01c <0.01c 0cB <0.01c <0.01 <0.0001 Control <0.01c <0.01c <0.01cB <0.01c 5.15bB <0.01c <0.01c 10.94aA <0.01c <0.01c <0.01 <0.0001 Staves <0.01b <0.01b <0.01bB <0.01b 5.41aA <0.01b <0.01b <0.01bB <0.01b <0.01 <0.0001 SE <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01

POA - - <0.0001 - <0.0001 - - <0.0001 - -

Table 8. 4-EP and 4-EG content in examined wines (mg/L); values presented as least square means; SE – standard error of mean; PCYL

– statistical probability of commercial yeast and LAB addition; POA – statistical probability of oak addition; mean values with a different letter (A, B, C) within the column per grape variety are statistically significantly different (P ≤ 0.05, significance of differences between oak addition); mean values with a different letter (a-g) within a row are significantly different (P ≤ 0.05, significance of differences between the used commercial yeast and LAB)

Yeast BDX Yeast BM4x4 Yeast ICVD21 Yeast Grape Cmpd. LAB Control SE PCYL variety ALPHA Control VP41 ALPHA Control VP41 ALPHA Control VP41 Oak type 4-EP Kratošija Chips <0.01eB 1.51aA 0.58cA <0.01eC 0.72bB <0.01eB 0.02eB 0.02eB 0.14dB 0.03 <0.0001 Control 0.01gA 1.56aA 0.39eB 0.48dB 0.54cC 0.54cA <0.01gB 0.61bA 0.24fA <0.01gA 0.03 <0.0001 Staves <0.01eB <0.01eB <0.01eC 0.62cA 0.89aA <0.01eB <0.01eC 0.73bC 0.27dA 0.02 <0.0001 SE <0.01 0.06 0.02 0.02 0.04 0.02 <0.01 0.03 0.01 <0.01

POA <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 - Vranac Chips <0.01c <0.01c 0.44bA <0.01c 2.25aB <0.01c <0.01c <0.01cB <0.01c 0.04 <0.0001 Control <0.01b <0.01b <0.01bB <0.01b 2.06aB <0.01b <0.01b 2.13aA <0.01b <0.01b 0.05 <0.0001 Staves <0.01b <0.01b <0.01bB <0.01b 2.84aA <0.01b <0.01b <0.01bB <0.01b 0.05 <0.0001 SE <0.01 <0.01 0.01 <0.01 0.12 <0.01 <0.01 0.06 <0.01 <0.01

POA - - <0.0001 - 0.0005 - - <0.0001 - - 4-EG Kratošija Chips <0.01e 0.23bA 0.11cA <0.01eC 0.26aBA <0.01eB <0.01e <0.01eA 0.04dB 0.01 <0.0001 Control <0.01f 0.21bA 0.08dB 0.14cB 0.24aB 0.23bA <0.01f 0.14cA 0.06eA <0.01fA 0.01 <0.0001 Staves <0.01e <0.01eB <0.01eC 0.20bA 0.29aA <0.01eB <0.01e 0.13cB 0.07dA 0.01 <0.0001 SE <0.01 0.01 <0.01 0.01 0.01 0.01 <0.01 0.01 0.00 <0.01

POA - <0.0001 <0.0001 <0.0001 0.0109 <0.0001 - <0.0001 <0.0001 - Vranac Chips <0.01c <0.01c 0.06bA <0.01c 0.45aB <0.01c <0.01c <0.01cB <0.01c 0.01 <0.0001 Control <0.01c <0.01c <0.01cB <0.01c 0.40aC <0.01c <0.01c 0.20bA <0.01c <0.01c 0.01 <0.0001 Staves <0.01b <0.01b <0.01bB <0.01b 0.52aA <0.01b <0.01b <0.01bB <0.01b 0.01 <0.0001 SE <0.01 <0.01 0.00 <0.01 0.02 <0.01 <0.01 0.01 <0.01 <0.01

POA - - <0.0001 - 0.0017 - - <0.0001 - -

in all wines fermented with BM4x4 yeast (without (with addition of VP41 LAB and chips), low content of addition of LAB, and with and without addition of oak VP was detected. Within these three Vranac wines alternatives). In one Vranac wine with BDX yeast fermented with BM4x4 yeast (without addition of

344 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020)

LAB), the significantly high 4-EP content of 2.84 mg/L higher the need for SO2, in this couple of Kratošija and 4-EG (0.52 mg/L) was achieved when staves wines with free SO2 approx. 30 mg/L, pH was > 3.70. were added. Where chips were added, this content Additionally, research from [39] suggested that not was 4-EP (2.25 mg/L) and 4-EG (0.45 mg/L), while only concentrations of free or total SO2 are important, wines without oak addition, contained 2.06 mg/L 4-EP but rather their ratios. And, according to their res- and 0.40 mg/L 4-EG. These wines with high VP con- earch, the recommended ratio of free-to-total SO2 is tent had a high alcohol content (>13.5 vol.%) and low 0.4, represented as enough to ensure wine microbial pH value (<3.4), showing that these parameters stability. In this study, after aging, the ratio was differ- cannot be used for prediction of VP. ent in varietal wines. In Vranac wines the ratio varied B. bruxellensis growth and VP production (e.g., from 0.28 to 0.67, on average 0.39±0.11, while in Kra- 4-EP and 4-EG), can be affected by synergism of tošija wines the ratio on average was the lowest ethanol content, SO2, temperature and pH [28,37]. (0.26±0.06) and ranged from 0.18 to 0.39. Finally, in Tolerance to ethanol depends on the strain of B. Cabernet Sauvignon wines the variation of the ratio bruxellensis involved. Previous studies showed that was the lowest (0.03), on average 0.32±0.03 and B. bruxellensis could grow at ethanol concentrations ranged from 0.18 to 0.39. of 14.5 vol.%, which is above the concentration of The level of reducing sugars (RS) is important, Vranac and Kratošija wines, where VP were detected. as it is a nutrition source for Brettanomyces yeasts. In With the exception of two wine samples, the concen- Kratošija wines RS content was below 3.00 g/L, tration of total SO2 was below 100 mg/L, the concen- showing that B. bruxellensis can survive under low tration necessary for the inhibition of yeast [38]. At RS content and that it can utilise other energy high ethanol concentrations in combination with sul- sources [40], therefore even low RS concentrations phites, B. bruxellensis enters a viable but non-cul- are not a limiting factor for Brettanomyces growth. turable (VBNC) state [28]. Based on the chemical This is a key factor in their survival along with pro- parameters of Kratošija wines with VP content it can duction and storage phases under favourable external be concluded that VP will be formed in wines with the conditions (e.g., decrease of free sulphite content alcohol content from 12.5 to 13.5 vol.% and based on during ageing and maturation of wine) [4,35]. Besides, this parameter the formation of VP could not be pre- the presence of viable Brettanomyces in red wines dicted. which have been bottled for more than 50 years has The effect of commercial yeast, LAB and oak been reported [4]. alternatives on pH of wines showed no statistically significant differences between grape varieties and CONCLUSIONS different oenological additives. The pH of the wines ranged from 3.24 to 3.75, with the pH range of the Based on the results, it can be concluded that wines with VP detected from 3.47 to 3.72. Production there were statistically significant differences between of VP could not be predicted, based on alcohol content. commercial yeast, LAB and oak alternatives on VP The most common additive that inhibits the formation in Montenegrin red wines, however, statistically significant differences in VP content were found growth of Brettanomyces is sulphur dioxide (SO2) between samples with or without addition of oak alter- [12]. In this research SO2 was added before bottling natives. Based on the results, it is impossible to iden- in order to keep concentrations of free SO2 in the tify which treatment is suitable and further studies are range of 30–35 mg/L. Total and free SO2 content in examined wines, after four years (vintage 2012, needed. unpublished data) of bottle aging was significantly Regarding chemical parameters of wines with lower than that proposed for inhibiting any undesired VP content, it is concluded that VP will be formed in Brettanomyces spp. yeasts in the wine. Based on the wines even with high alcohol content, low pH and with results after three years of bottle aging it is still a free SO2 concentrations at bottling of about 30-35 mg/L. Therefore, VP formation could not be predicted challenge to maintain the SO2 concentrations stable over long periods even under anaerobic conditions. It based on these parameters. However, in this res- is also interesting that the highest content of VP was earch, Kratošija as grape variety from the Monte- negrin terroir during 2013 vintage showed the highest found in wines where concentration of free SO2 was a slightly above 30 mg/L, suggesting that in this case, potential to VP formation, compared to Vranac and Cabernet Sauvignon. Brettanomyces were not sensitive to free SO2. How- ever, as it is known that for microbial stability of wine, pH is also important. The higher the pH of wine, the

345 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020)

REFERENCES [20] N.S. Brizuela, B.M. Bravo-Ferrada, M. Ángeles Pozo- -Bayón, L. Semorile, E. E. Tymczyszyn, Food Res. Int. [1] D. Wedral, R. Shewfelt, J. Frank, LWT Food Sci. Technol. 106 (2017) 22-28 43 (2010) 1474-147 [21] M. S. Cappello, G. Zapparoli, A. Logrieco, E.J. [2] N. Rodrigues, G. Gonçalves, S. Pereira-da-Silva, M. Bartowsky, Int. J. Food Microbiol. 243 (2017) 16-27 Malfeito-Ferreira, V. Loureiro, J. Appl. Microbiol. 90 [22] V. Maraš, M. Tomić, V. Kodžulović, S. Šućur, J. Raičević, (2001) 588-599 D. Raičević, M. Čizmović, Agro-Knowl. 13 (2012) 103-112 [3] I. Silva, M.F. Campos, T. Hogg, J.A. Couto, Int. J. Food [23] V. Maraš, V. Božović, S. Giannetto, M. Crespan, OENO Microbiol. 145 (2011) 471-475 One 48 (2014) 87-97 [4] V. Renouf, M.C. Perello, G. De Revel, A. Lonvaud-Funel, [24] Commission regulation (EEC) No. 2676/90 determining. Am. J. Enol. Vitic. 58 (2007) 379-386 Official Journal 272 (1990) p. 192 [5] J. S. Sáez, C.A. Lopes, V.C. Kirs, M.P. Sangorrín, Lett. [25] M.M. Giusti, R.E. Wrolstad, Curr. Protoc. 1 (2001) F1.2.1– Appl. Microbiol. 51 (2010) 170-176 -F1.2.13 [6] A. Barata, D. Pagliara, T. Piccininno, F. Tarantino, W. [26] H. Prosen, B. Pendry, Phytochem. Lett. 16 (2016) 311-323 Ciardulli, M. Malfeito-Ferreira, V. Loureiro, FEMS Yeast [27] W. J. Du Toit, I. S. Pretorius, A. Lonvaud-Funel, J. Appl. Res. 8 (2008) 1097-1102 Microbiol. 98 (2005) 862-871 [7] S. Benito, F. Palomero, A. Morata, F. Calderón, J.A. [28] C.G. Edwards, T.A. Oswald, Lett. Appl. Microbiol. 66 Suárez-Lepe, J. Food Sci. 74 (2009) M15-M22 (2018) 71-76 [8] D. Chassagne, C. Charpentier, M. Guilloux-Benatier, H. [29] S. Benito, A. Morata, F. Palomero, M.C. González, J.A. Alexandre, M. Feuillat, J. Int. Sci. Vigne Vin. 35 (2001) Suárez-Lepe, Food Chem. 124 (2011) 15-23 157-164 [30] R. Pajović, D. Raičević, T. Popović, P. Sivilotti, K. Lisjak, [9] J. A. Suárez-Lepe, A. Morata, Trends Food Sci. Technol. A. Vanzo, S. Afr. J. Enol. Vitic. 35 (2014) 139-148 23 (2012) 39-50 [31] S. Radonjić, T. Košmerl, A. Ota, H. Prosen, V. Maraš, L. [10] C. Berbegal, G. Spano, M. Fragasso, F. Grieco, P. Demšar, T. Polak, Chem. Ind. Chem. Eng. Q. 25 (2019) Russo, V. Capozzi, Appl. Microbiol. Biotechnol. 102 309-319 (2018) 569-576 [32] A. Lima, C. Oliveira, C. Santos, F.M. Campos, J.A. [11] C. Berbegal, C. Garofalo, P. Russo, S. Pati, V. Capozzi, Couto, Eur. Food Res. Technol. 244 (2018) 1985-1994 G. Spano, Fermentation 3 (2017) 1-12 [33] R. Gawel, Brettanomyces Character in Wine, in Aus- [12] R. Suárez, J. A. Suárez-Lepe, A. Morata, F. Calderón, tralian Society of Wine Education National Convention, Food Chem. 102 (2007) 10-21 Hunter Valley, 2004 [13] P. Silva, H. Cardoso, H. Gerós, Am. J. Enol. Vitic. 55 [34] A.P. Pollnitz, K.H. Pardon, M.A. Sefton, J. Chromatogr., A (2004) 65-72 874 (2000) 101-109 [14] R. Hooper, A. Marks, Aust. N. Z. Grapegrower Winemak. [35] J. Steensels, L. Daenen, P. Malcorps, G. Derdelinckx, H. 337 (1992) 25-26 Verachtert, K.J. Verstrepen, Int. J. Food Microbiol. 206 [15] J. A. Couto, F. M. Campos, A.R. Figueiredo, T.A. Hogg, (2015) 24-38 Am. J. Enol. Vitic. 57 (2006) 166-171 [36] S. Rayne, N.J. Eggers, J. Environ. Sci. Health 42 (2007) [16] D. Knockaert, K. Raes, C. Wille, K. Struijs, J. Van Camp, 887-897 J. Sci. Food Agric. 92 (2012) 2291-2296 [37] M.E. Sturm, F. Arroyo-López, A. Garrido-Fernández, A. [17] A. Devi, K.A. Anu-Appaiah, Food Chem. 268 (2018) 101- Querol, L. Mercado, M. Ramirez, M. Combina, Int. J. –109 Food Microbiol. 170 (2014) 83-90 [18] P. Filannino, M. Gobbetti, M. De Angelis, R. Di Cagno, [38] W.R. Sponholz, in Wine Microbiology and Biotechnology, Appl. Environ. Microbiol. 80 (2014) 7574-7582 G. H. Fleet (Ed.), Hardwood Academic Publishers. Chur, [19] M.G. Madsen, N.K. Edwards, M.A. Petersen, L. Mok- 1993, pp. 395-420 wena, J.H. Swiegers, N. Arneborg, Am. J. Enol. Vitic. 68 [39] E. Robinson, P. Godden, AWRI Technical Review 145 (2017) 23-29 (2003) 7-11 [40] L. Conterno, C.M. L. Joseph, T.J. Arvik, T. Henick-Kling, L.F. Bisson, Am. J. Enol. Vitic. 57 (2006) 139-147.

346 S. RADONJIĆ et al.: INCIDENCE OF VOLATILE PHENOLS IN MONTENEGRIN… Chem. Ind. Chem. Eng. Q. 26 (0) 000−000 (2020)

SANJA RADONJIĆ1 ISPARLJIVI FENOLI U CRNOGORSKIM CRVENIM HELENA PROSEN2 VINIMA: VRANCU, KRATOŠIJI I KABERNE VESNA MARAŠ1 3 SOVINJONU LEA DEMŠAR 3 TATJANA KOŠMERL Glavni cilj ovog rada bio je da se utvrdi uticaj tri različita komercijalna kvasca i dve mleč- 1Research and Development Sector, nokiselinske bakterije, kao i upotrebu hrasta (iver i štapići), na formiranje isparljivih “13. Jul Plantaže“ a.d., Podgorica, fenola u crnogorskim starim crnim vinima (vranac, kratošija i kaberne sovinjon) berbe Montenegro 2013. godine. Utvrđeni su standardni hemijski parametri vina, kako bi se uspostavio 2Faculty for Chemistry and Chemical odnos sa VP. Posle tri godine odležavanja u boci, utvrđeni su hemijski parametri vina

Technology, University of Ljubljana, (sadržaj alkohola, zaostali šećeri, pH, slobodni i ukupni SO2 i antocijani) i isparljivi fenoli Ljubljana, Slovenia (4-etilfenol i 4-etil gvajakol). Isparljivi fenoli pronađeni su u 18 uzoraka kratošije i 5 uzo- 3 Biotechnical Faculty, University of raka vranca. Isparljivi fenoli nisu otkriveni u vinima kaberne sovinjon. Kratošija, kao Ljubljana, Ljubljana, Slovenia sorta grožđa iz crnogorskog regiona, tokom berbe 2013. godine imala je najveći potencijal za formiranje isparljivih fenola, dok je najveći sadržaj 4-etilfenola (2,84 mg/l) i 4-etil- NAUČNI RAD gvajakola (0,52 mg/l) utvrđen u vinu vranac. Pokazano je da će se isparljivi fenoli stva- rati u vinima čak i sa visokim sadržajem alkohola (> 12,50 vol.%), niskim pH (< 3,30) i

koncentracijom slobodnog SO2 pri flaširanju od 30-35 mg/l.

Ključne reči: crveno vino, kvasac, mlečnokiselinske bakterije, alternative hrastu, isparljivi fenoli.

347

Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020) CI&CEQ

P.C. NNAJI1 Luffa cylindrica SEED: BIOMASS FOR B.I. OKOLO1 2 WASTEWATER TREAMENT, SLUDGE O.D. ONUKWULI GENERATION STUDY AT OPTIMUM 1Department of Chemical CONDITIONS Engineering, Michael Okpara University, Umudike, Nigeria Article Highlights 2Department of Chemical • Luffa cylindrica seed powder contains reasonably high amounts of crude protein Engineering, Nnamdi Azikiwe • Biomass fibrous nature has an active coagulation-flocculation site University, Awka, Nigeria • pH and biomass dosage have a dominant influence on the process • High color removal efficiency at low pH SCIENTIFIC PAPER • COD and color/TSS reduction >94%

UDC 628.3.034.2:58:66 Abstract A BoX-Behnken design implementing response surface methodology was employed to investigate the sludge generation and removal of pollutants from teXtile wastewater using Luffa cylindrica seed powder (LCSP) in its natural form as biomass in the coagulation-flocculation process. With color/total suspended solids (CTSS) removal, chemical oXygen demand (COD) removal and sludge volume indeX (SVI) as the three responses, the three quadratic models of three variables including LCSP dosage, solution pH and stirring time were developed. The optimal conditions determined by the combination of the three desirable responses were 1.4 g/l LCSP, solution pH 2 and 30 min stirring time, resulting in 87.83% CTSS removal, 96.07% COD removal, and 98.29 ml/g SVI. The use of LCSP eXhibited great potential for teXtile wastewater treatment and thus can contribute to a greener environment. Keywords: biomass, chemical oXygen demand, Luffa cylindrica, sludge volume indeX, teXtile wastewater.

The increasing population and consequent ind- treatment. Local inhabitants rely on this river because ustrial eXplosion are at their peak throughout the of their water source, and the treatment facilities are globe, particularly in developing countries such as highly restricted. TeXtile wastewater pollutants are Nigeria. Increased dye manufacturing has led to the usually produced from caustic soda, detergents, proliferation of industrial wastewater generation, par- starch, waX, urea, ammonia, pigments, and dyes that ticularly from teXtile industries, to satisfy the teXtile boost its biological oXygen demand (BOD), chemical requirements of the population. The primary cause of oXygen demand (COD), solid content and toXicity [3]. this wastewater generation is the use of large volume The decomposition of organic pollutants from dye of water during processing, which is accompanied by materials can generate toXic substances that are chemical compleXes and particulate elements in the known to have mutagenic effects and are not bio- effluent streams [1,2]. This large amount of waste- degradable due to their high molecular mass and water generally finds its way into our waterways like compleX molecular structures [2,4,5]. the Aba River in Abia State, Nigeria, with little or no Removal of these pollutants requires the implementation of multiple treatment techniques, including

membranes separation, aerobic and anaerobic deg- Correspondence: P.C. Nnaji, Department of Chemical Engineer- ing, Michael Okpara University, Umudike, Nigeria. radation with different microorganisms, chemical E-mail: [email protected]; [email protected] oXidation, coagulation-flocculation and reverse osmo- Paper received: 23 June, 2019 sis [6], chromatography, lime precipitation and modi- Paper revised: 27 January, 2020 Paper accepted: 16 March, 2020 fied bleaching sequence [7]. Coagulation-flocculation stands out as the most feasible primary treatment https://doi.org/10.2298/CICEQ190623012N

349 P.C. NNAJI et al.: L. cylindrica SEED: BIOMASS FOR WASTEWATER… Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020) alternative for removing pollutants from dye-based the relationship between different measured res- wastewater despite the development of other tech- ponses and the essential variables of input [2]. In the niques. This is due to its simple on-site application, literature [1,2,9] there are reports on the implement- high efficiency of treatment, simplicity as well as low ation of RSM for wastewater treatment through coa- cost of assembly and operation [1]. gulation-flocculation with natural biomass. However, The coagulation-flocculation process is achieved less attention has been paid to optimizing the method by adding a coagulant to the wastewater in order to of coagulation-flocculation using Luffa cylindrica seed destabilize and neutralize the color/colloid dispersion powder (LCSP), an underused seed in Nigeria, as and the subsequent agglomeration of the resulting biomass for removing pollutants from teXtile waste- individual particles [1]. Chemical coagulants such as water. This is where the significance of this work ferric chloride, aluminum sulphate, etc. have been comes to light. commonly used to remove a broad variety of waste- L. cylindrica grows with limited use throughout water pollutants [8]. However, there is an intrinsic Nigeria. L. cylindrica was used as filter sponge and disadvantage in the brownish coloring of machinery adsorptive biomass support [13] but the seed has little by iron salts and secondary contamination issues use in wastewater treatment. The objective of this connected with aluminum salts [2,9]. Associated study is to optimize the process of coagulation-floc- health issues such as Alzheimer’s disease [1,9], gen- culation and to investigate the interactive influences eration of large sludge volumes resulting in enormous of the eXperimental variables, including solution pH, disposal costs and inefficiency of aluminum salt in the dosage of LCSP and stirring time. To this end, low-temperature waters [10] have been reported. teXtile wastewater was selected as the target waste- The search and use of biomass for wastewater water to be treated using RSM-optimized coagulation- treatment have gained importance in reducing these flocculation technique. COD, CTSS removal from established hazards and increasing concern for envi- wastewater and sludge volume indeX (SVI) have ronmental issues associated with the use of conven- been selected as responses. tional coagulants. Natural biomass used for wastewater treatment is easily accessible, environmentally MATERIALS AND METHODS friendly, efficient, secure for both humans and ani- The raw teXtile wastewater was gathered from mals and contributes to the green environment [1,2]. the teXtile plant in Aba, Nigeria. Grab sampling Natural materials such as Moringa olifera, tannins, method was used for wastewater collection. A plastic Detarium microcarpum, etc. have been eXplored for container of twenty-five liters was used to collect the the removal of pollutants [1,11,12]. Wastewater samples and was stored in the laboratory at 4 °C. treated with natural biomass using a coagulation-floc- Analytical grade was used for all reagents. culation method does not pose a danger to biological organisms as opposed to synthetic coagulants. The Characterization and chemical analysis of Luffa generated sludge can be treated through biological cylindrica seed powder means and used as a soil conditioner [2]. Luffa cylindrica sponges were collected in and The coagulation-flocculation process is inf- around Amawom in Ikwuano L.G.A of Abia State, luenced by temperature, solution pH, quality of waste- Nigeria, and dried. The seeds of L. cylindrica were water, the concentration of coagulants, and type of separated from the dried sponges that had matured. coagulant among other variables. Optimizing these Separate seeds were washed, dried and homogen- factors can significantly enhance the effectiveness of ized to obtain LCSP and placed in a container that the method. Apart from the time-consuming nature of was airtight. The proXimate LCSP analysis was the conventional one-factor-at-a-time eXperiments, it is not possible to obtain the eXact option since the Table 1. ProXimate analysis results of biomass interactions between variables are ignored. In deve- S/n Parameter LCSP loping the RSM jar test, a three-factor BoX-Behnken design (BBD) implementing RSM using Design 1. Moisture content (%) 9.30 EXpert 10.0 was used. The suggested RSM will 2. Ash content (%) 2.16 determine the effects and interactive impacts of 3. Fat content (%) 6.80 individual factors. 4. Crude protein (%) 21.88 RSM, a collection of statistical principles was 5. Carbohydrate (%) 60.46 suggested for developing eXperiment design models, 6. Crude fiber (%) 8.70 assessing the impacts of different factors and looking 7. Calorific value (kJ/kg) 360.56 for optimum variable conditions. RSM also quantifies 8. Dried moisture (%) 90.70

350 P.C. NNAJI et al.: L. cylindrica SEED: BIOMASS FOR WASTEWATER… Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020) based on standard methods [14] and presented in Coagulation-flocculation process Table 1. Using an infrared spectrometer (Agilent The coagulation-flocculation process was per- Technologies) with a resolution range of 4000–650 formed using LCSP 0.2, 0.28 and 0.36 g in 200 ml -1 -1 cm and 30 scans at 8 cm with 16 background teXtile wastewater at pH 2, 6 and 10 to give concen- scans, FT-IR spectra were acquired. A scanning tration of 1.0, 1.4 and 1.8 g/l of LCS, respectively. electron microscope (Phenon-World, MVE Using 0.1 M sulfuric acid and 0.1 M sodium hydro- 016477830) acquired the sample’s morphological Xide, respectively, pH adjustment was accomplished. features. A modified jar test method (using magnetic stirrer, Characterization of teXtile wastewater D-91126 Schwabach, MR Hei-standard) was used with a fast stirring of 2 min at 150 rpm and a slow agit- By standard methods [15], the characterization ation of 5, 15 and 30 min at 25 rpm. Before the agit- of teXtile wastewater has been determined. The char- ation, the biomass was added. The solution was care- acterization was carried out and presented in Table 2 fully poured into a 250 ml cylinder at the end of the at the National Soil, Plant, Fertilizer and Water Labor- slow stirring and allowed to settle for 300 min. After atory, Umudike, Nigeria. Mettler Toledo Delta 320 pH settling, each cylinder’s supernatant was used to eva- Meter, DDS 307 conductivity meter and UNICO 1100 luate CTSS and COD to determine the removal of spectrophotometer were used to determine the sol- CTSS and COD. While the rest of the supernatant was ution pH, electrical conductivity and CTSS. used by conventional techniques to achieve SVI [15]. Table 2. Characterization results of teXtile wastewater BoX-Behnken design (BBD) of eXperiment

S/n Parameter TeXtile wastewater In developing the RSM jar test, BBD implement- 1. pH 6.45 ing RSM from Design EXpert 10.0 was used. BBD is 2. E.C (μΩ/cm) 392.16 a three-level design of quantitative variables with all 3. SDP (mg/l) 5638.2 factors. The layout varied over three levels each of 4. BOD (mg/l) 491.4 the numerical variables; a high (+1), low (-1) and mid 5. COD (mg/l) 920.15 (0). The eXperiment needed a total of 17 runs (Table 6. TOC (mg/l) 588.10 3). The response factors represented as Y were 7. Lead (mg/l) 0.049 CTSS removal percentage, COD removal percentage and SVI (g/ml). The chosen variables for the study 8. Nickel (mg/l) 0.014 were the dosage of LCSP (X ), solution pH (X ) and 9. Chromium (mg/l) 0.0027 1 2 stirring time (X3). Table 4 provided the range and

Table 3. BBD for COD, CTSS removal and SVI for teXtile wastewater

Run Coded Uncoded Responses

X1 X2 X3 Dosage (g/l) pH Stirring time (min) CTSS Removal (%) COD Removal (%) SVI (ml/g) 1. 0 0 0 1.4 6 15.0 70.93 89.89 51.49 2. -1 0 -1 1.0 6 5.0 59.62 90.10 27.4 3. 1 -1 0 1.8 2 15.0 75.53 97.59 94.0 4. 0 -1 -1 1.4 2 5.0 74.18 97.10 94.22 5. 0 1 -1 1.4 10 5.0 71.98 83.30 26.47 6. -1 0 1 1.0 6 30.0 64.28 84.34 35.29 7. -1 1 0 1.0 10 15.0 66.33 84.01 27.57 8. 1 0 -1 1.8 6 5.0 72.45 87.80 38.55 9. -1 -1 0 1.0 2 15.0 73.16 97.59 78.38 10. 0 0 0 1.4 6 15.0 63.04 89.44 50.91 11. 0 0 0 1.4 6 15.0 66.26 88.89 51.49 12. 0 0 0 1.4 6 15.0 72.11 89.01 52.11 13. 0 0 0 1.4 6 15.0 71.43 89.44 51.09 14. 0 1 1 1.4 10 30.0 65.39 85.30 26.80 15. 0 -1 1 1.4 2 30.0 87.83 96.07 98.29 16. 1 1 0 1.8 10 25.0 75.00 88.93 26.09 17. 1 0 1 1.8 6 30.0 64.18 89.8 38.55

351 P.C. NNAJI et al.: L. cylindrica SEED: BIOMASS FOR WASTEWATER… Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020) concentrations used in the research. This strategy is contains crude protein of 28.45% [19], it has been to fit a quadratic polynomial equation model [2,9,16]: noted that crude protein is the active element for coa-

kk k gulation-flocculation. =+ +2 + To make the greatest use of it, it is essential to y bbXbXbXXoiiiiiijij  (1) ii==11 ij understand the nature and property of a material. For the prediction of chemical interactions, peak shift and where y is the variable response to be modeled; X i, intensity are essential in FT-IR absorbance spectro- and X are the independent variables influencing y, b , j o scopy. While peak shift indicates a change in a func- b , b and b the offset terms, the i-th linear coefficient, i ii ij tional groups’ chemical environment, bandwidth the ii-th quadratic coefficient and the ij-th interaction appearance and disappearance point at reactions inv- coefficient, respectively. olving the relevant functional groups [2]. Figure 1a shows the eXistence of the polymeric -OH functional Table 4. EXperimental design levels of chosen variables group at 3276.3 cm-1, methyl C-H and methylene C-H Coded values levels Variable asymmetric stretching group at 2951.1 and 2914.8 Lower limit (-1) Middle (0) Up Limit (+1) cm-1, respectively. 1707.1 cm-1 could be an indication

Biomass dosage, X1 (g/l) 1.0 1.4 1.8 of carboXylic ketone stretching [10], while 1636.3, -1 pH, X2 2 6 10 1543.1 and 1408.9 cm could be attributed to olefin,

Stirring time, X3 (min) 5.0 15.0 30.0 aromatic ring stretch and methyl asymmetric bends, respectively. It was possible to assign spectra 1226.3 RESULTS AND DISCUSSION cm-1 and strong 1036.2 cm-1 to skeletal C-C vibration and primary C-N stretch amine. For the attachment of Characterization of Luffa cylindrica seed powder colloidal particles and some dissolved ions, the pre- The results presented in Table 1 show a fairly sence of OH, carboXyl stretching and other groups high level of crude protein that is similar to the acces- could serve as active sites [2,12]. Figure 1b, the sible literature. The available literature indicates that sludge spectrum (biomass and coagulated CTSS), -1 defatted LCSP has crude protein of 42.17-70.65% that shows peak shifts around 2800-3300 cm band- [17], 45.06-50.06% [18] and while non-defatted LCSP width may be due to shifts in the chemical environ-

Figure 1. FTIR spectra of: a) Luffa cylindrica seed powder: b) sludge (biomass and coagulated CTSS).

352 P.C. NNAJI et al.: L. cylindrica SEED: BIOMASS FOR WASTEWATER… Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020)

ment of the polymeric-OH group. The presence of Y1 (CTSS removal) = 68.75 + 4.22X1 – 4.0X2 + -1 + 1.68X3 + 1.57X1X2 -5.73X1X3 – 5.06X2X3 – more bandwidth points between 0–1800 cm could be 2 2 2 due to reactions involving carboXylic, olefin and aro- –4.23X1 + 7.98X2 – 1.89X3 (2) matic groups. The result could not have been uncon- Y2 (COD removal) = 90.34 – 0.19X1 -3.75X2 + nected to the presence of settled CTSS coagulated +0.073X3 – 0.81X1X2 + 0.19X1X3 + 1.17X2X3 – 2 2 2 sludge. - 0.19X1 + 3.80X2 + 0.97X3 (3)

Biomass surface morphology was observed Y3 (SVI) = 51.42 + 3.57X1 - 32.25X2 + 1.54X3 – 2 using SEM assessment. The fibrous nature of LCSP - 4.28X1X2 - 1.97X1X3 - 0.94X2X3 – 10.70X1 + 2 2 [13] with some fissures and pores [20] was revealed + 15.79X2 - 5.97X3 (4) by two magnifications of 200 and 1000 μm described where X1, X2 and X3 were the dosage of LCSP, in Figure 2a. This is an indication of the eXistence of solution pH and stirring time, respectively,. The one- a macroporous structure that introduces coagulation- -factor coefficient represents the effect of the specific flocculation and adsorption active sites [21]. The mor- variable, whereas the two-factor coefficient and the phology of the sludge (biomass and coagulated second-order coefficient represent the interaction CTSS) in two 200 and 1000 μm magnifications between the two variables and the quadratic impact. obtained after coagulation-flocculation shown in Fig- A positive sign before the conditions is synergistic, ure 2b showed rough and irregular teXture, which while a negative sign is an antagonistic effect [2,16]. could be attributed to the coagulation-flocculation pro- Following removal of non-significant interaction terms cess involving certain chemical reactions which break- as shown in Table 5, Eqs. (5)–(7) have been produced: down the eXisting structure to allow floc formation. Y1 (CTTS removal) = 68.75 + 4.22X1 – 4.0X2 + 2 +1.68X3 + 1.57X1X2 – 5.06X2X3 – 4.23X1 + 2 2 + 7.98X2 – 1.89 X3 (5)

Y2 (COD removal) = 90.34 – 0.19X1 -3.75X2 + 2 + 0.073X3 + 0.19X1X3 + 1.17X2X3 – 0.19X1 + 2 2 + 3.80X2 + 0.97 X3 (6)

Y3 (SVI) = 51.42 + 3.57X1 -32.25X2 +1.54X3 – 2 - 4.28X1 X2 - 1.97X1X3 - 0.94X2X3 – 10.70X1 + 2 2 + 15.79X2 - 5.97X3 (7) Model adequacy Model summary statistics and variance analysis (ANOVA) were used to determine the adequacy of the model to identify the factors influencing the variables of response and thus to determine the most significant parameters. In Tables 5 and 6, these were presented. In Table 6 p-value for CTSS removal was 0.0173, while for the selected model the COD removal and SVI value were lower than 0.01. Considering the CTSS removal response (Y ), it is clear that the Figure 2. SEM micrograph of the: a) biomass and b) sludge 1 (biomass and coagulated CTSS); magnification 200X, top and linear terms for LCSP dosage (X1) and solution pH 1000X, bottom). (X2) have a moderate effect on the removal of the dye as shown by its f-values, whereas the stirring time

Response surface fitting by BoX-Behnken design factor (X3) has a negligible effect due to the low 2 (BBD) f-value of 1.36. The quadratic term (X3 ) showed a mild effect due to the negligible effect of (X ). The The three responses, Y for CTSS removal (%), 3 1 relationship between dosage–pH (X X ) showed a sig- Y for COD removal (%) and Y for SVI (ml/g) were 1 2 2 3 nificant effect, while pH–stirring time (X X ) and dos- associated with three variables: LCSP dosage (X ), 2 3 1 age–stirring time (X X ) suggested a mild effect. This solution pH (X ) and stirring time (X ), using quadratic 1 3 2 3 could be due to the effect of concentration of the polynomial equation as described in Eq. (1). The reactant and pH on reactions that lead to coagulation- models of second-order regression from the eXperi- -flocculation. mental data were displayed in Eqs. (2)–(4): Likewise, the solution pH factor (X2) for linear

and interaction term (X2X3) has shown a dominant

353 P.C. NNAJI et al.: L. cylindrica SEED: BIOMASS FOR WASTEWATER… Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020)

Table 5. ANOVA results for three responses (Y1, Y2 and Y3)

Response Source SS DF MS f-value Prob>f Remark

Y1(Color/TSS), % Model 878.49 9 97.61 5.87 0.0147 Suggested

Linear X1 142.55 1 142.55 8.57 0.0221

X2 128.0 1 128.0 7.69 0.0276

X3 22.61 1 22.61 1.36 0.2819 2 Pure quadratic X1 9.92 1 9.92 0.60 0.4653 2 X2 131.45 1 131.45 7.90 0.0261 2 X3 102.41 1 102.41 6.15 0.0422

Interaction X1X2 75.37 1 75.37 4.53 0.0709

X2X3 268.25 1 268.25 16.12 0.0051

X1X3 15.05 1 15.05 0.90 0.3733 Residual 116.49 7 16.64 Lack of fit 54.46 3 18.15 1.17 0.1950 Not significant Pure error 2.03 4 15.51 Cor. Total 994.98 16

Y2(COD), % Model 188.06 9 20.90 87.60 <0.0001 Suggested

Linear X1 0.30 1 0.30 1.28 0.2960

X2 112.65 1 112.65 472.25 <0.0001

X3 0.042 1 0.042 0.18 0.6872 2 Pure quadratic X1 2.62 1 2.62 11.0 0.0128 2 X2 0.14 1 0.14 0.57 0.4734 2 X3 5.52 1 5.52 23.15 0.0019

Interaction X1X2 0.15 1 0.15 0.61 0.4601

X2X3 60.93 1 60.93 255.42 <0.0001

X1X3 3.99 1 3.99 16.75 0.0046 Residual 1.67 7 0.24 Lack of fit 1.10 3 0.37 2.60 0.1897 Not significant Pure error 0.57 4 0.14 Cor. Total 189.73 16

Y3(SVI) Model 10120.1 9 1124.5 143.06 <0.0001 Suggested

Linear X1 101.89 1 101.89 12.96 0.0087

X2 8317.9 1 8317.9 1058.2 <0.0001

X3 18.88 1 18.88 2.40 0.1651 2 Pure quadratic X1 73.10 1 73.10 9.30 0.0186 2 X2 15.56 1 15.56 1.98 0.2022 2 X3 3.50 1 3.50 0.44 0.5261

Interaction X1X2 482.31 1 482.31 61.36 0.0001

X2X3 1050.42 1 1050.4 133.64 <0.0001

X1X3 140.07 1 140.07 17.82 0.0039 Residual 55.02 7 7.86 Lack of fit 54.17 3 18.06 4.48 0.1150 Not significant Pure error 0.85 4 0.21 Cor. Total 10175.1 16

effect with very high f-values of 472.25 and 255.42, unconnected to the effect of pH on different reactions respectively, for the COD removal response. In addit- which resulted in COD reduction. Because of the ion, significant effects have been indicated by the dominant effect of dosage and pH on the reaction 2 quadratic term for LCSP dosage (X1 ), stirring time leading to COD elimination, the dosage–pH, and pH– 2 (X3 ), and the interaction term (X1X3). The decrease in -stirring time interactions showed significant influ- COD removal as pH increases could not have been ence. Similar effects for an SVI response have also

354 P.C. NNAJI et al.: L. cylindrica SEED: BIOMASS FOR WASTEWATER… Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020)

Table 6. Model summary statistics for three responses (Y1, Y2 and Y3)

Response Source f-value Prob>f Std. Dev. R2 Adj. R2 PRESS Remark

Y1(Color/TSS), % Linear 1.81 0.1950 7.35 0.2946 0.1319 1405.57 2FI 1.77 0.2155 6.77 0.5397 0.2634 2089.26 Quadratic 6.84 0.0173 4.08 0.8829 0.7324 968.33 Suggested Cubic 1.17 0.4250 3.94 0.9377 0.7506 Aliased

Y2(COD), % Linear 6.38 0.0063 2.43 0.5956 0.5022 137.55 2FI 0.40 0.7538 2.63 0.6392 0.4228 255.79 Quadratic 93.31 <0.0001 0.49 0.9912 0.9799 18.54 Suggested Cubic 2.60 0.1897 0.38 0.9970 0.9881 Aliased

Y3(SVI) Linear 21.06 <0.0001 11.56 0.8293 0.7900 3633.52 2FI 0.19 0.9029 12.82 0.8384 0.7414 8442.53 Quadratic 67.40 <0.0001 2.80 0.9946 0.9876 867.98 Suggested Cubic 84.48 0.0005 0.46 0.9999 0.9997 Aliased

been observed. The linear terms (X1) and (X2) sug- percent, the models were statistically significant, gested significant SVI response effects, while (X3) meaning that there is only a 5%-chance that the showed a negligible effect above 0.1 with a p-value. f-value was due to noise. The model is not significant 2 In quadratic terms only (X1 ) was important, while the if the p-value is above 0.1 [9]. three interaction terms clearly showed significant Process analysis effects. This is consistent with the research reported earlier [22,23]. The model summary statistics showed Table 6 explains the linear (X1, X2, X3), quadratic 2 2 2 2 determination coefficient R , 0.8829, 0.9912 and (X1 , X2 , X3 ) and interaction (X1X2, X2X3, X1X3) 0.9946, respectively, for CTSS, COD removal and effects of the parameters. The response, Y1, revealed 2 2 SVI. This means that the independent variables that X1, X2, X2 , X3 and X2X3 are meaningful terms 2 explained 88.29, 99.12 and 99.46% of the variation with p < 0.05, whereas X3, X1 , and X1X3 are not for CTSS, COD removal and SVI and also implied meaningful terms. Similarly, Y2 and Y3 responses to 2 2 that the empirical models could not explain 11.71, contain significant terms of X2, X1 , X3 , X2X3, X1X3, 2 0.88 and 0.54% of their variations, respectively. The and X1, X2, X1 , X1X2, X2X3, X1X3. Although non-sig- 2 2 2 2 elevated R values stated that the model was well nificant are X1, X3, X2 , X1X2 and X3, X2 , X3 , respect- fitted to the response [9]. From the results, it can also ively. For Y1, Y2, and Y3, respectively, the lack of fit be seen that the experimental data shows a desirable f-values of 1.17, 2.60 and 4.48 implies that the lack of and acceptable agreement with the proximity of the fit is not significant compared to the pure error. Only 2 2 R to the adjusted R . This similarity indicates that the 19.50, 18.97 and 11.50%, respectively, for Y1, Y2 and quadratic models have been satisfactorily modified to Y3, are likely to occur due to noise for the lack of fit experimental data as observed with COD removal f-values this large. There is a good non-significant and SVI responses, with the difference between R2 lack of fit, the model must fit. and adjusted R2, stated as 0.0113 and 0.007, respect- Effect of variables on CTSS removal efficiency (Y1) ively [9,16]. However, a significant gap between R2 2 Figure 3a shows the effect of the two individual and the adjusted R shows the presence of insignific- LCSP dosage and solution pH factors on removing ant terms and/or moderate effects of factors on the CTSS based on one factor at a time. The plot in Fig- response variables as seen with CTSS response, with 2 2 ure 4 shows the individual effect on the percentage of a difference between R and the adjusted R recorded CTSS removal of LCSP dosage, solution pH and stir- as 0.1504 [23]. ring time-based on a total of 17 BB-design of three Due to the moderate and high f-value of 5.87 variables; LCSP dosage (X ), solution pH (X ) and and 87.6 for CTSS, COD removal and 143.06 for SVI, 1 2 stirring time (X ) as shown in Table 4. the second-order regression for CTSS and COD rem- 3 The curve slope in Figure 3a dropped sharply oval efficiency shows that the models were signific- from solution pH 2 to 8 and slightly increased at pH ant. Similarly, for quadratic regression models, the 10. This can be attributed to the rapid coagulation- p-value that provides an indication of the significance -flocculation in solution pH 2 [1,9,24]. The contour plot of a model in relation to the f-value was less than (2D) in Figure 4 showed the optimum pH 2, 1.3-1.4 g/l 0.05. This indicated that for a confidence level of 95

355 P.C. NNAJI et al.: L. cylindrica SEED: BIOMASS FOR WASTEWATER… Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020)

a) b) c) Figure 3. One-factor-at-a-time plots of: a) % CTSS removal; b) % COD removal; c) SVI (ml/g) vs. pH.

Figure 4. Contour plots of CTSS removal from textile wastewater: a) dosage vs. pH; b) stirring time vs. dosage; c) stirring time vs. pH.

dosage and 30 min stirring. The two-dimensional con- Effect of variables on COD removal efficiency (Y2) tours showed a considerable curvature in the contour Figure 3b showed the effect of two individual curves, implying the interdependence of these three LCSP dosage and solution pH factors on COD rem- factors [25]. For the contour plot of solution pH and oval based on one factor at a time, while Figure 5 LCSP dosage, however, the curvature is more pro- showed the individual effect of LCSP dosage, solution nounced, indicating them as dominant factors. pH, and stirring time on COD removal based on a

Figure 5. Contour plots of COD removal from textile wastewater: a) dosage vs. pH; b) stirring time vs. dosage; c) stirring time vs. pH.

356 P.C. NNAJI et al.: L. cylindrica SEED: BIOMASS FOR WASTEWATER… Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020) total of 17 BB-design of these three variables. The COD; 1.4 g/l LCSP, pH 2 and 30 min for SVI. As a trend is comparable to the trend of removing CTSS. combination, the optimal conditions for the three The curve slope dropped significantly from pH 2 to 10 CTSS removal, COD removal, and SVI responses with an optimal 1.4 g/l dosage. After 15 min of stirring were 1.4 g/l LCSP, pH 2 and 30 min stirring time, time, the contour chart stated optimum points at pH 2 yielding 87.83 and 96.07%, and 98.79 ml/g, respect- and dosage 1.0 and 1.8 g/l. Also, the contour showed ively, for CTSS, COD and SVI. a fair curvature in the curve, showing the interdepen- In conclusion, the results obtained indicated that dence of the three factors. LCSP has a strong potential for treatment of textile wastewater via coagulation-flocculation technique. Effect of variables on SVI (Y ) 3 This treatment will be achieved at low cost, and will Similarly, the effect of two individual LCSP contribute to a greener environment. dosage and solution pH variables on SVI based on standard one-factor-at-a-time shown in Figure 3c REFERENCES showed comparable trends as in removal of CTSS and COD. The contour plot based on 17 BBD of three [1] B.I. Okolo, P.C. Nnaji, O.D. Onukwuli, J. Environ. Chem. variables: LCSP dosage, solution pH and stirring time Eng. 4 (2016) 992-1001 shown in Figure 6 showed sensible curvature in the [2] T.F.S. Maísa, A.A. Cibele, A. Elizangela, B.S. Lídia, contour curve of all variables combinations. This is a K.F.S.F. Thábata, D.M. Daniele, M.C. Gizilene, C.G. Juliana, J. Taiwan Inst. Chem. Eng. 10 (2016) 1-10 clear sign of the variables’ interdependence and posi- [3] N. Muralimohan, T. Palanisamy, Asian J. Chem. 26 tive interaction [25]. (2014) 911-914 [4] S.M. Hadi, M.K.H. Al-Mashhadni, M.Y. Eisa, Chem. Ind. CONCLUSION Chem. Eng. Q. 25 (2019) 39-46 The LCSP spectrum disclosed waveform with [5] S.S. Konstantinovic, M.G. Kodric, R. Nicic, D.M. Djor- djevic, Chem. Ind. Chem. Eng. Q. 25 (2019) 11-19 peaks of complicated polysaccharides features. [6] S.B. Olugbenga, M.L. Bukola, J.A. Olamide, E. Vanain, Further assessment showed a compound with some Chem. Speciation Bioavailability 29 (2017) 120-134 fissures and pores with fibrous nature. To study the [7] I. Muhammad, B. Tahir, I. Naz, A. Naeem, A.K. Ruaf, S. real applicability of LCSP as natural biomass, BBD Aamir, Arab. J. Chem. 10 (2017) S2307-S2318 implementing RSM optimization was conducted. Vari- [8] W.W. Pei, T.T. Tjoon, A.R. Nik, N. Nik, Water Qual. ance analysis (ANOVA) disclosed the data fit the Resear. J. Can. 42 (2007) 54-62 quadratic model and the concentration of LCSP [9] B.I. Okolo, P.C. Nnaji, E.O. Oke, K.F. Adekunle, C.S. showed important effectiveness in removing CTSS Ume, O.D. Onukwuli, NIJOTECH 36(2017) 1104-1113 and COD. The effect of LCSP dosage in sludge [10] C.O. Aniagor, M.C. Menkiti, J. Environ. Chem. Eng. 6 generation was also stated by ANOVA. (2018) 2105-2118 The best condition for textile wastewater was [11] J.U. Ani, N.J.N. Nnaji, O.D. Onukwuli, C.O.B. Okoye, J. 1.4 g/l LCSP, solution pH 2 and 30 min for removal of Harzad. Mater. 243 (2012) 59-66 CTSS; 1.0 g/l LCSP, pH 2 and 15 min for removal of [12] A.A. Adewale, V.P. Fabiano, Beni-Suef Univ. J. Basic

Figure 6. Contour plots of SVI for textile wastewater: a) dosage vs. pH; b) stirring time vs. dosage; c) stirring time vs. pH.

357 P.C. NNAJI et al.: L. cylindrica SEED: BIOMASS FOR WASTEWATER… Chem. Ind. Chem. Eng. Q. 26 (4) 349−358 (2020)

Appl. Sci. 6 (2017) 118-126 [19] [H.S. Rabab, Middle East J. Appl. Sci. 7 (2017) 613-625 [13] A.L. Cleber, A.M. Aryane, C.G.J. Affonso, Rev. Ambiente [20] I.O. Oboh, E.O. Ahuyor, T.O.K. Audu, InTech, (2011) Agua 9 (2014) 212-223 195-212 [14] AOAC, Official Methods of Analysis, 16th Edition, [21] A. Rais, H. Shaziya, Groundwater Sustain. Dev. 1 (2015) Association of Official Analytical Chemist, Gaitherburg 146-154 (2005) [22] V.K. Sangal, V. Kumar, I.M. Mishra, Comp. Chem. Eng. [15] AWWA, APHA, WEF, Standard Method for the 40 (2012) 33-40 nd Examination of Water and Wastewater, 22 ed., New [23] A. Baharlouei, E. Jalilnejad, M. Sirousazar, Chem. Eng. York, 2012 Comm. 205 (2018) 1537–1554 [16] S.C. Kim, J. Indust. Eng. Chem. 38 (2016) 93-102 [24] M.C. Menkiti, O.D. Onukwuli, AlChE J. 156 (2012) 303- [17] A.S. Abitigun, A.O. Ashogbon, Enthnobot. Leafl. 14 –313 (2010) 225-235 [25] I. Fendri, L. Khannous, A. Timoumi, N. Gharsallah, R. [18] O. Afolabi, J.A. Adegbite, M.J. Ogunji, Int. J. Inno. Sci. Gdoura, Afr. J. Biotechnol. 12 (2013) 4819–4826. Eng. Tech. 4 (2017) 356-368

P.C. NNAJI1 SEME Luffa cilindrica: BIOMASA ZA TRETMAN B.I. OKOLO1 OTPADNIH VODA - ISTRAŽIVANJE GENERISANJA 2 O.D. ONUKWULI MULJA U OPTIMALNIM USLOVIMA 1Department of Chemical Engineering, Michael Okpara University, Umudike, Za istraživanje generisanja mulja i uklanjanje zagađivača iz otpadnih voda tekstilne Nigeria industrije pomoću praha semena Luffa cilindrica kao pomoćnog sredstva u procesu koa- 2 Department of Chemical Engineering, gulacije-flokulacije korišćen je Boks-Benkenov eksperimentalni plan zajedno sa metodo- Nnamdi Azikiwe University, Awka, logijom površine odziva. Proučavan je uticaj količine semena, pH i vremena mešanja na Nigeria stepen uklanjanja boje/ukupnih suspendovanih čvrstih materija i hemijske potrošnje kiseonika i indeks zapremine mulja korišćenjem su tri kvadratna modela. Optimalni NAUČNI RAD uslovi određeni kombinacijom tri poželjna odgovora bili su 1,4 g semena/l, pH 2 i vreme mešanja od 30 min, pri čemu je uklanjeno 87,83% boje/ukupnih suspendovanih čvrstih materija i 96,07% hemijske potrošnje kiseonika i postignut indeks zapremine mulja od 98,29 ml/g. Prema tome, samleveno seme L. cilindrica je pokazalo veliki potencijal za prečišćavanje otpadnih voda tekstilne industrije.

Ključne reči: biomasa, hemijska potrošnja kiseonika, Luffa cilindrica, indeks zapremine mulja, otpadne vode tekstilne industrije.

358 Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 359−367 (2020) CI&CEQ

KRISTINA JANČAITIENĖ INFLUENCE OF CELLULOSE ADDITIVE ON RASA ŠLINKŠIENĖ THE GRANULATION PROCESS OF Kaunas University of Technology, POTASSIUM DIHYDROGEN PHOSPHATE Department of Physical and Inorganic Chemistry, Kaunas, Article Highlights Lithuania • Fluidized bed and rotary granulators cannot be used to obtain granulated PDP • Drum-granulator-dryer can be used for PDP granulation SCIENTIFIC PAPER • PDP cannot be granulated by using water • MC addition improves the most important properties of the fertilizer UDC 631.8:66.099.2:54 Abstract

KH2PO4, which was manufactured using conversion of KCl and NH4H2PO4, is a concentrated crystalline chlorine-free phosphorus and potassium fertilizer. It is

usually used as a component of a liquid complex fertilizer, because KH2PO4 crystals melt very easily, and have very high hygroscopicity and caking effect. Granulated products are considerably more convenient than powders, but

KH2PO4 crystals are pure, hardly agglomerate, therefore they need a proper binder. This study aims to investigate the influence of cellulose additive and other different conditions on the granulation process and on the properties of the granulated product. Potassium dihydrogen phosphate was granulated using three different types of granulators (rotary, a fluidized bed and a drum) and by changing the amount of water used for irrigation purposes. The achieved results indicate that in order to obtain granulated potassium dihydrogen phosphate with optimal properties, the use of water does not suffice; therefore, another binder (cellulose) was used in order to improve physical and mechanical properties of the granules. It was determined that cellulose additive (5%) improves some properties of the fertilizer, such as the amount of marketable fraction, SGN, pH, and also reduces the hygroscopicity of the fertilizer granules about 2 times. But cellulose additive does not improve the static strength of granules and bulk density. Keywords: potassium dihydrogen phosphate, cellulose, granulation, fertilizer.

Preparation of various substances in granulated of granulation processes that ensure formation of form is of great interest to the producers of fertilizers, particles of desired size is a topical problem. synthetic detergents, building mixes, food additives, The solution of such problems always involves etc. Granulated products are considerably more the determination of the influence exerted on the convenient than powders in transfer, dosing, storage particle-size distribution of the product by process and transportation [1,2]. Therefore, the development parameters and design characteristics of the equipment. In this case, it is possible to control the particle- size distribution, form and strength of granules. How- ever, the relationships of granulation in apparatuses Correspondence: K. Jančaitienė, Kaunas University of Techno- of different types are not yet understood to a sufficient logy, Department of Physical and Inorganic Chemistry, Radvilėnų degree, despite the fact that studies in this field have rd. 19, LT-50254 Kaunas, Lithuania. E-mail: [email protected] been carried out for more than 40 years. This is pri- Paper received: 29 October, 2019 marily caused by high complexity and multistage Paper revised: 5 February, 2020 character of the granulation process. Granulation Paper accepted: 18 March, 2020 includes introduction of a binder, its distribution https://doi.org/10.2298/CICEQ191029013J

359 K. JANČAITIENĖ, R. ŠLINKŠIENĖ: INFLUENCE OF CELLULOSE ADDITIVE… Chem. Ind. Chem. Eng. Q. 26 (4) 359−367 (2020) throughout the working volume of the apparatus, con- together. Wet granulation technologies available inc- tact of the binder with the granule surface, wetting of lude roller compaction, spray drying, fluid bed granul- the surface and spilling over it, sticking of finer par- ation, high shear mixing, steam granulation, reverse ticles, and collisions with other granules and with wet granulation, thermal adhesion granulation, and internal parts of the apparatus [3]. All the above moisture-activated dry granulation. Selecting the type phenomena are random and cannot be described of process requires thorough knowledge of the phys- quantitatively on the level of a separate granule. icochemical properties of the raw materials, exci- Therefore, granulation is one of few processes of pients, required marketable product and release pro- chemical technology whose simulation cannot be per- perties [7-11]. formed using the approach based on analysis of an Wet granulation process is a popular way for elementary event, followed by transfer of the results manufacturers to produce various product mixtures of this analysis to the whole set of dispersed particles. that consist of many different ingredients. This Depending on the industry, granular material method is often applied in industries such as drug and properties and request of granular product, various pharmaceutical, food and drink, household chemicals, ways and equipment can be selected and used. One and bulk fertilizer manufacturing, etc. In each ind- of the most frequently used pieces of equipment is a ustry, the manufacturers use the right granulation pro- high-speed horizontal cylindrical granulator, which cess to mix selected materials and excipients. Every exhibits high productive capacity at relatively simple manufacturer aims to get homogeneous and reliable design and hence shows much promise for wide use products. Fertilizer manufacturers most commonly in industry. Such a granulator includes a cylindrical use wet granulating process and suitable equipment casing equipped with tubes for product loading and to homogenously mix different minerals and chemical unloading, and with nozzles for injecting the liquid components to produce high-density granules. These phase acting as a binder [1]. granules must be of the suitable marketable size and The conversion of fiddly powders into larger par- right chemical composition. Granulation depends on a ticles is called granulation. The use of powders intro- number of independent parameters that have a sig- duces various difficulties such as inhalation, flow nificant influence on the process and on the final gra- problems, product losses and even explosion risks in nular product properties. It is determined that the pro- chemical, pharmaceutical, and food industries [4]. To cess, and product properties, have varying sensitivi- eliminate the use of powders in a process, granulation ties to changes in the conditions of the process [8-10]. can be used. It can introduce a lot of advantages inc- In addition, the requirements for granules differ luding reduced caking, improved flow properties, inc- from industry to industry and from application to reased bulk density, control of surface-to-volume ratio application. For recycling products, the size and the and slowing down solubility. Therefore, granulation, shape of the granules are not very important. On the or agglomeration, increases the value of a product, other hand, products like fertilizers need to be the and is a very appropriate process [5,6]. precise size, roundness, strength and the effectivity. The granulation process transforms small par- These properties are highly dependent not only on ticles (powders) into free-flowing, dust-free granules raw materials but also on the way of granulation. that are easy to compress. Granulation has many Thus, it is a common practice in the fertilizer industry benefits. If you compare the flow ability of fine pow- to use rotation drum and disc granulators [7-8,12]. ders like flour or starch to sugar crystals, then it is Particle size is the parameter that needs to be clear that sugar flows much better. The granulation controlled in many granulation procedures and is process can be divided into two separate types. The most widely used for particle description [13-15]. Par- first type - wet granulation - is the process of intercon- ticle size and particle size distribution are very import- necting different size particles together using liquid ant properties when handling and using fertilizers. solution or adhesives as a binder. The second type - Usually, fertilizer bulk blending requires granules dry granulation - requires no special binder and any which are 1-5 mm in size. Particles of this size are liquid for granules to form. Wet granulation is carried also an excellent source of secondary nutrients to be out by adding a liquid to the powder. As a result, the used for direct application [16]. liquid forms a bridge to bind small powder particles A granule also holds together different compo- and granules are formed. When granules are manu- nents in a matrix so that a certain ratio of different factured industrially, a mixer, a fluidized bed or any components is provided at all times. Granulation rows other fitting unit operation could be used for spraying that have low plasticity, such as dolomite, ash and in order to bring the solid powder and a liquid others, are difficult to granulate in the absence of

360 K. JANČAITIENĖ, R. ŠLINKŠIENĖ: INFLUENCE OF CELLULOSE ADDITIVE… Chem. Ind. Chem. Eng. Q. 26 (4) 359−367 (2020) additives. Therefore, various binders are often used cellulose (MC) was used. The resultant granules were to improve agglomeration [17,18]. dried for approximately 12 h at 60 °C, and then their It is known that product properties such as gra- physical and chemical properties were assessed. nulometric composition and crush strength depend on Granulated fertilizer potassium dihydrogen raw materials and the kinds of additives used, size phosphate was fractioned by using RETSCH-made and shape of the particles, granulation method and woven sieves, and the shares of fractions (%) were other features of the granulation process. The effect determined by weighing them with electronic scales of process parameters on the crush strength of gra- (weighing precision of 0.001 g) [24]. nular fertilizers is described in [19]. The amount of moisture was measured with the In addition, reducing dust of the product (fertil- electronic moisture analyzer HG53. It utilizes the ther- izers) is very important to farmers when they bring out mogravimetric principle, i.e., its activity is based on the fertilizer into the field. Due to their coarse form, the decrease of weight under heating until the sample the granules can be spread in the desired range with reaches the final stable weight [24]. the farmers’ machines. If it were powder, no trajectory By using special equipment, (IPG-2) static would be possible and the wind might blow the fer- strength of PDP granules was determined and the tilizer to another field. This would cause the farmer a average was calculated [25]. more or less unwanted loss [7]. Bulk density (loose and packaged) and pH of Knowledge about accurate predictions of the granulated fertilizers were measured by using stan- process behaviour brings great benefits in the transfer dard procedures [24,26]. of processes from the laboratory to industrial fields. The Stokes number of the granules is lower than Therefore, this study aims to investigate and evaluate the critical value, which was calculated by the equat- the influence of process conditions (moisture content, ion [27]: types of additives and binders, using recycle etc.) on ρω 2 = 8 rR the granulation procedure and granulated product St v (1) properties (granulometric composition, static strengths 9µ of granules, Stokes number etc.) [8]. where ρ is the granule density (kg m-3); r is the effective granule size (m); ω is the granulator speed (s-1); EXPERIMENTAL R2 is the granulator radius (m); µ is the binder viscosity (kg m-1 s-1). Apparatus and reagents Average particle size of the fraction (d50) is a Potassium chloride (KCl, 99-100.5% Sigma-Ald- very important indicator of fertilizer technology effi- rich), ammonium dihydrophosphate (NH4H2PO4, ciency and product sales. It is indicated as the size, 99.0% Fluka Analytical), microcrystalline cellulose which is smaller than 50% of the product mass and is (Sigma-Aldrich, 9004–32–2) [20] and distilled water bigger than 50% of mass. It was calculated [28]: were used. ()50 −C Potassium dihydrogen phosphate (PDP) was dZ=+n () Z − Z (2) 50nnn− + 1 granulated in the laboratory using a drum-granulator- ()CCnn+1 dryer which was made according to the parameters of industrial equipment [21]. A horizontal drum granul- where Zn is the nominal sieve mesh size in mm, ator with blades fixed on the wall was arranged inside whose cumulative approaches, but does not exceed the casing. The blades were arranged in such a fash- 50% of the weight; Zn+1 is the nominal sieve mesh ion that they favor the transport of the product along size in mm, whose cumulative approaches and the apparatus axis toward the unloading zone and exceeds 50% of weight; Cn is the cumulative weight in ensure mixing and growth of the particles. The blades % on sieve n; Cn+1 is the cumulative weight (%) on inside the drum clean the internal surface of the sieve n+1. casing to remove the adhered product without stop- SGN (size guide number – the diameter, expres- ping the granulator. A simulative drum-type granul- sed as mm multiply by 100) was calculated from the ator-dryer was used to perform a modified wet granul- results of granulometric composition. The higher the ation method that is described in previous works [22]. number, the greater number of particles that are close By using the synthesized potassium dihydrogen phos- in size to the given SGN [29,30]: phate (fraction < 1 mm) [23], samples were granul- = SGN100 d 50 (3) ated in the laboratory conditions and 10 samples were obtained. To aid granulation, microcrystalline

361 K. JANČAITIENĖ, R. ŠLINKŠIENĖ: INFLUENCE OF CELLULOSE ADDITIVE… Chem. Ind. Chem. Eng. Q. 26 (4) 359−367 (2020)

For scan electron microscopy (SEM), the FEI ation process in laboratory conditions were deter- Quanta 200 FEG electronic microscope was used at mined. Potassium dihydrogen phosphate was granul- magnification rates from 10 to 500,000 [31]. ated by using 3 different types of granulators: rotary, a The TA.XT plus texture analyzer from Stable fluidized bed and a drum. Potassium phosphate gra- Micro Systems Ltd. (Godalming, UK), was used in nulation was investigated by changing the amount of order to characterize the stiffness and strength of the water used for irrigation purposes, as well as by add- granules. Individual granules were loaded into a cylin- ing a binder (cellulose). According to the obtained drical stainless-steel tool (5 mm in diameter) and rot- results (marketable fraction i.e., size of granules 1- ated at a constant test speed of 0.01 mm/s up to the -3.15 mm was not found), the fluidized bed and rotary deformation extent of 0.3 mm [32]. granulators cannot be used to obtain granulated pot- To evaluate the static strength results, the rel- assium dihydrophosphate because the granulometric ative (RSD), standard (SD) and absolute (ASD) devi- composition of the granulated product does not meet ations were calculated at 95% probability [33]. the necessary requirements for fertilizers. Compared to the product obtained by using a drum granulator, RESULTS AND DISCUSSION this product’s granules (which were obtained by using fluidized bed and rotary granulators) were smaller and The equilibrium of potassium chloride and had a powder-like appearance. ammonium dihydrophosphate in solid and liquid Therefore, this article presents only the results phases was analyzed under isothermal conditions at obtained when a drum-type granulator was used. All ° temperatures of 20, 40, 60 and 80 C. the samples were cooled and fractioned, and then, The equilibrium between the solid (the main the granulometric composition, the granule moisture compound is KH2PO4) and liquid (the main compound and bulk density were determined. The static is NH4Cl) phases was observed by measuring the strengths of granules of 1-2 mm and 2-3.15 mm fract- refractive index which stabilized when the final equi- ions and the 10% solution pH values were measured. librium had been reached. The results of the first 7 granulated samples are The phases were separated from each other by presented in Table 1. filtering through the Buchner filter and analyzed by With 21.02% moisture being present in the employing methods of chemical and instrumental source materials, the obtained granulated product analysis. The chemical composition of the solid phase was denoted by the optimal granulometric compo- was determined with chemical methods and was pre- sition (Sample 4) as its marketable fraction (1-3.15 sented in previous work [34]. Potassium dihydrogen mm) constitutes 40.56%. Hence this sample was phosphate in the solid phase is a crystalline powdery investigated to establish other properties of com- material that is hygroscopic and has a clogged pound fertilizers: pH of 10% solution, 1-2 mm and 2- bundle, so it needs to be granulated to improve its -3.15 mm diameter granule static strength, bulk physical properties. It is therefore logical that particle (freely poured and multiplexed) density and the analysis in this investigation should examine the Stokes number. The product was also characterized changes in granule size during the granulation. Three by calculating SGN (Table 2). techniques were used to analyze particle size [35]. From the data presented above, it is evident that In the experiment, the influence of the amount of the static strength of the granules is relatively low (7.3 additive MC used and the moisture content of the raw N/gran.), and that the 10% fertilizer solution is slightly mixture on the product properties were investigated. acidic (pH is 3.7). Loose bulk density varied in range From the results, optimal parameters of the granul-

Table 1. Physicochemical properties of granulated PDP with water

Raw material Granulometric composition, % Granule Sample No. moisture, % <1 mm 1-2 mm 2-3.15 mm 3.15-5 mm >5 mm moisture, % 1 10.08 94.8 1.60 3.60 – – 0.71 2 15.09 88.63 8.32 2.69 0.09 – 0.67 3 17.50 87.20 7.21 5.40 0.91 – 1.71 4 21.02 58.83 18.47 22.09 0.32 0.29 1.26 5 22.50 91.19 2.86 5.00 0.06 – 0.53 6 26.85 87.10 3.4 9.45 0.14 – 1.44 7 30.21 83.85 6.15 9.74 0.27 – 1.55

362 K. JANČAITIENĖ, R. ŠLINKŠIENĖ: INFLUENCE OF CELLULOSE ADDITIVE… Chem. Ind. Chem. Eng. Q. 26 (4) 359−367 (2020)

Table 2. Parameters of product granulated only with water; * – SD = 1.62, ASD = 0.76, RSD = 0.0052; ** – SD = 2.14, ASD = 1.00, RSD = 0.0072.

Diameter of pH Static strength Bulk density of granules, kg/m3 SGN Sample No. Stokes number granules (10% solution) N/gran. Loose Packaged mm 4 1-2 mm 3.7 7.27* 778 805 0.02 152 2-3.15 mm 3.7 7.25** 765 790 0.03

from 765 to 778 kg/m3; packaged bulk density varied in the 2-3.15 mm granule fraction. The weight of the in range from 790 to 805 kg/m3 which is acceptable granules (Figure 1b), depending on their size in the for standard of bulk fertilizers. SGN 152 of granulated 2-3.15 mm fraction, ranges from 5.7 to 27 mg in product is within the required SGN 125-150 range weight and granules of 11 mg in weight predominate. that is typically used on golf fields, low cut sports turfs The granulated PDP is non-plastic, which means that and sometimes in combined fertilizer products. It is it is not resistant to compression because, as it is often produced as a homogeneous particle but can be shown in Figure 1c, a force from 1.2 to 5.2 N is suffi- produced as a blend. Appropriate SGN is certainly an cient to crush the granules. When compressed, the important consideration when choosing your products size of the granules decreases from 0.05 to 0.6 mm. [36]. Stokes number of 0.02-0.03, in accordance with Such a fairly large change in the size of the pellets the authors’ (Walker G.D., Holland C.R.) data, means when compressing them with the force ranging from that the formation of granules was caused by the 1.2 to 5.2 N can be explained by the fact that aniso- adhesion and growth of the particles of the source metric granules were formed by granulating PDP with material [28]. using water only. In order to explore the properties (Figure 1) of Pellet uniformity assessment was made by tak- granulated PDP in greater detail, TA.XT plus texture ing photos of the fertilizer using the technique of analyzer with Exponent software was employed to scanning electron microscopy. The obtained photo is research the distribution of 2-3.15 mm granules in the presented in Figure 2. sample (50 granules of either size category were used). The presented photograph shows that the gra- Analysis of the distribution of granulated PDP in nules have irregular spherical forms, which does not terms of size (Figure 1a) shows that granules approx- meet the requirements for fertilizers and is not very imately 2.3 mm in size occurred with 40% frequency convenient for bulk fertilizer equipment. The obtained

Figure 1. Properties of granules (2–3.15 mm) of the granulated product, which was obtained by wetting potassium dihydrogen phosphate only with water. Distribution by: a – diameter; b – weight; c – breakage force; d – compressibility.

363 K. JANČAITIENĖ, R. ŠLINKŠIENĖ: INFLUENCE OF CELLULOSE ADDITIVE… Chem. Ind. Chem. Eng. Q. 26 (4) 359−367 (2020) results indicate that, in order to obtain granulated pot- Based on the data presented in Table 3, it is assium dihydrogen phosphate with optimal pro- evident that cellulose as an additive greatly improves perties, the use of water does not suffice. In order to the granulometric composition of potassium dihydro- improve the physical and mechanical properties of the gen phosphate. By using about 21% moisture and 1% granules it is necessary to use other binders as well. MC, marketable fraction (1-3.15 mm) yield was Analysis of scientific research works on the binding approximately 47%. When the same amount of mois- materials that can be used in the granulation tech- ture was used and MC amount was increased up to nology suggests that if the objective is to obtain maxi- 2%, correspondingly about 58% of the marketable mally pure potassium dihydrogen phosphate, cellu- fraction was obtained. The granulated product with lose should be chosen as the binder as it contains no the best granulometric composition and 71.41% of the additional nutrients consumed by plants [37]. In addit- marketable fraction was obtained when 5% MC was ion, cellulose is denoted by its cohesive properties. used (Sample 10). According to the established optimal granulometric composition, the static strength, the bulk (loose and packaged) density, and pH of 10% solution, for 1-2 mm and 2-3.15 mm diameter granule samples were measured. The Stokes number and SGN were also calculated. Parameters of product granulated with various content of MC and water are presented in Table 4. From the data presented above it is evident that some properties have not changed: static strength of the granules is approximately 7.4 N/gran., loose bulk density varied within the range of 770-785 kg/m3, packaged bulk density within the range of 790-810 kg/m3 and Stokes number varied between 0.02 and 0.04. However, SGN 224 of granulated product is in accordance with SGN 200+ that is typically used on landscape turf, golf rough and other standard cut turf and always produced as a blended product [36]. The Figure 2. SEM photo of a product, granulated only with water (no microcrystalline cellulose). 10% fertilizer solution is less acidic (pH is 4.3) and this change is due to MC, because pH of 10% sol- Microcrystalline cellulose was used for granul- ution is 6.7. ation [37]. MC was found to be the most suitable In order to further analyze the granular PDP pro- additive for production of high-quality fertilizers, vari- perties, cumulative frequency and distribution fre- ous mixtures of raw materials. Data on the use of cel- quency were calculated according to diameter, weight lulose as the binding material and employing water as of a granule, breakage force and compressibility (Fig- the moisturizer is presented in Table 3. ure 3).

Table 3. Physicochemical properties of granulated PDP with MC and water

Sample No. Granulation conditions Granulometric composition, % Gran. moisture Moisture content in Cellulose content <1 mm 1-2 mm 2-3.15 mm 3.15-5 mm >5 mm % the raw m., % in the raw m., % 8 21.08 1 51.52 34.83 11.81 1.72 - 2.00 9 21.14 2 33.37 32.96 24.54 8.15 0.92 1.86 10 21.39 5 20.11 36.36 35.50 5.75 2.00 1.56

Table 4. Parameters of product granulated with MC and water; * - SD = 2.48, ASD = ±1.16, RSD = 0.0078; ** - SD = 2.38, ASD = ±1.11, RSD = 0.0076

3 Diameter of pH Static strength Bulk density of granules, kg/m Stokes SGN Sample No. granules (10% solution) N/gran. Loose Packaged number mm 10 1-2 mm 4.3 7.38* 785 0.02 0.02 224 2-3.15 mm 4.3 7.32** 770 0.04 0.04

364 K. JANČAITIENĖ, R. ŠLINKŠIENĖ: INFLUENCE OF CELLULOSE ADDITIVE… Chem. Ind. Chem. Eng. Q. 26 (4) 359−367 (2020)

Figure 3. Properties of the granules (2–3.15 mm) of the granular product which was obtained by wetting potassium dihydrogen phosphate with water and adding 5% MC. Distribution by: a – diameter; b – weight; c – breakage force; d – compressibility.

Analysis of the distribution of granulated PDP in production of more spherically shaped granules in terms of size (Figure 3a) shows that granules of 2.3 comparison to the granules formed when granulating mm in size occurred with 40% distribution frequency with water only. in granule distribution of 2-3.15 mm. The weight of the Comparing the data obtained with potassium granules (Figure 3b), depending on their size in the dihydrogen phosphate granulated only with water and 2-3.15 mm fraction, ranges from 6.2 to 18.2 mg and with the cellulose additive, it has been determined approximately 10 mg granules predominate. As shown that the cellulose additive improves the most import- in Figure 3c, a force ranging from 1.2 to 7.2 N is ant properties of the fertilizer, such as granulometric needed to crush the granules. composition (i.e., the amount of the marketable frac- Granule uniformity was assessed by using the tion), pH, SGN but it does not increase the granule fertilizer photo obtained by employing scanning elec- strength and bulk density. SEM images also show tron microscopy technique, which is presented in Fig- that by using MC the granules have a more spherical ure 4. shape. Apart from the listed properties, hygroscopicity - another very important parameter for fertilizer granules - was measured by observing the alterations of the weight of the fertilizer sample when water vapor absorption was taking place in 1-2 mm and 2-3.15 mm size granules. Water vapor absorption studies showed that when granules were stored above a

saturated NaNO2 solution, the absorption equilibrium was achieved after 100 h. Respectively, when granulated products were stored above water, the intensive process took place over 400 h, and subsequently slowed down. The curves in Figure 5, I and II, show that the granules above saturated NaNO2 solution absorbed very small (0.17-0.32%), and when above Figure 4. SEM photo of product granulated with MC and water. water absorbed very large (55-95%) amounts of water The photo presented in Figure 4 makes evident vapor. The maximum amount of absorbed water that the use of cellulose for granulation resulted in vapor was reached after 100 h of storage of samples

365 K. JANČAITIENĖ, R. ŠLINKŠIENĖ: INFLUENCE OF CELLULOSE ADDITIVE… Chem. Ind. Chem. Eng. Q. 26 (4) 359−367 (2020)

Figure 5. Mass increasing of the samples depending on the duration, when the samples were kept above: I – NaNO2 (temperature

20-22 °C and moisture 73-75%); II – H2O (temperature 20-22 °C and moisture 96-98%).

above saturated NaNO2 solution and after 700 h without cellulose was reached after 700 h of storing above water. Comparing the maximal absorbed samples above water. amount of water vapor of the PDP granules without Acknowledgment cellulose and with 5% cellulose, it can be stated that the addition of cellulose reduces the hygroscopicity We want to thank Prof. Dr.-Ing. Sergiy Antonyuk about 2 times (in both cases - above water and above and his scientific group for the possibility to carry out some of these experiments in their laboratories. saturated NaNO2). These results are very important when evaluating the quality of fertilizer granules. REFERENCES CONCLUSIONS [1] O.M. Flisyuk, N.A. Martsulevich, T.N. Shininov, Russ. J. It was determined that the drum granulator-dryer Appl. Chem. 89 (2016) 603–608 can be used for granulation of potassium dihydrogen [2] P.V. Klassen, I.G. Grishaev, I.V. Shomin, Granulirovanie (Granulation), Moscow, 1991, p. 125 phosphate. The obtained results (content of product- [3] L. Briens, R.D. Logan, AAPS J. 12 (2011) 1358–1365 ion fraction of 40.54% and static strength of the gra- [4] S.M. Iveson, J.D. Litster, K. Hapgood, B.J. Ennis, Powder nules of 7.25-7.27 N/gran., SGN 152 of granulated Technol. 117 (2001) 3–39 product) indicate that in order to obtain granulated [5] B.J. Ennis, J.D. Litster, in Perry’s Chemical Engineers’ KH PO with optimal properties, the use of water 2 4 Handbook, R. Perry, D. Green (Eds), McGraw-Hill, New (approximately 21% moisture of raw materials) does York, 1997, p. 29 not suffice. In this case, the product granules have [6] P.R. Mort, S.W. Capeci, J.W. Holder, Powder Technol. irregular spherical shapes. The addition of cellulose 117 (2001) 173–176 (5%) when raw material moisture is approximately [7] C. Klein, Maschinenfabrik Gustav Eirich, Hardheim, 21% improves the most important properties of the (2012), https://docplayer.net/46281896-Granulation-chris- fertilizer, such as the amount of the marketable fract- tina-klein-maschinenfabrik-gustav-eirich-hardheim-7-1- ion (71.41%), the pH of 10% fertilizer solution (4.3 2012.html (accessed 10 April 2019) pH), SGN 224 of granulated product and the granules [8] S. Shanmugam, BioImpacts 5 (2015) 55-63 have a more spherical shape. However, static [9] K. Ax, H. Feise, R. Sochon, M. Hounslow, A. Salman, strength of granules measures (7.32-7.38 N/gran.), Powder Technol. 179 (2008) 190-194 loose (770-785 kg/m3) and packaged (790-810 kg/m3) [10] A. Braumann, M.J. Goodson, M. Kraft, P.R. Mort, Chem. bulk density, and the Stokes number (0.02-0.04) Eng. Sci. 62 (2007) 4717-4728 barely increased. The results of hygroscopicity show [11] L. Briens, Wet Granulation: The Ultimate Guide for that the addition of cellulose (5%) to PDP reduces the Beginners and Professionals, https://www.saintytec.com/ /wet-granulation-the-ultimate-guide-for-beginners-and- hygroscopicity of fertilizer granules about 2 times. The professionals (accessed 9 April 2018) maximum amount of absorbed water vapor (0.17% [12] S.M. Raj Kumar, R. Malayalamurthi, R. Marappan, Int. J. PDP with cellulose and 0.32% without cellulose) was Appl. Eng. Res. 55 (2015) 3811-3818 reached after storing samples above saturated

NaNO2 solution after 100 h. Respectively, 55-95%

366 K. JANČAITIENĖ, R. ŠLINKŠIENĖ: INFLUENCE OF CELLULOSE ADDITIVE… Chem. Ind. Chem. Eng. Q. 26 (4) 359−367 (2020)

[13] N. Jones, Research Project, https://pdfs.semantics- [27] M. Kimura, Testing methods for fertilizers, cholar.org/72fc/cca7fddb41b688025f212877b93fc288e32 www.famic.go.jp/ffis/fert/obj/TestingMethodsForFertilizers 0.pdf (accessed 21 September 2018) 2013 (accessed 16 January 2018) [14] S. Watano, Powder Technol. 117 (2001) 163-172 [28] G.M. Walker, C.R. Holland, M.N. Ahmad, J.N. Fox, A.G. [15] G.I. Tardos, M. I. Khan, P.R. Mort, Powder Technol. 94 Kells, Powder Technol. 107 (2000) 282–288 (1997) 245–258 [29] G.E.N. Lance, Handbook of Solid Fertilizer Blending, [16] M. Sviklas, R. Šlinkšienė, Mater. Sci. 13 (2007) 65-69 Code of Good Practice for Quality, EFBA/AFCOME [17] A. Abdullayev, M.F. Bekheet, D.A.H. Hanaor, A. Gurlo, Quality Committee, 2013, p. 231 Membranes 9 (2019) 1-31 [30] R. Jordan, Making sense of the numbers, https:// [18] N.Y. Baghdadi, R.K. Shah, A.B. Albadarin, Ch. Mang- //www.turfandrec.com/making-sense-of-the-numbers- 2954/ (accessed 7 January 2020) wandi, Chem. Eng. Res. Des. 132 (2018) 1070-1081 [31] Radiological & Environmental Management Division of [19] G.M. Walker, H.E. Moursy, C.R. Holland, Powder Tech- Environmental Health and Public, Safety Scanning nol. 132 (2003) 81–84 Electron Microscope, https://www.purdue.edu/ehps/rem/ [20] A.S. Huber, Pharmatrans, Sanaq AG. MCC, http:// /laboratory/equipment%20safety/Research%20Equipmen //www.pharmatrans-sanaq.com/products/multifunctional- t/sem.html (accessed 6 November 2018) excipients/mcc-sanaq-burst/ (accessed 5 December 2018) [32] L. Fries, S. Antonyuk, S. Heinrich, G. Niederreitera, S. [21] R. Paleckienė, A.M. Sviklas, R. Šlinkšienė, V. Štreimikis, Palzera, Particuology 12 (2014) 13–24 Pol. J. Environ. Stud. 21 (2012) 993–999 [33] Average, Standard Deviation and Relative Standard [22] V. Štreimikis, Doctoral dissertation, Technology, Kaunas, Deviation, www.chem.tamu.edu/class/fyp/keeney/ 2015, p. 25 /stddev.pdf (accessed 13 June 2018) [23] K. Jančaitienė, R. Šlinkšienė, Chemija 25 (2014) 89-95 [34] K. Jančaitienė, R. Šlinkšienė, Pol. J. Chem. Technol. 18 [24] M. Aizak, Regulation EC, No. 2003/2003 of the European (2016) 1–7 Parliament and of the Council of 13 October 2003 [35] K. Jančaitienė, R. Šlinkšienė, Conversion between KCl Relating to Fertilizers, Official Journal (2003), p. 22. and NH4H2PO4 and their products, Chemistry and [25] J.W. Hofstee, Physical Properties of fertilizer in relation to chemical technology, Kaunas, 2017, p. 78 handling and spreading, Wageningen University, Wage- [36] Choosing the right granular fertilizer, Turf Fertilizer, ningen, 1993, p. 27 https://www.lebanonturf.com/education-center/turf- [26] M.R. Tabari, Fertilizers — Determination of bulk density fertilizer/choosing-the-right-granular-fertilizer (accessed 6 (loose) of fine-grained fertilizers, https://www.iso.org/stan- January 2020) dard/14757.html (accessed 8 January 2020) [37] M.R. Brown, I.M. Saxena, Cellulose: Molecular and Struc- tural Biology, Springer, Amsterdam, 2007, p. 32.

KRISTINA JANČAITIENĖ UTICAJ CELULOZE NA PROCES GRANULISANJA RASA ŠLINKŠIENĖ KALIJUM-DIHIDROGEN-FOSFATA Kaunas University of Technology,

Department of Physical and Inorganic KH2PO4, dobijen reakcijom KCl i NH4H2PO4, je K i P đubrivo u obliku kristala, bez sadr- Chemistry, Kaunas, Lithuania žaja hlora. Obično se koristi kao komponenta tečnog složenog đubriva, jer se kristali

KH2PO4 vrlo lako rastvaraju, imaju vrlo visoku higroskopnost i efekat stvrdnjavanja. Gra- NAUČNI RAD nulisani proizvodi su znatno pogodniji od prahova, ali kristali KH2PO4 su čisti, slabo grade aglomerate, pa im je potrebno odgovarajuće vezivo. Cilj je istražiti uticaj celuloze i drugih uslova na proces granulacije i svojstva granulisanog proizvoda. Granulacija

KH2PO4 je izvršena pomoću tri različite vrste granulatora (rotacioni, fluidizacioni i dobo- šasti) uz promenu količine vode. Postignuti rezultati pokazuju da za dobijanje granuli-

sanog KH2PO4 sa optimalnim svojstvima upotreba vode nije dovoljna. Zbog toga je korišćeno drugo vezivo (celuloza) kako bi se poboljšala fizička i mehanička svojstva granula. Utvrđeno je da dodatak celuloze (5%) poboljšava neka svojstva đubriva, poput količine frakcije koja se može prodati, veličine granula i pH, dok se smanjuje higroskopnost granula đubriva oko 2 puta. Međutim, dodatak celulozi ne poboljšava statičku čvrstoću granula i zapreminsku gustinu.

Ključne reči: kalijum-dihidrogen-fosfat, celuloza, granulacija, đubrivo.

367

Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 369−376 (2020) CI&CEQ

BRUNA MURIEL F. COSTA STUDY OF THE DRYING KINETICS OF THE CAROLINA MONTEIRO MALT BAGASSE IN A PNEUMATIC COELHO TRANSPORTER CÁSSIA RIBEIRO SOUZA

GABRIELA DRUMOND A. Article Highlights DUARTE • Kinetic models of drying for malt bagasse in a pneumatic transporter were studied MARIA BERNADETE PINTO • Average energy for liquid diffusion in the drying of malt bagasse was 44.30 kJ/mol • VÍTOR J. O. SILVA The layout of 7 m reached humidity of 12% in 6 min and was more efficient • Drying of malt bagasse showed good fit using the Henderson-Pabis Modified model MARCELO S. BATISTA Abstract Chemical Engineering Department, Federal University of São João Del Malt bagasse is a by-product of the brewing industry that has high moisture, Rei, Campus Alto Paraopeba, making it very unstable and susceptible to fast microbial deterioration. This Minas Gerais, Brazil work evaluated drying kinetic models of malt bagasse during pneumatic transport with air flow at 30, 45 and 60 °C and layouts of 4.5 and 7.0 m. The results SCIENTIFIC PAPER showed that the decrease of moisture from malt bagasse was favored at higher air temperature due to the higher diffusion coefficient. In the ranges UDC 544.4:663.4:66 measured, the values of the effective moisture diffusivity and heat transfer coefficient were obtained between 2.05×10−10 to 12.74×10−10 m2/s and 175 to 363 W/(m2 K), respectively. Average energy for liquid diffusion in the malt bagasse drying process was 44.30 kJ/mol. Pneumatic transport with air flow at 60 °C and layout of 7.0 m reached rapidly the final moisture of 12% (w.b.), which may reduce transport costs and allow for long periods of stable storage for malt bagasse. The statistical tests results showed that the experimental data presented excellent fit using the modified Henderson-Pabis model, in the temperature range for both layouts. Keywords: brewing industry, flash pneumatic dryer, drying kinetic models, malt bagasse.

Beer is the third most popular drink in the world that Brazil produces an average of 13 billion L of beer after water and tea. In 2017, world beer production per year, consequently, about 3 million t of malt reached about 1.95 billion hectoliters and the global bagasse has been produced per year [5]. leading countries are China, the United States and The disposal of malt bagasse into the environ- Brazil [1]. Beer is made from water, malt, hops and ment leads to damage to ecosystems owing to their yeast as basic ingredients. The malted barley grain composition being rich in organic matter [6]. On the after the mashing and filtration process is the malt other hand, malt bagasse can be used for food pro- bagasse and constitutes the most abundant residue duction of animal origin, since it consists of 70% of (≈85%) of the brewing industry [2,3]. It is estimated fibers, 24% of proteins, and also minerals and vita- that for every 100 L of beer produced, approximately mins [7]. Bagasses are good sources of lignoceric 20 kg of malt bagasse was generated [4]. It is known and cerotic acids [8]. Thus, due to the high nutritional value and its low cost, malt bagasse has also been studied and recommended for animal and human Correspondence: M.S. Batista, Departamento de Engenharia consumption [9,10]. However, the use and storage of Química, Rodovia MG 443, km 5, Ouro Branco, Minas Gerais, wet malt bagasse is not recommended because of its 36420-000, Brazil. E-mail: [email protected] high moisture content (>70 wt.%) and the presence of Paper received: 27 August, 2019 fermentable sugars [11]. These facts make the resi- Paper revised: 5 March, 2020 Paper accepted: 19 March, 2020 due unstable and rapidly deteriorate within a 7- to 10- -day period [2]. For this reason, there is a clear need https://doi.org/10.2298/CICEQ190827014C

369 B.M.F. COSTA et al.: STUDY OF THE DRYING KINETICS OF THE MALT… Chem. Ind. Chem. Eng. Q. 26 (4) 369−376 (2020) to reduce moisture to prevent deterioration of malt rature, malt bagasse feed, air speed, operating time bagasse. and pipe size. Drying and handling for storage are important steps to preserve the quality of the malt bagasse. The EXPERIMENTAL fixed bed drier, a widely used model for the pre-drying or drying of grains, requires a great deal of physical The raw material used in this research was col- effort to be stirred, especially at the beginning of lected at the fermentation stage from a local brewery. drying, when the product is still high in moisture con- Part of the material's water was removed using a hyd- tent [12]. In addition, the increase in apparent drying raulic press with a maximum force of 1.41 t. It works time may decrease the economic gain and the quality by applying force to a small piston that transmits of the malt bagasse. On the other hand, pneumatic pressure to a larger piston that can be used on the transport can be widely used in many industrial pro- surface. As it has a direct relationship with the applic- cesses because it has advantages such as low acqui- ation area, molds of specific sizes were used in this sition cost, maintenance and operation, clean and step. The moisture reduction was calculated accord- efficient system with little contamination of the envi- ing to Eq. (1): ronment and product, as well as ease of automation Initial mass− Final mass U ()%100= (1) and versatility in assembly [13,14]. In addition, the Initial mass pneumatic transport can perform the movement and simultaneous drying of the malt bagasse, reducing The granulometric analyses were carried out to the number of operations. determine the value of medium Sauter diameter by Thus, this work proposes a pneumatic flash dry- using vibrating sieves for 20 min. Figure 1 shows the ing for the malt bagasse so that it can be reused in dimensions of the pneumatic transporter constructed other forms such as in the flour production to food ind- and used in this work. Two layouts were constructed ustry. For this, the drying behavior for the malt bag- with PVC tubes of 4.5 and 7.0 m in length and 100 asse during pneumatic transport using air flow at 30, mm in diameter. 45 and 60 °C was studied. In addition, evaluated Digital balance, dimmer and bridge H (to control parameters of heat and mass transfer and drying kin- solid feed) were used in the experiments. An anemo- etic models were evaluated. The operational perform- meter was used for speed measurement and thermo- ance of the dryer was evaluated in function of tempe- meters for temperature measurement. Were also

Figure 1. Pneumatic transporter with a layout of 7.0 m. The pneumatic transporter constisted of a blower (1), a set of resistors (2), a rotary valve type feeder (3), PVC pipes (5), two thermometers (6), two manometers (4 and 7) and a catch bag (8).

370 B.M.F. COSTA et al.: STUDY OF THE DRYING KINETICS OF THE MALT… Chem. Ind. Chem. Eng. Q. 26 (4) 369−376 (2020) used a blower which offered air at the speed of 40 Table 1 shows the mathematical models fre- m/s. The air temperature was adjusted for 30, 45 and quently used to represent the drying of malt bagasse. 60 °C. At the beginning of each experiment the air The proposed models for simulating the drying kin- velocity was adjusted through the dimmer and the etics data were evaluated by means of statistical charging malt bagasse through the rotary valve PWM tests, including determination of correlation coeffi- controller. The malt bagasse was injected into the cient, F-test and root mean square error (RMSE). transport duct with an average initial flow of 0.006 kg/s. After the malt bagasse transport, it was directed RESULTS AND DISCUSSION into the collector located at the end of the dryer and recirculated in the system until the malt bagasse Table 2 shows the results of granulometric ana- mass remained constant. The heat transfer, mass lysis of pressed and dried malt bagasse, respectively. transfer and the loss of load are factors that directly Table 2 shows that practically all pressed malt bag- influence the decrease of moisture, which is why they asse (96.44%) has the same particle size range were evaluated in this work. The heat transference which makes sense, since the malt bagasse particles coefficient for the drying process was found through have been agglomerated in the pressing step, Eq. (2), called the Dittus-Boelter equation [15]. It was whereas the dry malt bagasse has a higher diversity used for fully developed turbulent flow in smooth of diameters. The smaller diameter range is due to channels (Re >10,000, L/D ≥ 10 and n = 0.4) when the fact that the malt was ground before the mashing heating, along with Eq. (3): process. = 4/5n = 0.4 Nup 0.023 Re Pr (2) Table 2. Granulometric analysis of pressed and dry malt bagasse = hd Nu (3) D D Pressed malt Dry malt k Mesh 1 2 mm mm bagasse (%) bagasse (%)

The effective diffusion (Def) was obtained by the >6 3.360 3.360 2.38 0.00 settlement of the mathematical model described by 6.0-12 3.360 1.650 96.44 77.99 Eq. (4): 12.0-32.0 1.650 0.567 0.95 14.55 ∞ 32.0-48.0 0.567 0.328 0.00 4.48 − − π t ut() ue 61 (²nD ²ef ) RH== e r (4) 48.0-60.0 0.328 0.250 0.00 1.49 uu− π ²  n 2 oe n =0 60.0-100.0 0.250 0.149 0.24 0.75 The relation between the diffusion coefficient <100 0.149 0.149 0.00 0.75 and absolute temperature was found by Arrhenius equation (Eq. (5)): The medium diameter of dried bagasse particle was 1.47 mm and of the pressed was 2.42 mm. This E − result was already expected because when the malt DDe= RTk (5) 0 was hydrated it had shown an increase in its volume. The load loss is given by the Darcy-Weisbach In both cases, the sieve that retained the largest equation [16]: amount of material was mesh 12. However, dry malt bagasse presented higher mass distribution in the 2 = LV other sieves when it was compared to the pressed hfL (6) Di 2g material. In the malt bagasse pressing process, an aver- hh=+ h (7) totalLL higher lower age initial mass of 300 g was used and an average of

Table 1. Mathematical models used to predict the drying of malt bagasse; RH = ratio of humidity, t = drying time (s), k, k0, k1 are drying coefficients (s-1), a, b, c are constants of the models and n = number of terms in the equation

Model designation Equation =−+− Two-Terms RH aexp() k01 t b exp( k t ) Henderson and Pabis RH=− aexp( kt ) =−+−+− Modified Henderson and Pabis RH aexp() kt b exp ( k01 t ) c exp( k t ) =− Lewis RHexp( k0 t ) Page RH=−exp( kt n )

371 B.M.F. COSTA et al.: STUDY OF THE DRYING KINETICS OF THE MALT… Chem. Ind. Chem. Eng. Q. 26 (4) 369−376 (2020)

26.64% moisture was removed from the raw material, related to mobility of water molecules from the inside using an average force of 1.41 t. This stage is of ext- to the surface of the malt bagasse [19]. reme importance since the malt bagasse after the Moreover, the length of the tube also influences blast process has very high moisture and it has super- the diffusion coefficient, as shown in Table 4. This is natant liquid. This process is very fast and also rem- due to the fact that the malt bagasse remained longer oves the excess of moisture from bagasse. Excess in the longest tube, which favored the moisture dif- moisture from bagasse can be reused in the brewing fusivity. industry in the mashing step [17]. At the same time, considering only this step is not enough so that malt Table 4. Effective diffusivity calculated at different temperatures bagasse can be stored under proper conditions and Length of PVC 10 Temperature (°C) D ×10 (m²/s) another process is still necessary. pipes, m ef Table 3 shows the relation between the heat 4.5 30 2.0562 transfer coefficient and system temperature. The heat 45 5.0978 transfer coefficient was calculated for entry speed of 60 10.420 39.9 m/s (h1) and output speed of 18.0 m/s (h2) in the 7.0 30 2.7273 pneumatic transporter, both measured by a digital 45 6.5625 anemometer. Note that the convective coefficient of 60 12.7425 heat transfer was little influenced by temperature. This happens because there was practically no change in the temperature inside the pneumatic transporter. Figure 2 shows that the relation between effective diffusion and air temperature is satisfactorily des- Table 3. Coefficients of heat transfer and temperatures cribed by the Arrhenius equation (Eq. (5)).

2 2 Temperature (°C) h1 (W/(m K)) h2 (W/(m K)) 30 363.56 192.32 45 339.71 179.70 60 331.68 175.45

It is important to mention the relation between air velocity and malt bagasse mass. If the amount of malt bagasse mass is too high and contains high humidity, it can cause difficulties in the flow and agglomeration inside the tube. On the other hand, the residence time of the solid is very short at high air velocity which does not decrease the moisture rate of Figure 2. Arrhenius representations for effective diffusion as function of temperature for layouts of 4.5 and 7.0 m. the malt bagasse. In addition, there may also be an increase in energy consumption. Therefore, air velo- Average energy for liquid diffusion in the malt city/malt bagasse relation is important for the heat bagasse drying process was 44.30 kJ/mol. This and mass transfer [18]. energy value is in agreement with the energy range In addition, the speed is directly proportional to (13-110 kJ/mol) found for agricultural products [20]. the coefficient of heat transfer (Table 3). This fact was This energy is considered an energetic barrier to be already expected for the reason that the calculation of overcome for product drying. the coefficient of the heat transfer is used like a Figures 3 and 4 show the influence of the air dependent variant of speed. flow on load loss for the layouts of 4.5 and 7.0 m, res- Effective diffusivity was obtained by adjusting pectively. It is worth mentioning that pressure drop in Eq. (4) using the correlation of humidity ratio and the horizontal flow is given only by the friction factor time. Table 4 shows the values of effective diffusivity between the fluids and the tube walls, so acceleration at 30, 45 and 60 °C for the layouts of 4.5 and 7.0 m. of the fluids can be neglected. In general, the air flow Note that the effective diffusion increases with directly influences the load loss, that is, the increase the increase of air temperature. This happens of air flow increases the load loss proportionally. because of the water viscosity (resistant to the move- Figure 5 shows the reduction of the humidity in ment) that decreases with the increasing of tempe- function of the total time of the drying cycle, consider- rature. In this way, the diffusion coefficient is directly ing all the feedbacks realized in the system for pneu-

372 B.M.F. COSTA et al.: STUDY OF THE DRYING KINETICS OF THE MALT… Chem. Ind. Chem. Eng. Q. 26 (4) 369−376 (2020) matic transport with layouts of 4.5 and 7.0 m, respect- 6.5% after 120 days [31]. For achieving the necessary ively. Usually, in literature, the drying curves have on heat dissipation and drying, the heated air circulation the x-axis time and in the y-axis ratio of humidity (RH) may be used to dry the bagasse below 20% moisture. [21-25]. Note that the increase of temperature con- As shown in Figure 5, the layouts of 4.5 and 7.0 m tributes to the reduction of drying time in both layouts, with air flow at 60 °C reduce the humidity content of so that is the reason why the process was more malt bagasse to 36 and 12% in about 6 min, respect- effective at the temperature of 60 °C. According to ively. In this last case, losses should be minimised Tzempelikos et al. [26], an increased temperature due to lower moisture used to store malt bagasse. accelerates the drying process, promoting moisture Thus, the layout of 7.0 m reached lower humidity ratio content reduction more quickly at the beginning of the in same time and it was more recommended for drying process, which consequently results in shorter storage. drying time. This behavior was expected and was also recorded by other authors when drying pumpkin [27], sweet potato [28] and lemon slices [29]. In these cases, the highest rates of water vaporization in a shorter drying time were obtained by raising the temperature.

(a)

Figure 3. Influence of load loss on air flow for the layout of 4.5 m.

(b)

Figure 5. Ratio of humidity as function of time for layouts of: a) 4.5 and b) 7.0 m.

The mathematical models of Page modified, Lewis, Henderson-Pabis, modified Henderson-Pab and two-terms were evaluated for analysis of drying curves, at the temperatures of 30, 45 and 60 °C and layouts of 4.5 and 7.0 m. The choice of the best model was based on statistical tests results (correl- Figure 4. Influence of load loss on air flow for the layout of 7.0 m. ation coefficient, F-test and RMSE - root mean square error), which are shown in Tables 5 and 6. Among the The moisture reduction leads to less loss of dry selected models, the Lewis model used only one matter during prolonged storage. Correa et al. [30] parameter in the calculation and was the least fitted made careful measurements and found that the loss for the experimental data. On the other hand, the in a closely packed stack of bales was 20-25% after modified Henderson-Pabis model presented the best 2 only 40 days. Drying the bagasse to 25% moisture fit quality because of higher values of R , the lowest reduced the loss to about 3% in 40 days, rising to only values of F-test and RMSE and represented an excel-

373 B.M.F. COSTA et al.: STUDY OF THE DRYING KINETICS OF THE MALT… Chem. Ind. Chem. Eng. Q. 26 (4) 369−376 (2020) lent tool for estimating the drying time during pneu- Table 6. Statistical results of mathematical models for the 7.0 matic transport of malt bagasse, using different tem- m layout peratures in both layouts (Figures 6 and 7). These Mathematical model Parameters 30 °C 45 °C 60 °C results are in agreement with works reported in the Two-Terms a 0.943 0.941 0.433 literature [32-34], which evaluated the behavior of b 2.717 4.8E10 0.565 waste drying as a function of time. k0 0.064 0.207 0.262 k 21.70 2.5E11 0.260 Table 5. Statistical results of mathematical models for the 4.5 1 2 m layout R 0.994 0.993 0.944 RMSE 0.012 0.014 0.069 Mathematical model Parameters 30 °C 45 °C 60 °C F-Test 98.08 46.56 12.68 Two-Terms a 0.959 0.947 -0.115 Henderson and Pabis A 0.967 0.941 0.997 b 13.59 0.000 1.526 K 0.084 0.208 0.261 k 0.064 0.066 24.59 2 0 R 0.994 0.993 0.944 k1 836.1 -1.400 0.245 RMSE 0.012 0.145 0.069 R2 0.998 0.998 0.979 F-Test 196.2 93.12 25.36 RMSE 0.006 0.028 0.047 Modified Henderson A 0.937 0.920 8.536 F-Test 250.4 182.6 24.34 and Pabis B 6.3E13 10.00 -12.52 Henderson and Pabis A 0.959 0.947 1.256 C 0.001 0.000 4.627 K 0.064 0.066 0.245 K 0.007 0.190 0.565 R2 0.998 0.999 0.979 k0 1.6E14 5.9E03 0.626 RMSE 0.006 0.006 0.047 k1 -0.671 -2.440 0.565 F-Test 500.9 365.2 48.69 2 R 0.997 0.996 0.980 Modified Henderson A 0.021 7.250 5.545 RMSE 0.008 0.011 0.041 and Pabis B 13.59 0.000 -4.428 F-Test 65.40 31.05 8.571 C 0.949 -6.303 -0.058 Lewis K 0.091 0.229 0.262 K 1.017 0.065 0.130 R2 0.989 0.987 0.944 k 826.0 -1.399 0.130 0 RMSE 0.016 0.020 0.069 k 0.062 0.065 -0.257 1 F-Test 392.2 186.1 50.72 2 R 0.999 0.999 0.997 Page K 0.104 0.257 0.184 RMSE 0.006 0.006 0.017 N 0.917 0.896 1.262 F-Test 167.0 121.7 16.34 R2 0.991 0.992 0.953 Lewis K 0.073 0.076 0.189 RMSE 0.014 0.016 0.063 R2 0.984 0.980 0.946 F-Test 196.1 93.11 25.45 RMSE 0.185 0.024 0.075

F-Test 1001 729.7 96.17 100% Page K 0.100 0.111 0.075 ___ 30ºC ----- 45ºC N 0.802 0.785 1.606 ...... 60ºC 80% R2 0.997 0.995 0.993 RMSE 0.008 0.011 0.028 60% F-Test 500.8 365.1 48.50

Prediction of drying rate and drying time are 40% important aspects of kinetic models and, based on Ratio of Humidity (RH) statistical results, the modified Henderson-Pabis 20% model can be used safely for this type of malt bagasse drying. The fact that the Henderson-Pabis model 02468 has three exponential parameters, which provides a Time (min) better mathematical approximation of the experimen- Figure 6. Adjustment of experimental data to modified tal drying kinetics, could be associated to the fact that Henderson-Pabis model of drying in layout of 4.5 m at 30, 45 the model better fitted the experimental data. and 60 °C.

374 B.M.F. COSTA et al.: STUDY OF THE DRYING KINETICS OF THE MALT… Chem. Ind. Chem. Eng. Q. 26 (4) 369−376 (2020)

100% u Initial humidity content ___ 30ºC 0 ----- 45ºC ue Equilibrium humidity content 80% ...... 60ºC ut() Average humidity content in time t time (s) 60% d characteristic dimension (m)

Def Effective diffusivity (m²/s) 40% D Diffusion coefficient (m²/s)

Do Arrhenius factor (m²/s)

Ratio of Humidity (RH) Humidity of Ratio 20% R Gas constant (J/mol·K) E Activation energy (J/mol) 0% 02468 Tk Absolute temperature (K) Time (min) hL Load loss (m of H2O) hLhigher Pipe load loss (m of H2O) Figure 7. Adjustment of experimental data to modified hLhigher Load loss on discontinuance pipe (m of H2O) Henderson-Pabis model of drying in layout of 7.0 m at 30, 45 f Attrition factor and 60 °C. L Pipe length (m) Di Internal diameter of the pipe (m) CONCLUSION Ѵ Average velocity (m/s) This work evaluated the drying of malt bagasse during pneumatic transport with air flow at 30, 45 and REFERENCES 60 °C and layouts of 4.5 and 7.0 m. The raw material [1] Statista, Beer production worldwide from 1998 to 2017 (in pre-treatment was very fast and removed 27% of billion hectoliters), https://www.statista.com/statistics/ humidity from the malt bagasse. The results showed /270275/worldwide-beer-production/ (accessed 20 June that the convective coefficient of heat transfer was 2019) little influenced by temperature. In both layouts, the [2] J. Buffington, Adv. Chem. Eng. Sci. 4 (2014) 308-318 decrease of malt bagasse humidity was favored at [3] T.O. Aho, P. Niemi, A.M. Aura, M. Orlandi, K. Poutanen, higher air temperatures. Average energy for liquid dif- J. Buchert, T. Tamminen, J. Agric. Food Chem. 64 (2016) fusion in the malt bagasse drying process was 44.30 812-820 kJ/mol. [4] T.R.S. Mathias, P.P. Mello, E.F.C. Sérvulo, J. Brew. The air flow directly influences the load loss, i.e., Distill. 5 (2014) 1-9 the increase of air flow increases proportionally the [5] K.C. Piacentini, L.O. Rocha, G.D. Savi, L.C. Queiroz, B. load loss. Load loss was also higher for the layout of Corrêa, Kvasny Prum. 64 (2018) 284-286 7.0 m due to longer piping length and connections. [6] G.A. Jacometti, L.R. Mello, P.H. Nascimento, A.C. The increase of air temperature also contributes to Sueiro, F. Yamashita, S. Mali, LWT - Food Sci. Technol. the reduction of drying time in both layouts. The lay- 62 (2015) 138-143 outs of 4.5 and 7.0 m reduced the humidity content of [7] S.I. Mussatto, G. Dragone, I.C. Roberto, J. Cereal Sci. 43 (2006) 1-14 malt bagasse to 36 and 12% at 60 °C in about 6 min, [8] A.E. Nnabugwu, A.P. Uchenna, J. Am. Coll. Nutr. 38 respectively. The statistical tests results showed that (2018) 132–140 the experimental data presented excellent fit for the [9] C.M. Ajila, S.K. Brar, M. Verma, R.D. Tyagi, S. Godbout, Modified Henderson-Pabis model, using different J.R. Valéro, Crit. Rev. Biotechnol. 32 (2012) 382-400 temperatures in both layouts. [10] A.M. Teixeira, N. Sékula, B. Muller, J.R.M.V. Bezerra, M. Acknowledgments Rigo, Amb. Guarap. 14 (2018) 439-448 [11] J.A. Robertson, K.J.A. Ianson, J. Treimo, C.B. Faulds, We would like to thank the Masmorra brewery T.F. Brocklehhurst, V.G.H. Eijsink, K.W. Waldron, Food for the supply of malt bagasse. Sci. Technol. 43 (2010) 890–896 Nomenclature [12] V.R. Sagar, P.S. Kumar, J. Food Sci. Technol. 47 (2010) 15-26 h Heat transfer coefficient [13] G.E. Klinzing, Powder Technol. 333 (2018) 78-90 k Thermal conductivity coefficient of PVC [14] P. Mošorinski, S. Prvulović, I. Palinkaš, J. Appl. Eng. Sci. Nu Nusselt number p 15 (2017) 218-224 Re Reynolds number [15] F. W. Dittus, L. M. K. Boelter. Publ. Eng. 2 (1930) 443- Pr Prandtl number –461 RH Rate of humidity

375 B.M.F. COSTA et al.: STUDY OF THE DRYING KINETICS OF THE MALT… Chem. Ind. Chem. Eng. Q. 26 (4) 369−376 (2020)

[16] F.P. Incropera, D.P. Dewitt, T.L. Bergan, A.S. Lavine, [26] D.A. Tzempelikos, A.P. Vouros, A.V. Bardakas, A.E. LTC, 2008, p. 645 Filios, D.P. Margaris, Case Stud. Therm. Eng. 3 (2014) [17] W.G. Venturini Filho, T. Nokimoto. Cienc. Tecnol. 79-85 Aliment. 19 (1999) 174-178 [27] R.P. Guiné, S. Pinho, M.J. Barroca, Food Bioprod. [18] C.A.Q. Geraldi. Est. Univ. Maringa (Dissertation), 2006, Process. 89 (2011) 422-428 p. 161 [28] K.O. Falade, O.J. Solademi, Int. J. Food Sci. Technol. 45 [19] M. Kashaninejad, A. Mortazavi, A. Safekordi, L.G. Tabil, (2010) 278-288 J. Food Eng. 78 (2007) 98-108 [29] J. Wang, C. Law, P.K. Nema, J. Zhao, Z. Liu, L. Deng, Z. [20] N.P. Zogzas, Z.B. Maroulis, D.M. Kouris, Drying Technol. Gao, H. Xiao, J. Food Eng. 224 (2018) 129-138 14 (1996) 2225-2253 [30] J.L. Correa, A.F. Vela, D.O. Grimaldo, J.B. Delgado, J. [21] L.C. Corrêa Filho, E.M. D’Andrea, F.A. Sousa, V.G. Appl. Res. Technol. 8 (2010) 365-375 Figueira , E.T. Andrade, A.P. Martinazzo, Agriambi. 9 [31] B.S. Purchase, S. Rosettenstein, D.V. Bezuidenhoudt, (2015) 797-802 Proc. Ann. Congr. - S. Afr. Sugar Technol. Assoc. 86 [22] A.M.G. Diógenes, A.A.J.M. Queiroz, R.M.F. Figueiredo, (2013) 495-513 D.C. Santos, Ver. Caatinga. 26 (2013) 71-80 [32] A. Vega-Gálvez, M. Miranda, L.P. Díaz, L. Lopez, K. [23] P.C. Corrêa, E.F. Araújo, P.C. Afonso Jr., Rev. Bras. Mil. Rodriguez, K.D. Scala, Bioresour. Technol. 101 (2010) Sorgo. 2 (2003) 110-119 7265-7270 [24] R.C. Reis, L.S. Barbosa, M.L. Lima, J.S. Reis, I.A. [33] A.S. Rossi, M.G. Faria, M.S. Pereira, C.H. Ataíde, Drying Devilla, D.P.R. Ascheri. Agriambi. 15 (2011) 347-353 Technol. 35 (2017) 1130-1140 [25] C.N.B. di Domenico, T.M. Conrad, Vivências. 11 (2015) [34] A. Vega-Gálvez, A. Andrés, E. Gonzalez, E. Notte-Cuello, 134-146 M. Chacana, R. Lemus-Mondaca, Anim. Feed Sci. Technol. 151 (2009) 268-279.

BRUNA MURIEL F. COSTA PROUČAVANJE KINETIKE SUŠENJA PIVSKOG CAROLINA MONTEIRO COELHO TROPA U PNEUMATSKOM TRANSPORTERU CÁSSIA RIBEIRO SOUZA GABRIELA DRUMOND A. Pivski trop je sporedni proizvod pivarske industrije, koji ima visoku vlagu, što je čini vrlo DUARTE nestabilnom i podložnoj brzom mikrobiološkom kvarenju. Ovim radom su ocenjeni kin- MARIA BERNADETE PINTO etički modeli sušenja pivskog tropa tokom pneumatskog transporta vazduhom na 30, 45 VÍTOR J. O. SILVA i 60 °C i rasporedom od 4,5 i 7,0 m. Rezultati su pokazali da viša temperatura vazduha MARCELO S. BATISTA pogoduje smanjenju vlažnosti pivskog tropa zbog većeg koeficijenta difuzije. U merenim opsezima, vrednosti efektivne difuznosti vlage i koeficijenta prenosa toplote su i Chemical Engineering Department, 2,05×10-10-12,74×10-10 m2/s, odnosno 175-363 V/(m2 K). Prosečna potrošnja energija u Federal University of São João Del Rei, procesu sušenja pivskog tropa je 44,30 kJ/mol. Pri pneumatskom transportu vazduhom Campus Alto Paraopeba, Minas temperature 60 °C i sa rasporedom od 7,0 m brzo se postiže konačni sadržaj vlage od Gerais, Brazil 12% (vlažna biomasa), što može smanjiti transportne troškove i omogućiti duge periode NAUČNI RAD stabilnog skladištenja pivskog tropa. Rezultati statističkih testova pokazali su da se eksperimentalni podaci odlično fituju modifikovanim Henderson-Pabisovim modelom, u pri- menjenom temperaturnom opsegu za oba rasporeda.

Ključne reči: pivarska industrija, fleš pneumatska sušara, kinetički modeli suše- nja, pivski trop.

376 Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 377−384 (2020) CI&CEQ

MARIJA THE RHEOLOGICAL PROPERTIES OF BODROŽA-SOLAROV1 2 WHEAT DOUGH CONTAINING NEVENKA RAJIĆ ZEOLITE RESIDUE LATO PEZO3 1 JOVANA KOJIĆ Article Highlights 1 JELENA KRULJ • Zeolite residues affect rheological properties of dough made from treated grains BOJANA FILIPČEV1 • Zeolite strengthened the dough made of conventional and spelt wheat • RADA JEVTIĆ MUČIBABIĆ1 Na-enriched zeolite exerted higher improving effect than parent zeolite • Movable cations in zeolite structure have a role in influencing dough performance 1University of Novi Sad, Institute of Food Technology, Novi Sad, Abstract Serbia The use of natural zeolite – clinoptilolite, to protect wheat grain from storage 2University of Belgrade, Faculty of insects within environmentally-friendly storage techniques can lead to the pre- Technology and Metallurgy, sence of small amounts of zeolite residues in flour. This study investigated the Belgrade, Serbia influence of as-received zeolite clinoptilolite (Z) and sodium-rich clinoptilolite 3 University of Belgrade, Institute of (NaZ) in wheat dough on the dough rheological properties. Zeolites were General and Physical Chemistry, added to dough at 0.5-1.5 wt.% flour basis level, which is a range expected to Belgrade, Serbia remain in the grain (flour) after treatment to control storage pests. The effects SCIENTIFIC PAPER were studied in two types of wheat, conventional (Triticum aestivum) and spelt (T. aestivum spp. spelta) because they initially differ in rheological properties. UDC 633.11:549.67:664:66 NaZ was used to discern whether the presence of increased concentration of Na+ in the zeolite was able to exert a higher strengthening effect as compared to as-received zeolite (Z). NaZ exerted the highest dough strengthening effect which was mainly reflected as decreased dough softening and increased water absorption. The fact that the presence of NaZ was the most effective factor in improving the dough rheological profile suggested that the presence of movable cations in the zeolite lattice might have a pronounced role in the mechanism by which zeolite affects dough behaviour. Keywords: zeolite, wheat, spelt, dough, mixing properties.

Application of inert dusts in environmental- (rock phosphates, lime, limestone, salt), earths and friendly grain storage strategies has gained renewed ashes (powdered clay, diatomaceous earth), and syn- interest in recent times. Inert dusts are possible alter- thetic silica [2]. Natural zeolites are alkaline alumin- natives to chemical insecticides and are attractive for ium silicates which are most similar to diatomaceous use on organic grain, mostly due to their low mam- earth and can be classified in the same group – dusts malian toxicity and chemical inactivity. The Codex with natural silicates [3]. Kljajić et al. (2011) [4] rep- Alimentarius standards [1] have approved the use of orted similar insecticidal activity of natural zeolites to inert dusts instead of chemicals for post-harvest diatomaceous earth in stored wheat and proposed handling of organically produced grains. Various their use in the protection of wheat against most types of inert dusts have been used: mineral dusts prevalent insects such as rice weevil, lesser grain borer and red flour beetle. However, if inert dust is applied in storage facil- Correspondence: L. Pezo, University of Belgrade, Institute of ities, it can be expected that a certain amount of dust General and Physical Chemistry, Studentski Trg 12/V, 11000 Belgrade, Serbia. residue will remain in the treated grains. The effect of E-mail: [email protected] these remains on the quality of stored grains and end- Paper received: 8 July, 2019 products has to be considered. According to literature Paper revised: 3 April, 2020 Paper accepted: 7 April, 2020 data, around 2% of a diatomaceous earth preparation remained in wheat flour after the grain treatment [5]. https://doi.org/10.2298/CICEQ190708015B

377 M. BODROŽA-SOLAROV et al.: THE RHEOLOGICAL PROPERTIES OF WHEAT… Chem. Ind. Chem. Eng. Q. 26 (4) 377−384 (2020)

Korunić (2016) [6] gave an overview of undesirable and dough proteins affecting their aggregation. Noort effects associated with the use of diatomaceous et al. (2012) [15], investigated the partial replacement earths directly mixed with grains. Among the greatest of NaCl with different salts in wheat dough and disadvantages, the negative impact on grain flow- reported that KCl and Na2SO4 salts strengthened the ability and grain bulk density was accentuated. How- dough whereas CaCl2 and MgCl2 weakened the ever, Bodroža-Solarov et al. (2012) [7] reported some dough properties. He concluded that cations distinct- positive effects of inert dust treatments (diatomace- ively separated from the sodium and those placed at ous earth and zeolite) applied on infested and non- the end in the lyotropic series of ions (Ca2+ and Mg2+) infested wheat grains, which were mainly reflected have a weaker stabilization effect on protein config- through partial improvement of dough rheological uration. In contrast, cations, physically and chemically parameters like moisture absorption and dough more similar to sodium (Li+, K+), have identical action energy (extensigraph area). This finding was quite in dough like Na+. Effects of chloride salts of Li, Na unexpected and its mechanism has been unrevealed. and K on the rheological properties of wheat dough The explanation of this effect might be related to the were studied by Tuhumury et al. (2014) [16]. They specific composition of natural zeolites and presence observed that, regardless of cationic type, salt addit- of metal cations in its structure. ion increased dough strength and stability. The effect Natural zeolites are aluminoslicate minerals with of smaller cations (Li+

378 M. BODROŽA-SOLAROV et al.: THE RHEOLOGICAL PROPERTIES OF WHEAT… Chem. Ind. Chem. Eng. Q. 26 (4) 377−384 (2020) to dough in different amounts (in a range expected to Prior to further use Z and NaZ were finely remain in the flour after grain treatment with inert ground (Pulverisette 6, Fritsch/1 h at 250 rpm) to dust) and their effects on wheat dough rheological obtain the particle size of 0.5-1.5 μm. properties were compared. Elemental analysis of the clinoptilolite phase in the zeolitic tuff sample was performed using the MATERIAL AND METHODS energy dispersive X-ray spectroscopy (EDXS) analysis and Jeol JSM-6610LV. For the EDXS analysis Material the samples were carefully prepared by embedding In this investigation, flour of two subspecies of grains in an epoxy film, polishing crystallites, cutting Triticum aestivum: common wheat (CW) and Spelt them with a fine grid diamond cut and coating them (SW) were used. Grain material was collected after with gold. In this manner an intersection view of the the harvest in 2016. Wheat grain samples were tem- crystallite grains is obtained which allows for a det- pered by adjusting to 15 g/100 g moisture according ailed EDXS analysis of major mineral phases. A typ- to AACC method 26-95 [22]. The tempered wheat ical SEM photo is given in Figure 1. samples were ground in a Bühler 202 mill according Dough mixing properties to AACC approved methods 26-21A and 26-30A [22]. Z obtained from the sedimentary Zlatokop depo- Flour sample (4 g) was mixed with water con- sit (Vranjska Banja, Serbia) was used in the experi- taining different amounts of Z or NaZ (0.5, 1.0 and 1.5 ments. A previous detailed X-ray powder diffraction mass%) at 63 rpm using a Newport Micro-dough LAB analysis showed that the zeolitic tuff contains 72.6% mixer (Perten Instruments, Australia). The dough con- zeolite - clinoptilolite, 14.6% feldspar – plagioclase, sistency was recorded using the DLW version and 12.8% quartz [8]. Z was converted into Na-rich 1.0.0.56 software. To assess the dough mixing pro- form (NaZ) by treating Z with 2 mol dm-3 of NaCl at 60 perties, standard farinological parameters were col- °C for 5 days. Prior to its further use, the NaZ was lected during mixing such as water absorption (WA), filtered off, washed with distilled water until a negative arrival time (AT), departure time (DT), dough stability test on chloride ions and dried at 105 °C. The con- (DS) and degree of dough softening (DSO). The version does not affect the crystallinity, which was amount of water required to achieve a dough con- confirmed by an X-ray powder diffraction (XRPD) of sistency of 500 B.U. represented WA. AT is the time NaZ. The chemical composition of the clinoptilolite from the first addition of water and the point at which phase in NaZ expressed by oxide mass% is as fol- the top of the curve first intersects the consistency line at 500 B.U. DT is the interval between the initial lows: Z: SiO2 – 65.7, Al2O3 – 13.2, Fe2O3 – 1.04, Na2O – moment and the point at which the top of the curve 0.95, K2O – 1.33, CaO – 1.41 and MgO – 1.41, loss of leaves the 500 B.U. DS represents the time the dough ignition - 12.86; NaZ: SiO2 – 66.6, Al2O3 – 12.9, Fe2O3 – can be mixed before its consistency weakens. DSO is 1.04, Na2O – 10.3, K2O – 0.14, CaO and MgO – 0.04, loss of ignition - 8.96. the difference in dough consistency from the 500 B.U.

Figure 1. a) Typical SEM photo of NaZ. The shape and colour contrast indicate the presence of different mineral phases which were examined by EDXS in detail. The phases marked 1, 2 and 3 belong to feldspar, quartz and clinoptilolite, respectively; b) XRPD pattern of NaZ.

379 M. BODROŽA-SOLAROV et al.: THE RHEOLOGICAL PROPERTIES OF WHEAT… Chem. Ind. Chem. Eng. Q. 26 (4) 377−384 (2020)

line to the centre of the curve measured at 15 min of wheat (CW or SW); X2 - Z content; X3 - NaZ con- from the addition of the water. tent. All statistical analyses were carried out using Statistical analyses StatSoft Statistica 10.0® software. Descriptive statistical analyses of all the obtained results were expressed by means, standard deviation RESULTS AND DISCUSSION (SD), for each sample. Collected data were analysed by ANOVA and significant differences were calculated The assessment of the traditional rheological according to post-hoc Tukey’s test (HSD) at p<0.05 profile of dough was performed by measurements on significance level i.e., 95% confidence limit. the Microdough Lab mixer under conditions that The principal component analysis (PCA) was mimicked standard farinological analyses. The effects performed on mean dough data sets in order to of two forms of zeolite (Z and NaZ) and two wheat classify and discriminate the different samples. A pat- types (conventional T. aestivum and spelt wheat) on tern recognition technique has been applied within dough properties were analysed using two statistical descriptors to characterize and differentiate all vari- techniques, i.e., principal component analysis and eties of the samples. second order polynomial model. The second order polynomial (SOP) model was The results on the effects of Z and NZ addition used to fit the experimental data. Five mathematical to CW and SW are presented in Tables 1 and 2. models of the following form were developed to relate The CW and SW doughs differed regarding five responses (Y) and three variables (X): rheological profile as expected; CW was a stronger dough. SW had lower stability and short departure 33 3 =+ββ + β2 + β times, but the dough softening was at a similar level. Ykk0  kiiXX kiii  kijij XX , ii==11 iji ==+ 1,1 Rheological profile reflects the protein quality of the k = 1-5 (1) flours which markedly affects their baking/cooking performance and the quality of end-product. Protein β β β β where: k0, ki, kii, kij are constant regression coef- quality parameters of wheat flour, as well as protein ficients; Yk, either WA, DT, AT, DS or DSO; X1 - type content, showed a significant relationship with hard-

Table 1. Dough rheological properties of CW samples treated with natural and sodium-rich zeolite; a-d - different letters in subscript in the columns indicate that there is a significant difference at p < 0.05, according to Tukey’s HSD test. Data are presented as a mean±standard deviation of 3 replicates; CW-conventional wheat; Z-natural zeolite; NaZ-sodium-rich zeolite; WA- water absorption; AT–arrival time; DT –departure time; DS – dough stability; DSO –dough softening

Treatment WA (%) AT (min) DT (min) DS (min) DSO (B.U.) 1.5% NaZ 51.200±0.161c 0.633±0.009c 5.767±0.122b 5.200±0.098ac 118.333±0.559a 1.0% NaZ 50.867±0.064bc 0.567±0.008a 5.533±0.101ab 5.367±0.094a 120.000±1.076a 0.5% NaZ 50.767±0.107bc 0.567±0.012a 5.500±0.035ab 5.700±0.163d 120.333±2.819a 1.5% Z 50.033±0.070a 0.533±0.004b 5.567±0.142ab 5.267±0.068a 117.333±1.120a 1.0% Z 50.350±0.376ab 0.550±0.003ab 5.450±0.107a 5.400±0.147ad 119.000±2.010a 0.5% Z 50.533±0.214ab 0.567±0.011a 6.100±0.099d 4.933±0.046bc 120.000±0.723a Control 50.100±0.205a 0.533±0.004b 5.067±0.055c 4.700±0.047b 129.000±2.090b

Table 2. Dough rheological properties of SW samples treated with natural and sodium-rich zeolite; a-f - different letters in subscript in the columns indicate that there is a significant difference at p < 0.05, according to Tukey’s HSD test. Data are presented as a mean±standard deviation of 3 replicates; SW- spelt wheat; WA- water absorption; AT–arrival time; DT –departure time; DS – dough stability; DSO –dough softening

Treatment WA (%) AT (min) DT (min) DS (min) DSO (B.U.) 1.5% NaZ 52.533±0.565b 0.467±0.006a 1.667±0.017b 1.533±0.038f 110.333±1.622a 1.0% NaZ 52.033±0.214ab 0.433±0.005d 1.667±0.029b 1.333±0.020a 111.667±1.449a 0.5% NaZ 51.200±0.785ab 0.467±0.007a 1.700±0.038b 1.333±0.015a 112.000±1.224a 1.5% Z 51.233±0.434ab 0.400±0.007c 1.500±0.042ac 1.233±0.017d 116.333±0.933b 1.0% Z 50.800±0.150a 0.467±0.007a 1.567±0.018c 1.400±0.019e 117.667±1.566b 0.5% Z 51.267±0.537ab 0.500±0.007b 1.467±0.008a 1.033±0.018c 116.667±0.747b Control 50.833±0.357a 0.500±0.003b 1.467±0.026a 0.867±0.013b 130.667±2.104c

380 M. BODROŽA-SOLAROV et al.: THE RHEOLOGICAL PROPERTIES OF WHEAT… Chem. Ind. Chem. Eng. Q. 26 (4) 377−384 (2020) ness, cohesiveness, springiness, adhesiveness, doses deteriorated the sensory properties of bread chewiness, and gumminess of the cooked noodles whereas zeolite dose up to 3% yielded bread with [23]. The addition of two fractions of gluten protein acceptable quality [24]. In the present study, sensory (glutenin and gliadin) produced opposite effects on evaluation of breads was not displayed as it was not dough behaviour: glutenin improved the mixing pro- suitable to discern differences among breads pre- perties whereas gliadins decreased dough stability pared with flours that contained any of the studied and increased softening of the dough [24]. However, zeolite forms (Z or NaZ) at the applied doses (0.5- dough rheological properties also depend on the –1.5% flour basis). starch pasting properties, as shown in the work of Principal component analysis (PCA) Mudgil et al. [25]; various added ingredients added to dough (like hydrocolloids) that affect starch gelatiniz- The PCA of the presented data explained that ation may modify the properties of the dough system. the first two principal components accounted for Analysis of variance showed that statistically 90.31% of the total variance (70.56 and 19.75%, res- significant differences existed in most cases, as pectively) in the five-variable space (WA, DT, AT, DS expected. In both CW and SW samples (Tables 1 and and DSO). Considering the map of the PCA per- 2), the addition of both zeolite forms significantly and formed on the data, DT (which explained 25.1% of gradually increased dough stability and decreased total variance, based on correlations), AT (22.1%), dough softening. This enhancing effect was more pro- DS (24.3%) and DSO (11.0%) exhibited positive nounced in weak SW dough, which is in line with the scores according to first principal component, whereas finding of McCann, Li (2013) [19], that NaCl had a WA (which contributed 17.5 of total variance) showed more efficient strengthening effect in weaker (low pro- a negative score values according to the first principal tein) flour. component (Figure 2). The positive contribution to the WA was significantly increased in the NaZ treat- second principal component calculation was obs- ments, while the arrival time in CW doughs (Table 1) erved for DSO (52.2% of total variance, based on was significantly increased in the treatments with correlations), while the most evident negative impact 1.5% NaZ whereas Z did not contribute to significant was observed by WA (21.3%) and DS (11.1%). change in this parameter. In SW doughs, arrival times The position of CW and SW wheat samples were decreased by Z and NaZ addition as compared treated with Z and NaZ in the multivariate factor to the SW control. Departure times were increased, space of the first two Principal Components is dis- more pronounced in case of NaZ addition. In the played in Figure 2. PCA showed a clear separation study of Bodroža-Solarov et al. (2012) [7], it was rep- between different wheat types as well as between orted that various inert dust treatments caused a sig- differently treated samples. The scores for SW control nificant increase in dough water absorption in both and CW control were separated due to differences in low and high vitreous grains, whereas dough soften- the initial dough rheological profile. The scores rel- ing and flour quality number were not significantly ated to treated dough were distinctively arranged in affected. Marked dough improving effects were regis- two areas showing separation due to wheat type. tered in the case of insect-infested low vitreous wheat, Within each area, samples treated with Z were separ- which was explained as a consequence of greater ated from those treated with NaZ. The map of PCA accumulation of inert dust and the rise of the relative graphic showed that the first principal component proportion of protein and crude fibre in the damaged described the differentiation among the wheat culti- kernels [7]. Increased farinograph development time vars, while the second principal component described and extensigraph resistance (measured at 45 and 90 the variations in type of zeolite and also zeolite con- min) proved higher dough mixing strength of samples tent between samples. According to ANOVA data in which 0.3 g/kg of commercially available and regis- discussed above, NaZ dough showed better improve- tered diatomaceous earth preparation, “Protect-It”, ment in the rheological profile which is clearly rev- was added [23]. However, this improvement did not ealed in the PCA biplot. lead to improved baking performance, as far as loaf The points shown in the PCA graphics, which volume was considered. There is little information on are geometrically close to each other, indicate the the effect of zeolite addition on bread quality in the similarity of patterns that represent these points. The scientific literature. In a patent that describes the orientation of the vector describing the variable in fac- method of producing bread enhanced with zeolite to tor space indicates an increasing trend of these vari- improve its mineral pattern and therapeutic proper- ables, and the length of the vector is proportional to ties, it was mentioned that zeolite applied at 5% the square of the correlation values between the fit-

381 M. BODROŽA-SOLAROV et al.: THE RHEOLOGICAL PROPERTIES OF WHEAT… Chem. Ind. Chem. Eng. Q. 26 (4) 377−384 (2020)

Figure 2. PCA ordination of WA, DT, AT, DS and DSO of common and spelt wheat dough with added zeolite (Z or NaZ), based on component correlations. ting value for the variable and the variable itself. The The analysis revealed that the linear term of angles between corresponding variables indicate the wheat type was the most influential in the SOP model degree of their correlations (small angles correspond- for DT, AT and DS evaluation, statistically significant ing to high correlations). Having this in mind, it can be at p<0.01 level. The wheat type was very influential observed that high correlations exist between AT, DT variable for WA (p<0.01) and DSO (p<0.05), however, and DS which is expected as these parameters actu- substantially greater linear effect of NaZ content on ally all reflect dough stability and essentially carry the WA and DSO was observed, statistically significant at same information. DSO showed no correlation to p<0.01 level. The linear term of zeolite content influ- dough stability (DS) and WA was almost inversely enced AT and DSO, statistically significant at p<0.05 correlated to DSO indicating that the samples charac- level, while non-linear terms of Wheat type×NaZ and terized with higher WA were generally lower in DSO. Wheat type×Z have been found influential for AT calculation (p<0.05 level). These results clearly denote Second order polynomial model that NaZ was the most influential factor which affected The experimental data were fitted to the second WA and DSO. Natural zeolite did not influence WA order polynomial (SOP) model (Eq. (1)) to better rev- but did DSO. This confirms the hypothesis of the cap- eal the influence of zeolite type and wheat quality on ability of Na+ to significantly influence the improve- 2 the dough rheological profile. As confirmed by high R ment in dough rheology and implies that the presence values and insignificant lack-of-fit, the statistical of cations in the zeolite structure might be involved in accuracy of the models was verified (Table 3). the mechanism of affecting dough rheology. However,

Table 3. The effect of observed factors (wheat type, NaZ and Z content) in SOP models for prediction of responses (sum of squares); + - sgnificant at p<0.01 level; * - significant at p<0.05 level; ** - significant at p<0.10 level; error terms have been found statistically insignificant; df - degrees of freedom; NaZ – sodium-enriched zeolite content; Z - natural zeolite content; WA- water absorption; AT – arrival time; DT – departure time; DS – dough stability; DSO – dough softening; R2 - coefficient of determination

Response Term df WA AT DT DS DSO Wheat type 1 1.798+ 0.031+ 22.809+ 21.109+ 76.055* NaZ 1 2.067+ 0.001 0.160 0.327** 245.616+ NaZ2 1 0.001 0.001 0.000 0.173 30.624 Z 1 0.003 0.003* 0.017 0.252 164.103* Z2 1 0.072 0.001 0.187 0.111 55.412** Wheat type×NaZ 1 0.220** 0.005* 0.010 0.004 27.232 Wheat type×Z 1 0.079 0.003* 0.019 0.004 3.358 Error 6 0.318 0.003 0.371 0.419 75.118 R2 0.948 0.948 0.993 0.993 0.846

382 M. BODROŽA-SOLAROV et al.: THE RHEOLOGICAL PROPERTIES OF WHEAT… Chem. Ind. Chem. Eng. Q. 26 (4) 377−384 (2020) there also existed significant interactions between [4] T. Stadler, M. Buteler, D.K. Weaver, Pest. Manag. Sci. 66 wheat type, zeolite type and zeolite content in relation (2010) 577–579 to WA and AT which underlines the complexity of the [5] P. Kljajić, G. Andrić, M. Adamović, M. Pražić-Golić, J. Sci. problem. Obviously, the most affected parameter of Food Agric. 15 (2011) 1-15 dough quality was dough softening (DSO). Both zeo- [6] T. Stathers, R. Lamboll, B. M. Mvumi, Food Secur. 5 lite types were capable to significantly decrease DSO. (2013) 361-392 Lower degree of dough softening is important [7] Z. Korunić, Pestic. Phytomed. 31 (2016) 9-18 because it is an indicator of stronger and better mach- [8] http://www.pesting.org.rs/media/casopis/2016/no.1.2/31- inable dough. For example, Mailhot, Patton (1989) 1-2_9-18.pdf [28], recommended flours with DSO less than 75 B.U. [9] M. Bodroža-Solarov, P. Kljajić, G. Andrić, B. Filipčev, Lj. Dokić, J. Stored Prod. Res. 51 (2012) 61-68 as adequate for bread making. From the aspect of [10] N. Rajić, Đ. Stojaković, S. Jevtić, N. Zabukovec Logar, J. bread-making, good quality flours have higher water Kovač, V. Kaučić, J. Hazard. Mater. 172 (2009) 1450–1457 absorption, and higher dough stability. They require [11] T. Møretrø, E. Heir, L.L. Nesse, L.K. Vestby, S. Langsrud, longer mixing times and are more tolerant to over- Food Res. Int. 45 (2012) 532-544 mixing due to higher stability. [12] N. Dasgupta, S. Ranjan, D. Mundekkad, C. Ramalingam, R. Shanker, A. Kumar, Food Res. Int. 69 (2015) 381-400 CONCLUSION [13] Y. Kohno, Y. Shibata, N. Oyaizu, K. Yoda, M. Shibata, R. Matsushima, Micropor. Mesopor. Mat. 114 (2008) 373-379 The elementary hypothesis of the research that [14] N. Eroglu, M. Emekci, C.G. Athanassiou, J. Sci. Food Na+ ions from the zeolite lattice may contribute to the Agric. 97 (2017) 3487–3499 improvement of dough rheological properties has [15] R. Schmidt, P. Szakál, Nova Biotechnol. 7 (2007) 57-62 been confirmed. The findings in this study showed [16] E.S. El-Porai, A.E. Salama, A.M. Sharaf, A.I. Hegazy, that the zeolite residues in grain (flour) (up to 1.5%) M.G.E. Gadallah, Ann. Agric. Sci. 58 (2013) 51-59 were not detrimental to the dough rheological profile; [17] M.W.J. Noort, J.H.F. Bult, M. Stieger, J. Cereal. Sci. 55 on the contrary, they may contribute to the improve- (2012) 218-225 ment of dough properties especially in the case of [18] H.C.D. Tuhumury, D.M. Small, L. Day, J. Cereal. Sci. 60 weak doughs. This conclusion favours utilization of (2014) 229-237 natural zeolite – clinoptilolite as an environmental- [19] E.J. Lynch, F. Dal Bello, E.M. Sheehan, K.D. Cashman, friendly insecticide in cereal storage. E.K. Arendt, Food Res. Int. 42 (2009) 885-891 So far, it seems that from the standpoint of flour [20] M.C.E. Belz, L.A.M. Ryan, E.K. Arendt, Crit. Rev. Food technological quality, the utilization of zeolite as ins- Sci. 52 (2012) 514-524 ecticide in storage strategies should not be cons- [21] T.H. McCann, L. Day, J. Cereal. Sci. 57 (2013) 444-452 idered objectionable. However, additional investigat- [22] R.A. Miller, R.C. Hoseney, Cereal Foods World 53 (2008) ions are necessary to better resolve the effect of zeo- 4-6 lite composition (effect of other cations present in its [23] M.J.Correa, M.C. Añón, G.T. Pérez, C. Ferrero, Food structure) on dough properties in conjunction of vari- Res. Int. 43 (2010) 780–787 ous aspects of processing performance (bread, bis- [24] AACC, Approved Methods of the AACC (10th ed.). cuit, pasta-making properties) of treated grains. American Association of Cereal Chemists, St. Paul, MN,, 2003 Acknowledgement [25] S. Barak, D. Mudgil, B. S. Khatkar, Int. J. Food Prop. 17 This work is a result of Project No: 451-03- (2014) 731-740 -68/2020-14/200222 funded by the Ministry of Edu- [26] S. Barak, D. Mudgil, , B. S. Khatkar, Int. J. Food Prop. 17 cation, Science and Technological Development, (2014) 1428-1438 Republic of Serbia. [27] D. Mudgil, S. Barak, B. S. Khatkar, Int. J. Biol. Macromol. 93 (2016) 131-135 REFERENCES [28] T. Stadler, M. Buteler, D. K. Weaver, S. Sofie, J. Stored Prod. Res. 48 (2012) 81-90 [1] Codex Alimentarius Commission “Guidelines for the [29] Patent #2148914, Method for the production of bread. production, processing and marketing of organically The method relates to the production of bread intended produced foods”, 1999 for preventive nutrition. Class. IPC7: A21D8/02. Free [2] http://www.fao.org/input/download/standards/360/cxg_03 online intellectual property registry for Russia and the 2e.pdf USSR, 2018, http://russianpatents.com/214/ 2148914.html [3] H. van den Noortgate, S.P. Sree, N. Ostyn, B. Lagrain, M. [30] W.C. Mailhot, J.C. Patton, In Wheat: Chemistry and Roeffaers, T. Wenseleers, J.A. Martens, J. Pest. Sci. 92 Technology, Vol. II, Y. Pomeranz (Ed.), AACC, St. Paul,, (2019) 643-652 MN, 1988, p. 69.

383 M. BODROŽA-SOLAROV et al.: THE RHEOLOGICAL PROPERTIES OF WHEAT… Chem. Ind. Chem. Eng. Q. 26 (4) 377−384 (2020)

MARIJA BODROŽA-SOLAROV1 REOLOŠKE OSOBINE PŠENIČNOG TESTA SA NEVENKA RAJIĆ2 REZIDUAMA ZEOLITA LATO PEZO3 1 JOVANA KOJIĆ Upotreba prirodnog zeolita – klinoptilolita, u zaštiti pšeničnog zrna od insekata pri skla- 1 JELENA KRULJ dištenju je ekološki prihvatljiva tehnika, ali njegovo korišćenje može dovesti do prisustva BOJANA FILIPČEV1 malih ostataka zeolita u brašnu. U ovoj studiji ispitivan je uticaj prirodnog zeolita klino- RADA JEVTIĆ MUČIBABIĆ1 ptilolita (Z) i natrijumom obogaćenog klinoptilolita (NaZ), dodatih u pšenično brašno, na reološka svojstva testa. Zeolit je dodat u količini od 0,5 do 1,5% brašna, što je u skladu 1 Univerzitet u Novom Sadu, Naučni sa očekivanim rasponom vrednosti koji zaostaje u zrnu (brašnu) nakon obrade zrnene institut za prehrambene tehnologije u mase radi suzbijanja štetočina pri skladištenju žita. Efekat rezidua zeolita na reološke Novom Sadu, Bul. cara Lazara 1, osobine testa su proučavani za dve vrste pšenice, za konvencionalnu pšenicu (Triticum 21000 Novi Sad, Srbija aestivum) i speltu (T. aestivum spp. Spelta), uzimajući u obzir da postoji razlika u reo- 2Univerzitet u Beogradu, Tehnološko- loškim svojstvima brašna ovih žita. Uticaj povišene koncentracije Na+ u NaZ zeolitu metalurški fakultet, Karnegijeva 4, 11000 Beograd, Srbija dovodi do poboljšanja reoloških svojstava brašna u poređenju sa zaostalim prirodnim 3Univerzitet u Beogradu, Institut za zeolitom Z. Dodatak NaZ zeolita je ispoljio najveći efekat ojačavanja testa, što se ogle- opštu i fizičku hemiju, Studentski Trg dalo u smanjenom stepenu omekšavanja testa i povećanoj moći upijanjanja vode. Činje- 12/V, 11000 Beograd, Srbija nica da je prisustvo NaZ najefikasniji faktor u poboljšanju reološkog profila testa nago- vestila je da bi prisustvo pokretnih katjona u rešetki zeolita moglo imati izraženu ulogu u NAUČNI RAD mehanizmu kojim zeolit utiče na reološko ponašanje testa.

Ključne reči: zeolite, pšenica, spelta, testo, reološka svojstva.

384 Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 385−393 (2020) CI&CEQ

RIM JOURAIPHY1* SOLID-LIQUID SEPARATION TECHNIQUES ISLAAM ANOUAR2 3 FOR EFFICIENT RECOVERY OF P2O5 HAMID MAZOUZ FROM PHOSPHORIC SLUDGE BADR EL KORTOBI4 4 RACHID BOULIF Article Highlights 5 AHMED MAHROU • The growth in production leads to an increase in phosphoric acid sludge generation AMINA IBNLFASSI1 • Phosphoric acid sludge contains more than 70% of 54% phosphoric acid • Physical separation techniques represent a promising solution for the recovery of 1 Environmental Sciences and P2O5 losses Development laboratory, Faculty of • In our study, press filtration is the most effective separation technique Science and Techniques, • Recovered acid can directly integrate with the storage tanks of phosphoric acid University HASSAN I, Settat, Morocco Abstract 2Laboratory of Analytical Chemistry Phosphoric acid sludge is one of the prominent problems in the phosphate ind-

and Physico-Chemistry of ustry. Its production co-occurs with large losses of P2O5 that affect the process Materials, Department of performance. Management and valorization of this waste is a key issue. For Chemistry, Faculty of Sciences the industrial application in solid-liquid separation, selecting the suitable tech- Ben M'Sik, Hassan II University, nique for an efficient separation is a fundamental factor. Most of the separation Casablanca, Morocco processes fail because of insufficient development effort in the laboratory 3OCP. SA Jorf Lasfar, El Jadida phase and misunderstanding of separation objectives. The aim of this work is Morocco 4CBS laboratory, University to deal with the P2O5 losses by studying their recovery from phosphoric sludge Mohammed VI Polytechnic, using a physical solid-liquid separation technique. Hence, laboratory tests Benguerir, Morocco have been conducted using vacuum filtration, centrifugation, and press filtra- 5Laboratory of Physics and tion. The comparison was based on three requirements: the liquid recovery Nuclear, Atomic, Molecular, rate for each method, the amount of P2O5 recovered, and the solid content Mechanic, and Energy, remaining in the recovered acid. Results showed that press filtration gave the

Department of Physics, Faculty of higher performance of P2O5 recovery, up to 88.90%, followed by vacuum filtra- Sciences; University Chouaïb tion with 69.81% and then centrifugation with 80.91%. Doukkali, El Jadida, Morocco Keywords: phosphoric acid sludge, solid-liquid separation, vacuum fil- SCIENTIFIC PAPER tration, centrifugation, press filtration, P2O5 recovery.

UDC 661.632:66.067.17:62

Phosphoric acid is the most produced acid in the Ca3(PO4)2 + 3H2SO4 +6H2O  3(CaSO4.2H2O) + world after sulfuric acid [1]. It is a necessary raw + 2H3PO4 material for various processes, especially for phosphate-based fertilizers. The production of phosphoric The above reaction produces phosphogypsum acid, by wet process, is carried out by the reaction of (PG) as solid waste and weak phosphoric acid 27- phosphate rock with a mineral acid, mainly the sul- –29% P2O5. The latter is concentrated to reach 50- furic acid, according to the following reaction [2]: –54% P2O5, commercial-grade, using forced circulation evaporation process [2]. Due to the evaporation

Correspondence: H. Mazouz, OCP. SA Jorf Lasfar, El Jadida of water, the solubility of the impurities decreases, Morocco. leading to the formation of a saturated solid. The E-mail: [email protected] latter gives as a result the phosphoric acid sludge Paper received: 25 December, 2019 Paper revised: 7 April, 2020 (PAS) after desaturation and decantation steps [3]. Paper accepted: 21 April, 2020 PAS management has attracted much attention *Present address: Engineering and industrial management lab- from production engineers and researchers in the oratory, Faculty of Science and Techniques, University HASSAN I, Settat, Morocco phosphate industries. It has a direct impact on production performance, promoting the recovery of P O https://doi.org/10.2298/CICEQ191225016J 2 5

385 R. JOURAIPHY et al.: SOLID-LIQUID SEPARATION TECHNIQUES… Chem. Ind. Chem. Eng. Q. 26 (4) 385−393 (2020) as one of the performance indicators of phosphoric To fill the gap, a benchmark work had been acid production. The complexity of PAS management established on the methods of separating the PAS is related to its characteristics as a muddy, slushy sludge in the phosphate industries. This investigation mass and sediment [4]. To remedy this, adapted showed that the most used separation techniques by techniques of management have emerged in various phosphate process industries are filter press and phosphate processing industries such as the sludge band filters [2,4,6]. lagoon, the double settling, the intermediate cleaning, This work aims to select the appropriate phys- or as a raw material for the production of the triple ical separation technique to recover from the PAS the superphosphate (TSP) [2]. Moreover, the OCP eng- acid expressed as P2O5 and a dry solid to be stored in ineers studied internally various solutions essentially, stacks or removed with gypsum. Three potentials recycling to the phosphoric acid process reactor [5], methods were compared: vacuum filtration, centrifug- or the PG filter, or a desaturation reactor [4]. These ation and press filtration. The comparison was based alternatives represent disadvantages when applied on on three criteria: the liquid recovery rate for each an industrial scale. For the first case, PG crystalliz- method, the amount of P2O5 recovered and the solid ation is affected, leading to small crystals formation, content remaining in the recovered acid. which increases P2O5 losses as PG syn-crystallize. In the case of recycling on the filter, it affects the phos- EXPERIMENTAL phoric acid filtration operation by plugging of the filter pores. For the third case, recycling to the desaturation Sludge characterization reactor leads to a solid deposition that plugs the agi- The sludge used in this study was collected from tator during the decantation step. the 54% P2O5 phosphoric acid decanter in the phos- Few publications were conducted to study the phoric acid production unit, where the concentrated recovery of P2O5 from PAS using a solid-liquid separ- acid was settled. The analyses were made on the ation technique. Allen and Berry [4], in their study on solid phase obtained after specific treatment adapted the purification of phosphoric acid, noted that the to the XDR method. This protocol consisted of the sludge from the clarifier is sent to an underflow filter recovery of the solid phase after a solid-liquid separ- specially designed for the phosphoric acid recovery. ation by centrifugation at a speed of 3500 rpm for 15 Furthermore, the PAS was centrifuged or treated by min. The recuperated solid was washed three times settling to separate the particles from the acid in San- to eliminate the acid traces: two times with water and ders and Kealy’s study on the utilization of phosphoric one time with a mixture of water and acetone in order acid sludge, and in Allaedini and Zhang’s research on to eliminate the phosphoric acid trace and the organic the treatment of phosphoric acid sludge for rare earth matter. The solid was recovered by centrifugation at a elements recovery [6,7]. Although, Becker proposed speed of 3500 rpm for 15 min, dried, and then ana- different separation equipment like Lamella tanks and lyzed using the XRD technique. Below, in Figures 1 filter centrifuge, and gave the reasons why the tradi- and 2, the particle size distribution and solid XRD of tional settling thank is no longer the universal answer the samples are presented. to the problem of PAS [2]. Nowadays, the most used According to the initial analysis carried out using management technique in the phosphate industries is Malvern Mastersizer 2000, the particles are below gravitational settling, using tanks to recover the acid 400 µm, with 50% being below 35.01 µm, as shown in estimated at more than 60% of the total mass of the Figure 1. The volume average of the PAS is 51.612 sludge [2]. However, the inadequacy of this technique results in the presence of fine particles in the recovered acid, which causes settling issues during the phosphoric acid separation and transport. Moreover, decanted sludge always has a high acid content which represented a loss of more than 1% in phosphoric acid production. So far, researches in this field have focused on describing the performance of the implemented techniques in the phosphoric acid production plants. No work in the literature focused on a comparative study between different separation techniques for P O recovery from PAS. 2 5 Figure 1. Particle size distribution of the phosphoric acid sludge.

386 R. JOURAIPHY et al.: SOLID-LIQUID SEPARATION TECHNIQUES… Chem. Ind. Chem. Eng. Q. 26 (4) 385−393 (2020)

Figure 2. XRD result of the phosphoric acid sludge.

µm. It was measured that D10, D20, D50and D90 of separation. Studies concerning the physical separ- the solid particles by volume were, respectively, 7.37, ation techniques showed that for the three chosen 11.89, 35.01 and 113.12 µm. techniques, the following parameters need to be con- The composition of the PAS was analyzed by sidered: yields, flow rate, the feedstock variation, clar- Bruker AXS D-8 diffractometer using Cu-Ka radiation ity of the liquid phase, and solids content in the liquid in Bragg–Brentano geometry (q–2q). The results are [8]. The present study aims to deal with the PAS given in Figure 2. According to the XRD data, the par- issue, by focusing only on applied requirements for ticles of phosphoric acid sludge are predominately industrial uses. The determination of these require- composed of the gypsum CaSO4⋅2H2O, the malladrite ments allows choosing first the optimal conditions for

Na2SiF6, sodium as a minor element, the anhydrite each separation method, and then the separation

Ca(SO4) that precipitates during concentration of the technique allowing maximum recovery. The perform- acid, and the chukhrovite Ca4Al2(SO4)F12(H2O)12. ance evaluation for the selected technique was based

The P2O5 analysis of the sludge and the recup- on the parameters below. erated acid was measured using a method based on The solid content of recovered acid (SC). The UV spectroscopy. The technique consists of forming a percentage of suspended particles remaining in the vanadium phosphomolybdate complex by the reaction acid after separation. The solid content was mea- of orthophosphate ions with an acid solution contain- sured according to the Standard Practice NF EN 872: ing molybdates and vanadates. The orthophosphates Determination of suspended matter − glass fiber filter are obtained after the digestion of sludge/phosphoric filtration method. acid samples with a perchloric acid at high tempera- The separation rate of the acid (SRAcid). The tures. The absorbance of the solutions complex is mass of acid recovered (m Acid(R)) from the initial measured using UV spectroscopy to determine the sludge mass (m sludge (s)). This criterion is deter-

P2O5 concentration. mined according to the following relation: The physical characterization of the raw PAS = mAcid(R) (2) solid is presented in Table 1. SRAcid mSudge(S)

Table 1. Physical characteristics of phosphoric acid sludge The recovery rate of P2O5 (RRP2O5). The most Sludge Tempera- P O Con- Solid essential criterion in this work, since P2O5 is the per- 2 5 Density Viscosity sample ture centration Content formance criterion of the phosphate processing ind- Well ustry. It describes the amount of P2O5 recovered (m 40 °C 56% 1.9 27 cP 30% mixed P2O5 (r)) from the initial amount of P2O5 in the sludge (m P2O5 (s)). It is calculated by multiplying the con- Separation settings for the selected techniques centration of P2O5 by the mass of the recuperated acid/solid. The equation can be written as: In general, the first step in the solid-liquid separation process is to define the requirements of the

387 R. JOURAIPHY et al.: SOLID-LIQUID SEPARATION TECHNIQUES… Chem. Ind. Chem. Eng. Q. 26 (4) 385−393 (2020)

mXmsc(R)(R)(RL)(1− (RL)) and 70 µm and two pressures: 0.6 and 0.8 bar. The RR ==PO25 PO 25 (3) PO25 − air pushed the sludge towards the compartment of the mXmscPO(S)( PO S)(S)(1 (S)) 25 25 separation. The compacting phase of the cake was To maximize the recuperation of the acid, the not taken into consideration in the test. The two termi- mass of recovered acid (m(RL)) must be high, and its nals of the separation compartment allow the reco- solid content (SC) must be low. very of the acid. Tests were repeated four times using a clean filter in order to identify the incertitude of the Separation conditions for the selected techniques measurement. After the separation, the mass of each To investigate the effects of selected para- separated phase was measured. The selection of the meters, and their relative importance to P2O5 reco- porosity of the media was made according to the par- very, experimental studies were carried out as des- ticle size distribution and available sizes. cribed below. Figure 3 summarizes the organization of the Vacuum Filtration paper. For vacuum filtration, the sludge was introduced to the Buchner. The dispositive was filled with the sludge sample, and the pump was then started and adjusted to 0.6 bar. The vacuum was maintained until the surface appeared dry to avoid the effects of channeling and cracking. The tests were run for three filter media with three different porosities: 69.46, 57.06, and 110.88 µm under the same vacuum. The selection of the porosity of the media was made according to the particle size distribution and available pore sizes. These membranes, in polypropylene, obtained from Khosla Filters, are known to have excellent chemical resistance and a wide range of operating performance. During the separation, the filtration time is recorded. The mass of each separated phase and the solid content of the acid were measured after the separation. For incertitude identification, the test was repeated four times using a clean filter. Centrifugation Figure 3. Methodology flow chart of the PAS separation. In the case of centrifugation using Hermle Z446 RESULTS AND DISCUSSION K laboratory centrifuge, samples were introduced to the centrifuge tubes. The speed of separation varied To define the conditions under which the separ- between 2000 and 3500 rpm, and the rotation time ation is made with maximum efficiency, preliminary between 3 min and 20 min. The tests were first car- laboratory tests were carried out before choosing the ried for samples of 75 g to fix the optimal settings and optimal settings. then repeated with a quantity of 400 g of sludge, under five chosen optimal settings. After the separ- Vacuum filtration ation, the mass of each separated phase and the In this case, the efficiency of the solid-liquid sep- solid content of the acid were measured. Centrifug- aration was evaluated for different filter media with ation tests were repeated four times to identify the the same base (polypropylene) and the same weave incertitude of the measurement. (twill). The filtration test conducted was based on four Press filtration criteria: the filtration time, the cake mass, the liquid mass, and the solid content. Table 2 presents the For the filter press, a laboratory press filter of L. results. Choquenet was used. The sample was injected to the A first comparison of the liquid-solid separation tube. With the aid of a condenser, the air was intro- results, using vacuum filtration, showed that Media 1 duced into the tube of the filter press. Tests were allowed a filtration in the minimum of time with a high carried out with two types of membranes in polyester mass of the 50-54% phosphoric acid but a high per- obtained from Khosla, with two media porosities: 50

388 R. JOURAIPHY et al.: SOLID-LIQUID SEPARATION TECHNIQUES… Chem. Ind. Chem. Eng. Q. 26 (4) 385−393 (2020) centage of solid content. It is probably due to the pore solid, XRD results showed that the composition rem- size, which directly affects the pressure drop and par- ains the same despite the separation (Figure 2). ticle penetration performance [9]. As the ratio of the The same separation was carried out by El-Shall particles exceeding 110 μm constitutes only 10% [11] using a Buchner-type filter of a 4.6 inch diameter (D90 = 113.12), the pores of Media 1 are not tight with a polypropylene filter cloth of 200 mesh aperture, enough to retain the majority of the particles, contrary equivalent to 47 µm, with a vacuum of 22 Hg (0.74 to Media 2, that has small enough pores of 57.06 µm, bar). It showed a phosphoric acid recovery yield of allowing a suitable filtrate clarity with SC of 1.2%. The 70%. By using a media with 47 µm, the pores may be low solid content of the recovered acid is also due to smaller than needed for allowing for the liquid to the formation of a deposited layer, made of particles penetrate, causing the plugging of the media pores of varying sizes that capture the fine particles [10]. In and decrease in the flow rate [12]. Compared to our fact, the accumulation of solids in the cake increases study, using a filter cloth of 57.06 µm allowed us to the pressure drop for a given flow rate, increasing the reach a phosphoric acid recovery yield of 80.91%. filtration time: 72 s compared to 57 s for Media 1. Compared to the vacuum filtration of the weak phosphoric acid, as presented by Kruger [13], working Table 2. Results of the vacuum filtration sludge using three on the effect of ionic impurities on the crystallization types of media of gypsum in wet-process phosphoric acid, the separ- Parameter Media1 Media 2 Media 3 ation gave the same separation rate. The author used Pores size (µm) 110.88 57.06 69.46 a vacuum of 60 kPa, which is equal to the vacuum Filtration time (s) 57 72 64 used in our study. The result could be interpreted by the closeness of the viscosities values of the two Cake mass (g) 138.92 146.87 144.72 mixtures: the viscosity of the digested slurry is 50-60 Liquid mass (g) 245.5 241.93 243.42 cP at 75 °C, and the viscosity of the phosphoric acid Solid content (%) 3.4 1.2 2.8 sludge is 40-50 cP at 40 °C.

To evaluate the effectiveness of the selected Centrifugation settings, the separation rate and the recovery rate of Tests were carried out first by using samples of P2O5 had been calculated. The obtained results are 75 g to select the five optimal settings. Results are given in Figure 4. shown in Figure 5. In Figure 6, to make the test more effective, only the five optimal settings that gave the most significant values of SR and RR were selected. The five selected optimal settings are presented in Figure 6. The best performance was achieved at speed of

3500 rpm for 7 min (420 s), with 69.83% of P2O5 recovered as 50-54% phosphoric acid. The recovered solid XRD results showed that the composition remains the same despite the separation (Figure 2). According to Svarovsky, the particle size and the solid concentration are the primary properties to consider in a particular solid-liquid separation system [14]. Tak- ing these two criteria as a guide, Svarovky modified the diagram of Lloyd and Ward to show the range of Figure 4. Results of the vacuum filtration with optimal settings. solid-liquid separation technique for different particle sizes and concentrations. The diagram shows that for The Media 2, characterized by a pore size of PAS, containing more than 50% of particles with a 57.06 µm and a twill weave, gave the best result with size larger than 50 µm, the centrifugation is not the a percentage of P2O5 recovered of 80.91%. According most effective technique. to Eq. (3), the solid content of the acid is an important This technique was used by Salem [15] studying criterion. Media 2 produced an acid with 1.2% of solid the phosphoric acid purification sludge (PAPS). The content compared to Media 1 and 3 that produced PAPS has an average particle size of D50 = 37.94 µm,

3.4% and 2.8% solid content, respectively. As for the similar to the PAS average D50 = 35.01 µm. Further- more, the composition of the PAPS is nearly the same

389 R. JOURAIPHY et al.: SOLID-LIQUID SEPARATION TECHNIQUES… Chem. Ind. Chem. Eng. Q. 26 (4) 385−393 (2020)

Figure 5. Results of the sludge centrifugation: a) 2000; b) 2500; c) 3000; d) 3500 rpm.

Figure 6. Results of the sludge centrifugation with optimal settings: a) 3000 rpm; b) 3500 rpm. as the PAS. They are mainly composite of gypsum The results of the current study could be extra-

CaSO4⋅1/2H2O, malladrite Na2SiF6, anhydrite and polated to the production of phosphoric acid by the quartz, except for the chukhrovite, which is present hydrochloric attack. Pereira [16] used centrifugation only in the PAS. Salem used centrifugation separation as a technique of separation to study the advantages of 3000 rpm for 20 min to recover free purified phos- of this way to produce phosphoric acid. The attack phoric acid. In the same conditions, we reached a mixture was separated by a decanter centrifuge for 30

P2O5 recovery rate of 69.1%, Although, using the cen- min. Thus, the results can help to ameliorate the sep- trifugation with 3500 rpm for 7 min allowed a P2O5 aration of the reaction slurry. recovery rate of 69.9%. It allowed a significant rec- Press filtration overy rate and a gain in the time of separation (7 min compared to 20 min). This difference in time separ- Figure 7 summarizes the obtained results. ation could be explained mainly by the difference of The results showed that the filter press separ- viscosity between the two products, while it’s men- ation is optimal under a pressure of 0.8 bar using tioned that particle size and content are almost the Media 2 (52.42 µm). It permitted the highest recovery same. yield of P2O5 with 88.90% as 50-54% phosphoric acid. This may be due to the use of a polyester media.

390 R. JOURAIPHY et al.: SOLID-LIQUID SEPARATION TECHNIQUES… Chem. Ind. Chem. Eng. Q. 26 (4) 385−393 (2020)

According to K2 Technology [17], this media is char- The different parameters in Figure 8 are as acterized by its mechanical strength and high cap- follows. M1: Media 1(69.47 µm); M2: Media 2 (52.42 acity to withstand acidic medium. The resistance and µm); P1: Pressure 1 (0.6 bar); P2: Pressure 2 (0.8 bar). pore size of Media 2 allowed the formation of a por- In the case of centrifugation, a yield of 69.83% is ous cake that permits continued liquid flow with low very significant for a separation but remains in the solid content. With regard to the recovered solid, XRD case of PAS separation the least expressive. Accord- results showed that the composition remains the ing to Svarovky [14,18], providing a diagram illustrat- same despite the separation (Figure 2). ing the liquid-solid separation technology to select, centrifugation is expressive only in two cases: - Fine particles <5 μm at a low concentration; - An average diameter varying in 5 and 50 µm at a high concentration. For the PAS, more than 50% of the particles are larger than 35 µm. Thus, centrifugation remains the technique allowing for the lower recovery, about 69.83%. In the case of vacuum filtration, during the separation of a viscous mixture such as the PAS, the filtrate quality differs during the cycle by the clogging of the media. Therefore, the pressure drop increases [19]. For press filtration, the pressure variation is

poorly noticed. Even with high viscosity, the pressure Figure 7. Results of the press filtration: M1-Media 1(69.47 µm); squeezes out the liquid in the pores as it is pressed M2-Media 2(52.42 µm); P1-Pressure 1 (0.6 bar); by the upper portions [20] giving the highest recovery P2-Pressure 2 (0.8 bar). rate of 88.90%. The performance of the press filter in the sep- CONCLUSION aration of PAS could be enhanced by increasing the pressure. The press filter can generate a pressure Phosphoric acid sludge is a crucial issue for the drop of more than 1 bar, which is the theoretical limit phosphate industry because of the large amount of of the vacuum filter [14]. acid loss as P2O5. Tests for the selection of accurate Separation techniques comparison physical separation method, allowing the recovery of the maximum acid, and therefore P O , showed that The performance of the three separation tech- 2 5 the most appropriate separation technique is press niques was compared using the separation rate and filtration. This technique allowed a yield of 88.90% for the recovery rate requirements. Figure 8 summarizes a sludge characterized by Newtonian rheological the obtained results. behavior with a solid content of 30%, particle size below 400 µm and a viscosity of 25-30 cP, compared to centrifuge with 69.83% and vacuum filtration with 80.91%. Recovered acid can directly integrate the storage tanks of phosphoric acid. It will be interesting after this work to test the possibility of improving the efficiency of this separation by using separation aids. Nomenclature PG: Phosphogypsum PAS: Phosphoric acid sludge SC: Solid content of recovered acid SR: Separation rate of the acid

RR: Recovery rate of P2O5 PAPS: Phosphoric acid purification sludge Figure 8. Recovery rates and separation rates comparison.

391 R. JOURAIPHY et al.: SOLID-LIQUID SEPARATION TECHNIQUES… Chem. Ind. Chem. Eng. Q. 26 (4) 385−393 (2020)

REFERENCES [12] S. Zare, A. Kargari, Emerging Technologies for Sustain- able Desalination Handbook, Veera Gnaneswar Gure Ed, [1] W. Mecibah, R. Delimi, M. Gabli, Rev. Sci.Technol., Butterworth-Heinemann (2018), p. 558 Synthèse 24 (2012) 84-90 [13] A. Kruger, W.W. Focke, Z. Kwela, R. Fowles, Ind. Eng. [2] P. Becker, Phosphates and phosphoric acid: Raw Mat- Chem. Res. 40 (2001) 1364-1369 erials, technology, and economics of the wet-process, [14] L. Svarovsky, Solid-liquid separation part I, 4th ed., 2nd ed., Mercel Dekker, New York, 1989, p. 585 Butterworth-Heinemann, 2001, p. 554 [3] R. Gilmour, Phosphoric acid: Purification, uses, techno- [15] M. Salem, R. Souissi, F. Souissi, N. Abbed, J. Moutte, logy and economics, CRC PressTaylor & Francis Group, Waste Manage. 83 (2019) 46-56 New York, 2014, p. 334 [16] F. Pereira, Phosphoric acid production by hydrochloric [4] H.L. Allen, W.W. Berry, International Minerals & Chemical attack of phosphate ores with reduction of environmental Corporation), US N°22 083 (1981) nuisances and recovery of rare earths as by-products, [5] C. Fayard, (Timac SA), EP 0134732A1 (1983) Ecole nationale supérieure des mines de Saint-Etiennes, [6] M.D. Sanders, J. Kealy, (Swift & company), US No 443 2003, p. 230 644 (1969) [17] K2 Technologie, Filtring Canvas - Filter media for indus- [7] G.I. Allaedin, P. Zhang, Int. J. Surf. Eng. Interdiscip. trial filtration, About Filter cloth and filter cloth made of Mater. Sci. 7 (2019) 1842-1852 nylon, polyester, polypropylene, polyethylene and PTFE, https://www.k2tec.com/media-filtrant/toile-filtrante/ [8] D. Gertenbach, B.L. Cooper, Solid-liquid separation in (Accessed 24 July 2019) pilot-and commercial–scale operations, Paper 428c, Pre- sentation at the AICHE National Meeting, Salt Lake City, [18] S. Gravel, Experimental study of high pressure filtration of UT, 2010 suspended mining waste residues, University of Québec Chicoutimi, 2015, p. 454 [9] Y. Liu, B. Cheng, N. Wang, W. Kang, W. Zhang, K. Xing, W. Yang, J. Appl. Polymer. Sci. 124 (2012) 296-301 [19] R.J. Wakeman, E.S. Tarleton, Solid-Liquid Separation: Scale-up of Industrial Equipment, Elsevier Ltd., Oxford [10] J.S. Chang, S. Vigneswaran, J.K. Kandasamy, L.J. Tsai, 2005, p. 452 Sep. Sci. Technol. 43 (2008) 1771-1784 [20] A. Barua, Doctoral thesis, Department of Chemical Eng- [11] H. El-Shall, E.A. Abdel-Aal, Decreasing iron content in ineering & Chemical Technology Imperial College of Sci- wet-process phosphoric acid, Kraen J. Stewart (Ed.), The ence, Technology and Medicine, London, 2014, p. 233. Florida Institute of Phosphate Research, Bartow, FL, 2001, p. 58

392 R. JOURAIPHY et al.: SOLID-LIQUID SEPARATION TECHNIQUES… Chem. Ind. Chem. Eng. Q. 26 (4) 385−393 (2020)

RIM JOURAIPHY1 TEHNIKE SEPARACIJE ČVRSTO-TEČNO ZA ISLAAM ANOUAR2 EFIKASNO IZDVAJANJE P O IZ FOSFORNOG 3 2 5 HAMID MAZOUZ MULJA BADR EL KORTOBI4 4 RACHID BOULIF Mulj fosforne kiseline jedan je od istaknutih problema u industriji fosfata. Njegova AHMED MAHROU5 proizvodnja dovodi do velikih gubitaka P2O5 koji utiču na učinak procesa. Upravljanje i 1 AMINA IBNLFASSI valorizacija ovog otpada je ključno pitanje. Za industrijsku primenu u razdvajanju čvrsto-

1 tečno, izbor odgovarajuće tehnike za efikasno razdvajanje je osnovni faktor. Većina Environmental Sciences and Development laboratory, Faculty of procesa razdvajanja nije uspešna zbog nedovoljnih istraživanja u laboratoriji i Science and Techniques, University nerazumevanja ciljeva razdvajanja. Cilj ovog rada je suočavanje sa gubicima P2O5 HASSAN I, Settat, Morocco proučavanjem njihovog izdvajanja iz fosfornog mulja korišćenjem tehnike fizičkog 2Laboratory of Analytical Chemistry and razdvajanja čvrsto-tečno. Stoga su u laboratorijskim istraživanjima korišćene filtracije Physico-Chemistry of Materials, pod vakuumom i pritiskom i centrifugisanje. Poređenje se zasnivalo na tri kriterijuma:

Department of Chemistry, Faculty of brzini izdvajanja tečnosti, količine izdvojenog P2O5 i sadržaju čvrste supstance koja Sciences Ben M'Sik, Hassan II zaostaje u izdvojenoj kiselini. Rezultati su pokazali da je filtracija pod pritiskom dala

University, Casablanca, Morocco veće izdvajanje P2O5, do 88,90%, zatim filtracija pod vakuumom sa 69,81%, i na kraju 3 OCP. SA Jorf Lasfar, El Jadida centrifugisanje sa 80,91%. Morocco 4CBS laboratory, University Ključne reči: mulj fosforne kiseline; odvajanje čvrsto-tečno; vakuum filtracija; centrifugiranje; filtracija pod pritiskom; izdvajanje P O . Mohammed VI Polytechnic, Benguerir, 2 5 Morocco 5Laboratory of Physics and Nuclear, Atomic, Molecular, Mechanic, and Energy, Department of Physics, Faculty of Sciences; University Chouaïb Doukkali, El Jadida, Morocco

NAUČNI RAD

393

Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 395−399 (2020) CI&CEQ

JESÚS M. CASAS DROWNING-OUT CRYSTALLIZATION OF JOSUÉ LAGOS SODIUM MOLYBDATE IN AQUEOUS– Universidad Técnica Federico –ETHANOL SOLUTIONS Santa María, Valparaíso, Chile Article Highlights SCIENTIFIC PAPER • Sodium molybdate crystallization from aqueous solutions was studied • Sodium molybdate solubility was studied in water-ethanol solutions at 300 K UDC 546.33:546.77:66:661.722 • Na2MoO4 crystallization promoted by ethanol shows low energy and low water consumption

Abstract

The drowning-out crystallization process of sodium molybdate (Na2MoO4) was studied in water-ethanol solutions at room temperature. Sodium molybdate was separated from the solution in well-crystalline particles using less water and energy compared with the industrial processes that use evaporative crystallization. Results showed that crystallization of sodium molybdate di-hydrate was achieved after 20 or 40 min in one or two operating stages for super-saturation control with the addition of 25 or 50 vol.% ethanol, respectively. Crystal- lization with ethanol could reduce the operating costs for about 32% with respect to the conventional evaporative crystallization method, and the exhausted ethanol in the aqueous solution could be recovered by distillation and then recycled into the process. Keywords: crystallization, drowning-out, ethanol, molybdenum, sodium molybdate.

Molybdenum is an element classified as a tran- chemical industry, as catalysts, pigments, thermally sition metal with low concentration (10 mg/t) in nature. stable colouring agents, engine coolants, corrosion Commercial molybdenum recoveries are obtained inhibitors, fertilizer micronutrients, flame and smoke from some massive deposits of molybdenite (MoS2) suppressants, and in semiconductor and electronic ores or from molybdenite concentrates obtained from manufacturing. Sodium molybdates (Na2MoO4) are copper-moly sulphide ores in which the molybdenum used in the production of fertilizers, catalysts, fire ret- grades range from 0.005 to 0.18% [1,2]. Molybdenum ardants, pigments, and for corrosion inhibition in is being increasingly used in industry [3,4], especially water-cooling systems [1,2,4–6]. in alloys and high-performance steels applications. Molybdenum and molybdates are produced by Molybdenum can be alloyed with other metals to form metallurgical processes after a series of concentration corrosion-, thermal-, and wear-resistant materials and and refining operations. Ammonium molybdate has also is used as a protective coating on metals. been the main compound produced, which is syn- Molybdenum compounds with multiple valences thesized from the leach solutions obtained in copper, (-2 to +5) and mainly molybdates (valence +6) are lead or molybdenum industries, and also from solut- also used as reagents in analytical chemistry and the ions obtained from the recycling of molybdenum-containing wastes. In Chile, the Molymet company produces ammonium heptamolybdate, AHM, Correspondence: J.M. Casas, Universidad Técnica Federico [(NH ) Mo O .4H O] and ammonium dimolybdate, Santa María., Avda. España 1680, Valparaíso, Chile, Postal Code 4 6 7 24 2 2340000. ADM, [(NH4)2Mo2O7] by ammoniacal leaching of cal- E-mail: [email protected] cines (technical grade molybdenum trioxide, MoO3). Paper received: 29 December, 2019 Then, the purified molybdenum trioxide is obtained Paper revised: 25 March, 2020 Paper accepted: 6 May, 2020 from calcination of molybdates. Metallic molybdenum https://doi.org/10.2298/CICEQ191029017C is obtained in Germany from molybdenum trioxide by

395 J.M. CASAS, J. LAGOS: DROWNING-OUT CRYSTALLIZATION… Chem. Ind. Chem. Eng. Q. 26 (4) 395−399 (2020) a high temperature reduction using gaseous hydro- °C. The equilibrium time was verified by preliminary gen [1,2,4,6,7]. assays and measurements after 10 min did not exhi- Sodium molybdate can be produced by an eva- bit changes in the mass of sodium molybdate crystals porative crystallization process (90−100 °C), mixing formed. sodium hydroxide and molybdenum trioxide in an Equilibrium conditions were obtained readily, aqueous solution. Then, sodium molybdate is crystal- after 10 min, according to preliminary assays. lized by thermal evaporation of water with a high The final crystallization experiments were car- energy cost (about 6.74 MJ/kg Na2MoO4 [8]). The ried out at room temperature in a stirred small glass molybdenum product generated by evaporative crys- reactor using two operating batch stages, with 25 and tallization is a powder with fine particle sizes and pre- 50 vol.% ethanol, respectively. After 40 min of resi- sents low crystallinity and different particle morpho- dence time, the final solution was filtered through a logies [8]. Rumble 2018 [9] had reported that sodium GV 0.22 µm Millipore filter and then analysed. The molybdate is very soluble in water; its solubility inc- collected solids were washed with acetone in order to reased from 38.8 to 41.7 wt.% as the temperature remove the impregnated solution and then dried at 40 increased from 10 to 60 °C. °C in an electric oven. Then, the prepared solid sample The drowning-out crystallization of dissolved was homogenized and characterized by SEM and XRD. metals, promoted by the addition of organic compounds has been studied for a long time [10–12]. RESULTS AND DISCUSSION Ethanol has a ready market availability, it is a cheap reagent with low toxicity, and has been selected in Solubilities of sodium molybdate in aqueous- ° this work to improve the crystallization of highly sol- ethanol solutions at 27 C are presented in Figure 1. uble sodium molybdate. A new drowning-out molyb- This figure also presents previous results of crystal- ° date crystallization process, proposed in this study, is lization experiments obtained after 1 h at 27 C from 5 expected to reduce the water and energy cost and to 50 vol.% ethanol. consumption, compared to conventional industrial 300 processes, and also permits reagent recovery, resulting in a better cost-effective process. 250

200 EXPERIMENTAL METHOD 150 A sample of commercial molybdenum trioxide 100 MoO3 (purity 57%) was provided by Molymet SA; the main impurities contained in this sample were: Cu 50 Molybdenum Concentration (g/L) 0.5, S 0.1, P 0.05, Pb 0.05, K2O 0.2 (values in g/t). 0 The sodium molybdate compound used in this 0 5 10 15 20 25 30 35 40 45 50 55 study was synthesized by evaporating, at 80 °C, 1 L Ethanol Concentration, (vol.%) of an aqueous solution containing 40 g NaOH and Figure 1. Solubility of Na2MoO4 in water-ethanol solutions 27 °C. 118 g MoO3. Then, the synthesized compound was contacted with acetone to remove the impregnated The solubility of sodium molybdate in the solution, dried at 40 °C and then characterized by assayed solutions decreased markedly as the ethanol SEM (JEOL JSM-5410 scanning electron micro- concentration was increased. Molybdenum solubility scope) and XRD (CuKα radiation of a Siemens® in 5 and 50 vol.% ethanol were 256 and 37.8 g/L,

D5000 diffractometer), which confirmed that sodium respectively. The solubility of Na2MoO4 in aqueous molybdate di-hydrate was the only crystallized phase. solutions firstly was diminished by about 21% in 5 Distilled water, 99% NaOH p.a. pellets and vol.% ethanol and finally was reduced by 89% in 50

99.5% C2H5OH p.a. from Merck were used. vol.% ethanol.

The molybdenum concentrations of the filtered Ethanol acts as a mixed-solvent in Na2MoO4- samples were determined by a colorimetric method H2O solutions and produces a decrease in the inter- (8189, HACH). action forces between water molecules and dissolved

The solubility of Na2MoO4 was measured in species, due to the decrease in the dielectric constant water-ethanol solutions (5−50 vol.% ethanol) by and density of the solution. This promotes ionic asso- + 2- immersing, for a period of one hour, sealed tubes in a ciation between Na and MoO4 species, favoring the thermo-regulated water bath (PolyScience) at 27±2 sodium molybdate crystallization [13].

396 J.M. CASAS, J. LAGOS: DROWNING-OUT CRYSTALLIZATION… Chem. Ind. Chem. Eng. Q. 26 (4) 395−399 (2020)

The obtained phases of sodium molybdate were 300 analysed and showed a high purity and no other rel- 250 evant substance or element was found. Figure 2 pre- 200 sents the micrographs of sodium molybdate phases 1st Stage generated by evaporative crystallization and by crys- 150 tallization assisted by ethanol, at 0 and 50 vol.% 100 ethanol, respectively. 2nd Stage

The crystallized phase showed more consistent 50 morphology and larger particle sizes compared with Molybdenum Concentration (g/L) molybdate produced from evaporative crystallization. 0 0 5 10 15 20 25 30 35 40 45 50 55 This difference could be by attributed to evaporation Ethanol Concentration, (vol.%) temperatures, that produces differences in the eva- Figure 3. Operating diagram of two crystallization stages for poration and its respective crystallization rates. Crys- sodium molybdate in H2O-EtOH, at room temperature. Expe- tals formed with large particle sizes facilitates the rimental results after 20 min of crystallization in each stage. solid liquid separation. Elemental and DRX analyses show that the crystallized phase was Na2MoO4⋅2H2O. The crystallization process generated from res- Figure 3 presents an operating diagram pro- ults presented by Figure 3 could help to define a posed for the crystallization of sodium molybdate from conceptual process for producing sodium molybdate aqueous solution. Starting from a solution containing by crystallization with ethanol. The final ethanol sol- 225 g/L Mo, two crystallization stages are required in ution could be distilled, in the range of 80–86 °C, for order to obtain a performance higher than 80%, after recovering the solvent. 40 min of total crystallization time. In the first stage, Table 1 presents a summary of operating cost adding ethanol to 25 vol.%, molybdenum precipitates differences estimated for evaporative crystallization up to 121 g/L Mo. Further addition of ethanol to 50 and crystallization with ethanol of sodium molybdate. vol.% (second stage) reduces the concentration up to These calculations considered mass and energy bal- 44 g/L Mo. ances and unitary costs of purified ethanol (US$/m3 1.000), power (US$/MWh 100), and water (US$/m3 12).

Figure 2. SEM micrographs with identification of elements for sodium molybdate solids obtained: a-b) by evaporative crystallization; c-d) by crystallization assisted with ethanol.

397 J.M. CASAS, J. LAGOS: DROWNING-OUT CRYSTALLIZATION… Chem. Ind. Chem. Eng. Q. 26 (4) 395−399 (2020)

Table 1. Operating cost differences between crystallization with This work was funded by the CONICYT Chilean ethanol and evaporative crystallization (values were calculated Government agency via FONDECYT project 106 from mass and energy balances, available as supplementary 1160. Thanks goes to MOLYMET S.A. for providing material from the authors, and the unitary prices for electrical the valuable MoO used in the present work and to energy (US$/MWh 100), demineralized water (US$/m3 12) and 3 ethanol (US$/m3 1000) were obtained from the Chilean mar- the Metallurgy and Materials Department of the Uni- kets [14–16] versidad Federico Santa María, as well as the Mining Engineering Department of the Universidad de Chile, Cost Item Crystallization Evaporative for their support. (US$/kg Na2MoO4) with ethanol crystallization Power (for heating) 0.25 0.51 REFERENCES Purified water (evaporation) – 0.02 Ethanol (losses) 0.03 – [1] A. Sutulov, International Molybdenum Encyclopaedia, Total 0.28 0.53 Products, Uses and Trade, Vol. 3, Intermet Publications, Santiago de Chile, 189-200 (1980), pp. 13-22,149 The crystallization process with ethanol at room [2] F. Habashi, Principles of Extractive Metallurgy, Amalgam and Electro Metallurgy. Laval University, Métallurgie Ext- temperature exhibited about 50% lower operating ractive Québec, Vol. 4, Quebec, Canada, 1998, pp. 1361- cost of water and energy, relative to the conventional –1402 evaporative crystallization process performed, for [3] S.J. Kropschot, Molybdenum — A key component of metal example, at 92 °C using vacuum conditions. alloys: U.S. Geological Survey Fact Sheet, 2009–3106, 2010, 2p., available at http://pubs.usgs.gov/fs/2009/3106 CONCLUSIONS (accessed 25 October 2019) [4] P.C.H. Mitchell, Chemical Applications of Molybdenum, Solubility of sodium molybdate in aqueous sol- 2004, www.imoa.info (accessed 25 October 2019) ution decreased markedly with the presence of etha- [5] K.-H. Tytko, W.-D. Fleischmann, D. Grasm, E. Warkentin, ⋅ nol. Experimental solubilities of Na2MoO4 2H2O, mea- in Gmelin Handbook of Inorganic Chemistry: Mo Molyb- sured at 27 °C, in 5 and 50 vol.% ethanol were 256 denum Supplement Vol. B 4, 8th ed., Hartmut Katscher, and 37.8 g/L Mo, respectively. Aqueous molybdenum Wolfgang Kurtz, Friedrich Schröder, Springer-Verlag solubility can be reduced by about 89% at 50 vol.% GmbH, Berlin, 1985, pp. 51-74 ethanol and this reagent was an effective mixed-sol- [6] C. Gupta, Extractive Metallurgy of Molybdenum, CRC, vent that induced the crystallization. Press, Boca Raton, FL, 1992, pp. 4-11 [7] Molymet S.A., Applications of Molybdenum; Chemical The crystallized phase had more consistent Industries, www.molymet.cl (accessed 25 October 2019) morphology and a larger particle size than the molyb- [8] J. Lagos, Thesis of Environmental Chemistry, Univer- date produced from evaporative crystallization. Mass sidad de Chile, 2009 (unpublished results) balancing the system and DRX analysis show that [9] J.R. Rumble, Aqueous solubility of inorganic compounds ⋅ crystallized phase was Na2MoO4 2H2O. at various temperatures, in CRC Handbook of Chemistry Finally, the proposed process for sodium molyb- & Physics, 99th ed., CRC Press, Boca Raton, FL,, 2018, date recovery consists of 1 or 2 crystallization stages pp. 8-116 at room temperature, where the super-saturation and [10] A. Seidell, W.F. Linke, Solubilities: Inorganic and Metal- molybdenum recovery could be controlled by the -Organic Compounds: A Compilation of Solubility Data addition of ethanol (drowning-out crystallization pro- from the Periodical, Vol. 1, Am. Chem. Soc., Washington DC, 1958, p. 1272 cess). The crystallization promoted by ethanol results [11] T.A. Graber, J.W. Morales, P.A. Robles, H.R. Galleguil- in about 50% lower operating costs for water and los, M.E. Taboada, Cryst. Res. Technol. 43 (2008) 616- energy, compared with the conventional evaporative –625 crystallization method. The spent ethanol aqueous [12] J.M. Casas, E. Sepúlveda, L. Bravo, L. Cifuentes, Hydro- solution obtained after the separation of crystals could metallurgy 113–114 (2012) 192–194 be distilled in the range of 81–86 °C, in order to rec- [13] M.A. Abolghassemi-Fakhree, D.R. Delgado, F. Martínez, over the ethanol for recycling within the process. A. Jouyban, AAPS PharmSciTech 11 (2010) 1726-1729 [14] National Energy Council, Market price of electrical Supplementary material energy, www.cne.cl (accessed 25 October 2019) Additional data are available from corresponding [15] J. Balbontín, Water interconnection study: opportunities author upon request. and challenges for Chile, Report, SMICEChile, 2019, https://consejominero.cl (accessed 25 October 2019) Acknowledgments [16] Ethanol Prices, GlobalPetrolPrices, https://es.globalpetrolprices.com (accessed 25 October 2019)

398 J.M. CASAS, J. LAGOS: DROWNING-OUT CRYSTALLIZATION… Chem. Ind. Chem. Eng. Q. 26 (4) 395−399 (2020)

[17] CRC Handbook of Chemistry and Physics, D.R. Lide 380,986-987,2427. (Ed.), 85th ed., CRC Press, Boca Raton, FL, 2004, pp.

JESÚS M. CASAS KRISTALIZACIJA NATRIJUM-MOLIBDATA JOSUÉ LAGOS PRECIPITACIJOM U VODENO-ETANOLNIM Universidad Técnica Federico Santa RASTVORIMA María, Valparaíso, Chile Proučavan je proces kristalizacije natrijum-molibdata (Na MoO ) precipitacijom u NAUČNI RAD 2 4 rastvorima vode i etanola na sobnoj temperaturi. Natrijum molibdat je odvojen od rastvora u obliku kristalnih čestica koristeći manje vode i energije u poređenju sa industrijskim procesima koji koriste evaporativnu kristalizaciju. Rezultati su pokazali da je kristalizacija natrijum-molibdat-dihidrata postignuta nakon 20 ili 40 min, u jednoj ili dve operativne faze za kontrolu super-zasićenja sa dodatkom 25 ili 50 vol% etanola. Kristalizacija etanolom može smanjiti operativne troškove za oko 32% u odnosu na konvencionalnu metodu evaporativne kristalizacije, a iscrpljeni etanol u vodenom rastvoru može se predestilisati i ponovo koristiti.

Ključne reči: kristalizacija, precipitacija, etanol, molibden, natrijum-molibdat.

399

Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020) CI&CEQ

MAJA ČOLNIK BIODEGRADABLE POLYMERS, CURRENT MAŠA KNEZ HRNČIČ TRENDS OF RESEARCH AND THEIR MOJCA ŠKERGET APPLICATIONS, A REVIEW ŽELJKO KNEZ

Laboratory for Separation Article Highlights Processes and Product Design, • An overview of the different biodegradable polymers and their properties Faculty of Chemistry and Chemical • The possibility of production of biodegradable plastics from renewable or non-renew- Engineering, University of Maribor, able sources • Maribor, Slovenia Current scientific research, applications and global production of bioplastic are considered • The properties of polymers obtained from biomass are presented in details REVIEW PAPER

Abstract UDC 678.7:60:502/504 Biodegradable polymers have been developing rapidly in the last years and are widely used today in the fields of pharmacy, clinical biomedicine, cosmetic, medical, and packing industries, tissue engineering, agriculture and other areas. The interest in biodegradable polymers has been increasing, mainly due to rising oil prices, which is the basic feedstock of plastic derived from petroleum, and also due to the problem of the removal of waste plastics that accu- mulate in the environment. Biodegradable polymers have many advantages in contrast to synthetic polymers and can be decomposed in the environment to non-hazardous substances. Biodegradable polymers are classified into two classes based on their synthesis, i.e., synthetic and natural polymers. They are derived either from petroleum resources or from biological resources. The following review presents an overview of the different biodegradable polymers and their properties, current scientific research, applications, global production of bioplastic and replacement of conventional plastic. Keywords: biodegradable, polymers, natural origin, bioplastics.

Synthetic plastic was considered to be the mat- degradable polymer materials significantly increase erial of the future in the previous century, but today it and consequently can contribute to the decrease of represents a major environmental problem [1]. The environmental problems concerning waste polymer main disadvantages of synthetic polymers are the materials [3]. Research on renewable resources is lengthy decomposition process and the production focused on the use of corn, soy, sugarcane, potato, based on non-renewable raw materials. According to rice or wheat, and seeds rich in oil or fermentation expert estimation, as much as 7% of world oil and gas products as raw materials to produce biopolymeric supplies are used for polymers production [2]. An materials [4]. The use of biomass to produce biopoly- alternative possibility is presented by polymers that mers has a great advantage, since biomass-derived are biodegradable or made from renewable sources. polymers are biodegradable and relatively easy to Thus, the production and use of bio-based and bio- recycle [3]. One of the main disadvantages of biodegradable polymers obtained from renewable sources is Correspondence: Ž. Knez, Laboratory for Separation Processes their fast degradation rate, due to their dominant hyd- and Product Design, Faculty of Chemistry and Chemical Eng- rophilic character and, in some cases, inadequate ineering, University of Maribor, Smetanova 17, SI-2000 Maribor, mechanical properties, particularly in wet environ- Slovenia. E-mail: [email protected] ments [5]. Despite their disadvantages, biodegradable Paper received: 10 December, 2019 polymers offer a wide range of advantages since Paper revised: 4 May, 2020 many plant materials are used to make them. This Paper accepted: 6 May, 2020 means that it is no longer necessary to use chemical https://doi.org/10.2298/CICEQ191210018C

401 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020) fillers contained in synthetic plastic products. More- materials depends on their chemical structure and not over, during the production cycle of biodegradable the carbon source of the polymer [8]. According to the plastics, energy consumption is reduced while the forecast [9,10], the global bioplastics production cap- decomposition does not lead to unnecessary release acity is set to increase from around 2 million t in 2018 of CO2 into the environment. The use of biodegrad- to approximately 2.6 million t in 2023 (Figure 1) [9]. able products could reduce dependence on fossil Currently, bioplastics represent about 1% of 335 mil- fuels and thus contribute to a cleaner environment lion tons of plastic produced annually. Bioplastics are and more and more research is therefore being done an increasing, innovative industry that offers solutions in the field of new biodegradable polymers and their for a sustainable plastics economy and that plays a improved properties. key role in the transformation to a bio-based circular The aim of this article is to review the recent economy [9,10]. Yet, despite these advantages, the research on the most used biodegradable polymers, data shows that the overall growth of the global bio- to examine their application possibilities, and the pos- plastics industry is currently being slowed down by sibility to decrease, or even eliminate, the usage of low oil prices and a lack of political support for the synthetic plastics, which have already caused major bio-based economy [9,10]. environmental problems. Bio-based non-biodegradable plastics currently make up for around 48% (1 million t) of the global bio- BIOPLASTIC - BIODEGRADABLE POLYMERS plastics production capacities (Table 1) [9]. Just like for the bio-based polymers, the global production of Bioplastic materials are a family of materials biodegradable polymers is also increasing (Table 1). with different properties and applications that can be The production of PLA (polylactic acid) and PHA bio-based, biodegradable or both. Biodegradable (polyhydroxyalkanoates) is currently growing. Reg- plastics are made from renewable or non-renewable arding the global production capacities of bioplastic sources, which completely degrade in the environ- by material type, the PLA share is predicted to inc- ment through a thermochemical process into CO2, rease from 10.3 to 13.2%, while the PHA share is pro- methane, water, biomass and inorganic compounds jected to increase from 1.4 to 5.8%. The production of that microorganisms can easily mineralize [6,7], while other biodegradable polymers should remain the bio-based polymers contain organic carbon of renew- same or decrease [10]. able feedstock and are not necessarily biodegrad- In the following, biodegradable polymers are able. Moreover, bio-based polymers can be synthe- classified as either bio-based or petrochemical-based. sized from naturally occurring materials as well as The material is mostly biodegradable by nature and from natural substances that have been polymerized produced from biomass (plants, animals or microorg- into high molecular weight materials by chemical anisms) such as polysaccharides (e.g., starch, cellu- and/or biological methods [8]. Bio-based materials lose, lignin and chitin), proteins (e.g., gelatine, casein, have a similar molecular structure to petrochemical wheat gluten, silk and wool) and lipids (e.g., plant oils polymers. Above all, the biodegradability of a plastic and animal fats) [11,12]. The major natural polymers

Figure 1. Global production capacities of bioplastics [10].

402 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020) are starch, cellulose and soy protein. Natural rubber Furthermore, the latest research on the most as well as certain polyesters, either produced by mic- common biodegradable polymers, the addition of roorganisms or plants (PHA), and poly-3-hydroxybut- copolymers, additives or fibers, which can improve yrate (PHB) or synthesized from bio-derived mono- the mechanical and physical properties of polymers mers (PLA), fall into this category [11,12]. Petrochem- and possible applications of biodegradable polymers ical-based biodegradable polymers such as aliphatic are shown in Tables 2-9. Thermoplastic starch (TPS) polyesters (e.g., polyglycolic acid (PGA), polybutylene is still the most widely used material for the research succinate (PBS) and polycaprolactone (PCL)), aro- and production of biodegradable polymers. Namely, matic copolyesters (e.g., polybutylene succinate tere- TPS has been the first biopolymer on the market, but phthalate (PBAT)) and poly(vinyl alcohol) (PVA) are currently two groups of biodegradable polymers have produced by synthesis from monomers derived from the highest market potential. petrochemical refining, which possess a certain deg- The first is polylactic acid (PLA), produced by ree of characteristic biodegradability (Figure 2) [11,12]. fermenting carbohydrates, and the second important Particularly, PBS is already industrially produced from group of biodegradable polymers are polyhydroxyalk- renewable resources by PTT MCC Biochem Com- anoates (PHA), synthesized through bacterially pany and it is called BioPBSTM. guided fermentation processes. Their products are widely used in different fields. Table 1. Global production capacities of bioplastic by material type in 2018 and 2023 [10] POLYMERS OBTAINED FROM BIOMASS Material type 2018 (%) 2023 (%) Starch Biodegradable 43.2 44.5 PBAT 7.2 4.2 Starch is the major form of carbohydrate storage PBS 4.6 4.1 in green plants and is considered the second largest PLA 10.3 13.2 biomass produced on earth. Starch is a cheap and easily available raw material, which is present in large PHA 1.4 5.8 quantities in potatoes, maize, rice and wheat [4,13]. In Starch blends 18.2 15.8 natural form, starch is not meltable and cannot be Other (biodegradable) 1.5 1.4 processed as a thermoplastic [14]. Starch granules Bio-based/non-degradable 56.8 55.5 can be thermoplasticized through a gelatinization PET 26.6 20.5 process, where the granules are disrupted and the PA 11.6 11.0 ordered crystalline structure is lost under the influ- a PEF 0.0 2.9 ence of plasticizers, shears and heat [14]. The result- PE 9.5 11.3 ant melt-processable starch is called thermoplastic 1 PP 0.0 2.0 starch (TPS) [15]. Other (biobased/non-biodegradable) 10.1 7.8 Bioplastics based on starch are suitable for the a Bio-based PP and PEF are currently in development and pre- production of packaging, in agriculture, for medical dicted to be available in commercial scale in 2023

Figure 2. Classifications of biodegradable polymers [12].

403 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020) and cosmetic application [16-19]. Due to its low cost wastewater [25]. Besides, chitosan is a promising and high availability [20], starch attracts increasing material to produce food packaging materials, and in attention and interest of researchers and industries comparison to starch due to its more hydrophobic worldwide. Xie et al. [21] studied a completely new nature represents an attractive combination of price, biodegradable sisal fiber–starch packing composite, abundance and thermoplastic behavior [23]. Further, which exhibited good biodegradability. This compo- the researchers successfully prepared mixtures of site is a promising replacement for expandable poly- corn starch (TPS) and chitosan (TPC) by extrusion styrene (EPS) as packing material, especially under and they produced more thermally stable films with large compression loads [21] (Table 2). In comparison potential application in the packaging industry [23]. to sisal fiber, which is the most used natural fiber in Rice and maize are another widespread raw the world, cassava starch is also often used as matrix materials in the manufacture of biodegradable films material. In addition, Medina-Jaramillo et al. [22] [26]. The biodegradable films from hydroxypropylated investigated the active and smart biodegradable films rice starch (HPRS) provided an increase in elongation from cassava starch and glycerol with different natural at break, water vapor permeability, film solubility and extracts such as green tea and basil by casting. The transparency, while the tensile strength value dec- high content of phenolic compounds in the extracts reased when the propylene oxide proportion inc- led to significant antioxidant activity of the films. The reased [26]. Lopez et al. [27] studied the mechanical films were thermally stable up to 240 °C and deg- properties, oxygen permeability and water vapor, as raded in soil in two weeks [22]. Other researcher com- well as thermo-sealing capacity of composite films bined the cassava starch with biomass or biomass from thermo-compressed films of TPS with talc nano- extract of microalgae Heterochlorella luteoviridis and particles. With the addition of talc (3 mass%) to TPC, Dunaliella tertiolecta [18]. Incorporation of microalgae the water vapor and oxygen permeability were red- biomass resulted in biodegradable and highly soluble uced [27]. films, making it difficult to apply in moist foods, while Tacca leontopetaloides, also known as Poly- the films which contained 2.0% of H. luteoviridis ext- nesian arrowroot starch, is a wild perennial herb [28]. ract were appropriate for application in foods with This starch is an important food source for many high water content [18]. These prepared biodegrad- Pacific Islander cultures [28]. The potential of T. leon- able films with antioxidant properties can used for sal- topetaloides starch plasticized using glycerol and mon packaging, as they provide protection against crude palm oil (CPO) was investigated by Makhtar et lipid oxidation [18]. Several researchers investigated al. [28]. The mixture becomes highly thermally resist- the use of starch and chitosan in the production of ant up to 430 °C when it is plasticized with CPO. The biofilms. Chitosan is nontoxic, biodegradable and has thermal behavior of glycerol TPS was quite similar to antimicrobial activity [23]. Thus, chitosan as a biopoly- the conventional bioplastic which supports the utiliz- mer has a wide range application in drug delivery sys- ation of T. leontopetaloides starch for bioplastic deve- tems, in the areas of biomedicine, membranes, hydro- lopment [28]. In further study, starch was used also gels, adhesives, fuel cells, as a surface conditioner, for other applications. Namely, the disposed pres- for tissue engineering, etc. [24]. It has also shown a sure-sensitive adhesive tape widely used in daily life great potential for removing pollutants from water and has been contaminating the environment and has

Table 2. Starch as biodegradable polymers, addition of copolymers, additives/fibers and their potential applications

Biodegradable Material Co-polymer Additives/fibers Application Ref. polymer Starch Corn starch – Sisal fiber Packing material [21] Starch Cassava starch – Glycerol, natural extracts such as Food packaging [22] green tea Starch Cassava starch Biomass extract of microalgae – Salmon packaging [18] Starch Corn starch Chitosan – Packaging [23] Starch Native rice starch – Propylene oxide Packaging [26] Starch Corn starch – Talc nanoparticles Packaging bags [27] Starch Polynesian – Glycerol and crude palm oil Bioplastic development [28] arrowroot starch Starch Potato starch – Glycerol, monohydrate citric acid Medical tapes and [29] biomedical electrodes

404 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020) produced vast amounts of non-degradable trash. physical properties, or a filler, such as starch and PLA Czech et al. [29] studied the advanced biodegradable [31]. NFC has an incredible potential as reinforcement pressure-sensitive double-coated tape containing material in nanocomposites for many different uses, starch carrier and water-soluble partially degradable such as foams and adhesives. Missoum et al. [33] modified pressure sensitive adhesive. They observed prepared NFC and their derivatives with using three the excellent tack and peel adhesion of these newly chemical surface modification strategies. Antibacterial constructed biodegradable self-adhesive tapes, and activities of all samples were investigated against two high thermal shear strength [29]. The complete bio- kinds of Gram+ bacteria (Staphylococcus aureus) and degradability of starch carrier and partial biodegrad- Gram- bacteria (Klebsiella pneumoniae) [33]. They ability of modified acrylic PSA were confirmed [29]. also strongly enhanced the photo-catalytic antimicro-

This environmentally friendly technology based on the bial effect of TiO2 additive (Table 3). This study shows starch has great potentials for diverse applications that it is better to use grafted NFC either alone or such as the paper industry for manufacturing of eco- functionalized with TiO2 if anti-bacterial properties are logical biodegradable products, the production of desired [33]. water-soluble biodegradable labels, medical tapes The cellulose backbone is known to be easily and biomedical electrodes [29]. biodegradable in different biodegradation conditions New research has confirmed that starch is a and environments [33]. Also, biocomposites based on highly effective natural source and, depending on the natural cellulose fibers (CF) and hydroxyethyl cellu- starch source and processing conditions, a thermo- lose (HEC) were produced in the form of green pack- plastic material with different properties suitable for aging films [34]. The effect of the different single- various applications and industries can be produced. component plasticizers on the mechanical and dynamic thermomechanical properties of the films was Cellulose investigated. [34]. Moreover, the softening effect of Cellulose is the most widespread natural organic the two-component plasticizer based on deep eutectic 12 polymer, representing about 1.5×10 t of total annual solvents (DESs) was addressed. It was found that DESs biomass production and is considered an almost non- are highly promising plasticizers for cellulose-based exhaustive source of raw material for an increased biocomposites with similar or even better plasticizing need for environmentally friendly products [30]. The effects compared to conventional plasticizers [34]. primary source of cellulose is the existing lignocellu- In the following, cellulose was also used as mat- losic source in forests [30]. Commercial cellulose pro- erial for a new type of biodegradable cigarette filters duction is focused on sources such as wood and to accelerate their disappearance after disposal. naturally very pure cotton sources [30,31]. Cellulose About 4.5 trillion cigarette butts are discarded every fibers are used to produce biodegradable packaging year. Joly et al. [35] compared the decomposition of materials, and in cosmetic products without synthetic cellulose and plastic cigarettes filters, either intact or polymers or strong composites together with biodeg- smoked, on the soil surface or within a composting radable plastics [32]. Cellulose derivatives that have bin over a six-month field decomposition experiment been used in bioplastic synthesis were cellulose nano [35]. It was found that conventional plastic filters take crystals (CNC) [31], nanofibrillated cellulose (NFC), 7.5-14 years to disappear, in the compost and on the cellulose acetate butyrate, cellulose acetate and bio- soil surface, respectively, in contrast to cellulose fil- PE [32]. The derivative of the cellulose was mixed ters, which take 2.3-13 years to disappear, in the with other biopolymers matrices in order to improve compost and on the soil surface, respectively. [35].

Table 3. Cellulose as biodegradable polymers, addition of copolymers, additives/fibers and their potential applications

Biodegradable polymer Material Co-polymer Additives–fibers Application Ref.

Cellulose Wood pulp – TiO2 Paper, packaging and composites [33] Cellulose Softwood kraft pulp – Natural cellulose fibers, Green packaging films [34] glycerol, propylene carbonate and ethylene carbonate Cellulose Cellulose – – Cigarette filters [35] Cellulose Sugarcane bagasse – Hemicellulose (5%) of bagasse High value plastics [36] Cellulose Native cellulose Chitosan – Food packaging [37] (cotton linter) Cellulose Empty fruit bunch Cassava starch Glycerol Plastic bags and food packaging [38]

405 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020)

Sugarcane bagasse was fractionated to cellu- plastics. Blending of SP polymers with biodegradable lose, hemicellulose and lignin by a proprietary steam polymers is a natural choice to overcome the said explosion process followed by downstream purificat- lack of SP-based polymers. The polymers used for ions [36]. The cellulose was acetylated under hetero- mixing with SP are hydrophobic and therefore cannot geneous conditions to obtain cellulose acetates. The establish a strong bond with SP and a compatibilizer novel feature of this study was the utilization of the such as maleic anhydride is required in the mixture hemicellulose of bagasse as an internal plasticizer [36]. [40,41]. By adding it in a smaller amount, it is possible The results show how lignocellulosic agricultural wastes to improve the mechanical properties, the resistance can be utilized to produce high value plastics [36]. to moisture and the viability due to the increased Wu et al. [37] studied a facile way to prepare interactions within their compounds [40,41]. highly transparent antimicrobial films by grafting chito- Many reports have been published about bio- san as copolymer onto the oxidized cellulose matrix, medical applications of pure SP or blending with other that had a high transparent property of above 80% polymers or macromolecules [42-49]. SP also is used transmittance, excellent antioxidant activity and anti- in infant formulas and in baked, meat, and dairy pro- microbial properties against Escherichia coli and Sta- ducts [50]. The use of SP as a film-forming agent can phylococcus aureus [37]. Using a sausage model, it add value to soybeans by creating new channels for was shown that the composites exhibited better per- marketing SP [51]. formance than traditional polyethylene packaging Tulamandi et al. [52] found that with the addition material and demonstrated good potential as food of gelatin to papaya puree, the films have shown sig- packaging materials [37]. nificant increase in color properties, tensile strength Empty fruit bunch (EFB) is available abundantly and seal strength [52].Whereas, with the addition of in Indonesia as a side product of CPO production defatted SP along with gelatin to the papaya puree, [38]. Potential cellulose amount in EFB is 11.50 mil- the films have shown significant increase in elong- lion t [38]. For that reason, Isroi et al. [38] investigated ation, water permeability, water contact angle and bioplastic production based on cellulose from EFB. decrease in water solubility (Table 4). Cellulose was isolated by sodium hydroxide methods Non-biodegradable polyvinyl alcohol (PVA) deri- and bleached using sodium hypochlorite. Purity of ved from petroleum is the primary sizing agent due to obtained cellulose was 97%. For bioplastic synthesis, its excellent sizing performance on polyester-contain- glycerol as plasticizer and cassava starch as matrix ing yarns, especially in increasingly prevailing high- were used. In this research, bioplastic sheets were speed weaving [53]. However, due to poor biodeg- successfully produced by casting method. In future radability, PVA causes serious environmental pollut- prospects, bioplastic from EFB cellulose can be used ion, and thus, should be substituted with more envi- for plastic bags and food packaging [38]. ronmentally friendly polymers [53]. For this reason, Yang et al. [53] developed soy sizes from SP treated Soy proteins with glycerol and the biodegradable triol that was also Soy is a cheap and renewable source of bio- obtained from soy. It was observed that the soy sizes polymers, which has a great potential to replace pet- had good film properties, adhesion to polyester and rochemical polymers in many applications. Soy pro- abrasion resistance close to PVA. In summary, the tein (SP) is commercially available in three various fully biodegradable soy sizes have the potential to SP concentrations: soy flour (54%), soy concentrate substitute PVA for sustainable textile processing [53]. (65-72%) and soy isolate (90%) [39]. Also, Zhao et al. [54] grafted soy protein with acrylic Molecules of SP include 20 different amino acid and cast into biodegradable films as substitutes acids with strong inter- and intramolecular interact- of non-biodegradable PVA films. Acrylic acid grafting ions. Such interactions make SP unmeltable and provided SP films with biodegradability, flexibility, and therefore it is impossible to process SP in the form of adhesion to yarns substantially higher than PVA, a thermoplastic polymer, unless a sufficient amount of while water solubility and abrasion resistance was plasticizer is used [40,41]. The use of a significant similar to PVA, leading to high potential applications amount of plasticizer results in low mechanical proof the grafted SP in the fields of water soluble pack- perties of SP plastics. On the other hand, when the aging films and slashing to substitute PVA [54]. plasticizers migrate away from the SP plastics during Reddy et al. [55] developed completely biodeg- storage or service, the materials become very brittle radable SP composites reinforced with jute fibers with [40,41]. Moreover, the hydrophilicity of SP and the used water without any chemicals as plasticizer. It plasticizers leads to low moisture resistance of SP was found that the developed SP composites have

406 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020)

Table 4. SP as biodegradable polymers, addition of copolymers, additives/fibers and their potential applications

Biodegradable Material Co-polymer Additives/fibers Application Ref. polymer SP Defatted soy protein Gelatin, papaya puree Glycerol Packaging [52] (Shakthi soyas) SP Soy proteins – Glycerol Textile processing [53] SP Soy proteins – Acrylic acid Packaging [54] SP Soy proteins – Jute fiber Various applications [55] SP Soy protein isolate Cellulose nanocrystals Glycerol Active food packaging material [57] (CNCs) and Cedrus deodara pine needle extract SP Soy protein isolate – Virgin coconut oil, glycerol Industrial purposes [58] SP Soy protein isolate – Zinc sulphate monohydrate, Horticultural crop applications [56] glycerol SP Soy protein isolate – Nanosized bioactive glass, Wound-dressing applications [59] glycerol

excellent flexural strength, tensile strength and tensile as a matrix for wound-dressing applications. The modulus, much higher than polypropylene (PP) com- effect of the addition of bioactive glass nanoparticles posites, even at high relative humidity (90%) [55]. on blood clotting was assessed [59]. The composite Soy protein isolate (SPI) seems to be an ade- films could meet the essential requirements for an quate resource for the manufacture of natural-based appropriate wound dressing with additional favorable superabsorbent materials due to its hydrophilic char- properties such as hemostatic capability, mechanical acter and excellent processability when combined properties and significant cell cytocompatibility [59]. with a plasticizer [56]. Yu et al. [57] studied SP-based BIODEGRADABLE POLYMERS OBTAINED VIA films incorporated with CNCs and Cedrus deodara MICROBIAL PRODUCTION pine needle extract (PNE) as potential food packaging material. When a high content of PNE was incorpor- Polyhydroxyalkanoates (PHAs) ated in the films, the water vapor permeability was PHAs are classified as natural aliphatic biopoly- decreased due to the reduction of hydrophilic dom- esters, synthesized by many different bacteria as ains in the film matrix. Moreover, the PNE-added films intracellular carbon and energy storage materials. contained phenolic compounds and displayed strong PHAs have the ability to combine more than 150 antioxidant activities [57]. Carpiné et al. [58] pre- monomers to produce materials with very different sented an alternative by incorporating virgin coconut properties and functionalities [60,61]. Mechanical and oil (VCO, hydrophobic component with good oxidative biological compatibility can be altered by mixing, stability) into SPI films. In addition, as natural surfact- altering the surface, or by combining PHA with other ant, the Yucca schidigera (YS) saponin was applied. polymers, enzymes or inorganic materials, allowing Morphological analyses indicated that the incorpor- them a wider spectrum of use [60]. The main areas of ation of VCO or yucca extract caused notable PHA use include packaging (containers and films), changes in the structure of the SPI films. Therefore, coatings, pharmaceutical and medical applications this data should help to better understand the role of (wound dressings, medical devices, orthopedic pins, lipids and surfactants in protein-based films for future stents, nerve guides and bone marrow scaffolds). industrial purposes [58]. Jiménez-Rosado et al. [56] Besides, PHAs are used as hardeners in cosmetic studied the SP-based bioplastic matrices which were products, hygiene products, toners and adhesives, loaded with zinc sulphate monohydrate for potential electronic issues and golf balls [62]. PHA plastics are applications in horticulture. It was confirmed that sig- in contrast to other types of bioplastics (e.g., PLA), nificant amounts of an essential micronutrient for a UV stable, can withstand temperatures up to 180 °C plant (Zn) can be incorporated into bioplastic mat- and are poorly water-permeable [60]. rices, modifying water absorption, mechanical and Some specific forms of PHA are not present microstructural properties [56]. only in microorganisms, but are also found in plants, For medical purposes, Tansaz et al. [59] inves- animals, and even in humans as a component of tigated the fabrication of SP isolate/nanoscale bioact- membrane tissues in the form of polyhydroxybutyrate ive glass composite films by solvent casting method

407 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020)

(PHB), which is the most representative member of (Table 5) [66]. Wound management, coronary angio- this family [63]. PHB is a rare example of a hydro- plasty, nerve regeneration, bone tissue engineering, phobic polymer that is biocompatible and biodegrad- cardiac tissue engineering and drug delivery are able. It has a high melting point (170-180 °C) and a some examples of biomedical applications where crystalline degree close to 80%. PHB is rigid and PHA-based materials have been explored [67]. PHB brittle, with the degree of fragility being dependent on has been found to have low toxicity, since it degrades recrystallization, the glass transition temperature, and in vivo to d-3-hydroxybutyric acid, a normal cons- the microstructure of the polymer. The longer the time tituent of human blood. Applications of these poly- of its storage, the more fragile it becomes. Thermal mers previously tested or in phase of development instability during continuous polymer processing at include controlled drug release, artificial skin, and high temperatures and the tendency to crack and heart valves, as well as in industrial applications such fracture limit its wider use. The combination of high as paramedical disposables [68,69]. Subsequent crystallinity and low density of nucleation sites is ref- applications in the packaging and cosmetic industries lected in the formation of large spherulitic crystals that included the production of a wide variety of bottles, influence the spread of cracks and the fragility of the tubs, bags and wrappings. material. In order to improve mechanical and thermal Progress in tissue engineering is dependent on properties, different nucleation agents are added. the availability of suitable biomaterials. To overcome Among the common procedures for the modification the brittleness of poly(3-hydroxybutyrate) P(3HB), and of the original properties of PHB, it is certainly the widen its biomedical applications, plasticizing of mixing of the basic polymer component with other P(3HB) with oligomeric substances of related struc- polymers or with additives that is the most common. ture was investigated by Lukasiewicz et al. [66]. A PHB can be processed using conventional techno- biosynthesized medium-chain-length polyhydroxyalk- logical processes for the processing of plastics, which anoate (mcl-PHA) copolymer, the plasticizer precur- makes its use widely available [64,65]. sor, was obtained using vegetable waste frying oil as PHAs have emerged as highly promising bio- the sole carbon source [66]. Addition of oligomeric materials both for bulk and biomedical applications mcl-PHA to P(3HB) resulted in softer and more flex-

Table 5. PHA as biodegradable polymers, addition of copolymers, additives/fibers and their potential applications

Biodegradable Material Co-polymer Additives–fibers Application Ref. polymer PHA Bacillus subtilis OK2 and P(3HB) Acetic acid Soft tissue engineering [66] Pseudomonas mendocina applications CH50 PHA–PHA-g-MA–TPF PHA – Treated palm fiber 3D printing filaments [70] Maleic anhydride PHB PHB Starch – A coating material on [71] paper or cardboard used for food packaging P3HB Commercial P3HB – Maleinized linseed oil (MLO) Packaging [72] and an epoxidized fatty acid ester (EFAE) PHA Gordonia polyisoprenivorans – Poly(cis-1,4-isoprene) Recycling method for [73] VH2, Ralstonia eutropha, rubber waste Pseudomonas aeruginosa PHA Spirulina sp. LEB-18 – Sodium hypochlorite, Industrial [74] Microalgae methanol PHA Commercial PHA NCC, chitosan Tween 80 Industrial wastewater [75] PHA Peanut oil and Cupriavidus – Sodium hypochlorite Food packaging, [76] necator H16 solution biotechnology industry PHA Bacterially synthesized – GO, LAQ Food packaging [77] applications PHB PHB industrial Coconut fiber in – Packaging [78] nature and coconut fiber treated

408 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020) ible materials based entirely on PHAs. The proposed methods significantly influenced the molecular mass, approach for P(3HB) plasticization has the potential degree of crystallinity, and monomeric composition of for the generation of more pliable biomaterials based PHAs, showing that the extraction method is critical in on P(3HB) which can find application in unique soft obtaining polymers with desired characteristics for tissue engineering applications where a balance industrial applications [74]. The PHAs extracted from between stiffness, tensile strength and ductility is the Spirulina sp. were composed largely of 11-hydro- required [66]. xyhexadecanoate monomers and hydroxytetradec- The biodegradability, mechanical properties and anoate, which is a scientific novelty because these fabrication of three-dimensional (3D) printing fila- building blocks are constituents of completely new ments of composite materials made from maleic polymers [74]. Further, in the study of Soon et al. [75] anhydride-grafted polyhydroxyalkanoate (PHA-g-MA) the nanocellulose and chitosan were incorporated into and coupling agent-treated palm fiber (TPF) were the PHA composite with the use of the Pickering studied by Wu et al. [70]. TPF was successfully com- emulsion-electrospinning method and the material bined with PHA-g-MA by an esterification procedure was then tested as adsorbent for the removal of dye and 3D printing filaments were fabricated from the from industrial wastewater. The results showed that composites [70]. The strong interaction of the PF with dye (Congo red) removal percentage in the case of the PHA matrix in the PHA-g-MA/TPF composites led nanocellulose addition was much better (30.9%) than to improved tensile properties and had lower water in the case of chitosan addition (10.5%) [75]. In the absorption than the corresponding PHA/PF mem- work of Perez-Arauz et al. [76], peanut oil as a carbon branes, and the aqueous release solutions had good source for production of PHA films was used for the cell viability [70]. When incubated in soil, the biodeg- first time. The novel PHA films showed positive mech- radation rate of PHA-g-MA/TPF was lower than that of anical, physical and barrier properties so the material PHA/ PF, while still higher than that of pure PHA. Due could be suitable for food packaging applications and to their low cost and excellent characteristics, the in the biotechnology industry [76]. Xu et al. [77] prepared composites can be used as the biodegrad- created the well-designed multifunctional and robust able material of 3D printing filaments [70]. In order to PHA/GO-g-LAQ (long alkyl chain quaternary salt improve properties and cost reduction, Godbole et al. functionalized graphene oxide) nanocomposites with [71] studied the compatibility of PHB with starch. The superior gas barrier, heat resistant and inherent anti- results revealed that blend films had a single glass bacterial performances, which could be ideal for eco- transition temperature for all the proportions of PHB friendly food packaging [77]. and starch tested [71]. The nature of all combinations The influence of the coconut fiber used as reinf- was found to be crystalline. These blend films might orcement in PHB composites was studied by da Silva also be used as a coating material on paper or card- Moura et al. [78]. The microstructure showed a good board used for food packaging [71]. Garcia-Garcia et interfacial adhesion between the PHB and coconut al. [72] explored the potential of vegetable oil‐derived fiber [78]. plasticizers (maleinized linseed oil) and an epoxidized fatty acid ester as environmentally friendly solutions BIODEGRADBLE POLYMERS CHEMICALLY for P3HB industrial formulations with improved tough- SYNTHESIZED USING MONOMERS OBTAINED ness [72]. The optimum balance between ductile pro- FROM AGRO-RESOURCES perties is achieved with low plasticizer content for both plasticizer types [72]. Polylactic acid (PLA) Andler et al. [73] studied the biosynthesis of PLA is a synthetic biodegradable polyester with PHA utilizing poly(cis-1,4-isoprene) as the sole car- a monomer, lactic acid (LA), derived from natural bon source in order to find recycling methods for rub- resources. The most used raw material is maize or ber waste materials and alternative carbon sources other industrial plants with a high starch content. In for the synthesis of PHA [73]. Costa et al. [74] demon- the fermentation processes of hydrocarbons, Lacto- strated that the methods used for extracting PHAs bacillus lactic acid bacteria or Rhizopus oryzae fungi from the cyanobacterium Spirulina sp. LEB-18 show are used. The fermentation process requires the pre- different efficiencies in the extraction of the polymers. sence of bacterial strains and sufficient amounts of The use of sodium hypochlorite in the initial extraction carbon (glucose, sucrose or lactose), nitrogen (yeast stage increases polymer accumulation, while the use extract, peptides), and mineral elements for their of methanol at the end of the process is important for active functioning. The final product of bacterial syn- obtaining higher purity PHAs [74]. The extraction thesis is an optical isomer of lactic acid L- (+), which

409 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020) can be converted to polymeric form of poly-(L-lactic able polymeric blends based on poly(3-hydroxybut- acid) (PLLA) of low molecular weight in subsequent yrate) (PHB) and PLA by melt mixing. They con- polycondensation reactions. The properties of the cluded that plasticized PHB/PLA (30/70) and PHB/ PLA depend on its molecular structure, crystallinity, /PLA (40/60) blends, may be considered as sus- the size of the spherulitic crystals, morphology, and tainable alternatives to current non-natural and non- the level of arrangement of the polymer chains. The biodegradable materials for food packaging films, stereochemical structure of PLA determines its crys- considering their flexibility, transparency, possibilities talline arrangement. Depending on the selected syn- for processing at the industrial level and compo- thesis, polymerization process, the final form of the stability in terms of their final disposition [84]. By PLA can be completely amorphous (with 50-93% con- embedding nanochitosan in a PLA matrix using poly- tent of L-lactic acid) or semicrystalline (with >93% ethylene glycol (PEG) as a cross linking agent and content of L-lactic acid) [79,80]. The glass transition PVA as plasticizer, the composite film was developed temperature of the PLA homopolymer is 55 °C and by Fathima et al. [85]. The interactions between the the melting temperature is about 175 °C. Under con- PLA/NCS and PEG had a significant effect on the ten- trolled composting conditions, the material is com- sile strength and the heat-sealing properties. Anti- pletely biodegradable and can be processed using microbial properties of PLA/NCS films have been con- standard plastic processing techniques. To improve firmed against aerobic microorganisms and, con- process properties, lactide is often copolymerized sequently, the PLA/NCS composite films can be used with an enantiomer of the opposite configuration. This for packaging of fresh prawn to increase their shelf represents a copolymerized form of PLA, referred to life [85]. Arrieta et al. [86] prepared flexible electro- poly-(DL-lactide) (PDLLA) [4,81]. spun PLA-PHB biocomposites with chitosan or cat- The use of PLA polymers has great potential for echin as active film materials. Due to well-known anti- automotive and packaging industries and medicine. It oxidant activity of both loaded components and good can be used in the production of flexible and water- qualities of PLA-PHB biocomposites, these are pro- proof films, bottles, labels and cups, bakery pack- mising materials for biodegradable film applications aging and disposable dishes. In biomedicine, it is (agricultural mulch films, films for food packaging) used as a binding material in tissue engineering [81]. [86]. In a brand-new in vitro study, Abasian et al. [87]

However, its poor thermal stability, rigidity, intrinsic have successfully incorporated the NaX/Fe3O4 nano- brittleness, low crystallization rate and high cost limit particles and doxorubicin (DOX) into the PLA/chitosan its large-scale applications [82]. The introduction of nanofibers. These prepared nanofibers show excel- boron nitride (BN) as a reinforcement inorganic mat- lent potential in local chemotherapy of carcinoma erial in PLA may signiﬁcantly improve their properties, tumors. Namely, in vitro studies have shown that the making them more suitable for packaging of electro- maximum killing percentage of human carcinoma nic products. Namely, BN has superior thermal con- cells was 82% after 7 days [87]. ductivity and mechanical properties. The prepared Swaroop et al. [88] produced biofilms by reinfor- PLA/BN composite material has improved mechanical cement of magnesium oxide (MgO) nanoparticles in properties and it is suitable for packaging in industries PLA biopolymer using the solvent casting method. (Table 6) [83]. D'Amico et al. [84] prepared biodegrad- PLA/MgO films exhibited superior antibacterial effi-

Table 6. PLA as biodegradable polymers, addition of copolymers, additives/fibers and their potential applications

Biodegradable polymer Material Co-polymer Additives/fibers Application Ref. PLA, PLA–BN PLA resin – Boron nitride Packaging in industries [83] PLA Industrial PLA PHB Tributyrin (TBL) Food packaging [84] PLA PLA resin Nanochitosan Polyethylene glycol, polyvinyl Packing of fresh prawn [85] alcohol PLA Industrial PLA PHB Chitosan, catechin, acetyl-tri-n- Agricultural, food packaging [86] -butyl citrate (ATBC) PLA Industrial PLA chitosan DOX Local chemotherapy [87] Trifluoroacetic acid PLA Commercially – MgO nanoparticles Food packaging [88] available PLA PLA Industrial PLA ABS Cardanol 3D printing [89] MPEG-PLA Industrial MPEG-PLA – Verapamil and doxorubicin Ovarian cancer [90]

410 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020) cacy and they caused progressive damage and death modulus and rigidity of PCL limit its many applications of nearly 46% of E. coli bacterial culture after a 12 h [93]. High molecular weight PCL is not suitable as an treatment [88]. The produced films are transparent, oxygen barrier material. Despite many attractive pro- capable of screening UV radiations and exhibit sup- perties (slow degradation rate, high permeability, erior antibacterial efficacy what makes them an excel- easy availability and good mechanical properties), the lent food packaging material [88]. applications of PCL have also been limited by its rel- Moreover, the researchers investigated the atively high cost [94,95]. effect of cardanol-based compatibilizers, which can Nevertheless, PCL is very common in the pro- react by reactive extrusion with PLA and ABS (acrylo- duction of food packaging and, because of its excel- nitrile-butadiene-styrene). The modified chains of lent biocompatibility, it is explored as a structural mat- both PLA and ABS with cardanol derivatives during erial in tissue engineering (e.g., microstatic PCL reactive extrusion would intersperse between PLA foams are used for tissue regeneration and stem cell and ABS, reducing the interfacial tension and imp- transplantation) [96]. roving the interfacial adhesion, leading to the out- PCL has been repeatedly and successfully used standing compatibilization of the blend [89]. For medi- for the production of nanofibers [97]. Choi et al. [98] cal application, Zheng et al. [90] co-encapsulated were among the first that produced nanofibers from verapamil and DOX in methoxy poly(ethylene glycol)- PCL/collagen for cell regeneration of skeletal muscles -poly (L-lactic acid) (MPEG-PLA) nanoparticles to rev- [98] (Table 7). Zamani et al. [99] successfully incor- erse the drug resistance in ovarian cancer [90]. These porated metronidazole benzoate (MET) in PCL nano- MPEG-PLA nanoparticles reveal high biocompatibility fibers for periodontal diseases [99]. In the work of and superior safety. They also demonstrated the Potrč et al. [100], PCL nanofibers were incorporated enhanced anticancer effects of verapamil and DOX in poorly water-soluble agents such as resveratrol, loaded nanoparticles, with a sufficient evaluation of carvedilol and ibuprofen, thereby enabling prolonged safety in vivo [90]. release from the delivery form [100]. Zupančič et al. [101] produced PCL nanofibers with built resveratrol BIODEGRADABLE POLYMERS DERIVED FROM in order to treat periodontal disease [101]. Sessini et PETROLEUM al. [102] studied the influence of the addition of polyester-grafted-cellulose nanocrystals on the shape- Polycaprolactone (PCL) memory properties of biodegradable PLA/PCL nano- PCL is a semicrystalline aliphatic polyester, composites [102]. Joo et al. [103] obtained biocom- which is well known for its slow biodegradation rate patible biodegradable PU/PCL blends by blending. and high biocompatibility. PCL is produced by ring- Such polymer blends with suitable mechanical pro- opening polymerization of ε-caprolactone in the pre- perties and shape-memory behavior can be used in sence of various anionic and cationic catalysts [91]. surgical sutures or other medical devices [103]. Xu et Due to its slow biodegradation, it is suitable for pro- al. [104] studied the feasibility of utilizing PCL as a duction of prolonged-release delivery systems. The biocompatible and biodegradable binding agent to decommissioning takes place over two phases. First, fabricate electrospun three-dimensional (3D) scaf- the hydrolytic cleavage of the ester bond is carried folds. The obtained 3D scaffolds are soft while elastic, out, and in the second phase, intracellular degrad- and they possess interconnected and hierarchically ation occurs into the non-toxic metabolites that are structured pores with sizes ranging from sub-microns excreted directly from the body or after the metabolic to hundreds of microns; hence, they are morpho- change in the Krebs cycle [91]. Unlike most other bio- logically similar to natural extracellular matrices degradable polyesters, PCL does not locally lower the (ECMs), thus well suited for cell functions and tissue pH in the area of its degradation [91]. formation [104]. These various thermoplastic poly- The main problem is its hydrophobicity, which mers could be fabricated into 3D nanofibrous scaf- can be solved by combining it with other natural or folds/structures by first making blend nanofibers with synthetic polymers, thus improving its mechanical PCL, followed by processing via the thermally ind- properties and biocompatibility [92]. Besides, PCL uced (nanofiber) self-agglomeration (TISA) method, ° has low glass transition temperature (-60 C) and a and finally being thermally stabilized [104]. low melting point (60 °C), but presents an ideal biodegradable matrix for natural antimicrobial agents as Poly(butylene succinate) (PBS) it can be processed at low temperature. Moreover, the PBS is synthesized by the polycondensation toughness of PCL is superior to that of most bio- reaction between succinic acid and butanediol. The degradable polymers. On the other hand, lower

411 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020)

Table 7. PCL as biodegradable polymers, addition of copolymers, additives/fibers and their potential applications

Biodegradable Material Co-polymer Additives/fibers Application Ref. polymer PCL Industrial PLC Collagen / Cell regeneration of skeletal muscles [98] PCL Industrial PLC / Metronidazole benzoate (MET) Periodontal diseases [99] PCL Industrial PLC / Resveratrol, carvedilol and Different fields [100] ibuprofen PCL Industrial PLC / Resveratrol Periodontal disease [101] PCL Industrial PLC PLA CNC Biomedicine, food packaging [102] PCL Industrial PLC PU / Medical devices [103] PCL Industrial PLC Cellulose acetate (CA) / Tissue engineering [104]

reaction takes place in two steps. In the first part of for the development of various application of PBS sintering, the esterification takes place between the composites in different fields [109] (Table 8). Shi et al. diacids and the diols. Then, polycondensation takes [110] studied the preparation, characterization, and place under high temperature conditions to form high biodegradation of PBS/cellulose triacetate (CT) blends. molecular weight PBS [105]. PBS and CT were blended with different ratios using PBS has a relatively low biodegradation rate chloroform as a solvent. This finding indicated good due to its high crystallization rate and high crystal- hydrophilicity [110]. The weight loss became close to linity. It has good thermal stability and mechanical 90% after 16 h of degradation for PBS/CT10 [110]. properties. Furthermore, PBS naturally decomposes Therefore, the blend economized the PBS material, to nontoxic and harmless products such as water and was environment-friendly, and showed preferable

CO2 [106]. To promote the physical properties, to solid-state properties [110]. extend the application field, and to increase the bio- Sun et al. [111] investigated non-isothermal degradability of PBS, numerous approaches have crystallization of biopolyesters of PBS formed via in been used, such as physical blending, copolymer- situ polymerization in the presence of poly(vinyl but- ization, or formation of composites. Owing to the yral) (PVB). The effects of the PVB content and mole- excellent processability of PBS, it can be processed cular weights on non-isothermal crystallization and using conventional polyolefin equipment in the range mechanical properties were examined in detail [111]. of 160-200 °C. Injection, extrusion or blow molding Adding PVB greatly reduced spherulitic sizes and are suitable for the processing of PBS [107]. increased the peak temperature of crystallization The applications of PBS are still increasing in [111]. An optimal content was observed for a minimal many areas. In the field of packaging, PBS can be size of spherulites, and the PVB with a low molecular processed into foils, bags, or food and cosmetic pack- weight led to a smaller spherulitic size than its count- aging. In agriculture, PBS is used to produce mul- erpart of a high molecular weight did [111]. A mole- ching foils or delayed-release materials for pesticides cular model was proposed to explain accelerated nuc- and fertilizers. PBS is also increasingly used in fishing leation and spherical growth from PVB’s structural materials, forestry, construction, or other areas where perspective. The nucleated PBS exhibited a consider- processing and recycling of materials is problematic able improvement in mechanical properties [111]. The [106]. In the medical field, PBS can be used as a design and synthesis of a new PBS ionomer con- biodegradable drug for encapsulation systems and is taining a novel bisfunctional phosphinate monomer, also investigated for implants [108]. potassium salt of 10H-phenoxaphosphine-2,8-dicar- Production of PBS is more expensive than of boxylic acid,10-hydroxy-,2,8-dihydroxyethyl ester,10- synthetic polymers, hence it can be mixed with cheap -oxide (DHPPO-K) was studied by Xu et al. [112]. natural fibers or fillings (wood). Jiang et al. [109] pre- PBS ionomers containing aromatic cationic groups pared a wood plastic composite, where were the were synthesized by condensation polymerization of

CaCO3 acted as a reinforcing component, and alu- succinic acid and 1,4-butanediol in the presence of minum hypophosphite, ammonium polyphosphate DHPPO-K [112]. The crystallization properties, rheo- and calcium hypophosphite as flame-retardant com- logical properties, dynamic mechanical properties, ponents were separately incorporated into PBS mat- oxygen permeability, thermal conductivity and phys- rix. The investigation of mechanical and thermal pro- ical properties were thoroughly investigated consider- perties of PBS composites, can present a foundation ing their structure-property relationships [112].

412 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020)

Table 8. Research PBS as biodegradable polymers, addition of copolymers, additives/fibers and their potential applications

Biodegradable Material Co-polymer Additives/fibers Application Ref. polymer

PBS Industrial PBS Decayed wood CaCO3 Different fields [109] AHP, APP and CaHP PBS Industrial PBS Cellulose triacetate – Different fields [110] PBS Synthesized PBS with the content Poly (vinyl butyral) Tetrabutyl titanate Biochemical engineering [111] of butyral group (PVB) PBS ionomer Synthesized PBS ionomer by – – Package material, such as [112] condensation polymerization of film, foams succinic acid, 1,4-butanediol and DHPPO-K BioPBS BioPBSTM PLA – Different industries and [114] applications BioPBS BioPBSTM – Apple pomace Packaging [115] BioPBS BioPBSTM – Grape pomace Food packaging [116]

BioPBS BioPBSTM Cellulose CDI, PHA, EST, APTMS, Construction applications [117] (FZ71PB) MAH, PMDI TM BioPBS BioPBS – CHCl3 Biomedical or tissue [118] (FZ91PM) DMSO engineering application DMF PBSA BioPBS™ Collagen – Agriculture applications [119] (FD92PM)

PTT Public Company Limited (PTT) and Mitsu- extrusion-injection. Both byproducts can be utilized as bishi Chemical Corporation (MCC) produce not only fillers and used to decrease the amount of polymer in petroleum-based PBS but also bio-based PBS a bio-based composite blend. The results indicated (BioPBSTM). Both plastics have the same properties that high levels of GP and AP can be successfully and processability, the only difference is their source incorporated into a polymer matrix and improve its material. BioPBSTM is produced from succinic acid mechanical and thermo-mechanical properties. So, derived from natural resources (sugarcane, cassava prepared BioPBS-based composites could be used and corn) and 1,4-butanediol. BioPBSTM products for the production of disposable or single-use and have a high heat deflection temperature point and sustainable biodegradable applications [115,116]. high seal strength in heat sealing. The polymer is Platnieks et al. [117] prepared BioPBS/micro cellu- highly compatible with PLA. Bio-based PBSA con- lose (MCC) woody-like composites, by using a melt tains a higher percentage of amorphous structure, is blending of 70 wt.% of MCC processed from bleached stickier and has higher adhesion strength than stan- softwood. Besides, in order to enhance dispersion dard PBS. Properties of BioPBSTM are similar to those and compatibility, MCC was modified by using carbo- of LDPE. Possible commercial applications for diimide (CDI), polyhydroxy amides (PHA), alkyl ester BioPBSTM are paper lamination, multi-layer bags, and (EST), (3-aminopropyl) trimethoxysilane (APTMS), films (plastic bags, mulch films) [113]. maleic acid anhydride (MAH), and polymeric diphe- Qahtani et al. [114] used BioPBS and PLA for nylmethane diisocyanate (PMDI). The chemically making more sustainable 3D-printed PLA/BioPBS modified MCC composites lead to very enhanced blends. It was found that, with the blending ratio 90/10 mechanical, thermo-mechanical and thermal proper- of PLA and BioPBS, higher tensile and impact ties of the PBS/MCC composites. This composite strength were achieved than by the neat PLA, while material can be suitable for various construction further addition of BioPBS lead to improved tough- applications, including profiles, decks and housing ness of the 3D blends and increased the viscosity appliances [117]. Unlike the others, Cooper et al. [114]. Furthermore, BioPBS was combined by varying [118] created highly porous electrospun fibers, where contents of apple pomace (AP) [115] and grape pom- BioPBS were dissolved in three different solutions: ace (GP) [116], byproducts from wine and apple juice chloroform, chloroform/N,N-dimethylformamide (DMF), industries, and biocomposites were fabricated by melt or chloroform/ /dimethyl sulfoxide (DMSO). Prepared

413 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020) in this way, the electrospun BioPBS fibers have high through reactive extrusion with PBAT and the antimic- porosity at the micro- and nanoscale and conse- robial thermoplastic starch (ATPS) in the presence of quently have suitable mechanical properties for pot- the coupling agent, 2,2′-(1,3-phenylene)-bis(2-oxazo- ential in wound healing and soft tissue engineering line) (PBO). The antimicrobial PBAT films with excel- [118]. In the work of Seggiani et al. [119], bio-based lent and rapid antimicrobial activity were obtained by PBSA and raw hydrolysed collagen (HC), by-products using a blown film extrusion system. PBAT films have of the tannery industry, were used to produce ther- a great potential to be used in the prevention of dis- moplastic blends. Different PBSA/HC blends with ease infections in hospitals and in fields requiring 5-20 wt.% of HC were processed by extrusion, film rapid and highly efficient antimicrobial activity [123]. blowing and injection molding. The PBSA/HC-based blends with higher content of HC (10-20 wt.%) showed BIODEGRADABLE ALTERNATIVES TO good mechanical properties and present promising CONVENTIONAL PLASTICS candidates to produce biodegradable products for agricultural applications [119]. Many bioplastics have mechanical properties equivalent to those of their conventional counterparts Poly(butylene-adipate-co-terephthalate) (PBAT) (e.g., PP, PS and PE) and can be processed using PBAT is an aliphatic aromatic copolyester that technologies widely used in the polymer industry can be fully degraded in a few weeks [120]. It has (e.g., compounding, film processing and molding) good mechanical properties, such as adequate tough- [10]. Their use has been found in many short service ness and tear resistance, but it has a low barrier char- life applications where biodegradability is a key adv- acter to water vapor, oxygen and carbon dioxide [120]. antageous feature, including consumer packaging PBAT can be used to produce fibers, textile, (e.g., trays, pots, films and bottles in food packaging), films and packaging materials. It has also been used convenience food disposables (e.g., cutlery/table- for the preparation of antimicrobial films that could be ware), bags (shopping, garden or domestic waste), used in food packaging to inhibit bacterial growth, agriculture mulch films, personal care disposals (e.g., helping to preserve food products safely [121]. In nappies) and even golf tees [10]. Bioplastic polymers such films, PBAT serves as the bulk of the film with have also been used in more durable applications the antimicrobial agent being incorporated during pro- such as in textiles, consumer goods, automotive cessing. Sangroniz et al. [120] studied PBAT/PHAE parts, and building and construction where the focus blends for packaging applications. It was found that is on the use of renewable (bio)resources and any immiscible blends were obtained with very good com- inherent biodegradability properties need to be sup- patibility. The mechanical performance and the per- pressed or controlled by careful design [124]. meability to water vapor, limonene and carbon dioxide PLA is eco-friendly, it is 100% bio-based and were measured to assess the suitability of these biodegradable, but only under certain conditions, and blends for packaging applications [120] (Table 9). it is industrially compostable [125]. PLA is also a bio- Tavares et al. [122] studied the PBAT-kraft lignin (KL) compatible material and should not produce carcino- blend films, which were obtained by extrusion. Multi- genic or toxic effects in local tissue treatment. Com- layer composites were obtained by lamination, in pared to the other biopolymers such as PCL, it has a which PBAT-KL blend films were bonded to a poly- better thermal processibility [125]. PLA is also relat- ethylene layer using polyurethane adhesive [122]. ively hydrophobic and has a slow degradation rate. PBAT-KL could be a synergetic solution for food Table 10 shows that PLA has similar mechanical pro- packaging that can integrate the reuse of industrial perties to PET polymer. Elastic modules and tensile waste, mechanical properties, flexibility, peeling res- strength are comparable to PET [125]. istance, and the well-known biodegradability of PBAT High molecular weight PCL polymer has mech- [122]. anical properties similar to PE (HDPE and LDPE), In the work of Wei et al. [123], the non-leaching processing a tensile strength of 12-30 MPa and a antimicrobial biodegradable PBAT was prepared break extension of 400-900% [3,105]. Disadvantages

Table 9. Research of PBAT as biodegradable polymers, addition of copolymers, additives/fibers and their potential applications

Biodegradable polymer Material Co-polymer Additives/fibers Application Ref. PBAT Industrial PBAT PHAE – Packaging applications [120] PBAT Industrial PBAT Kraft lignin – Food packaging [122] PBAT Industrial PBAT ATPS, PHGH PBO Prevention of disease infection in the hospital [123]

414 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020)

Table 10. Comparison of the properties of biodegradable and synthetic polymers

Glass transition (°C) Melting temperature (°C) Tensile strength Strain at break Flexural modulus Polymer Tg Tm MPa % MPa PHB [105] 15 175 40 6 3500-4000 PLA [89,125] 55 165 59 7 5000 BioPBS [113] -22 115 30 170 630 PBS [105,109] -32 114 57 700 656 LDPE [105] -120 110 35 176 400 HDPE [105] -120 129 29 1070 650 PP [105] 5 163 44 1370 800 PBAT (Ecoflex) [105] - 110-115 32-36 580-800 - PVC [130] 82 100-260 20-76 150-400 1300-4690 PET [131,132] 73,5 250 80 36 2410 PCL [3,105] -60 60-65 23 400-900 - PS [133] 100 240 30-60 1.5-2.3 3200

of PCL are very low glass transition temperature (-60 CONCLUSION °C) and low melting point (60-65 °C), which prohibits its application at elevated temperatures [3,105]. More than its origin, the chemical structure of Therefore, PCL is often blended with other polymers, the biopolymer is the key factor that determines its such as PP, PC, polyethylene oxide (PEO) and starch, biodegradability. Indeed, the future outlook for dev- to produce composites with desired properties [89]. elopment in the field of biopolymer materials is pro- Due to the plastic-like properties and biodegrad- mising; the future of each polymer is dependent on its ability of PHAs, these polymers present a potential competitiveness but also society’s ability to pay for it. replacement for non-degradable PE, PP [126]. PHB Biodegradable plastics made from renewable res- homopolymers possess several physical properties, ources can retain all the benefits of petroleum-based e.g., glass transition temperature (15 °C), crystallinity plastic without the negative environmental impact. (80%), and tensile strength (40 MPa), similar to those The discovery and implementation of plastic made of PP. However, PHB is significantly more brittle than from natural resources is an enormous leap into the PP (strain at break 6 vs. 400%) [123]. PBS is bio- future for different branches of industry. In order to degradable and currently mostly fossil-based but could successfully replace biodegradable polymers with in theory be 100% bio-based. Its mechanical pro- current plastics, cooperation is needed not only with perties resemble those of PE and PP, while its melt- industry and academics, but also with various dis- ing point is between 90-120 °C, and glass transition ciplines such as chemistry, engineering, material sci- temperature around -45 to -10 °C [3,127,128]. PBS ence, biogeochemistry and climate science. This will has good processability, better than that of PLA [105]. require plenty of time and a key multidisciplinary dev- Furthermore, the properties of BioPBSTM are similar to elopment will be needed. those of LDPE [113]. PBAT is also currently mostly Acknowledgments fossil-based. Its commercial name is Ecoflex®, pre- ® The authors would like to acknowledge the pared by BASF (Germany). Ecoflex F Blend C1200 Slovenian Research Agency (ARRS) for ﬁnancing the has properties similar to PE-LD because of its high research within Programme P2-0046. molecular weight and its long-chain branched molecular structure [105]. REFERENCES Cellulose and starch are alternatives for PP and PS. Cellulose acetate (CA), which is the most import- [1] M. Raja, A. Murali, J. Mater. Sci. Eng., B 1 (2011) 86–89 ant cellulose derivative, has tensile strength compar- [2] T. Helmer Pedersen, F. Conti, Waste Manage. 68 (2017) able to PS [129]. Starch blends are completely bio- 24–31 degradable and 25-100% bio-based, when starch is [3] I. Vroman, L. Tighzert, Materials 2 (2009) 307–44 added to one or several biodegradable polymers [4] M.B. Kannan, in Surface Modification of Magnesium and [129]. its Alloys for Biomedical Applications, T.S.N.S. Naraya- nan, I.S. Park, M.H. Lee (Eds.), Woodhead Publishing, Sawston, 2015, p. 355

415 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020)

[5] L. Yu, K. Dean, L. Li, Prog. Polym. Sci. 31 (2006) 576– [28] N.S.M. Makhtar, M.F.M. Rais, M.N.M. Rodhi, N. Bujang, –602 M. Musa, K.H.K. Hamid, Procedia Eng. 68 (2013) 385– [6] S. Lambert, M. Wagner, Chem. Soc. Rev. 46 (2017) –391 6855–6871 [29] Z. Czech, K. Wilpiszewska, B. Tyliszczak, X. Jiang, Y. [7] J.P. Greene, in Sustainable Plastics: Environmental Bai, L. Shao, Int. J. Adhes. Adhes. 44 (2013) 195–199 Assessments of Biobased, Biodegradable, and Recycled [30] H.P.S. Abdul Khalil, A.H. Bhat, A.F. Ireana Yusra, Plastics, John Wiley & Sons, New York, 2014, p. 336 Carbohydr. Polym. 87 (2012) 963–979 [8] K. Sudesh, T. Iwata, Clean – Soil Air Water 36 (2008) [31] M.P. Arrieta, E. Fortunati, F. Dominici, J. López, J.M. 433–442 Kenny, Carbohydr. Polym. 121 (2015) 265–275 [9] H. Rosenheim, I. De, S. Hyvedemm, Bioplastics market [32] L. Shen, E. Worrell, M. Patel, Biofuels Bioprod. Bioref- data 2018, https://www.european-bioplastics.org/wp-con- ining 4 (2010) 25–40 tent/uploads/2016/02/Report_Bioplastics-Market- [33] K. Missoum, P. Sadocco, J. Causio, M.N. Belgacem, J. Data_2018.pdf (accessed 9 June 2019) Bras, Mater. Sci. Eng., C 45 (2014) 477–483 [10] H. Rosenheim, I. De, S. Hyvedemm, Bioplastics market [34] J. Sirviö, M. Visanko, J. Ukkola, H. Liimatainen, Ind. data 2017, https://docs.european-bioplastics.org/public- Crops Prod. 122 (2018) 513–521 ations/market_data/2017/Report_Bioplastics_Market_Dat [35] F.-X. Joly, M. Coulis, Waste Manage. 72 (2018) 349–353 a_2017.pdf (accessed 9 June 2019) [36] H.M. Shaikh, K.V. Pandare, G. Nair, A.J. Varma, [11] R. Smith, Biodegradable polymers for industrial applic- Carbohydr. Polym. 76 (2009) 23–29 ations, Woodhead Publishing, Sawston, 2005, p. 552 [37] Y. Wu, X. Luo, W. Li, R. Song, J. Li, Y. Li, B. Li, S. Liu, [12] M.A. Elsawy, K.-H. Kim, J.-W. Park, A. Deep, Renew. Food Chem. 197 (2016) 250–256 Sustain. Energy Rev. 79 (2017) 1346–1352 [38] A. Isroi, T. Cifriadi, N.A. Panji, K. Wibowo, K. Syamsu, [13] A.B. Balaji, H. Pakalapati, M. Khalid, R. Walvekar, H. IOP Conf. Ser.: Earth Environ. Sci. 65 (2017) 012011 Siddiqui, in Biodegradable and Biocompatible Polymer [39] J. Zhang, P. Mungara, J. Jane, Polymer 42 (2001) 2569– Composites: Processing, Properties and Applications, –2578 N.G. Shimpi Ed., Woodhead Publishing, Duxford, 2017, p. 3 [40] J. John, M. Bhattacharya, Polym. Int. 48 (1999) 1165– –1172 [14] P. Lescher, Starch/Staerke 61 (2009) 43–45 [41] W. Liu, A.K. Mohanty, P. Askeland, L.T. Drzal, M. Misra, [15] J. Rydz, M. Musioł, B. Zawidlak-Węgrzyńska, W. Sikor- Polymer 45 (2004) 7589–7596 ska, in Biopolymers for Food Design, A.M. Grumezescu, A.M. Holban (Eds.), Elservier, Amsterdam, 2018, p. 431 [42] S. Tansaz, A.R. Boccaccini, J. Biomed. Mater. Res., A 104 (2016) 553–569 [16] J. Jane, J. Macromol. Sci., A 32 (1995) 751–757 [43] S. Tansaz, L. Liverani, L. Vester, A.R. Boccaccini, Mater. [17] A.J. Gunorubon, Res. J. Eng. Appl. Sci. 1 (2012) 219–214 Lett., C 199 (2017) 143–146 [18] M. Carissimi, S. Flôres, R. Rech, Algal Res. 32 (2018) [44] M. Santin, C. Morris, G. Standen, L. Nicolais, L. Ambro- 201–209 sio, Biomacromolecules 8 (2007) 2706–2711 [19] R.V. Gadhave, A. Das, P.A. Mahanwar, P.T. Gadekar, [45] C. Morris, J. Thorpe, L. Ambrosio, M. Santin, J. Nutr. 136 Open J. Polym. Chem. 8 (2018) 21-33 (2006) 1166–1170 [20] K. Petersen, P. Nielsen, G. Bertelsen, M. Lawther, M. B. [46] A. Merolli, L. Nicolais, L. Ambrosio, M. Santin, Biomed. Olsen, N. Nilsson, G. Mortensen, Trends Food Sci. Mater. 5 (2010) 015008 Technol. 10 (1999) 52–68 [47] G.A. Silva, C.M. Vaz, O.P. Coutinho, A.M. Cunha, R.L. [21] Q. Xie, F. Li, J. Li, L. Wang, Y. Li, C. Zhang, J. Xu, S. Reis, J. Mater. Sci.: Mater. Med. 14 (2003) 1055–1066 Chen, Carbohydr. Polym. 189 (2018) 56–64 [48] Z. Peles, M. Zilberman, Acta Biomater. 8 (2012) 209–217 [22] C. Medina-Jaramillo, O. Ochoa-Yepes, C. Bernal, L. Famá, Carbohydr. Polym. 176 (2017) 187–94 [49] S. Tansaz, A.K. Durmann, R. Detsch, A.R. Boccaccini, J. Appl. Polym. Sci. 134 (2017) 1-9 [23] J.F. Mendes, R.T. Paschoalin, V.B. Carmona, A.R. Sena Neto, A.C.P. Marques, J.M. Marconcini, L.H.C. Mattoso, [50] S.A. Witherly, Am. J. Clin. Nutr. 51 (1990) 705–706 E.S. Medeiros, J.E. Oliveira, Carbohydr. Polym. 137 [51] N. Cao, Y. Fu, J. He, Food Hydrocolloids 21 (2007) 1153– (2016) 452–458 –1162 [24] S. Kumar, A. Mukherjee, J. Dutta, Trends Food Sci. [52] S. Tulamandi, V. Rangaraju, S. Rizvi, R. S. Singhal, S. Technol. 97 (2020) 196–209 Kr. Chattopadhyay, N. Chandra Saha, Food Packag. [25] H. Honarkar, M. Barikani, Monatsh. Chem. 140 (2009) 1- Shelf Life 10 (2016) 60–71 –18 [53] M. Yang, H. Xu, X. Hou, J. Zhang, Y. Yang, J. Environ. [26] T. Woggum, P. Sirivongpaisal, T. Wittaya, Food Manage. 188 (2017) 26–31 Hydrocolloids 50 (2015) 54–64 [54] Y. Zhao, H. Xu, B. Mu, L. Xu, Y. Yang, Polym. Degrad. [27] O.V. López, L.A. Castillo, M.A. García, M.A. Villar, S.E. Stab. 133 (2016) 75–84 Barbosa, Food Hydrocolloids 43 (2015) 18–24 [55] N. Reddy, Y. Yang, Ind. Crops Prod. 33 (2011) 35–41

416 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020)

[56] M. Jiménez-Rosado, V. Pérez-Puyana, F. Cordobés, A. [81] K. Madhavan Nampoothiri, N.R. Nair, R.P. John, Romero, A. Guerrero, Ind. Crops Prod. 121 (2018) 345– Bioresour. Technol. 101 (2010) 8493–8501 –351 [82] A.K. Abay, M.B. Gebeyehu, H.K. Lin, P.C. Lin, J.Y. Lee, [57] Z. Yu, L. Sun, W. Wang, W. Zeng, A. Mustapha, M. Lin, C.M. Wu, R.I. Murakami, T.C. Chiang, J. Polym. Res. 23 Ind. Crops Prod. 112 (2018) 412–419 (2016) 1-12 [58] D. Carpiné, J.L.A. Dagostin E.F. de Andrade, L.C. Bertan, [83] B. Bindhu, R. Renisha, L. Roberts, T.O. Varghese, M.R. Mafra, Ind. Crops Prod. C 83 (2016) 364–371 Polym. Test. 66 (2018) 172–177 [59] S. Tansaz, M. Schulte, U. Kneser, D. Mohn, W. Stark, [84] D.A. D’Amico, M.L. Iglesias Montes, L.B. Manfredi, V.P. J.A. Roether, I. Cicha, A.R. Boccaccini, Eur. Polym. J. Cyras, Polym. Test. 49 (2016) 22–28 106 (2018) 232–41 [85] P.E. Fathima, S.K. Panda, P.M. Ashraf, T.O. Varghese, J. [60] S. Chanprateep, J. Biosci. Bioeng. 110 (2010) 621–632 Bindu, Int. J. Biol. Macromol. 117 (2018) 1002–1010 [61] Z. Li, J. Yang, X.J. Loh, NPG Asia Mater. 8 (2016) [86] M.P. Arrieta, J. López, D. López, J.M. Kenny, L. Peponi, e265.R Polym. Degrad. Stab. 132 (2016) 145–156 [62] R. Tarrahi, Z. Fathi, M.Ö. Seydibeyoğlu, E. Doustkhah, A. [87] P. Abasian, M. Radmansouri, M. Habibi Jouybari, M.V. Khataee, Int. J. Biol. Macromol. 146 (2020) 596–619 Ghasemi, A. Mohammadi, M. Irani, F.S. Jazi, Int. J. Biol. [63] A. Shrivastav, H.Y. Kim, Y.R. Kim, BioMed Res. Int. 2013 Macromol. 121 (2019) 398–406 (2013) 1-12 [88] C. Swaroop, M. Shukla, Int. J. Biol. Macromol. 113 (2018) [64] S.Y. Lee, Biotechnol. Bioeng. 49 (1996) 1–14 729–736 [65] M. Avella, E. Martuscelli, M. Raimo, J. Mater. Sci. 35 [89] A. Rigoussen, J.M. Raquez, P. Dubois, P. Verge, Eur. (2000) 523–545 Polym. J. 114 (2019) 118-126 [66] B. Lukasiewicz, P. Basnett, R. Nigmatullin, R. Matharu, [90] W. Zheng, M. Li, Y. Lin, X. Zhan, Biomed. Pharmacother. J.C. Knowles, I. Roy, Acta Biomater. 71 (2018) 225–234 108 (2018) 565–573 [67] G.Q. Chen, Chem. Soc. Rev. 38 (2009) 2434–2446 [91] M.A. Woodruff, D.W. Hutmacher, Prog. Polym. Sci. 35 [68] Y. Doi, Y. Kanesawa, M. Kunioka, T. Saito, Macromole- (2010) 1217–1256 cules 23 (1990) 26–31 [92] M. Abedalwafa, F. Wang, L. Wang, C. Li, Rev. Adv. [69] R. Sodian, S.P. Hoerstrup, J.S. Sperling, D.P. Martin, S. Mater. Sci. 34 (2013) 123-140 Daebritz, J.E.J. Mayer, J.P. Vacanti, ASAIO J. 46 (2000) [93] S. Khalid, L. Yu, M. Feng, L. Meng, Y. Bai, A. Ali, H. Liu, 107-110 L. Chen, Food Packag. Shelf Life 18 (2018) 71–79 [70] C.S. Wu, H.-T. Liao, Y.-X. Cai, Polym. Degrad. Stab., C [94] K.K. Sadasivuni, J.J. Cabibihan, D. Ponnamma, M.A. 140 (2017) 55–63 AlMaadeed, J. Kim, in Biopolymer Composites in [71] S. Godbole, S. Gote, M. Latkar, T. Chakrabarti, Bio- Electronics, Elsevier, Amsterdam, 2016, p. 544 resour. Technol. 86 (2003) 33–37 [95] V. Sessini, I. Navarro-Baena, M.P. Arrieta, F. Dominici, D. [72] D. Garcia‐Garcia, O. Fenollar, V. Fombuena, J. López, L. Torre, J.M. Kenny, P. Dubois, J.-M. Raquez, L. Peponi, Polym. Degrad. Stab. 152 (2018) 126–138 Lopez‐Martinez, R. Balart, Macromol. Mater. Eng. 302 (2017) 1-12 [96] B. Dariš, Acta Medico-Biotech. 9 (2016) 42-48 [73] R. Andler, R. Vivod, A. Steinbüchel, J. Biosci. Bioeng. [97] Z. McEachin, K. Lozano, J. Appl. Polym. Sci. 126 (2012) 127 (2019) 360–365 473–479 [74] S.S. Costa, A.L. Miranda, D. de J. Assis, C.O. Souza, [98] J.S. Choi, S.J. Lee, G.J. Christ, A. Atala, J.J. Yoo, M.G. de Morais, J.A.V. Costa, J.I. Druzian, Algal Res. 33 Biomaterials 29 (2008) 2899–2906 (2018) 231–238 [99] M. Zamani, M. Morshed, J. Varshosaz, M. Jannesari, Eur. [75] C.Y. Soon, N.A. Rahman, Y.B. Tee, R.A. Talib, C.H. Tan, J. Pharm. Biopharm. 75 (2010) 179–185 K. Abdan, E.W.C. Chan, J. Mater. Res. Technol. 8 (2019) [100] T. Potrč, S. Baumgartner, R. Roškar, O. Planinšek, Z. 5091–5102 Lavrič, J. Kristl, P. Kocbek, Eur. J. Pharm. Sci. 75 (2015) [76] A.O. Pérez-Arauz, A.E. Aguilar-Rabiela, A. Vargas- 101–113 Torres, A.-I. Rodríguez-Hernández, N. Chavarría-Hernán- [101] Š. Zupančič, S. Baumgartner, Z. Lavrič, M. Petelin, J. dez, B. Vergara-Porras, M.R. López-Cuellar, Food Kristl, J. Drug Deliv. Sci. Technol. 30 (2015) 408–416 Packag. Shelf Life 20 (2019) 1-8 [102] V. Sessini, I. Navarro-Baena, M.P. Arrieta, F. Dominici D. [77] P. Xu, W. Yang, D. Niu, M. Yu, M. Du, W. Dong, M. Chen, López, L. Torre, J.M. Kenny, P. Dubois, J.M. Raquez, L. P. Jan Lemstra, P. Ma, Chem. Eng. J. 382 (2020) 1-11 Peponi, Polym. Degrad. Stabil. 152 (2018) 126-138 [78] A. da Silva Moura, R. Demori, R.M. Leão, C.L. Crescente [103] Y.S. Joo, J.R. Cha, M.S. Gong, Mater. Sci. Eng., C 91 Frankenberg, R.M. Campomanes Santana, Mater. Today (2018) 426–435 Commun. 18 (2019) 191–198 [104] T. Xu, Z. Liang, B. Ding, Q. Feng, H. Fong, Polymer 151 [79] L. Jiang, M.P. Wolcott, J. Zhang, Biomacromolecules 7 (2018) 299–306 (2006) 199–207 [105] L. Jiang, J. Zhang, in Applied Plastics Engineering Hand- [80] M. Yasuniwa, S. Tsubakihara, Y. Sugimoto, C. Nakafuku, book (2nd ed.), M. Kutz (Ed.), William Andrew Publishing, J. Polym. Sci., B 42 (2004) 25–32 Norwich, NY, 2017, p. 127

417 M. ČOLNIK et al.: BIODEGRADABLE POLYMERS, CURRENT TRENDS… Chem. Ind. Chem. Eng. Q. 26 (4) 401−418 (2020)

[106] J. Xu, B.-H. Guo, in Plastics from Bacteria: Natural Func- [120] A. Sangroniz, L. Sangroniz, N. Aranburu, M. Fernández, tions and Applications, G.G.Q. Chen (Ed.), Springer A. Santamaria, M. Iriarte, A. Etxeberria, Eur. Polym. J. Berlin, 2010, p. 347 105 (2018) 348–358 [107] Y. Tokiwa, B. Calabia, C. Ugwu, S. Aiba, Y. Tokiwa, B.P. [121] L. Bastarrachea, S. Dhawan, S.S. Sablani, J.H. Mah, Calabia, C.U. Ugwu, S. Aiba, Int. J. Mol. Sci. 10 (2009) D.H. Kang, J. Zhang, J. Tang, J. Food Sci. 75 (2010) 215– 3722–3742 –224 [108] C. Brunner, E. Baran, E. Duarte Pinho, R.L. Reis, N. [122] L.B. Tavares, N.M. Ito, M.C. Salvadori, D.J. dos Santos, Neves, Colloids Surfaces, B 84 (2011) 498–507 D.S. Rosa, Polym. Test. 67 (2018) 169–176 [109] S. Jiang, Y. Yang, S. Ge, Z. Zhang, W. Peng, Arab. J. [123] D. Wei, H. Wang, Z. Ziaee, F. Chibante, A. Zheg, H. Xiao, Chem. 11 (2018) 844–857 Mater. Sci. Eng. C 58 (2016) 986–991 [110] K. Shi, Y. Liu, X. Hu, T. Su, P. Li, Z. Wang, Int. J. Biol. [124] J.H. Song, R.J. Murphy, R. Narayan, G.B.H. Davies, Macromol. 114 (2018) 373–380 Philos. Trans. R. Soc., B 364 (2009) 2127–2139 [111] H. Sun, Y. Luo, B. Yang, H. Zhang, J. Huang, Polymer [125] [S. Farah, D.G. Anderson, R. Langer, Adv. Drug Deliv. 153 (2018) 378–390 Rev. 107 (2016) 367–392 [112] Y. Xu, S. Zhang, P. Wang, J. Wang, Polymer 154 (2018) [126] M.L. Ortiz-Hernandez, E. Sanchez-Salinas, E. Dantan- 258–271 Gonzalez, M.L. Castrejon-Godınez, in Biodegradation: [113] BioPBS, http://www.pttmcc.com/new/faq.php (accessed Life of Science, R. Chamy, F. Rosenkranz (Eds.), Intech, 15 March 2020) Rijeka, 2013, p. 380 [114] M. Qahtani, F. Wu, M. Misra, S. Gregori, D. Mielewski, A. [127] X. Wang, J. Zhou, L. Li, Eur. Polym. J. 43 (2007) 3163– Mohanty, ACS Sustain. Chem. Eng. 7 (2019) 14460- –3170 –14470 [128] G.Z. Papageorgiou, D.S. Achilias, D.N. Bikiaris, [115] M.C. Picard, A. Rodriguez-Uribe, M. Thimmanagari, M. Macromol. Chem. Phys. 208 (2007) 1250-1264 Misra, A.K. Mohanty, Waste Biomass Valor. (2019) 1-13 [129] F. Aeschelmann, M. Carus, Ind. Biotechnol. 11 (2015) [116] A. Gowman, T. Wang, A. Rodriguez-Uribe, A.K. Mohanty, 154–159 M. Misra, ACS Omega 3 (2018) 15205–15216 [130] T.R. Crompton, Physical Testing of Plastics, Smithers [117] O. Platnieks, S. Gaidukovs, A. Barkane, G. Gaidukova, L. Rapra Technology, 2012 Grase, V.K. Thakur, I. Filipova, V. Fridrihsone, M. Skute, [131] B. Demirel, A. Yaraş, H. Elçiçek, BAÜ Fen Bil. Enst. M. Laka, Molecules 25 (2020) 1-18 Dergisi Cilt. 13 (2016) 26–35 [118] C.J. Cooper, A.K. Mohanty, M. Misra, ACS Omega 3 [132] G. Kfoury, J.M. Raquez, F. Hassouna, J. Odent, V. Toni- (2018) 5547–5557 azzo, D. Ruch, P. Dubois, Front Chem. 1 (2013) 1-46 [119] M. Seggiani, V. Gigante, P. Cinelli, M.-B. Coltelli, M. [133] Properties of Polystyrene, https://polymerdatabase.com/ Sandroni, I. Anguillesi, A. Lazzeri, Polym. Test. 77 (2019) /polymer%20classes/Polystyrene%20type.html 1-10 (accessed, 5 May 2019).

MAJA ČOLNIK BIORAZGRADIVI POLIMERI, TRENUTNI TRENDOVI MAŠA KNEZ HRNČIČ ISTRAŽIVANJA I NJIHOVA PRIMENA: PREGLED MOJCA ŠKERGET ŽELJKO KNEZ Biorazgradivi polimeri su se brzo razvijali poslednjih godina i danas se široko koriste u Laboratory for Separation Processes farmaciji, kliničkoj biomedicini, kozmetici, medicini, industriji pakovanja, inženjerstvu and Product Design, Faculty of tkiva, poljoprivredi i drugim oblastima. Interes za biorazgradivim polimerima raste, Chemistry and Chemical Engineering, uglavnom zbog porasta cena nafte, koja je osnovna sirovina za dobijanje plastike na University of Maribor, Maribor, Slovenia bazi nafte, kao i zbog problema uklanjanja otpadne plastike koja se akumulira u životnoj sredini. Biorazgradivi polimeri imaju mnogo prednosti u odnosu na sintetičks polimere i PREGLEDNI RAD mogu se razgraditi u životnoj sredini do bezopasnih materija. Biorazgradivi polimeri se klasifikuju u dve klase na osnovu njihove sinteze, tj. sintetičke i prirodne polimere. Dobi- jaju se iz naftnih ili bioloških resursa. Ovaj pregledni rad predstavlja pregled različitih biorazgradivih polimera i njihovih svojstava, trenutna naučna istraživanja, primene, globalnu proizvodnju bioplastike i zamenu konvencionalne plastike.

Ključne reči: biorazgradivi, polimeri, prirodno poreklo, bioplastika.

418 Available on line at Association of the Chemical Engineers of Serbia AChE

Chemical Industry & Chemical Engineering Quarterly www.ache.org.rs/CICEQ Chem. Ind. Chem. Eng. Q. 26 (4) 419 (2020) CI&CEQ

NILAY GIZLI Corrigendum to ENHANCED SORPTION OF MERVE ARABACI CU(II) IONS FROM AQUEOUS IONS FROM Ege University, Faculty of AQUEOUS SOLUTION BY IONIC LIQUID Engineering, Chemical IMPREGNATED NANO-SILIC AND Engineering Department, Izmir, Turkey NANO-ALUMINA PARTICLES

CORRIGENDUM (Chem. Ind. Chem. Eng. Q. 23(2) (2017) 207-216)

The authors regret that Figure 6, Table 3, and 2) Table 4, page 214, has been replaced by the some statements of the originally published article, following table: must be corrected while the other parts of the article remain unchanged. The corrections are as follows: Table 4. Equilibrium isotherm parameters and regression coef- Figure 6, page 214, has been replaced by the ficients following figure: Nanoparticle Model Coefficient S1 S2 S3 -1 Langmuir qmax / mg.g 6.65 7.47 11.66 b / L mg-1 0.61 0.46 0.39 r2 0.99 0.99 0.99 -1 Freundlich kF / mg.g 2.06 2.41 3.88 n 2.76 2.95 2.84 r2 0.88 0.95 0.87

3) On page 214, right column, sixth line from above, the values “2.61, 2.25 and 8.25 mg/g” has been replaced by “16.65, 7.47 and 11.66 mg/g.” 4) On page 215, left column, first line from above, the values “0.001 to 0.021” has been replaced by Figure 6. Equilibrium behaviour of the nanoparticle impregnated “0.72 to 0.05.” with IL at a ratio of 0.5:1. 5) Under Conclusions, page 215, left column, second paragraph, third line from below, the value ‘8.25” has been replaced by “11.66.” The authors apologize for any inconvenience caused by these inadvertent mistakes.

Correspondence: N. Gizli, Ege University, Faculty of Engineering, Chemical Engineering Department, 35100 Bornova, Izmir, Turkey. E-mail: [email protected] https://doi.org/10.2298/CICEQ210101001G

419

Journal of the Association of Chemical Engineers of Serbia, Belgrade, Serbia

CI&CEQ Vol. 26 Contents: Issues 1–4 YEAR 2020

No. 1 Hori, Erika Ohta Watanabe, Evaluation of supercritical carbon dioxide extraction to obtain bioactive compounds from Vernonia amygdalina Delile leaves ...... 113 Dragan Živančev, Bojan Jocković, Novica Mladenov, Alek- sandra Torbica, Miona Belović, Branka Mijić, Jordana Carolina B. Carvalho, Esdras P. Carvalho, Mauro A.S.S. Ninkov, The effects of wheat cultivars on the production Ravagnani, Combined neural networks and predictive of different types of wheat flours of precisely defined control for heat exchanger networks operation ...... 125 magnesium content ...... 1 Longhao Li, Yongshou Dai, An adaptive soft sensor det- Pejman Roohi, Esmaeil Fatehifar, Enhanced treatment of erioration evaluation and model updating method for 2-methylpropane-2-thiol contaminated soil using time-varying chemical processes ...... 135 magno-modified Fenton process ...... 9 Jovan Ćirić, Nataša Joković, Slavica Ilić, Sandra Konstan- Karla Raphaela, Braga de Melo, Gabriela Cantarelli Lopes, tinović, Dragiša Savić, Vlada Veljković, Production of Dayana de Gusmão Coêlho, João Inácio Soletti, Liquid- lactic acid by Enterococcus faecalis on waste glycerol liquid equilibrium for systems composed by biodiesel from biodiesel production ...... 151 from Catolé oil (Syagrus cearensis), methanol and Jasmina Vitas, Stefan Vukmanović, Jelena Čakarević, glycerol ...... 21 Ljiljana Popović, Radomir Malbaša, Kombucha fer- Ana Elisa Achiles, Vádila Giovana Guerra, Performance of mentation of six medicinal herbs: Chemical profile and a cyclone scrubber in removal of fine particulate matter ... 31 biological activity ...... 157 Marija Lješević, Jelena Milić, Gordana Gojgić-Cvijović, Tat- Aleksandra Bogdanovic, Vanja Tadic, Slobodan Petrovic, jana Šolević Knudsen, Mila Ilić, Jelena Avdalovic, Dejan Skala, Supercritical CO2 Extraction of steroidal Miroslav M. Vrvić, Evaluation of assays for screening sapogenins from fenugreek (Trigonella foenum- polycyclic aromatic hydrocarbon-degrading potential of -graecum L.) seed ...... 171 bacteria ...... 41 Hicham Kemmoukhe, Slavica Terzić, Mirjana Dimić, Danica Mohd Azahar Mohd Ariff, Norfazilah Abdullah, Optimization Simić, Zijah Burzić, Ljiljana Jelisavac, Influence of the of reflux extraction for cat’s whiskers leaves extract octogen quality and production scale on characteristics using response surface methodology ...... 49 of granulated plastic bonded explosive ...... 183 Eduardo Ramos Braga, George de Souza Mustafa, Danilo Anibal José Barrios Quant, Jose Andrés Perez Mendoza, de Aguiar Pontes, Luiz Antônio Magalhães Pontes, John Telotte, A comprehensive analysis of critical Economic analysis and technicalities of acrylic acid variables used to design a spray dryer of whey protein production from crude glycerol ...... 59 concentrates ...... 191 Erhan Sulejmani, Muhamet Demiri, Volatile compounds of Azmi Seyhun Kipcak, İbrahim Doymaz, Microwave and Macedonian fermented sausage as affected by ripening infrared drying kinetics and energy consumption of process using SPME/GC-MS ...... 71 cherry tomatoes ...... 203 Marija Jokanović, Bojana Ikonić, Predrag Ikonić, Vladimir Book Review: Aleksandar M. Spasic, Rheology of Tomović, Tatjana Peulić, Branislav Šojić, Snežana Emulsions: Electrohydrodynamics Principles ...... 213 Škaljac, Maja Ivić, Jelena Ivetić, Towards reproducibility of traditional fermented sausages: Texture profile analyses and modelling ...... 79 No. 3 Evgeny Akulinin, Oleg Golubyatnikov, Dmitry Dvoretsky, Stanislav Dvoretsky, Optimization and analysis of Luana Ventura Baia, Lucia R. Raddi de Araujo, Carlos pressure swing adsorption process for oxygen Guerra Pereira, Wallace Carvalho de Souza, Marco production from air under uncertainty ...... 89 Antonio Gaya de Figueiredo, Adsorption as alternative process in the preliminary production of automotive additive ...... 215 No. 2 Jun Niu, Shuo Liu, Jing-Yu Xu, Investigation on separation performance of vane-type gas-liquid tube separator ...... 227 Muhammad Sayuti, Reza Putra, Muhammad Yusuf, The Anđelka Tomašević, Aleksandar Marinković, Dušan Mijin, characterisation of magnetic materials extracted from Marina Radišić, Slavica Porobić, Nevena Prlainović, Aceh iron sand ...... 105 Slavica Gašić, A study of photocatalytic degradation of Fabíola Medeiros da Costa, Caroline Ortega Terra Lemos, methomyl and its commercial product Lannate-90 ...... 237 Sarah Arvelos, Marcos Rodrigo Traczynski, Edson Antônio da Silva, Lúcio Cardozo-Filho, Carla Eponina

421

CI&CEQ Vol. 26 Contents: issues 1–4 YEAR 2020

A. Arunmani, T. Senthilkumar, Enhancing corrosion resist- Sanja Radonjić, Helena Prosen, Vesna Maraš, Lea Demšar, ance of A-tig welded UNS s32750 joint by optimizing its Tatjana Košmerl, Incidence of volatile phenols in technological parameters ...... 249 Montenegrin red wines: Vranac, Kratošija and Cabernet Mohammed Foukrach, Houari Ameur, Effect of impeller Sauvignon ...... 337 blade curvature on the hydrodynamics and power P.C. Nnaji, B.I. Okolo, O.D. Onukwuli, Luffa cylindrica seed: consumption in a stirred tank ...... 259 Biomass for wastewater treament, sludge generation Naveedul Hasan Syed, Simulations of multi-size solid study at optimum conditions ...... 349 particles in a modified fluidized bed separator ...... 267 Kristina Jančaitienė, Rasa Šlinkšienė, Influence of cellulose Thamayne Valadares de Oliveira, Renata Bachmann Guim- additive on the granulation process of potassium arães Valt, Rafael Bruno Vieira, Haroldo de Araújo dihydrogen phosphate ...... 359 Ponte, Maria José Jerônimo de Santana Ponte, Carlos Bruna Muriel F. Costa, Carolina Monteiro Coelho, Cássia Itsuo Yamamoto, Antônio Augusto Ulson de Souza, Ribeiro Souza, Gabriela Drumond A. Duarte, Maria Selene Maria de Arruda Guelli Ulson de Souza, Regen- Bernadete Pinto, Vítor J. O. Silva, Marcelo S. Batista, eration study of ecat-R as adsorbent for denitrogenation Study of the drying kinetics of the malt bagasse in a and desulfurization of diesel fuels ...... 277 pneumatic transporter ...... 369 Dejana Popovic, Nenad Mitrovic, Aleksandar Petrovic, Milos Marija Bodroža-Solarov, Nevenka Rajić, Lato Pezo, Jovana Milosevic, Nikola Momcilovic, Sustainable development Kojić, Jelena Krulj, Bojana Filipčev, Rada Jevtić Muči- of pressure equipment using 3D digital image babić, The rheological properties of wheat dough correlation method ...... 287 containing zeolite residue ...... 377 Sarah Arvelos, Thalles Senna Diógenes, Carla Eponina Rim Jouraiphy, Islaam Anouar, Hamid Mazouz, Badr El Kor-

Hori, Lucienne Lobato Romanielo, Adsorption of CO2, tobi, Rachid Boulif, Ahmed Mahrou, Amina Ibnlfassi,

N2, CH4 and their mixtures on silicalite: A critical eva- Solid-liquid separation techniques for efficient recovery luation of force fields...... 295 of P2O5 from phosphoric sludge ...... 385 Velibor Marinković, A novel desirability function for multi- Jesús M. Casas, Josué Lagos, Drowning-out crystallization -response optimization and its application in chemical of sodium molybdate in aqueous–ethanol solutions ...... 395 engineering ...... 309 Maja Čolnik, Maša Knez Hrnčič, Mojca Škerget, Željko Knez, Biodegradable polymers, current trends of research and their applications, a review ...... 401 No. 4 Corrigendum to Enhanced sorption of Cu(II) ions from aqueous ions from aqueous solution by ionic liquid Pongayi Ponnusamy Selvi, Rajoo Baskar, CO2 absorption in impregnated nano-silic and nano-alumina particles nanofluid with magnetic field ...... 321 (Chem. Ind. Chem. Eng. Q. 23(2) (2017) 207-216) ...... 419 D. Revathi, Dr.K. Saravanan, Experimental studies on Contents: Vol. 26, Issues 1–4, 2020 ...... 421 hydrodynamic aspects for mixing of non-Newtonian fluids in a Komax static mixer ...... 329 Author Index, Vol. 25, 2019 ...... 423

422 Journal of the Association of Chemical Engineers, Belgrade, Serbia

CI&CEQ Vol. 26 Author Index YEAR 2020

A D

A. Arunmani (3) 249 D. Revathi (4) 329 Ahmed Mahrou (4) 385 Danilo de Aguiar Pontes (1) 59 Amina Ibnlfassi (4) 385 Dayana de Gusmão Coêlho (1) 21 Ana Elisa Achiles (1) 31 Demšar Lea (4) 337 Anibal José Barrios Quant (2) 191 Dimić Mirjana (2) 183 Antônio Augusto Ulson de Souza (3) 277 Dmitry Dvoretsky (1) 89 Avdalović Jelena (1) 41 Dr.K. Saravanan (4) 329 Azmi Seyhun Kipcak (2) 203 E B Edson Antônio da Silva (2) 113 Badr El Kortobi (4) 385 Eduardo Ramos Braga (1) 59 Belović Miona (1) 1 Erhan Sulejmani (1) 71 Bodroža-Solarov Marija (4) 377 Erika Ohta Watanabe (2) 113 Bogdanović Aleksandra (2) 171 Esdras P. Carvalho (2) 125 Bruna Muriel F. Costa (4) 369 Esmaeil Fatehifar (1) 9 Burzić Zijah (2) 183 Evgeny Akulinin (1) 89

C F

Carla Eponina Hori (2) 113; (3) 295 Fabíola Medeiros da Costa (2) 113 Carlos Guerra Pereira (3) 215 Filipčev Bojana (4) 377 Carlos Itsuo Yamamoto (3) 277 Carolina B. Carvalho (2) 125 G Carolina Monteiro Coelho (4) 369 Caroline Ortega Terra Lemos (2) 113 Gabriela Cantarelli Lopes (1) 21 Cássia Ribeiro Souza (4) 369 Gabriela Drumond A. Duarte (4) 369 Gašić Slavica (3) 237 Ć George de Souza Mustafa (1) 59 Gojgić-Cvijović Gordana (1) 41 Ćirić Jovan (2) 151 H Č Hamid Mazouz (4) 385 Čakarević Jelena (2) 157 Haroldo de Araújo Ponte (3) 277 Čolnik Maja (4) 401 Houari Ameur (3) 259

423

CI&CEQ Vol. 26 Author Index YEAR 2020

I Lj

İbrahim Doymaz (2) 203 Lješević Marija (1) 41 Ikonić Bojana (1) 79 Ikonić Predrag (1) 79 M Ilić Mila (1) 41 Ilić Slavica (2) 151 Malbaša Radomir (2) 157 Islaam Anouar (4) 385 Maraš Vesna (4) 337 Ivetić Jelena (1) 79 Marcelo S. Batista (4) 369 Ivić Maja (1) 79 Marco Antonio Gaya de Figueiredo (3) 215 Marcos Rodrigo Traczynski (2) 113 J Maria Bernadete Pinto (4) 369 Maria José Jerônimo de Santana Ponte (3) 277 Jelisavac Ljiljana (2) 183 Marinković Aleksandar (3) 237 Jesús M Casas (4) 395 Marinković Velibor (3) 309 Jevtić Mučibabić Rada (4) 377 Mauro A.S.S. Ravagnani (2) 125 Jing-Yu Xu (3) 227 Mijić Branka (1) 1 João Inácio Soletti (1) 21 Mijin Dušan (3) 237 Jocković Bojan (1) 1 Milić Jelena (1) 41 John Telotte (2) 191 Milošević Miloš (3) 287 Jokanović Marija (1) 79 Mitrović Nenad (3) 287 Joković Nataša (2) 151 Mladenov Novica (1) 1 Jose Andrés Perez Mendoza (2) 191 Mohammed Foukrach (3) 259 Josue Lagos (4) 395 Mohd Azahar Mohd Ariff (1) 49 Jun Niu (3) 227 Momčilović Nikola (3) 287 Muhamet Demiri (1) 71 K Muhammad Sayuti (2) 105

Muhammad Yusuf (2) 105 Karla Raphaela Braga de Melo (1) 21

Kemmoukhe Hicham (2) 183 N Knez Hrnčič Maša (4) 401

Knez Željko (4) 401 Naveedul Hasan Syed (3) 267 Kojić Jovana (4) 377 Ninkov Jordana (1) 1 Konstantinović Sandra (2) 151 Norfazilah Abdullah (1) 49 Košmerl Tatjana (4) 337 Nnaji P.C. (4) 349 Kristina Jančaitienė (4) 359

Krulj Jelena (4) 377 O

L Okolo B.I. (4) 349

Oleg Golubyatnikov (1) 89 Longhao Li (2) 135 Onukwuli O.D. (4) 349 Luana Ventura Baia (3) 215

Lucia R. Raddi de Araujo (3) 215 P Lucienne Lobato Romanielo (3) 295

Lúcio Cardozo-Filho (2) 113 Pejman Roohi (1) 9 Luiz Antônio Magalhães Pontes (1) 59

424 CI&CEQ Vol. 26 Author Index YEAR 2020

Petrović Aleksandar (3) 287 Š Petrović Slobodan (2) 171 Peulić Tatjana (1) 79 Škaljac Snežana (1) 79 Pezo Lato (4) 377 Škerget Mojca (4) 401 Pongayi Ponnusamy Selvi (4) 321 Šojić Branislav (1) 79 Popović Dejana (3) 287 Šolević Knudsen Tatjana (1) 41 Popović Ljiljana (2) 157 Porobić Slavica (3) 237 T Prlainović Nevena (3) 237 Prosen Helena (4) 337 T. Senthilkumar (3) 249 Tadić Vanja (2) 171 R Terzić Slavica (2) 183 Thalles Senna Diógenes (3) 295 Rachid Boulif (4) 385 Thamayne Valadares de Oliveira (3) 277 Radišić Marina (3) 237 Tomašević Anđelka (3) 237 Radonjić Sanja (4) 337 Tomović Vladimir (1) 79 Rafael Bruno Vieira (3) 277 Torbica Aleksandra (1) 1 Rajić Nevenka (4) 377 Rajoo Baskar (4) 321 V Rasa Šlinkšienė (4) 359 Renata Bachmann Guimarães Valt (3) 277 Vádila Giovana Guerra (1) 31 Reza Putra (2) 105 Veljković Vlada (2) 151 Rim Jouraiphy (4) 385 Vitas Jasmina (2) 157 Vítor J. O. Silva (4) 369 S Vrvić M. Miroslav (1) 41 Vukmanović Stefan (2) 157 Sarah Arvelos (2) 113; (3) 295 Savić Dragiša (2) 151 W Selene Maria de Arruda Guelli Ulson de Souza (3) 277 Wallace Carvalho de Souza (3) 215 Shuo Liu (3) 227 Simić Danica (2) 183 Y Skala Dejan (2) 171 Stanislav Dvoretsky (1) 89 Yongshou Dai (2) 135

Živančev Dragan (1) 1

425