<<

THE POTENTIAL USE OF MICRO- AND NANO- FIBRILLATED AS A REINFORCING ELEMENT IN ISKO KAJANTO*, MIKA KOSONEN

The use of nanofibrillar cellulose as a reinforcement agent in paper has been evaluated on a high-speed pilot . Using mill and mill process , two different grades of were tested at 1% and 2% addition levels together with 1% of cationic . The results indicate significant increase in tensile strength, enabling up to 8 g/m2 reduction. Paper had a slightly lower scattering coefficient and lower air permeability. Wire-section dewatering was reduced a little, with a loss of less than one percentage point of content, whereas dry matter after wet pressing even increased upon nanocellulose addition. The reason for this unexpected ob- servation is still unclear. Total retention remained constant after nanocellulose addition was started. ABSTRACT

INTRODUCTION Interest is increasing in micro- and nanofi- A common challenge in manufactur- brillar , or , which ing of nanofibrillar cellulose is high energy represent suprastructures of cellulose fi- consumption. High energy use increases bres. Much research effort has been ex- production cost, but more significantly, pended to demonstrate their usefulness as it is also an indication of low production a strengthening agent in paper and board capacity. Especially with purely mechani- [1], a viscosity modifier in -based sys- cal methods, energy levels of up to 20 tems [2] or as an -barrier material MWh/t of dry matter have been reported in packaging [3,4]. [1,6]. Recent research, using suitable pre- MIKA KOSONEN A wide variety of nanocellulose treatment methods, has enabled signifi- UPM-Kymmene, New families with different properties can be cant reductions in energy consumption Businesses and Development, Paloasemantie produced depending on the manufactur- [5, 7], making large-scale production more 19, 53200 Lappeenranta, ing method [5]. The main groups, shown realistic. The most common pre-treat- Finland in Fig. 1, are nanofibrillar cellulose, nano- ment methods are modifying fibre charge, crystalline cellulose, and bacterial cellu- either by TEMPO-mediated oxidization lose. The common factor is that in all of [8,9,10] or by carboxymethylation to a low these, the basic building block is the cel- DS [7,11,12,13], or weakening the fibres lulose elemental . In nanofibrillar cel- with enzymatic treatment [7,14]. The ad- lulose, both the crystalline and amorphous vantage of charge-modification methods regions are preserved, and the resulting is that the will also become smaller material consists of long flexible fibrils. than with mechanical methods only, thus Moreover, varying amounts of hemicel- improving nanocellulose properties. Fibre ISKO KAJANTO luloses remain in the material depending charge promotes fibrillation because it UPM-Kymmene, on the fibre source. In nanocrystalline cel- increases repulsion between the cellulose New Businesses and lulose, only the crystalline parts remain, microfibrils. Development, Tekniikantie 2 C, 02150 Espoo, and the resulting material consists of stiff As a paper additive, nanocelluloses Finland nanorods. act principally on internal bonding, leading *Contact: [email protected]

42 J-FOR Journal of Science & Technology for Products and Processes: VOL.2, NO.6, 2012 SPECIAL NANOTECHNOLOGY ISSUE

Eriksen [1] added 4% of mechanically Moreover, the dry solids after a standard manufactured nanocellulose into labo- wet-pressing procedure were two to four ratory TMP handsheets. He observed percentage points below the reference increased air resistance and a modest in- case. crease in tensile strength. Moreover, light With TEMPO-oxidized nanocellu- scattering decreased by -4 m2/kg. In par- lose, Heijnesson-Hulten reported drainage ticular, the quantitative changes were the time increases of 0 s–1.5 s (from a level following: of 2.5 s) with 100% CTMP pulp [13]. The drainage time increase tends to accelerate Fig. 1 - Grouping of nanocellulosic • Tensile strength + 7%–+21% with larger amounts of nanocellulose. materials. (+3 Nm/g–+ 7 Nm/g) Comparing these studies, it seems • Air resistance (Gurley), strong in- that drainage can be a bigger issue with to the following effects: crease, from 40 s to 200 s. chemical pulp, but the strengthening ef- fect of nanocellulose is also greater. • moderate increase in dry tensile Taipale [18] added 6% of mechani- A wet-pressing study simulating re- strength, larger with chemical than cally manufactured nanocellulose to labo- alistic wet pressing was reported in [20]. with mechanical pulp ratory handsheets. In his study, the pulp In this study, a dynamic pressing simula- • marked drop in air permeability was long-fibre bleached chemical pulp. He tor was used with TMP laboratory hand- • lower light scattering and thus noticed a strong increase in Scott Bond (+ sheets. The dry matter after wet pressing lower opacity 55%) and a good increase in tensile index was found to have been reduced by less • denser paper (up to + 15 Nm/g). than 0.5 percentage points, but drainage • increased hygroexpansivity. Similarly to Taipale, Manninen [19] was affected more. NFC reduced solids af- added 5% of Masuko-refined nanocellu- ter forming by approximately 1.5 percent- These will be reviewed in more detail lose to chemical pulp handsheets. She no- age points and increased drainage time by in the next section. ticed similar increases in tensile strength to up to 60%. The potential as a strengthening those seen by Taipale and also determined Retention of nanocellulose has not agent in paper is based on the fact that that nanocellulose increased the hygroex- been extensively studied. One main reason nanocellulose has a high surface area and pansion coefficient of freely dried sheets for this is that nanocellulose is difficult to therefore many sites for bond- by up to +30%. detect in paper because it is chemically ing. Nanocellulose has a character similar Heijnesson-Hulten [10,13] used similar to the pulp, and after drying, the to chemical pulp fines, except that because TEMPO-oxidized nanocellulose in labo- fibrils cannot be seen with microscopic of their much smaller size, the particles ratory CTMP handsheets. With addition methods. However, some concerns exist could be regarded as superfines. It is levels at 5%, there was an increase in Z- about how well such fine material is re- known from earlier research that cellulose strength by almost 200% (from 137 kPa tained in the web. It is known that fines fines improve bond strength and tensile to 387 kPa). In addition, the tensile index retention is typically smaller than the aver- strength, but they can have a detrimental increased noticeably, at +20%, although age retention value. effect on light scattering [15,16,17]. the absolute change was not as dramatic, In the literature cited above, the Despite the abundance of research at 8 Nm/g, as with the chemical pulps in amount of nanocellulose added has been into nanocellulose, all the published re- the Taipale and Manninen studies. Besides high, typically approximately 5%. Com- search with paper has been on a labora- the increases in strength, bulk decreased pared to this, the effects on, e.g., tensile tory scale. To determine the real potential by 13%. strength seem fairly small. This may be the of the material, its properties should be Processability issues have also been results of inefficient methods of adding demonstrated in industrial trials. As a step studied. The main effects of nanocel- nanocellulose or of a saturation effect at in this direction, the potential of nanocel- lulose are the poorer drainage of labora- lower amounts than those added. Mannin- lulose on a high-speed paper machine was tory handsheets and lower solids content en showed [19] that the effect of mechani- evaluated using a mill-type forming sec- after a laboratory wet-pressing procedure. cally prepared nanocellulose increased tion and press configuration. Retention can also be an issue, because it steadily until 5% addition, and then the ef- is well known that fines retention is lower fect leveled off. With grades manufactured EARLIER RESEARCH than fibre retention. using chemical pre-treatment, the satura- In her study, Manninen [19] noticed tion point may be at some other level. Several have been published on the that with a 5% addition of nanocellulose effect of nanocellulose on paper proper- into chemical pulp, the drainage time of ties. laboratory handsheets almost doubled.

J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.6, 2012 43 EXPERIMENTAL METHODS for longer fibrils. • One-nip shoe press, shoe length The pulp used was machine-chest 350 mm Materials chemical pulp from a . The pulp • Maximum line load 2000 kN/m As a paper strength additive, nanofibril- is a mixture of Finnish SW/HW with no • No drying section. lar cellulose is a better choice than nano- filler. Because mill pulp was used, - ital crystalline cellulose. The flexibility of the ready contained some additives, namely a The former was run in hybrid former fibrils promotes better conformity to the small amount of wet-end starch from the configuration with a one-nip shoe press. fibre surfaces and therefore a larger bond- machine circulation. ed area. The amorphous regions and the For a more realistic study, recircu- Variables during the trial present also provide more lated water from the paper mill was also The paper composition was as follows: groups for hydrogen bonding. Moreover, used. It was collected from the mill on the because of its higher yield compared to day before the trial. • Basis weight 50–65 g/m2 nanocrystalline cellulose, nanofibrillar Cationic potato starch, Raisamyl • Nanocellulose type AS or KS cellulose is a potentially cheaper material 50021 (DS = 0,035), 1% of the pulp dry • Nanocellulose dosage 0%, 1%, to manufacture. Therefore, as in all ear- weight, was used in all trials. The starch or 2% lier research on nanocellulose as a paper- was the same in all trials. Starch was Paper machine strengthening agent, nanofibrillar cellulose cooked on site and dosed in batches to the • Press load (1250, 1500, 1750, was chosen for this research. stock during trials. 2000 kN/m) In this work, the properties of two A two-component retention system o 1500 kN/m was used as a refer- types of nanocellulose manufactured at adapted from that used in the paper mill ence value UPM’s pilot facilities were tested. Different was used. Dosage of the retention aids • Paper machine speed (600, 800, pre-treatment methods using some com- was not modified with nanocellulose ad- 900 m/min) bination of the steps shown in Fig. 2 were dition. o 600 m/min was used as a refer- used to make the two grades. The main To be economical, the amount of ence value. characteristics of the materials are shown nanocellulose should be small, preferably in Table 1. The measurements were made at the addition levels of other strength ad- Solids content samples were collect- at a constant consistency which is less than ditives. Therefore, dosages of 1% - 2% ed after the wire section, after the press the manufacturing consistency to achieve were selected. section, and from the reel. Paper was comparable viscosity and turbidity values. Nanocellulose was diluted to 0.5% reeled after the press section, and paper Viscosity was measured with a Brookfield consistency before addition. samples were dried before measuring. Pa- RVDV-III viscometer at 10 rpm using a per samples were dried with a Kodak-type Vane spindle. Turbidity was measured Methods drying drum. Sheets were placed between with a HACH P2100 Turbidometer us- Trials were done on a narrow pilot paper plotting boards, and the surface tempera- ing a nephelometric principle (90 angle machine capable of high operating speeds. ture of the drying cylinder was 100°C. between light source and detector). Strong The machine did not have a press section, mixing was used to dilute the samples to but it was equipped with the possibility of RESULTS the measurement consistency. preparing small reels. Generally, the lower the turbidity, the The specifications of the pilot machine General more nanoscaled the material is. However, were as follows: Good overall runnability was achieved flocculation of the fibrils may also affect during the trial. No paper strength- or turbidity, although the individual fibrils • Configurable wire section with dewatering-related web breaks took would be nanoscaled. A higher viscos- the possibility of running the machine place, total retention level was constant, ity generally indicates the same thing, but in gap former, hybrid former, or Four- and formation remained good with- viscosity also reflects the properties of the drinier mode out any additional adjustment need. fibril network. Viscosity is usually higher • Maximum web speed 2500 m/min Press- and former-section dewatering

TABLE 1 Main characteristics of the nanocelluloses used in this research.

Viscosity Turbidity Nanocellulose Consistency = 0,8% Consistency = 0,1%

Type AS 16 600 mPa 64 NTU

Fig. 2 - Possible routes of manufacturing nanofibrillar cellulose Type KS 8 000 mPa 63 NTU using pre-treatments.

44 J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.6, 2012 SPECIAL NANOTECHNOLOGY ISSUE

remained at an acceptable level and is dis- Press-section dewatering tures speed up dewatering. However, no cussed in the following section. Contrary to expectations, the dry solids differences in the temperatures recorded- content of the paper web increased when during the trial were noticed. Forming-section dewatering nanocellulose was added to the stock. The Water-retention value (WRV) of the Former-section dewatering was analyzed nanocellulose-containing paper web ar- furnish typically increases with nanocellu- based on solids-content samples taken at rived at a lower solids content to the press lose addition, and this should lead to lower the end of the former section. Based on section, and therefore press dewatering after-press solids content, which does not the results achieved, nanocellulose addi- must have been very effective. Figure 4 support the result obtained here. How- tion led to 0.5–1 percenetage-point loss shows press solids-content results with 65 ever, the single-nip shoe-press section is of dry matter after the wire section. Ini- g/m2 paper. a relatively new technology, and relatively tial dewatering slowed down, and a greater The mechanism for high press dewa- few paper machines are operating with this share of water was removed at the flat vac- tering of a nanocellulose-containing web unit. It is believed that possible reasons for uum boxes at the end of the wire section. was studied further to gain more under- the improved after-press solids content for Figure 3 shows the solids content after the standing of the reasons for this unexpected a nanocellulose-containing web are related wire section for 65 g/m2 paper. result. It is commonly known that increas- to press configuration and to the pres- ing web speeds and pressing tempera- sure pulse in the shoe press, resulting in

Fig. 4 - Solids content after the press section. The figure shows Fig. 3 - Solids content after the wire section. The figure shows the the results with 65 g/m2 paper. The points represent the average of results with 65 g/m2 paper. different press loads.

Fig. 5 - Nanocellulose-containing paper produced systematically Fig. 6 - Nanocellulose-containing paper produced systematically higher after-press solids than the reference case; the shoe press higher after-press solids than the reference case; higher speed functioned even better as speed was increased. slightly decreased after-press solids.

J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.6, 2012 45 differences in rewetting. It is also possible Paper properties and cross-direction strength to form an that at the reference point (0% nanocel- Paper properties were analyzed based on impression of total tensile-strength poten- lulose), for some reason, the dry matter reeled and sheet-form separately dried tial regardless of possible variations in the content became too low. samples. The effects of nanocellulose on tensile ratio of the paper. As a function of press load, the af- strength properties and air permeability As seen in Fig. 9, nanocellulose-con- ter-press solids behaved as expected (Figs. of the paper were of special interest, and taining paper had a higher tensile strength 5 and 6). the results will be presented in the follow- than the reference case. Based on these ing section. The results were also partly results, it was estimated that up to 8 g/m2 Retention compared with a practical application, and basis-weight reduction potential existed Total retention remained at a very high the basis-weight reduction potential was without compromising tensile strength. level during the trial, greater than 94%. evaluated. AS-type higher-viscosity nanocellulose The furnish used had no filler and was Generally, as shown in Fig. 8, nano- seemed to give slightly higher tensile val- based totally on beaten pine and birch cellulose-containing trials gave higher den- ues than KS-type nanocellulose. Kraft pulps. Nanocellulose addition sity values than the reference case. This is In the case of air permeability, as seemed to have had practically no effect based on the strong network-building and shown in Fig. 10, nanocellulose gave a on the achieved retention, as shown in Fig. bonding capabilities of nanocellulose. 20%–30% lower porosity than the refer- 7. Neither did nanocellulose type seem to Tensile strength was calculated as a ence case. Translated into basis-weight have any major effect on retention. geometric average of machine-direction reduction potential, approximately 8 g/m2

Fig. 8 - Nanocellulose addition led to higher sheet . Press load Fig. 7 - Total retention remained constant during the trial despite was 1500 kN/m for all trials, and nanocellulose addition level was increased nanocellulose content. 1%–2%.

Fig. 9 - Tensile strength, geometric average MD*CD. Press load was Fig. 10 - Bendtsen air permeability. Press load was 1500 kN/m for all 1500 kN/m for all trials, and nanocellulose addition level was 1%–2%. trials, and nanocellulose addition level was 1%–2%.

46 J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.6, 2012 SPECIAL NANOTECHNOLOGY ISSUE

lower basis weight with nanocellulose- A negative effect on opacity was observed. H., Maezawa, T., Shiiba, R., Isogai, A., containing paper would have given the The relevance of opacity is highly appli- “Oxygen-Barrier Properties of TEM- same air permeability as the reference case. cation-dependent and should be taken PO-Oxidised Cellulose Nanofibres,” Optical properties of paper became into account when testing nanocellulose Proceedings, 8th International Paper worse after nanocellulose addition. At a for different end-uses. The strength in- and Coatings Chemistry Symposium, constant basis weight, approximately two crease indicates beneficial use in pack- Stockholm, Sweden, 10–14 June, p. to three percentage points lower opac- aging papers and boards, where loss of 333 (2012). ity was measured. This opacity loss arose opacity is not as critical. and 4. Rodionova, G., Saito, T., Lenes, M., mainly from improved bonding, which re- writing papers can benefit if the increase Eriksen, O., Gregersen, O., Fuku- zumi, H., Isogai, A., “Mechanical and duced the light-scattering surface. in strength is used to increase filler level. Oxygen-Barrier Properties of Films This loss of opacity did not sup- The very marked increase in elastic mod- Prepared from Fibrillated Disper- port the basis-weight reduction target ulus could provide bending stiffness in sions of TEMPO-Oxidized Norway in the same way as the good tensile and structural (layered) products and also part- Spruce and Pulps,” Cellu- air permeability results. If full advantage ly compensate for the effect of possibly lose 19(3):705-711 (2012). had been taken of the improved strength lowered basis weight. 5. Klemm, D., Kramer, F., Moritz, S., properties to lower the basis weight, the Lindström,, T., Ankerfors, A., Gray, result would have been approximately six ACKNOWLEDGEMENTS D., Dorris, A., “Nanocelluloses: A percentage points lower opacity. New Family of Nature-Based Mate- Opacity and light-scattering results The authors would like to thank TEKES rials,” Angewandte Chemie, Interna- are shown in Figs. 11 and 12. for providing funding for the research. tional Edition 50, 5438–5466 (2011). 6. Spence, K.L., Venditti, R.A., Rojas, CONCLUSIONS REFERENCES O.J., Habibi, Y., Pawlak, J.J., “A Com- 1. Eriksen, O., Syverud, K., Gregersen, parative Study of Energy Consump- Nanocellulose addition to paper resulted O., “The Use of Microfibrillated Cel- tion and Physical Properties of Mi- in interesting paper properties. First of lulose Produced from Kraft Pulp as crofibrillated Cellulose Produced by all, nanocellulose affected paper strength Strength Enhancer in TMP Paper,” Different Processing Methods,” Cel- beneficially even at low dosage levels. Nordic Pulp and Paper Research Jour- lulose 18( 4):1097-1111 (2011). Both tested nanocelluloses gave practically nal 23(3):299-304 (2008). 7. Ankerfors, M., Microfibrillated Cel- identical results. 2. Xhanari, K., Syverud, K., Stenius, P., lulose: Energy-Efficient Preparation Techniques and Key Properties (Li- There were no runnability or major “ Stabilized by Microfibril- centiate Thesis). TRITA-CHE Report production-efficiency issues during the lated Cellulose: The Effect of Hy- drophobization, Concentration, and 2012:38, Stockholm, Sweden: Royal trials. Solids content measurements indi- Institute of Technology, (2012). cated much better press dewatering than O/W Ratio,” Dispersion Science and Technology 32(3):447–452 (2011). 8. Saito, T., Nishiyama, Y., Putaux, expected, for which the reason is not to- J.-L., Vignon, M., Isogai, A., tally clear. 3. Kumamoto, Y., Mukai, K., Kawajiri,

Fig. 11 - Opacity of the paper samples. Press load 1500 was kN/m for Fig. 12 - Light-scattering coefficient of the paper samples. Press load all trials, and the nanocellulose addition level was 1%–2%. Average of was 1500 kN/m for all trials, and the nanocellulose addition level was top-side and wire-side measurements. 1%–2%. Average of top-side and wire-side measurements.

J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.6, 2012 47 “Homogeneous Suspensions of Coating Chemistry Symposium, June Nazhad, M.M., “Effect of Chemical Individualized Microfibrils from 10–12 (2009). Pulp Fines on Filler Retention and Pa- TEMPO-Catalyzed Oxidation of Na- 13. Heijnesson-Hultén, A., “Chemically per Properties,” Appita 60(6):469-473 tive Cellulose,” Biomacromolecules Pre-Treated MFC Process, Manufac- (2007). 7(6):1687-1691 (2006) turing and Application,” PFI Seminar, 18. Taipale, T., Osterberg, M., Nykanen, 9. Loranger, E., Piché, A.-O., Daneault, 14–15 November (2012b). A., Ruokolainen, J., Laine, J., “Ef- C., “Influence of High-Shear Disper- 14. Siddiqui, N., Mills, R.H., Gardner, fect of Microfibrillated Cellulose and sion on the Production of Cellulose D.K., Bousfield, D., “Production and Fines on the Drainage of Kraft Pulp Nanofibres by Ultrasound-Assisted Characterization of Cellulose Nano- Suspension and Paper Strength,” Cel- TEMPO Oxidation of Kraft Pulp,” fibres from Pulp,” Journal of lulose 17(5):1005-1020 (2010). Nanomaterials 2, 286-297 (2012). Adhesion Science and Technology 19. Manninen, M. , Kajanto, I., Happo- 10. Hejnesson-Hultén, A., Basta, J., Sam- 25(6-7), 709-721 (2011). nen, J., Paltakari, J., “The Effect of uelsson, M., Greschik, T., Ander, P., 15. Retulainen, E., “Fibre Properties as Microfibrillated Cellulose Addition on Daniel, G., “The Influence of Micro- Control Variables in , Drying Shrinkage and Dimensional fibrillated Cellulose (MFC) on Paper Part 1: Fibre Properties of Key Im- Stability of Wood-Free Paper,” Nor- Strength and Surface Properties,” portance in the Network,” Paperi ja dic Pulp and Paper Research Journal Bioresources 7(3):3051-63 (2012a). Puu 78(4):187-194 (1996). 26(3):297-305 (2011). 11. Walecka, J.A., “An Investigation of 16. Retulainen, E., Nieminen, K., “Fibre 20. Hii, C., Gregersen, Ø.W., Chinga- Low Degree of Substitution Carboxy- Properties as Control Variables in Pa- Carrasco, G., Eriksen, Ø., “The Effect methyl Celluloses,” TAPPI Journal permaking, Part 2: Strengthening In- of MFC on the Pressability and Paper 39(7):458-463 (1956). terfibre Bonds and Reducing Gram- Properties of TMP- and GCC-Based 12. Lindström, T., Ankerfors, M., “Nano- mage,” Paperi ja Puu 78(5):305-312 Sheets,” Nordic Pulp and Paper Re- cellulose Development in Scandi- (1996). search Journal 27(2):388-396 (2012). navia,” 7th International Paper and 17. Lin, T., Yin, X., Retulainen, E.,

www.paptac.ca J-FOR CALL FOR PAPERS Traditional Areas: (covering all pulping processes, both wood and non-wood) Pulping, bleaching and papermaking fundamentals, processes and technologies Energy and chemical recovery fundamentals, processes and technologies Recycled fibre and recycling technology Development of sensors, analytical methods and process control logics Mill water and energy usages and optimization Environmental concerns and their mitigation Emerging Areas: Emerging forest-based products and their chains of added value Fundamentals of converting forest-based into and other Nanotechnology and other high added-value processes Development of chemical, biochemical and thermochemical processes for the forestry industry Integrating emerging and sustainable processes into the Harvesting and procurement of forest and other biomass feedstocks

J-FOR publishes peer-reviewed articles of the highest quality, dealing with the science and technology of traditional and emerging areas that are pertinent to the forest industry. PAPTAC’s preeminent and flagship publication, it incorporates a broad scope of target areas and brings together a wide range of scientific, technological and technical papers. To submit a paper, please visit www.paptac.ca or contact PAPTAC (514-392-0265 / [email protected]).

PAPTAC NEWS

48 J-FOR Journal of Science & Technology for Forest Products and Processes: VOL.2, NO.6, 2012