Recuperative Amino Acids Separation Through Cellulose Derivative Membranes with Microporous Polypropylene Fiber Matrix

membranes

Article Recuperative Amino Acids Separation through Cellulose Derivative Membranes with Microporous Polypropylene Fiber Matrix

1 1 2 1, Aurelia Cristina Nechifor , Andreia Pîrt, ac , Paul Constantin Albu , Alexandra Raluca Grosu * , 3 1 3 1 1 Florina Dumitru , Ioana Alina Dimulescu (Nica) , Ovidiu Oprea , Dumitru Pas, cu , Gheorghe Nechifor and Simona Gabriela Bungău 4

1 Analytical Chemistry and Environmental Engineering Department, University Politehnica of Bucharest, 1-7 Polizu St., 011061 Bucharest, Romania; [email protected] (A.C.N.); [email protected] (A.P.); [email protected] (I.A.D.); [email protected] (D.P.); [email protected] (G.N.) 2 IFIN Horia Hulubei, Radioisotopes and Radiation Metrology Department (DRMR), 30 Reactorului St., 023465 Măgurele, Romania; [email protected] 3 Department of Inorganic Chemistry, Physical Chemistry and Electrochemistry, University Politehnica of Bucharest, 1-7 Polizu St., 011061 Bucharest, Romania; d_ﬂ[email protected] (F.D.); [email protected] (O.O.) 4 Faculty of Medicine and Pharmacy, University of Oradea, Universită¸tiiSt., no.1, Bihor, 410087 Oradea, Romania; [email protected] * Correspondence: [email protected]

Citation: Nechifor, A.C.; Pîrt,ac, A.; Albu, P.C.; Grosu, A.R.; Dumitru, F.; Abstract: The separation, concentration and transport of the amino acids through membranes have Dimulescu (Nica), I.A.; Oprea, O.; been continuously developed due to the multitude of interest amino acids of interest and the sources

Pas, cu, D.; Nechifor, G.; Bung˘au,S.G. from which they must be recovered. At the same time, the types of membranes used in the sepa- Recuperative Amino Acids ration of the amino acids are the most diverse: liquids, ion exchangers, inorganic, polymeric or Separation through Cellulose composites. This paper addresses the recuperative separation of three amino acids (alanine, phe- Derivative Membranes with nylalanine, and methionine) using membranes from cellulosic derivatives in polypropylene ma-trix. Microporous Polypropylene Fiber The microfiltration membranes (polypropylene hollow fibers) were impregnated with solu-tions Matrix. Membranes 2021, 11, 429. of some cellulosic derivatives: cellulose acetate, 2-hydroxyethyl-cellulose, methyl 2-hydroxyethyl- https://doi.org/10.3390/ celluloseand sodium carboxymethyl-cellulose. The obtained membranes were characterized in terms membranes11060429 of the separation performance of the amino acids considered (retention, flux, and selectivity) and from a morphological and structural point of view: scanning electron microscopy (SEM), high resolution Academic Editor: Isabel C. Escobar SEM (HR-SEM), Fourier transform infrared spectroscopy (FT-IR), energy dispersive spectroscopy

Received: 29 April 2021 (EDS) and thermal gravimetric analyzer (TGA). The re-sults obtained show that phenylalanine has Accepted: 3 June 2021 the highest fluxes through all four types of mem-branes, followed by methionine and alanine. Of the Published: 5 June 2021 four kinds of membrane, the most suitable for recuperative separation of the considered amino acids are those based on cellulose acetate and methyl 2-hydroxyethyl-cellulose. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in Keywords: amino acids; cellulose derivatives; polypropylene hollow fibers; impregnated membranes; published maps and institutional affil- amino acid separation; membrane processes iations.

1. Introduction Copyright: © 2021 by the authors. Amino acids are substances whose biological role, when incorporated into protein Licensee MDPI, Basel, Switzerland. molecules, neurotransmitters, or hormones and in precursors of other molecules, has led to This article is an open access article a continuous development of fundamental and applied research [1,2]. The simultaneous distributed under the terms and presence in the molecule of amino acids, of a basic, amino group (-NH2) but also of an conditions of the Creative Commons acid group, carboxyl (-COOH), close spatially (Figure1a), gives these substances unique Attribution (CC BY) license (https:// physical-chemical and biological properties [3,4]. Of course, one of the important aspects of creativecommons.org/licenses/by/ this structure is given by the presence of electrical charges in amino acids [5–7], regardless 4.0/).

Membranes 2021, 11, 429. https://doi.org/10.3390/membranes11060429 https://www.mdpi.com/journal/membranes Membranes 2021, 11, x FOR PEER REVIEW 2 of 25

Membranes 2021, 11, 429 aspects of this structure is given by the presence of electrical charges in amino acids [5–7],2 of 24 regardless of the pH of the aqueous solution in which they are found (Figure 1b). Simul- taneously, the chemical properties of the amino and carboxyl groups make the amino acids the molecules of greatest interest in modifying, by grafting, the properties of various of the pH of the aqueous solution in which they are found (Figure1b). Simultaneously, organic and inorganic materials [8–11]. the chemical properties of the amino and carboxyl groups make the amino acids the All these characteristics, but also the multiple chemicals, biological and biomedical molecules of greatest interest in modifying, by grafting, the properties of various organic applications make it necessary to recover amino acids from any source, even when their and inorganic materials [8–11]. concentration is low [12–17].

(a) (b)

FigureFigure 1. 1. TheThe chemical chemical structure structure of ofamino amino acids acids (a)( anda) and the the ionic ionic forms forms depending depending on the on pH the of pH of aqueous solution in which they are found (b). aqueous solution in which they are found (b).

TheAll theseproperties characteristics, of an amino but acid also will the also multiple be influenced chemicals, by biological the R group, and biomedicalwhich can giveapplications the molecule make specific it necessary interactions: to recover hydrophobic, amino acids hydrophilic, from any source,or redox, even depending when their on theconcentration atoms that make is low it [ 12up– [18,19].17]. All these characteristics of amino acids allowed researchers to Theaddress properties the most of anvaried amino techniques acid will alsoand bemethods inﬂuenced of separation, by the R group, among which which can some give ofthe those molecule involving speciﬁc membranes interactions: are presented hydrophobic, in Table hydrophilic, 1 [20–59]. or redox, depending on the atoms that make it up [18,19]. All these characteristics of amino acids allowed researchers Tableto address 1. Membrane the most amino varied acids techniques separation andtechniques methods and ofmethods. separation, among which some of those involving membranes are presented in Table1[20–59]. Methods or Techniques Gradient Refs.

Table 1. MembraneDialysis amino (D), acids Hemodialysis separation techniques (HD) and methods. Δc [20–22] Ion-exchange membranes (IEM) Δc, ΔE [23–25] MethodsElectro dialysis or Techniques (ED) GradientΔE [26,27] Refs. DialysisElectro-ultrafiltration (D), Hemodialysis (EUF) (HD) ΔE,∆c Δp [28–33] [20–22] Ion-exchangeUltrafiltration membranes (UF) (IEM) ∆c,Δ∆pE[ [34–37]23–25] NanofiltrationElectro dialysis (ED)(NF) ∆ΔE[p [38–43]26,27] Molecularly imprintedElectro-ultrafiltration polymeric membranes (EUF) (MIPM) ∆E,Δ∆pp [28[44]–33 ] Liquid membranes (LM) Δc [45–47] Ultrafiltration (UF) ∆p [34–37] Other processes (Adsorption, Polyelectrolyte nanoparticles, Nanofiltration (NF) ∆p [38–43] Composite membranes, Combined processes, Δc, ΔE, Δp [48–59] Molecularly imprintedMolecular polymeric recognition) membranes (MIPM) ∆p [44] Liquid membranes (LM) ∆c [45–47] OtherSources processes of amino (Adsorption, acid are Polyelectrolyte among the most nanoparticles, diverse, but mainly hydrolysates in the food industryComposite have been membranes, addressed: Combined milk, processes,meat, fish, wine, or animal∆c, ∆E, skin∆p processing, [48–59] agriculture, and biotechnology.Molecular recognition) Each method or process of membrane separation, concentration, or purification of aminoSources advantages of amino and disadvantages acid are among and the must most be diverse, correlated but with mainly the hydrolysatesprocessing system, in the sofood as to industry meet both have selectivity been addressed: and productivity milk, meat, requirements fish, wine, [60]. or animal skin processing, agriculture, and biotechnology. Each method or process of membrane separation, concentration, or purification of amino advantages and disadvantages and must be correlated with the processing system, so as to meet both selectivity and productivity requirements [60]. Membranes 2021, 11, 429 3 of 24

However, we must mention some of the specific performances of the membrane processes, which make them extremely attractive for the separation of compounds of biological interest such as the amino acids [51–60]: - High selectivity; - Productivity guided by membrane design; - Accessible operating parameters; - Raising the scale without multiplying the problems of chemical engineering; - Reduced volume of installations; - Access to the treatment of coarse and viscous dispersed systems; - Minimizing of the consumption of chemical reagents and auxiliary materials; - Reduced investments for the realization of the auxiliary systems; - Complete automation; - Possibility to operate in isolate or hard to reach locations; - Favorable economic ratio, operating costs: product quality. From the perspective of the application of the presented membrane processes (Table1 ) for membrane separation of amino acids, they can be approached as follows: - high flows and productivity [23–27], - simplicity and robustness of the installations [28–43], - or their selectivity [44–47]. The membranes can also be chosen according to selectivity and flow criteria [61–63], but in terms of biocompatibility, accessibility and cost the cellulose based membranes remain the most studied and applied [64,65]. This paper addresses the recuperative separation of three amino acids (alanine, phenylalanine, and methionine) from synthetic solutions, using membranes from cellulosic derivatives (cellulose acetate, 2-hydroxyethyl-cellulose, and methyl 2-hydroxyethyl-cellulose or sodium carboxymethyl-cellulose) in polypropylene hollow fiber matrix.

2. Materials and Methods 2.1. Materials 2.1.1. Chemicals The materials used in the present work were of analytical purity. They were purchased from Merck (Merck KGaA, Darmstadt, Germany): sodium hydroxide (NaOH) and hydrochloric acid solution (HCl, 35%). The amino acids (alanine, phenylalanine and, methionine) and the cellulose derivatives (cellulose acetate (Product of USA), 2-hydroxyethyl-cellulose (product of USA), methyl 2-hydroxyethyl-cellulose (product of Belgium) and sodium carboxymethyl-cellulose (Prod- uct of USA)) were purchased from Aldrich Chemistry (Merck KGaA, Darmstadt, Germany). The puriﬁed water, characterized by a conductivity of 18.2 µS/cm, was obtained with a RO Millipore system (MilliQR Direct 8 RO Water Puriﬁcation System, Merck, Darmstadt, Germany).

2.1.2. Membrane Support The hollow ﬁbers polypropylene support membranes (PPM) were provided by GOST Ltd., Perugia, Italy. Their characteristics are presented in Figure2[66,67]. MembranesMembranes2021 2021, 11, 11, 429, x FOR PEER REVIEW 44 of of 24 25

FigureFigure 2. 2.The The characteristics characteristics of of the the hollow hollow fibers fibers polypropylene polypropylene support support membranes membranes (PPM). (PPM). 2.2. Impregnated Cellulose Derivatives Polypropylene Membrane Preparation (Cell-D-PPM) 2.2. Impregnated Cellulose Derivatives Polypropylene Membrane Preparation (Cell-D-PPM) 2.2.1. Obtaining the Cellulose Derivatives Solution 2.2.1. Obtaining the Cellulose Derivatives Solution The cellulosic derivatives (cellulose acetate, 2-hydroxyethyl-cellulose, methyl 2-hydroxyethyl-celluloseThe cellulosic derivatives or sodium (cellulose carboxymethyl-cellulose), acetate, 2-hydroxyethyl-cellulose, having the characteristics methyl 2-hy- indicateddroxyethyl-cellulose in Table2, were or sodium solubilized carboxymethyl-c in a massellulose), concentration having of the 2%, characteristics in a mixture indi- of methylenecated in Table chloride:methyl 2, were solubilized alcohol in = a 2:1. mass concentration of 2%, in a mixture of methylene chloride:methylFor this, the alcohol glass vessel = 2:1. in which the mixture of solvents and the corresponding amount of polymer were introduced, was placed in an ultrasonic bath (Elmasonic S, Elma SchmidbauerTable 2. The characteristics GmbH, Singen, of cellulosic Germany) derivatives for four under hours, test observing and the membranes the complete obtained. dispersion and obtainingCellulose the polymeric solution. The obtained solutions were filtered using a metal Molar Membrane sieve withDerivatives 40 µm × 40 µm meshes andChemical then placed Formula in closed containers for 24 h in order to Weight Symbol remove bubbles.(Cell-D) 2.2.2. Obtaining Cellulose Derivatives Impregnated on Polypropylene Fibers Membranes (Cell-D-PPM) cellulose acetate (CA) 50,000 CA-PPM For impregnation of support hollow fiber membranes, the polymer solution was placed in a glass cylindrical vessel with measured capacity and then the fiber bundle was immersed (Figure3) with the heads out of the solution and coupled to a preliminary vacuum pump [68,69]. The fibers were coupled to the vacuum pump to remove the air and when2-hydroxy-ethylcellu- the pressure difference increased, the pump switched-off automatically. The system 380,000 HEC-PPM was maintainedlose (HEC) in this state for 10 min, after which the impregnated fibers were removed above the polymer solution. After 30 min, the fibers were placed in the vacuum oven, at 60 ◦C, to remove the solvent mixture. After this stage, the fibers were placed in a 1 L vessel with deionized water and after about an hour, they could be used in the permeation process.methyl Four 2-hydroxy- types of membranes were obtained, symbolized according to Table2. ethyl-cellulose *) CHEC-PPM (MHEC)

sodium carboxymethyl-cellulose 90,000 NaCMC-PPM (NaCMC)

*) Viscosity (c = 2%, water, 20 °C) 15,000–20,500 cps.

Membranes 2021, 11, x FOR PEER REVIEW 4 of 25

Membranes 2021, 11, x FOR PEER REVIEW 4 of 25 Membranes 2021, 11, x FOR PEER REVIEW 4 of 25

Figure 2. The characteristics of the hollow fibers polypropylene support membranes (PPM).

Figure 2. The characteristics of the hollow fibers polypropylene support membranes (PPM). Figure2.2. Impregnated 2. The characteristics Cellulose of Derivatives the hollow Polyprfibers polypropyleneopylene Membrane support Preparation membranes (Cell-D-PPM) (PPM). 2.2. Impregnated Cellulose Derivatives Polypropylene Membrane Preparation (Cell-D-PPM) 2.2.2.2.1. Impregnated Obtaining Cellulose the Cellulose Derivatives Derivatives Polypropylene Solution Membrane Preparation (Cell-D-PPM) 2.2.1. Obtaining the Cellulose Derivatives Solution 2.2.1. ObtainingThe cellulosic the Cellulose derivatives Derivatives (cellulose Solution acetate, 2-hydroxyethyl-cellulose, methyl 2-hydroxyethyl-celluloseThe cellulosic derivatives or sodium (cellulose carboxymethyl-c acetate, 2-hydroxyethyl-cellulose,ellulose), having the characteristics methyl 2-hy- indi- droxyethyl-cellulosecatedThe in cellulosic Table 2, were derivatives or solubilized sodium (cellulose carboxymethyl-c in a mass acetat concentratione, ellulose),2-hydroxyethyl-cellulose, havingof 2%, in the a mixturecharacteristics methyl of methylene 2-hy- indi- droxyethyl-cellulosecatedchloride:methyl in Table 2, were alcohol or solubilized sodium = 2:1. carboxymethyl-c in a mass concentrationellulose), of having 2%, in the a mixture characteristics of methylene indi- Membranes 2021, 11, 429 catedchloride:methyl in Table 2, werealcohol solubilized = 2:1. in a mass concentration of 2%, in a mixture of methylene5 of 24 chloride:methylTable 2. The characteristics alcohol = 2:1. of cellulosic derivatives under test and the membranes obtained. Table 2. The characteristics of cellulosic derivatives under test and the membranes obtained. Table 2. TheCellulose characteristics of cellulosic derivatives under test and the membranes obtained. Table 2.CelluloseThe characteristics of cellulosic derivatives under test and the membranesMolar obtained.Membrane Derivatives Chemical Formula Molar Membrane DerivativesCellulose Chemical Formula Weight Symbol Cellulose(Cell-D) Molar Membrane Derivatives Chemical Formula Weight SymbolMembrane (Cell-D)Derivatives Chemical Formula MolarWeight Weight Symbol Symbol (Cell-D)(Cell-D)

cellulose acetate (CA) 50,000 CA-PPM cellulose acetate (CA) 50,000 CA-PPM cellulosecellulose acetate acetate (CA) (CA) 50,00050,000 CA-PPM CA-PPM

Membranes 2021, 11, x FOR PEER REVIEW 5 of 25

2-hydroxy-ethylcellu- For this, the glass vessel in which the mixture of solvents380,000 and the correspondingHEC-PPM 2-hydroxy-ethylcellu-2-hydroxy-ethylcelluloselose (HEC) amount2-hydroxy-ethylcellu- of polymer were introduced, was placed in an ultrasonic380,000 bath (ElmasonicHEC-PPM HEC-PPM S, Elma lose(HEC) (HEC) 380,000 HEC-PPM Schmidbauerlose (HEC) GmbH, Singen, Germany) for four hours, observing the complete dispersion and obtaining the polymeric solution. The obtained solutions were filtered using a metal sieve with 40 µm × 40 µm meshes and then placed in closed containers for 24 h in order to remove bubbles. methylmethyl 2-hydroxy- Membranes 2021, 11, x FOR PEER REVIEW2.2.2.2-hydroxyethyl-cellulosemethyl ethyl-celluloseObtaining 2-hydroxy- Cellulose Derivatives Impregnated on Polypropylene*)*) FibersCHEC-PPM CHEC-PPM Membranes5 of 25 methyl 2-hydroxy- (Cell-D-PPM)ethyl-cellulose(MHEC)(MHEC) *) CHEC-PPM ethyl-cellulose *) CHEC-PPM For(MHEC) impregnation of support hollow fiber membranes, the polymer solution was (MHEC) placed in a glass cylindrical vessel with measured capacity and then the fiber bundle was immersed (Figure 3) with the heads out of the solution and coupled to a preliminary vacuum pump [68,69]. The fibers were coupled to the vacuum pump to remove the air and whensodium the sodiumpressurecarboxyme- difference increased, the pump switched-off automatically. The system wascarboxymethyl-cellulose maintainedthyl-cellulose in this state for 10 min, after which the impregnated90,000 fibers NaCMC-PPM were NaCMC-PPM removed above the(NaCMC)(NaCMC) polymer solution. After 30 min, the fibers were placed in the vacuum oven, at 60 °C, to remove the solvent mixture. After this stage, the fibers were placed in a 1 L vessel with deionized water and after about an hour, they could be used in the permeation pro- *) Viscosity (c = 2%, water, 20 ◦C) 15,000–20,500 cps. *)cess. Viscosity Four (c types = 2%, of water, membranes 20 °C) 15,000–20,500 were obtained, cps. symbolized according to Table 2.

For this, the glass vessel in which the mixture of solvents and the corresponding amount of polymer were introduced, was placed in an ultrasonic bath (Elmasonic S, Elma Schmidbauer GmbH, Singen, Germany) for four hours, observing the complete dispersion and obtaining the polymeric solution. The obtained solutions were filtered using a metal sieve with 40 µm × 40 µm meshes and then placed in closed containers for 24 h in order to remove bubbles.

2.2.2. Obtaining Cellulose Derivatives Impregnated on Polypropylene Fibers Membranes (Cell-D-PPM) For impregnation of support hollow fiber membranes, the polymer solution was placed in a glass cylindrical vessel with measured capacity and then the fiber bundle was

immersed (Figure 3) with the heads out of the solution and coupled to a preliminary vac- (a) (b) uum pump [68,69]. The fibers were coupled to the vacuum pump to remove the air and Figure 3. Schematic depictionwhen of the impregnationpressure difference procedure: increased, (a) hollow the polypropylene pump switched-off ﬁberfiber bundles; automatically. and (b) scheme The system of the stages of the impregnation procedure. the stages of the impregnationwas procedure.maintained in this state for 10 min, after which the impregnated fibers were removed above the polymer solution. After 30 min, the fibers were placed in the vacuum oven, at 2.3. Permeation Procedures 602.3. °C, Permeation to remove Proceduresthe solvent mixture. After this stage, the fibers were placed in a 1 L vessel In the source phase (SP) the synthetic solution of the considered chemical species with deionizedIn the source water phase and after (SP) theabout synthetic an hour, solution they could of the be used considered in the permeation chemical species pro- (amino acids—Table 3), with a concentration of 0.150–0.250 mol/L, is introduced in the cess.(amino Four acids—Table types of membranes3), with a were concentration obtained, ofsymbolized 0.150–0.250 according mol/L, isto introducedTable 2. in the installation (Figure 4). The pH of the source phase is made with hydrochloric acid solution installation (Figure4). The pH of the source phase is made with hydrochloric acid solution 0.01 mol/L, deionized water, or sodium hydroxide 0.1 mol/L. The receiving phase (RP) is 0.01 mol/L, deionized water, or sodium hydroxide 0.1 mol/L. The receiving phase (RP) is formed in all cases from deionized water. The volume of the source phase is 10 L, and of the receiving phase is 1 L. The operating flows are the same for all experiments, for operational reasons already studied [69–71]. Thus, for the source phase the flow in the pertraction module is 5 L/min, and the flow of the receiving phase is 0.1 L/min. In order to determine the amino acids separation performances, a 1 mL of solution is taken periodically and analyzed for spectrophotometric analysis (CamSpec Spectrophotometer) [72–74]. For the receiving phase, the operation is performed through capillaries while for the source phase the operation is performed outside the capillaries (Figure 4).

(a) (b) Figure 3. Schematic depiction of the impregnation procedure: (a) hollow polypropylene fiber bundles; and (b) scheme of the stages of the impregnation procedure.

2.3. Permeation Procedures In the source phase (SP) the synthetic solution of the considered chemical species (amino acids—Table 3), with a concentration of 0.150–0.250 mol/L, is introduced in the installation (Figure 4). The pH of the source phase is made with hydrochloric acid solution 0.01 mol/L, deionized water, or sodium hydroxide 0.1 mol/L. The receiving phase (RP) is formed in all cases from deionized water. The volume of the source phase is 10 L, and of the receiving phase is 1 L. The operating flows are the same for all experiments, for oper-

Membranes 2021, 11, x FOR PEER REVIEW 6 of 25

Membranes 2021, 11, 429 ational reasons already studied [69–71]. Thus, for the source phase the flow in the pertrac-6 of 24 Membranes 2021, 11, x FOR PEER REVIEWtion module is 5 L/min, and the flow of the receiving phase is 0.1 L/min. In order to deter-6 of 25

ationalmine the reasons amino already acids separationstudied [69–71]. performances, Thus, for thea 1 sourcemL of solutionphase the is flow taken in periodicallythe pertrac- tionandformed moduleanalyzed in allis for5 cases L/min, spectrophotometric from and deionized the flow water.of analysis the Thereceiving (CamSpec volume phase of Spectrophotometer) the is source0.1 L/min. phase In isorder 10 [72–74]. L, to and deter- ofFor the the receiving phase, the operation is performed through capillaries while for the source ationalminereceiving the reasons amino phase already acids is 1 L.separation studied The operating [69–71]. performances, flows Thus, are for the athe 1 same mLsource of for solutionphase all experiments, the is flow taken in for periodicallythe operational pertrac- phase the operation is performed outside the capillaries (Figure 4). tionandreasons moduleanalyzed already is for 5 L/min, spectrophotometric studied and [the69– 71flow]. Thus,of analysis the forreceiving (CamSpec the source phase Spectrophotometer) phase is 0.1 theL/min. flow In in order the [72–74]. pertractionto deter- For module is 5 L/min, and the flow of the receiving phase is 0.1 L/min. In order to determine minethe receiving the amino phase, acids the separation operation performances, is performed athrough 1 mL of capillaries solution iswhile taken for periodically the source Tablethe amino 3. The acidscharacteristics separation of the performances, tested amino acids a 1 mL(at 25 of °C). solution is taken periodically and andphase analyzed the operation for spectrophotometric is performed outside analysis the capillaries(CamSpec (FigureSpectrophotometer) 4). [72–74]. For Membranes 2021, 11, x FOR PEER REVIEWanalyzed for spectrophotometric analysis (CamSpec Spectrophotometer) [72–74]. For6 theof 25 the receiving phase, theMolar operation Weight is perforSolubilitymed through Isoelectric capillaries Point whileAcidity for Constantsthe source Amino acid ChemicalTablereceiving 3. FormulaThe phase, characteristics the operation of the tested is performed amino acids through (at 25 capillaries°C). while for the source phase phase the operation is performed(g/mol) outside the(g/L) capillaries (Figure(Ip) 4). (pKa) the operation is performed outside the capillaries (Figure4). Table 3. The characteristicsMolar Weightof the tested Solubility amino acids Isoelectric (at 25 °C). Point Acidity Constants Amino acid ChemicalTable 3. Formula The characteristics of the tested amino acids (at 25 °C). Table 3. The characteristics(g/mol) of the tested amino(g/L) acids (at 25 ◦C).(Ip) (pKa) Molar Weight Solubility Isoelectric Point 2.35Acidity (carboxyl), Constants AlanineAmino acid Chemical Formula Molar89.09 Weight Solubility167.2 Isoelectric6.11 Point Acidity Constants Amino acid Chemical Formula Molar Weight(g/mol) Solubility(g/L) Isoelectric(Ip) Point 9.87Acidity (amino)(pKa) Constants Amino Acid Chemical Formula (g/mol) (g/L) (Ip) (pKa) (g/mol) (g/L) (Ip) 2.35 (carboxyl),(pKa) Alanine 89.09 167.2 6.11 9.872.35 (amino) (carboxyl), Alanine 89.09 167.2 6.11 9.87 (amino) 2.352.35 (carboxyl), (carboxyl), AlanineAlanine 89.0989.09 167.2 167.2 6.11 6.11 9.879.87 (amino) (amino) 2.58 (carboxyl), Phenylalanine 165.19 26.9 5.91 9.13 (amino) 2.58 (carboxyl), Phenylalanine 165.19 26.9 5.91 2.582.58 (carboxyl), (carboxyl), PhenylalaninePhenylalanine 165.19165.19 26.9 26.9 5.91 5.91 9.13 (amino) 9.139.13 (amino) (amino) 2.58 (carboxyl), Phenylalanine 165.19 26.9 5.91 9.13 (amino) MethionineMethionine or or Methionine2-amino-4- or 2.282.28 (carboxyl), (carboxyl), 2-amino-4-(methyl- 149.21149.21 56.6 5.74 5.74 2.28 (carboxyl), (methylthio)2-amino-4-(methyl- butanoic 149.21 56.6 5.74 9.219.21 (amino) (amino) thio)Methionine butanoicacid or acid 9.21 (amino) thio) butanoic acid 2.28 (carboxyl), 2-amino-4-(methyl- 149.21 56.6 5.74 9.21 (amino) thio)Methionine butanoic oracid 2.28 (carboxyl), 2-amino-4-(methyl- 149.21 56.6 5.74 9.21 (amino) thio) butanoic acid

FigureFigureFigure 4. 4.Depiction4.Depiction Depiction of oftheof thethe operational operationaloperational scheme schemescheme with withwith the thethe pertraction pertractionpertraction module: module:module: SP SP SP– source –– sourcesource phase, phase,phase, RS RS RS – receiving phase. 1. Hollow ﬁber pertraction module; 2. SP reservoirs; 3. RP reservoirs; 4. SP pump; – receiving– receiving phase. phase. 1. Hollow1. Hollow fiber fiber pertraction pertraction module; module; 2. SP2. SPreservoirs; reservoirs; 3. RP3. RP reservoirs; reservoirs; 4. SP4. SP pump;5pump;. RP 5 pump;. RP5. RP pump;6 pump;. T thermostat. 6. T6 .thermostat. T thermostat. Figure 4. Depiction of the operational scheme with the pertraction module: SP – source phase, RS – receiving phase. 1. Hollow fiber pertraction module; 2. SP reservoirs; 3. RP reservoirs; 4. SP TheThe ﬂuxes fluxes from from the the source source phase phase [75 [75]] were were determined determined against against the the measured measured permeate perme- pump;The 5. RP fluxes pump; from 6. T thethermostat. source phase [75] were determined against the measured perme- Figureatemassate mass 4.mass withinDepiction within within a determineda of determined a the determined operational time time time range,scheme range, range, applying with applying applying the pertraction the the followingthe following following module: equation: equation: SP equation: – source phase, RS – receiving phase. 1. Hollow fiber pertraction module; 2. SP reservoirs; 3. RP reservoirs; 4. SP The fluxes from the source phaseM [75] were determined against the measured perme- pump; 5. RP pump; 6. T thermostat.J =𝐽 = (mg (mg(m⁄(m2 s))2 s))or or(( mol((mol))(m⁄(m2 s2 ))s)) (1)(1) ate mass within a determined timeS·t range,∙ applying the following equation: 2 where:The fluxes M—permeate from the sourcemass (g phase or mol), [75] S— wereeffective determined surface against of the themembrane measured (m perme-), t—time ate (s)mass necessary within ato determined collect the permeate time range, volume. applying the following equation: The extraction efficiency (EE%) for the species of interest using the concentration of

the solutions [76] was calculated as follows:

𝑐 −𝑐 𝐸𝐸(%) = ∙ 100 (2) 𝑐

where: cf—final concentration of the solute (considered chemical species), co—initial concentration of solute (considered chemical species).

Membranes 2021, 11, 429 7 of 24

where: M—permeate mass (g or mol), S—effective surface of the membrane (m2), t—time (s) necessary to collect the permeate volume. The extraction efﬁciency (EE%) for the species of interest using the concentration of the solutions [76] was calculated as follows: c0 − c f EE(%) = ·100 (2) c0

where: cf—ﬁnal concentration of the solute (considered chemical species), c0—initial concentration of solute (considered chemical species). The same extraction efﬁciency can also be computed based upon the absorbance of the solutions, as in: (A − A ) EE(%) = 0 s ·100 (3) A0

where: A0—initial absorbance of sample solution, As—current absorbance of the sample.

2.4. Equipment The microscopy studies, scanning electron microscope (SEM) and high resolution SEM (HR SEM), were performed using a Hitachi S4500 system (Hitachi High-Technologies Europe GmbH, Mannheim, Germany). Thermal characterizations were performed using a Netzsch Thermal Analyzer (Netzsch—Gerätebau GmbH, Selb, Germany). The thermal analysis TG-DSC for the cellulose samples (~20 mg) was performed with a Netzsch STA 449C Jupiter apparatus. The samples were placed in an open crucible made of alumina and heated with 10 K·min−1 from room temperature up to 900 ◦C, under the ﬂow of 50 mL min−1 dried air. An empty alumina crucible was used as reference. Spectroscopy Bruker Tensor 27 FTIR with a Diamond Attenuated Total Reﬂection— ATR (Bruker) was used to study the interactions between the chemicals used in the membranes developed. FTIR analysis was recorded in the range of 500 to 4000 cm−1. UV-VIS analysis was performed on a Spectrometer CamSpec M550 (Spectronic CamSpec Ltd., Leeds, UK). Other devices used were as follows: ultrasonic bath (Elmasonic S, Elma Schmidbauer GmbH, Singen, Germany), vacuum oven (VIOLA—Shimadzu, Bucharest, Romania).

3. Results and Discussion The separation, concentration and purification of amino acids concerns researchers from various fields of activity: chemistry, biochemistry, engineering, medicine or the environment. Remarkable results were obtained in this field through membrane processes, but further research considers both the recovery of amino acids from poor sources and the creation of physical-chemically resistant biocompatible membranes—especially for sterilization or cleaning in order to reuse or maintain process performance. Obviously, the concern of cellulose derivatives usage for amino acid recovery through membrane processes it is understandable. These derivatives are compatible with the biological environment and biological chemical species, in our case amino acids and their bio-resources. Likewise, the cellulose derivatives have a natural, specific interaction with amino acids, and the chemical modifications of this interaction are very important. The experimental study addressed three amino acids, two of which are essential amino acids. The three amino acids have different molar weights and solubilities, but the side chain, R, is hydrophobic of three types: alkyl (alanine), aromatic (phenylalanine) and sulfur (methionine). The hydrophobic radical of amino acids can interact with the cellulosic membrane material. For this reason the four chosen derivatives have substituents with etheric groups (HEC, MHEC), ester (CA) or carboxylic groups (NaCMC). Membranes 2021, 11, 429 8 of 24

3.1. Membranes and Material Membranes Characterization 3.1.1. Thermal Characteristics Among the physical characteristics of the membranes and membrane materials pro- posed for study the mechanical ones are provided by the support matrix of polypropylene hollow ﬁber. But in the practice of membrane processes, the working temperature as well as the temperature of regeneration-washing or of sterilization is very important. Hence, Membranes 2021, 11, x FOR PEER REVIEWthe need to determine the thermal behavior of both membrane materials and impreg-8 of 25

nated membranes obtained (Figures5 and6, Tables4 and5 as well as the Supplementary Figures S1 and S2).

(a) (b)

Figure 5. 5. ThermalThermal diagrams diagrams for: for: ( (aa)) cellulose cellulose acetate acetate (CA); (CA); (b (b) )2–hydroxyethyl 2–hydroxyethyl cellulose cellulose (HEC); (HEC); ( (cc)) methyl methyl 2–hydroxyethyl– 2–hydroxyethyl– cellulose (MHEC); (d) sodium carboxymethyl–cellulose (NaCMC). cellulose (MHEC); (d) sodium carboxymethyl–cellulose (NaCMC).

TheThe cellulose cellulose acetate acetate test test (Figure (Figure 55a)a) isis thermallythermally stablestable upup toto 275275 °C.◦C. The test loses 2.47%2.47% of of its its initial mass, most most likely due to the water (moisture) absorbed, the process beingbeing accompanied accompanied by by a a weak weak endothermic endothermic effect effect at 64.4 °C.◦C. A A weak exothermic effect is alsoalso observed observed at at 251.8 251.8 °C,◦C, corresponding corresponding to to a a process process of of partial oxidation of the organic molecule.molecule. Degradative Degradative oxidation oxidation starts starts at at275 275 °C,◦ theC, the test test loosing loosing 73.24% 73.24% up to up 370 to °C. 370 The◦C. processThe process is accompanied is accompanied by two by separate two separate exothermic exothermic peaks, peaks, at 340.8 at and 340.8 354.1 and °C, 354.1 corre-◦C, spondingcorresponding to the todegradation the degradation of the acetate of the acetateand cellulose and cellulose group. The group. carbon The mass carbon remain- mass ingremaining after initial after degradation initial degradation is eliminate is eliminate in the inrange the range 370–520 370–520 °C, through◦C, through oxidation, oxidation, the processthe process being being accompanied accompanied by byan anexothermic exothermic intense, intense, wide wide effect, effect, with with two two maxima maxima at 442.1442.1 and and 458.2 458.2 °C.◦C. TheThe 2-hydroxyethyl cellulosecellulose test test (Figure (Figure5b) 5b) loses loses 3.69% 3.69% of its of initial its initial mass inmass the in range the rangeRT-200 RT-200◦C, probably °C, probably water absorbed water absorbed by the hydroxyethyl by the hydroxyethyl cellulose cellulose powder. Thepowder. process The is processaccompanied is accompanied by a weak by endothermic a weak endothermic effect, at 68.3 effect,◦C. at After 68.3 200°C. ◦AfterC the 200 main °C process the main of process of oxidative degradation takes place, the mass loose continuing slowly until 500 °C. During this interval, 71.65% of the initial mass is lost. The process is accompanied by several exothermic effects with maxima at 246.1, 309.5 and 381.3 °C. The methyl 2-hydroxyethyl cellulose test (Figure 5c) looses 4.11% of its initial mass, most likely the absorbed water, the process being accompanied by a weak endothermic effect, with a minimum at 71 °C. The degradative oxidation process took place in the interval 240–350 °C, the test loosing 68.84%. The process is accompanied by two separate, asymmetric, exothermic peaks at 309.5 and 327.8 °C, corresponding to the degradation of cellulose and to the lateral groups, respectively. The remaining carbon mass following the initial degradation is eliminated in the interval 350–560 °C, by oxidation, the process being

Membranes 2021, 11, 429 9 of 24

oxidative degradation takes place, the mass loose continuing slowly until 500 ◦C. During this interval, 71.65% of the initial mass is lost. The process is accompanied by several exothermic effects with maxima at 246.1, 309.5 and 381.3 ◦C. The methyl 2-hydroxyethyl cellulose test (Figure5c) looses 4.11% of its initial mass, most likely the absorbed water, the process being accompanied by a weak endothermic effect, with a minimum at 71 ◦C. The degradative oxidation process took place in the interval 240–350 ◦C, the test loosing 68.84%. The process is accompanied by two separate, asymmetric, exothermic peaks at 309.5 and 327.8 ◦C, corresponding to the degradation of cellulose and to the lateral groups, respectively. The remaining carbon mass following the initial degradation is eliminated in the interval 350–560 ◦C, by oxidation, the process being accompanied by an exothermic, intense, wide, asymmetric effect with the maximum at 499.9 ◦C. The sodium carboxymethyl-cellulose test (Figure5d) looses 9.01% of its initial mass in the interval RT-240 ◦C, probably having a higher water content absorbed by the carboxymethyl- cellulose powder. The process is accompanied by a weak endothermic effect, at 91.4 ◦C. The main oxidative degradation takes place in the range 240–295 ◦C (42.17% of its initial mass is eliminated), the mass loss continuing slowly until 565 ◦C (another 9.76% is still eliminated). The processes are accompanied by exothermic effects with maxima at 284.7 ◦C and 369.6 ◦C. After 565 ◦C, there is a mass loss of 17.22% accompanied by a strong exothermic effect, with a maximum at 581.2 ◦C, most likely due to the decomposition of sodium carbonate and the reaction with the mass of the crucible material (Al2O3).

Table 4. Main thermal characteristics of the used cellulose derivatives.

Mass Loss Mass Loss Sample Water Content Residual Mass 200–400 ◦C 400–800 ◦C CA 2.47% 76.35% 20.31% 1.24% HEC 3.71% 67.92% 23.17% 4.45% MHEC 4.11% 73.63% 19.06% 2.25% NaCMC 9.02% 47.84% 26.28% 17.22%

For impregnated membranes that have provided the best separation results (CA-PPM and MHEC-PPM), the thermal diagrams indicate behaviors close to the support fiber, but with specificities that must be taken into account, especially for operation, regeneration or thermal sterilization (Figure6 and Table5). The sample PPM (Figure6a) is relatively stable up to 180 ◦C, with 0.69% recorded mass loss. At the same time, a small endothermic effect is present on the DSC curve, with onset at 154.9 ◦C and a peak at 165.0 ◦C. This effect corresponds to the melting of the polypropylene. Between 180 and 410 ◦C the PPM fibers suffered an oxidative degradation, the recorded mass loss being 94.06%. The effects on DSC curve are a mix of endothermic (decomposing reactions) and exothermic (partial oxidation) effects, with the latest being more intense than former. In the 410–500 ◦C interval, the oxidation of the carbonaceous mass is achieved with a mass loss of 7.29%. The process is accompanied by a strong exothermic effect, asymmetric, with a maximum at 418.0 ◦C. Membranes 2021, 11, x FOR PEER REVIEW 9 of 25

accompanied by an exothermic, intense, wide, asymmetric effect with the maximum at 499.9 °C. The sodium carboxymethyl-cellulose test (Figure 5d) looses 9.01% of its initial mass in the interval RT-240 °C, probably having a higher water content absorbed by the carboxymethyl-cellulose powder. The process is accompanied by a weak endothermic effect, at 91.4 °C. The main oxidative degradation takes place in the range 240–295 °C (42.17% of its initial mass is eliminated), the mass loss continuing slowly until 565 °C (another 9.76% is still eliminated). The processes are accompanied by exothermic effects with maxima at 284.7 °C and 369.6 °C. After 565 °C, there is a mass loss of 17.22% accompanied by a strong exothermic effect, with a maximum at 581.2 °C, most likely due to the decomposition of sodium carbonate and the reaction with the mass of the crucible material (Al2O3).

Table 4. Main thermal characteristics of the used cellulose derivatives.

Sample Water Content Mass Loss 200–400 °C Mass Loss 400–800 °C Residual Mass CA 2.47% 76.35% 20.31% 1.24% HEC 3.71% 67.92% 23.17% 4.45% MHEC 4.11% 73.63% 19.06% 2.25% NaCMC 9.02% 47.84% 26.28% 17.22%

For impregnated membranes that have provided the best separation results (CA- PPM and MHEC-PPM), the thermal diagrams indicate behaviors close to the support fi- Membranes 2021, 11, 429 ber, but with specificities that must be taken into account, especially for operation, regen-10 of 24 eration or thermal sterilization (Figure 6 and Table 5).

Membranes 2021, 11, x FOR PEER REVIEW 10 of 25

(a) (b)

(e)

Figure 6. ThermalThermal diagrams diagrams for for hollow hollow fiber ﬁber membranes: membranes: (a)( aPPM;) PPM; (b) ( bCA–PPM;) CA–PPM; (c) (MHEC–PPM;c) MHEC–PPM; (d) (HEC–PPM;d) HEC–PPM; (e) NaCMC–PPM. (e) NaCMC–PPM.

The sample CA-PPMPPM (Figure (Figure 6a)6 isb) relatively is stable up stable to 200 up◦ toC, 180 the recorded°C, with mass0.69% loss recorded being mass0.26%. loss. At theAt samethe same time, time, a small a small endothermic endothermi effectc effect is present is present on the on DSC the curve,DSC curve,with onset with onsetat 154.3 at ◦154.9C and °C a peakand ata peak 166.4 at◦C. 165.0 This effect°C. This corresponds effect corresponds to the melting to the of themelting polypropy- of the polypropylene. Between 180 and 410 °C the PPM fibers suffered an oxidative degradation, the recorded mass loss being 94.06%. The effects on DSC curve are a mix of endothermic (decomposing reactions) and exothermic (partial oxidation) effects, with the latest being more intense than former. In the 410–500 °C interval, the oxidation of the carbonaceous mass is achieved with a mass loss of 7.29%. The process is accompanied by a strong exothermic effect, asymmetric, with a maximum at 418.0 °C. The sample CA-PPM (Figure 6b) is stable up to 200 °C, the recorded mass loss being 0.26%. At the same time, a small endothermic effect is present on the DSC curve, with onset at 154.3 °C and a peak at 166.4 °C. This effect corresponds to the melting of the polypropylene. Between 200 and 400 °C the sample presents a mass loss of 93.62%. The effects on DSC curve starts with an exothermic, broad one, with peaks at 240 and 327.5 °C, followed by a series of endothermic peaks. This indicates that first process is an oxidation of the polymeric fiber, followed by a series of decompositions. The carbonaceous mass obtained towards end decomposition starts to burn and a strong, sharp exothermic peak can be observed at 398.8 °C. The final oxidation process takes place quick up to 460 °C,

Membranes 2021, 11, 429 11 of 24

lene. Between 200 and 400 ◦C the sample presents a mass loss of 93.62%. The effects on DSC curve starts with an exothermic, broad one, with peaks at 240 and 327.5 ◦C, followed by a series of endothermic peaks. This indicates that first process is an oxidation of the polymeric fiber, followed by a series of decompositions. The carbonaceous mass obtained towards end decomposition starts to burn and a strong, sharp exothermic peak can be observed at 398.8 ◦C. The final oxidation process takes place quick up to 460 ◦C, with a recorded mass loss of 4.38%, and proceeds slowly after, with another 2.15% mass loss up to 900 ◦C. The sample MHEC-PPM (Figure6c) is stable up to 200 ◦C, the recorded mass loss being 0.26%. At the same time, a small endothermic effect is present on the DSC curve, with onset at 153.8 ◦C and a peak at 165.7 ◦C. This effect corresponds to the melting of the polypropylene. Between 200 and 400 ◦C the sample presents a mass loss of 92.42%. The first effect on DSC curve is exothermic, broad, asymmetric, with peaks at 235 and 295.5 ◦C, and indicates that some oxidation processes are responsible for the mass lost in the first part. There are multiple endothermic effects around 350 ◦C, which indicate the predominance of decomposition processes. The carbonaceous mass obtained by the end of the decomposition starts to burn and a strong, sharp exothermic peak can be observed at 394.2 ◦C. The final oxidation process takes place quickly up to 460 ◦C, with a recorded mass loss of 5.85%, and proceeds slowly after, with another 2.44% mass loss up to 900 ◦C. The sample HEC-PPM (Figure6d) is losing 3.26% of its initial mass up to 200 ◦C, indicating the presence of a small HEC quantity on fiber surface. In the same time, a small endothermic effect is present on the DSC curve, with onset at 156.7 ◦C and a peak at 166.2 ◦C. This effect corresponds to the melting of the polypropylene. Between 200 and 400 ◦C the sample exhibit a mass loss of 74.40%. The effects on DSC curve starts with a series of exothermic peaks at 209.8, 271.7, 343.6 and 393.1 ◦C and 327.5 ◦C indicating the presence of multiple degradative-oxidative processes. The carbonaceous mass obtained towards end decomposition starts to burn and a series of strong, sharp exothermic peaks can be observed at 417.6, 537.1 and 559.1 ◦C. In the final oxidation process, up to 600 ◦C, a mass loss of 10.00% is recorded. The sample NaCMC-PPM (Figure6e) is stable up to 200 ◦C, the recorded mass loss being 0.49%. A small endothermic effect is present on the DSC curve, with onset at 156.3 ◦C and a peak at 166.7 ◦C. This effect corresponds to the melting of the polypropylene fibers. Between 200 and 410 ◦C the sample presents a mass loss of 92.23%. The associated effects on DSC curve are exothermic, overlapped, with peaks at 303.4, 345.8, 374.9 and 404.9 ◦C, indicating the occurring of oxidation processes in this interval. The carbonaceous mass obtained starts to burn after 410 ◦C, when some strong, sharp exothermic peaks can be observed at 426.7 and 448.2 ◦C. In the final oxidation process the sample is losing 4.21% up to 900 ◦C.

Table 5. Main thermal characteristics of the hollow ﬁber membranes.

T PP Melting PP Melting Peak Decomposition Start 10% Sample (Temperature for 10% Onset (◦C) (◦C) (◦C) Mass Loss) (◦C) 1_PP_ﬁber (support) 154.9 165.0 180 258 2 CA-PPM 154.3 166.4 200 274 3 MHEC-PPM 153.8 165.7 200 266 4 HEC-PPM 156.7 166.2 80 246 5 NaCMC-PPM 156.3 166.7 200 281

3.1.2. Chemical Structure and Composition The chemical structure of the chosen cellulosic derivatives was highlighted through Fourier transform infrared spectroscopy (FT-IR), in which the functional groups that can Membranes 2021, 11, 429 12 of 24

interact speciﬁcally with the amino acids considered are observed (Figure7 and Figure S3 Membranes 2021, 11, x FOR PEER REVIEWand Table 6) Their characterization was obtained directly onto the solid samples, using12 theof 25 Bruker Tensor 27 with ATR diamond for the materials in this study.

FigureFigure 7.7. TheThe FourierFourier transform transform infrared infrared spectroscopy spectroscopy (FT–IR) (FT–IR) spectrum spectrum of the of impregnated the impregnated membranes. membranes. Table 6. The main structural characteristics identiﬁed through Fourier transform infrared spec- Table 6. The main structural characteristics identified through Fourier transform infrared spectroscopy (FT–IR). troscopy (FT–IR).

WaveWave Numbers Numbers of the InterestingInteresting Groups Groups for for the the Considered Considered Interactions (cm−1) CelluloseCellulose Interactions (cm−1) Derivatives ν O-C- δ-CH2 derivatives ν O-C- δ-CH2 δ-CH2 ν C=O ν C-H ν O-H δ O-C-O δ O-H δ-CH2 ν C=O ν C-H ν O-H δ O-C-O δ O-H CA 1034 1371 1410 1743 2960 s 3499 CA 1034 1371 1410 1743 2960 s 3499 HECHEC 10571057 1352 1412 1412 - -2987 2987 3374 3374 MHECMHEC 10551055 1371 1420 1420 - - 2896 2896 3422 3422 NaCMCNaCMC 10341034 1319 1418 1418 - - 2940 2940 3280 3280 νν == vibration;vibration;δ =δ deformation.= deformation. The chemical structure of the four cellulosic derivatives which impregnate the mem- The chemical structure of the four cellulosic derivatives which impregnate the membranes indicates possibilities of interaction with the amino acids or their ionic forms, both branes indicates possibilities of interaction with the amino acids or their ionic forms, both by hydrogen or dipole-dipole bonds (given by the carboxyl or amino groups) and by by hydrogen or dipole-dipole bonds (given by the carboxyl or amino groups) and by hy- hydrophobic interactions of hydrocarbon chains. This statement is strengthened by the drophobic interactions of hydrocarbon chains. This statement is strengthened by the char- character of the radical R (Figure1a) of the considered amino acids: aromatic hydrophobic acter of the radical R (Figure 1a) of the considered amino acids: aromatic hydrophobic (phenylalanine) or aliphatic hydrophobic (alanine and methionine). (phenylalanine) or aliphatic hydrophobic (alanine and methionine). Of course, depending on the pH of operation in the separation with the considered Of course, depending on the pH of operation in the separation with the considered membranes, the complexity of the amino acid—cellulose derivative interaction must be takenmembranes, into account. the complexity The superﬁcial of the amino composition acid—cellulose determined derivative through interaction energy dispersive must be spectroscopytaken into account. analysis The (EDAX) superficial of impregnated composition membranes determined Cell-D-PPM through energy compared dispersive with thespectroscopy PPM support analysis membrane (EDAX) shows of impregna the appearanceted membranes in different Cell-D-PPM concentrations compared of oxygen with atomsthe PPM on support the surface membrane (Table7 andshows Supplementary the appearance Figure in different S4). concentrations of oxygen atomsThis on the concentration surface (Table of the7 and oxygen supplementary atoms on Figure the surface S4). of the membrane will be responsible for the attraction of the amino acid, but the migration inside the impregnated membrane depends on the overall composition of the cellulose derivative used for impregnation.

Membranes 2021, 11, x FOR PEER REVIEW 13 of 25

Membranes 2021, 11, 429 Table 7. The surface composition of the prepared impregnated membranes. 13 of 24 Membranes/ PPM CA-PPM HEC-PPM MHEC-PPM NaCMC-PPM Composition Weight% Atomic% Weight% Atomic% Weight% Atomic% Weight% Atomic% Weight% Atomic% C K 93.49 Table95.03 7. The surface68.90 composition 77.87 of the88. prepared26 impregnated90.92 53.74 membranes. 60.74 87.86 90.79 O K 6.51 4.97 31.10 28.23 11.74 9.08 46.26 39.26 11.27 8.74 PPM CA-PPM HEC-PPM MHEC-PPM NaCMC-PPM Membranes/Na K ------0.88 0.47 Composition Weight% Atomic% Weight% Atomic% Weight% Atomic% Weight% Atomic% Weight% Atomic% C K 93.49 95.03This 68.90 concentration 77.87 of the 88.26 oxygen 90.92atoms on the 53.74 surface 60.74of the membrane 87.86 will 90.79 be re- O K 6.51 4.97sponsible 31.10 for the 28.23attraction 11.74of the amino 9.08 acid, but 46.26 the migration 39.26 inside 11.27 the impregnated 8.74 membrane depends on the overall composition of the cellulose derivative used for im- Na K ------0.88 0.47 pregnation.

3.1.3.3.1.3. Membrane Membrane Morphology Morphology TheThe surface surface morphology morphology of of the the considered considered samples samples was was analyzed analyzed using using a scanning a scan- electronning electron microscope microscope (SEM), (SEM), high-resolution high-resolution scanning scanning electron electron microscope microscope (HR-SEM). (HR-SEM). All samplesAll samples werewere properly properly dried dried prior priorto the to microscopy the microscopy analysis analysis and were and sufficiently were sufﬁciently coated withcoated a sputtered with a sputtered gold layer gold of layer 400 Å. of 400 Å. TheThe results results of the image analysis are pres presentedented in Figures 88–11–11 (see(see SupplementarySupplementary FiguresFigures S5–S8). S5–S8).

(a) (b) Membranes 2021, 11, x FOR PEER REVIEW 14 of 25 Figure 8. MorphologyMorphology CA–PPM membranes: ( (aa)) membrane membrane cross–section; cross–section; ( b) membrane surface.

(a) (b)

FigureFigure 9. Morphology 9. Morphology HEC–PPM HEC–PPM membranes: membranes: (a )(a membrane) membrane cross–section; cross–section; ( b(b)) membrane membrane surface.surface.

(a) (b) Figure 10. Morphology MHEC–PPM membrane: (a) membrane cross–section; (b) membrane surface.

(a) (b) Figure 11. Morphology NaCMC–PPM membranes: (a) view; (b) membrane surface.

Membranes 2021, 11, x FOR PEER REVIEW 14 of 25 Membranes 2021, 11, x FOR PEER REVIEW 14 of 25

Membranes 2021, 11, 429 14 of 24 (a) (b) (a) (b) Figure 9. Morphology HEC–PPM membranes: (a) membrane cross–section; (b) membrane surface. Figure 9. Morphology HEC–PPM membranes: (a) membrane cross–section; (b) membrane surface.

(a) (b) (a) (b) Figure 10. MorphologyMorphology MHEC–PPM membrane: ( (aa)) membrane membrane cross–section; cross–section; ( b) membrane surface. Figure 10. Morphology MHEC–PPM membrane: (a) membrane cross–section; (b) membrane surface.

(a) (b) (a) (b) Figure 11. Morphology NaCMC–PPM membranes: (a) view; (b) membrane surface. Figure 11. Morphology NaCMC–PPM membranes: ( a) view; (b) membrane surface.

From the images obtained through scanning electron microscopy (SEM) and high resolution scanning electron microscopy (HR-SEM) some common features of the obtained impregnated membranes can be highlighted, but also several specificities, as follows: • Impregnation of the membranes takes place superficially and adherently, without the cellulosic derivative reaching the inside of propylene hollow fiber membrane (Figures8a,9a and 10a). • The layer of cellulosic derivative from the surface of the membranes has about 5 µm (Supplementary Figures S5–S8a,b) being highlighted in the detail depicted in Figure 12. Membranes 2021, 11, x FOR PEER REVIEW 15 of 25 Membranes 2021, 11, x FOR PEER REVIEW 15 of 25

From the images obtained through scanning electron microscopy (SEM) and high resolutionFrom the scanning images electronobtained microscopy through scanni (HR-SEM)ng electron some microscopy common features (SEM) andof the high ob- resolutiontained impregnated scanning electron membranes microscopy can be highlighted,(HR-SEM) some but also common several features specificities, of the as ob- fol- tainedlows: impregnated membranes can be highlighted, but also several specificities, as follows: • Impregnation of the membranes takes place superficially and adherently, without • Impregnationthe cellulosic of derivative the membranes reaching takes the placinsidee superficially of propylene and hollow adherently, fiber membrane without the(Figure cellulosic 8a–10a). derivative reaching the inside of propylene hollow fiber membrane Membranes 2021, 11, 429 • (FigureThe layer 8a–10a). of cellulosic derivative from the surface of the membranes has about15 of5 µm 24 • The(supplementary layer of cellulosic Figures derivative S5–S8a,b) from being the surfacehighlighted of the in membranes the detail depicted has about in 5Figure µm (supplementary12. Figures S5–S8a,b) being highlighted in the detail depicted in Figure 12.

Figure 12. Detail of the superficial layer from the cellulose derivatives on the polypropylene hol- Figure 12. Detail of the superficial layer from the cellulose derivatives on the polypropylene hollow fiber. Figurelow fiber. 12. Detail of the superficial layer from the cellulose derivatives on the polypropylene hollow fiber. • The surface layer of cellulosic derivative has a microstructure specific to nanofiltration • membranesThe surface (Figures layer of8 b,cellulosic9b, 10b andderivative 11b), highlightedhas a microstructure in two significant specific to details nanofiltra- in • The surface layer of cellulosic derivative has a microstructure specific to nanofiltra- Figuretion membranes 13. (Figure 8b–11b), highlighted in two significant details in Figure 13. tion membranes (Figure 8b–11b), highlighted in two significant details in Figure 13.

(a) (b) (a) (b) Figure 13. Detail of the superficial layer from the cellulose derivatives: (a) cellulose acetate and (b) 2–hydroxyethyl cellulose. Figure 13. Detail of the superficial layer from the cellulose derivatives: (a) cellulose acetate and (b) 2–hydroxyethyl cellulose. Figure 13. Detail of the superﬁcial layer from the cellulose derivatives: (a) cellulose acetate and (b) 2–hydroxyethyl cellulose.

3.2. Performance of the Amino Acids Removal Process

The separation of amino acids is very important, and the use of membranes based on functionalized biopolymers involves the careful study of working parameters and especially of operational ones, such as pH and temperature. That is why the main operating parameters and their inﬂuence on the evolution of the target chemistry species separation are presented. The pH of the feeding solution was equal to 2 and 12, respectively, to have the certainty of the formation of the carboxyl anion, in the ﬁrst case and of the ammonium cation, in the Membranes 2021, 11, x FOR PEER REVIEW 16 of 25

3.2. Performance of the Amino Acids Removal Process The separation of amino acids is very important, and the use of membranes based on functionalized biopolymers involves the careful study of working parameters and especially of operational ones, such as pH and temperature. That is why the main operating parameters and their influence on the evolution of the target chemistry species separation Membranes 2021, 11, 429 are presented. 16 of 24 The pH of the feeding solution was equal to 2 and 12, respectively, to have the certainty of the formation of the carboxyl anion, in the first case and of the ammonium cation, insecond the second case (Figure case (Figure1b and 1b Table and3 Table). The 3). receiving The receiving phase phase consists consists of pure of water pure water (pH = (Ph 7) =so 7) that so that when when re-extracted re-extracted from from the membranethe membrane the the amino amino acid acid reaches reaches the the shape shape speciﬁc spe- cificto the to isoelectricthe isoelectric point point (Figure (Figure 14). 14). The The working working temperature temperature was was 25 25◦ °CC inin thethe whole membranemembrane system.

(a) (b)

FigureFigure 14. 14. TheThe transport transport mechanism mechanism of of the the amino amino acids acids at: at: (a ()a pH) pH = 2 = and 2 and pH pH = 12 = for 12 forthe thesource source phase (SP); and (b) pH = 7 (pure water) in the receiving phase (RP). phase (SP); and (b) pH = 7 (pure water) in the receiving phase (RP).

3.2.1.3.2.1. The The Effect Effect of of the the Composition Composition of th thee Impregnated Membrane on the Recovery of Amino Acids TheThe first ﬁrst pertraction pertraction tests tests of of the the chosen chosen amino amino acids, acids, alanine, alanine, phenylalanine phenylalanine and andme- thionine,methionine, considered considered the study the study of the of evolution the evolution of amino of amino acids acidsin source in source phase, phase, in a deter- in a mineddetermined interval interval of 120 of min 120 (Figures min (Figures 15 and 15 16). and 16). TheThe concentration of of the the three amino acids decreases rapidly rapidly in in the first ﬁrst 60 min of operation,operation, after after which which the the decrease decrease is slower, is slower, or a or level a level is created is created (Figures (Figures 15 and 15 16).and The 16). lowestThe lowest decreasedecrease raterate is in is all in cases all cases for foralan alanine,ine, followed followed by by methionine, methionine, and and phenylalanine. phenylalanine. However,However, some peculiarities stood out: •• The difference in transfer rate was larg largee between alanine and the other two amino acids, especially forfor HEC-PPM,HEC-PPM, MHEC-PPM MHEC-PPM and and NaCMC-PPM NaCMC-PPM membranes, membranes, at pHat pH = 12= and for CA-PPM and MHEC-PPM membranes, at pH = 2, Membranes 2021, 11, x FOR PEER REVIEW 12 and for CA-PPM and MHEC-PPM membranes, at pH = 2, 17 of 25 • • The transfer rates differed less in the case of CA-PPM at pH = 12 and for HEC-PPM and NaCMC at pH = 2. and NaCMC at pH = 2.

Figure 15. Cont.

Figure 15. Variation of source phase amino acid concentration depending on the operating time at pHSP = 12 and pHRP = 7: (a) CA–PPM; (b) HEC–PPM; (c) MHEC–PPM; (d) NaCMC–PPM. A = alanine (blue); M = methionine (green) and PA = phenylalanine (brown).

(a) (b)

Membranes 2021, 11, 0 17 of 25 Membranes 2021, 11, x FOR PEER REVIEW 17 of 25 MembranesMembranes 20212021,, 1111,, xx FORFOR PEERPEER REVIEWREVIEW 1717 ofof 2525

Membranes 2021, 11, 429 17 of 24

FigureVariation 15. 15. VariationVariation of source of ofsource source phase phase aminophase amin aminoacido acid concentration acid concentration concentration depending depending depending on on the the operating on operating the operating time time at at pH time pHSP at = 12 pH and andSP =pHpH 12RP and= =7: 7:( ) FigureFigure 15. 15. Variation of source phase amino acid concentration depending on the operating time at pHSPSP = 12 and pHRP = 7: a (Figurea) CA–PPM; 15. Variation (b) HEC–PPM; of source (phasec) MHEC–PPM; amino acid ( dconcentration) NaCMC–PPM. depending A = alanine on the (blue); operating M = timemethionine at pHSP (green)= 12 and and pH RPPA = 7:= CA–PPM;pH(a) RPCA–PPM; (b=) 7 HEC–PPM;:(a) CA–PPM;(b) HEC–PPM; (c) MHEC–PPM;(b) HEC–PPM; (c) MHEC–PPM; (d (c)) NaCMC–PPM. MHEC–PPM; (d) NaCMC–PPM. (dA) NaCMC–PPM.= alanine A = (blue); alanineA =M alanine(blue);= methionine (blue);M = methionineM (green)= methionine and (green)PA (green)= phenylalanineand PA and = phenylalanine(a) CA–PPM; ((brown).b) HEC–PPM; (c) MHEC–PPM; (d) NaCMC–PPM. A = alanine (blue); M = methionine (green) and PA = (brown).PAphenylalaninephenylalanine= phenylalanine (brown).(brown). (brown).

Membranes 2021, 11, x FOR PEER REVIEW 18 of 25

(a) (b) ((aa)) ((bb))

Figure 16. Cont.

Figure 16. Variation of source phase amino acid concentration depending on the operating time at pHSP = 2 and pHRP = 7: Figure 16. Variation of source phase amino acid concentration depending on the operating time at pHSP = 2 and pHRP = 7: (a()a) CA–PPM; CA–PPM; (b (b) HEC–PPM;) HEC–PPM; (c ()c MHEC–PPM;) MHEC–PPM; and and (d ()d NaCMC–PPM.) NaCMC–PPM.A =A alanine= alanine (blue); (blue);M M= methionine= methionine (green), (green), and andPA PA= = phenylalanine (brown). phenylalanine (brown).

ConsideringConsidering thethe amino acid acid flows flows (Table (Table 8),8), depending depending on on the the nature nature of of cellulose cellulose de- derivativerivative that that forms forms the the impregnated impregnated membrane membrane and and the the surface surface composition composition (Table (Table 7),7), it itcan can be be seen seen that that there there were were significant significant differ differencesences in in the the flow flow of of the the three three amino amino acids. acids.

Table 8. The fluxes of the cellulose derivatives membranes at pH = 2 and pH = 12 of the source phase (A = alanine; M = methionine, and PA = phenylalanine).

Membrane Flux (µMol/m2·s) pH Source Phase CA-PPM HEC-PPM MHEC-PPM NaCMC-PPM A M PA A M PA A M PA A M PA 2 83.3 186.1 241.6 180.5 222.5 244.5 69.4 208.5 227.8 194.4 229.0 230.8 12 236.1 277.0 300.4 83.2 235.6 305.4 111.4 280.5 305.5 97.2 236.1 278.0

The results obtained show that phenylalanine has the highest fluxes through all four types of membranes, followed by methionine and alanine. pH does not change this order but suggests the possibility of selective separation of the three amino acids, more obvious being the separation of alanine from phenylalanine. Of the four kinds of membrane, the most suitable for recuperative separation of the considered amino acids are those based on cellulose acetate and methyl 2-hydroxyethyl- cellulose.

3.2.2. The Effect of pH and Solubility of Amino Acids on the Flow through Impregnated Membranes The flow performances were also found in the efficiency of phenylalanine extraction, chosen for the highest transfer rates found in all previous experiments for all four types of membrane (Figure 17). Depending on the pH of the source phase, the following succession was obtained:

𝐸𝐸 = 𝐸𝐸 >𝐸𝐸 (4)

Membranes 2021, 11, 429 18 of 24

Table 8. The ﬂuxes of the cellulose derivatives membranes at pH = 2 and pH = 12 of the source phase (A = alanine; M = methionine, and PA = phenylalanine).

Membrane Flux (µMol/m2·s) pH Source CA-PPM HEC-PPM MHEC-PPM NaCMC-PPM Phase A M PA A M PA A M PA AM PA 2 83.3 186.1 241.6 180.5 222.5 244.5 69.4 208.5 227.8 194.4 229.0 230.8 12 236.1 277.0 300.4 83.2 235.6 305.4 111.4 280.5 305.5 97.2 236.1 278.0

The results obtained show that phenylalanine has the highest ﬂuxes through all four types of membranes, followed by methionine and alanine. pH does not change this order but suggests the possibility of selective separation of the three amino acids, more obvious being the separation of alanine from phenylalanine. Of the four kinds of membrane, the most suitable for recuperative separation of the considered amino acids are those based on cellulose acetate and methyl 2-hydroxyethyl- cellulose.

3.2.2. The Effect of pH and Solubility of Amino Acids on the Flow through Impregnated Membranes The ﬂow performances were also found in the efﬁciency of phenylalanine extraction, chosen for the highest transfer rates found in all previous experiments for all four types of membrane (Figure 17). Depending on the pH of the source phase, the following succession Membranes 2021, 11, x FOR PEER REVIEWwas obtained: 19 of 25

EEpH 12 = EEpH 2 > EEpH 7 (4)

FigureFigure 17.17. TheThe phenylalaninephenylalanine extraction extraction efﬁciency efficiency EE EE (%) (%) variation variation versus versus the the pH pH of theof the source source phase (SP)phase for (SP) the for prepared the prepared membranes. membranes.

DeterminationDetermination ofof thethe efficiencyefficiency ofof extractionextractionof ofthe the aminoamino acidsacids fromfrom sourcesourcesolutions solutions ◦ ofof pH = 12, 12, at at temperatures temperatures of of35, 35, 45 and 45 and 55 °C, 55 showsC, shows that the that separation the separation efficiency efficiency grows growswith increasing with increasing temperature temperature for all for four all types four typesof membranes of membranes (Figure (Figure 18). A18 ).more A more pro- pronouncednounced effect effect can canbe seen be seen in the in case the caseof HEC-PPM of HEC-PPM and NaCMC-PPM and NaCMC-PPM membranes, membranes, which whichcould couldcorrelate correlate with higher with higher hydrophilicity hydrophilicity of membranes of membranes based based on the on thetwo two cellulosic cellulosic de- derivatives.rivatives. However, However, the theusage usage of these of these impreg impregnationnation derivatives derivatives is limited is limited by the solubil- by the solubilityity of the polymeric of the polymeric layer in layerthe source in the solution, source solution, which would which lead would over leadtime overto significant time to significantlosses and lossesdrastic and decreases drastic in decreases membrane in membrane thickness, thickness,known in knownthe literature in the literatureof the mem- of thebranes membranes as ‘membrane as ‘membrane washing’ washing’ [77]. [77].

Figure 18. The phenylalanine extraction efficiency EE (%) variation versus the temperature of source phase (SP) for the prepared membranes.

The transfer rate of alanine is lower in all cases, remaining in high concentration in the source phase. This observation suggests the possibility of separating the three chosen amino acids by the appropriate choice of the cellulosic derivative. Thus, phenylalanine and methionine can be recovered from the receiving phase and alanine is concentrated in the source phase. One explanation that would answer to the results presented in Figures 13–15a would be that the solubility of alanine is much higher than that of the other two amino acids (Table 3), which would lead to an important retention of it in the source phase.

Membranes 2021, 11, x FOR PEER REVIEW 19 of 25

Figure 17. The phenylalanine extraction efficiency EE (%) variation versus the pH of the source phase (SP) for the prepared membranes.

Determination of the efficiency of extraction of the amino acids from source solutions of pH = 12, at temperatures of 35, 45 and 55 °C, shows that the separation efficiency grows with increasing temperature for all four types of membranes (Figure 18). A more pronounced effect can be seen in the case of HEC-PPM and NaCMC-PPM membranes, which could correlate with higher hydrophilicity of membranes based on the two cellulosic derivatives. However, the usage of these impregnation derivatives is limited by the solubil- Membranes 2021, 11, 429 ity of the polymeric layer in the source solution, which would lead over time to significant19 of 24 losses and drastic decreases in membrane thickness, known in the literature of the membranes as ‘membrane washing’ [77].

Figure 18. The phenylalanine extraction efficiency EE (%) variation versus the temperature of Figure 18. The phenylalanine extraction efﬁciency EE (%) variation versus the temperature of source source phase (SP) for the prepared membranes. phase (SP) for the prepared membranes.

The transfer rate of alanine is lowerlower inin allall cases,cases, remainingremaining inin highhigh concentrationconcentration inin the source phase.phase. This observation suggests the possibility ofof separatingseparating thethe threethree chosenchosen amino acidsacids byby the the appropriate appropriate choice choice of of the the cellulosic cellulosic derivative. derivative. Thus, Thus, phenylalanine phenylalanine and methionineand methionine can becan recovered be recovered from from the receivingthe receiving phase phase and and alanine alanine is concentrated is concentrated in the in sourcethe source phase. phase. One explanation that would answer to the results presented inin FiguresFigures 1313–15a, 14 and would 15a wouldbe that bethe that solubility the solubility of alanine of alanine is much is muchhigher higher than that than of that the ofother the othertwo amino two amino acids acids(Table (Table 3), which3), which would would lead leadto an to important an important retention retention of it in of the it in source the source phase. phase. The argument presented is even more valid as the transfer of phenylalanine is in all cases higher than that of methionine. Thus, it could be generalized that that the transfer ﬂow (J) of amino acids through the studied membranes grows in this order:

Jphenylalanine > Jmethionine > Jalanine (5)

which would thus be correlated with the solubility in pure water (So) of the amino acids:

Sophenylalanine < Somethionine < Soalanine (6)

The results can be explained by the fact that in aqueous solution the amino acids participate in proton exchange equilibria which are controlled by pH:

+ + HOOC-CHR-NH2 + H3O ↔ HOOC-CHR-NH3 + HOH (7)

− + HOOC-CHR-NH2 + HOH ↔ OOC-CHR-NH2 + H3O (8) Based on these balances, the acidity constants are deﬁned:

[HOOC − CHR − NH ][H O+] = 2 3 Ka1 + (9) HOOC − CHR − NH3 + [OOC − CHR − NH2 ][H3O ] Ka2 = (10) [HOOC − CHR − NH2] The degree of formation of these chemical species can be assessed with the following relationships: 1 α = 0 pK −pH pK +pK −2pH (11) 1 + 10 a2 + 10 a1 a2 Membranes 2021, 11, 429 20 of 24

1 α = 1 pK −pH pH−pK (12) 1 + 10 a1 + 10 a2 1 = α2 pH−pK 2pH−pK −pK (13) 1 + 10 a1 + 10 a1 a2

In which α0, α1 and α2 represent, respectively, the degrees of formation of the species – + OOC-CHR-NH2, HOOC-CHR-NH2 and HOOC-CHR-NH3 , in the aminophenol solution. The actual solubility (S) of the considered amino acid is a function of pH and the solubility in pure water (S0), which explains the behavior of the amino acid in the pertraction process:

S = S0 · f (α) (14)

Membranes 2021, 11, x FOR PEER REVIEW With these relations one can illustrate the speciation diagram for a generic amino21 acidof 25

with: pKa1 = 2.30 and pKa2 = 9.20 (Figure 19a):

(a) (b) Figure 19. The speciation diagram of an amino acid, as a function on pH: (a) diagram; and (b) the diagrams superimposed Figure 19. The speciation diagram of an amino acid, as a function on pH: (a) diagram; and (b) the diagrams superimposed over the extraction efficiency graph from Figure 17. over the extraction efﬁciency graph from Figure 17.

The results obtained in this paper areare wellwell illustratedillustrated byby overlappingoverlapping thethe diagramdiagram inin Figure 1717 with with that that of of the the chemical chemical speciation speciation of aminoof amino acids acids as function as function of pH of ( FigurepH (Figure 19a). Thus,19a). Thus, it can it be can observed be observed (Figure (Figure 19b) that19b) ifthat the if pH the of pH the of source the source phase phase varies varies from from 2 to 72 to and 7 and then then to 12, to 12, while while keeping keeping constantly constantly the the pH pH of of the the receiving receiving phase phase at at 7, 7, thenthen thethe extraction efﬁciencyefficiency follows a winding curve that is speciﬁcspecific to thethe existenceexistence ofof thethe threethree forms ofof thethe aminoamino acid.acid. Based on the analysis of the speciation diagram, thethe pHpH conditionsconditions forfor thethe formationformation of the differentdifferent chemicalchemical species of thethe studiedstudied aminoamino acidsacids cancan bebe anticipated,anticipated, soso asas toto obtain the optimal results inin separation.separation.

4. Conclusions Amino acidsacids areare substances substances whose whose chemical chemical and and biochemical biochemical impact impact are quiteare quite remark- re- able.markable. A special A special place place in the in study the study of the of amino the amino acids acids is occupied is occupied by their by recoverytheir recovery from variousfrom various sources, sources, through through a multitude a multitude of separation of separation methods methods and techniquesand techniques in which in which the membranesthe membranes occupy occupy a special a special place. place. The presentpresent paperpaper addressed addressed the the recuperative recuperative separation separation of threeof three amino amino acids acids (alanine, (ala- phenylalaninenine, phenylalanine and methionine)and methionine) from from synthetic synthetic solutions, solutions, using using membranes membranes from from cellu- cel- loselulose derivatives derivatives (cellulose (cellulose acetate, acetate, 2-hydroxyethyl-cellulose 2-hydroxyethyl-cellulose and methyl and 2-hydroxylethyl- methyl 2-hy- cellulosedroxylethyl-cellulose or sodium carboxymethyl-cellulose) or sodium carboxymethyl-cellulose) in the polypropylene in the polypropylene hollow ﬁber hollow matrix. fiber matrix.In order to determine the basic characteristics, the separation performances of the consideredIn order amino to determine acids (retention, the basic ﬂow characteristics, and selectivity) the and separation the morphological performances and of struc- the considered amino acids (retention, flow and selectivity) and the morphological and structural ones, the membranes and membrane materials were analyzed by specific techniques: scanning electron microscopy (SEM), high resolution SEM (HR-SEM), Fourier transform infrared spectroscopy (FT-IR), energy dispersive spectroscopy (EDS) and thermal-gravimetric analyzer (TGA). The support offered by polypropylene hollow fibers confers physical-chemical re- sistance, and cellulosic derivatives used for impregnation contribute to the separation performances of prepared impregnated membranes. The results of the separation were influenced by the pH of the source phase and the solubility of the amino acids considered, the best results as flow and extraction efficiency being obtained when the source phase had a pronounced acidic or basic pH and the receiving phase had a neutral pH (pure water). Among the amino acids, the highest transmembrane fluxes and extraction efficiency are offered by phenylalanine, then methionine and finally alanine. This succession is

Membranes 2021, 11, 429 21 of 24

tural ones, the membranes and membrane materials were analyzed by specific techniques: scanning electron microscopy (SEM), high resolution SEM (HR-SEM), Fourier transform infrared spectroscopy (FT-IR), energy dispersive spectroscopy (EDS) and thermal-gravimetric analyzer (TGA). The support offered by polypropylene hollow fibers confers physical-chemical re- sistance, and cellulosic derivatives used for impregnation contribute to the separation performances of prepared impregnated membranes. The results of the separation were influenced by the pH of the source phase and the solubility of the amino acids considered, the best results as flow and extraction efficiency being obtained when the source phase had a pronounced acidic or basic pH and the receiving phase had a neutral pH (pure water). Among the amino acids, the highest transmembrane fluxes and extraction efficiency are offered by phenylalanine, then methionine and finally alanine. This succession is closely correlated with the solubility of the amino acids studied in water, a high solubility in water leading to a low transfer rate and separation efficiency. By choosing the right cellulose derivative, phenylalanine and methionine can be recovered from the receiving phase, and alanine is concentrated in the source phase. The considered cellulosic derivatives (cellulose acetate, 2-hydroxyethyl-cellulose, methyl 2-hydroxyethyl-cellulose or sodium carboxymethyl-cellulose) have a similar behavior in the separation process, but following the experiments, cellulose acetate and methyl 2-hydroxyethyl-cellulose are recommended for the realization of impregnated membranes. Although when increasing the operating temperature from 25 to 55 ◦C the performance of membranes impregnated with 2-hydroxyethyl-cellulose and sodium carboxymethyl- cellulose increased, still the use of these membranes at high temperature raised the problem of degradation by the solubilization-washing phenomenon.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/membranes11060429/s1, Figure S1 and S2: The superposed thermal diagrams; Figure S3: FTIR spectra of the obtained membranes; Figure S4: The compositional surface (EDAX) of the support membrane and two impregnated membrane; Figures S5-S8: The results of the image analysis (SEM). Author Contributions: Conceptualization, A.C.N., A.R.G., S.G.B. and G.N.; methodology, S.G.B., P.C.A., A.R.G., F.D. and A.C.N.; validation, G.N., O.O. and A.R.G.; formal analysis, A.P., I.A.D., D.P. and A.C.N.; investigation, A.C.N., G.N., S.G.B., O.O., F.D., D.P., P.C.A., I.A.D. and A.R.G.; resources, G.N., S.G.B. and A.R.G.; data curation, A.P., I.A.D.; writing—original draft preparation, A.C.N., G.N., S.G.B. and A.R.G.; writing—review and editing, A.R.G.; supervision, G.N., S.G.B. and A.C.N. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Acknowledgments: The authors gratefully acknowledge the valuable help and friendly assistance of Eng. Roxana Tru¸scă for performing the scanning microscopy analysis. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References 1. Bolasco, P.; Caria, S.; Cupisti, A.; Secci, R.; Dioguardi, F.S. A novel amino acids oral supplementation in hemodialysis patients: A pilot study. Ren. Fail. 2011, 33, 1–5. [CrossRef] 2. Dabadé, D.S.; Jacxsens, L.; Miclotte, L.; Abatih, E.; Devlieghere, F.; De Meulenaer, B. Survey of multiple biogenic amines and correlation to microbiological quality and free amino acids in foods. Food Control. 2021, 120.[CrossRef] 3. Brancatelli, G.; Dalcanale, E.; Pinalli, R.; Geremia, S. Probing the structural determinants of amino acid recognition: X-ray studies of crystalline ditopic host-guest complexes of the positively charged amino acids, Arg, Lys, and His with a cavitand molecule. Molecules 2018, 23, 3368. [CrossRef] Membranes 2021, 11, 429 22 of 24

4. Martins, J.N.; Lima, J.C.; Basílio, N. Selective recognition of amino acids and peptides by small supramolecular receptors. Molecules 2020, 26, 106. [CrossRef][PubMed] 5. Sandoval-Lira, J.; Mondragón-Solórzano, G.; Lugo-Fuentes, L.I.; Barroso-Flores, J. Accurate estimation of pKb values for amino groups from surface electrostatic potential (VS,min) calculations: The isoelectric points of amino acids as a case study. J. Chem. Inf. Model. 2020, 60, 1445–1452. [CrossRef][PubMed] 6. Sahraei, A.A.; Larachi, F. Chemical transformation and dissociation of amino acids on metal sulfide surface: Insights from DFT into the effect of surface vacancies on alanine-sphalerite system. Appl. Surf. Sci. 2021, 540, 148304. [CrossRef] 7. Cordes, M.; Koettgen, A.; Jasper, C.; Jacques, O.; Boudebous, H.; Giese, B. Influence of amino acid side chains on long-distance electron transfer in peptides: Electron hopping via “stepping stones.” Angew. Chem. Int. Ed. 2008, 47, 3461–3463. [CrossRef] 8. Stückenschneider, K.; Merz, J.; Hanke, F.; Rozyczko, P.; Milman, V.; Schembecker, G. Amino-acid adsorption in MFI-type zeolites enabled by the pH-dependent ability to displace water. J. Phys. Chem. C 2013, 117, 18927–18935. [CrossRef] 9. Rimola, A.; Sodupe, M.; Ugliengo, P. Affinity scale for the interaction of amino acids with silica surfaces. J. Phys. Chem. C 2009, 113, 5741–5750. [CrossRef] 10. Singla, P.; Riyaz, M.; Singhal, S.; Goel, N. Theoretical study of adsorption of amino acids on graphene and BN sheet in gas and aqueous phase with empirical DFT dispersion N. correction. Phys. Chem. Chem. Phys. 2016, 18, 5597–5604. [CrossRef][PubMed] 11. Costa, D.; Savio, L.; Pradier, C.-M. Adsorption of amino acids and peptides on metal and oxide surfaces in water environment: A synthetic and prospective review. J. Phys. Chem. B 2016, 120, 7039–7052. [CrossRef] 12. Wendisch, V.F. Metabolic engineering advances and prospects for amino acid production. Metab. Eng. 2020, 58, 17–34. [CrossRef] 13. Hermann, T. Industrial production of amino acids by coryneform bacteria. J. Biotechnol. 2003, 104, 155–172. [CrossRef] 14. Aghajanyan, A.E.; Hambardzumyan, A.A.; Vardanyan, A.A.; Saghiyan, A.S. Desalting of neutral amino acids fermentative solutions by electrodialysis with ion-exchange membranes. Desalination 2008, 228, 237–244. [CrossRef] 15. Readi, O.K.; Mengers, H.; Wiratha, W.; Wessling, M.; Nijmeijer, K. On the isolation of single acidic amino acids for biorefinery applications using electrodialysis. J. Membr. Sci. 2011, 384, 166–175. [CrossRef] 16. Readi, O.K.; Gironès, M.; Wiratha, W.; Nijmeijer, K. On the isolation of single basic amino acids with electrodialysis for the production of biobased chemicals. Ind. Eng. Chem. Res. 2012, 52, 1069–1078. [CrossRef] 17. Mangano, V.; Gervasi, T.; Rotondo, A.; De Pasquale, P.; Dugo, G.; Macrì, F.; Salvo, A. Protein hydrolysates from anchovy waste: Purification and chemical characterization. Nat. Prod. Res. 2021, 35, 399–406. [CrossRef][PubMed] 18. Chen, C.W.; Whitlock, H.W. Molecular tweezers: A simple model of bifunctional intercalation. J. Am. Chem. Soc. 1978, 100, 4921–4922. [CrossRef] 19. Yemm, E.W.; Cocking, E.C.; Ricketts, R.E. The determination of amino-acids with ninhydrin. Analyst 1955, 80, 209–214. [CrossRef] 20. Ginn, H.E.; Frost, A.; Lacy, W.W. Nitrogen Balance in Hemodialysis Patients. Am. J. Clin. Nutr. 1968, 21, 385–393. [CrossRef] 21. Bolasco, P. Hemodialysis—Nutritional Flaws in Diagnosis and Prescriptions. Could Amino Acid Losses Be the Sharpest “Sword of Damocles”? Nutrients 2020, 12, 1773. [CrossRef][PubMed] 22. Hendriks, F.K.; Kooman, J.P.; Van Loon, L.J. Dietary protein interventions to improve nutritional status in end-stage renal disease patients undergoing hemodialysis. Curr. Opin. Clin. Nutr. Metab. Care 2021, 24, 79–87. [CrossRef][PubMed] 23. Martinez, D.; Sandeaux, R.; Sandeaux, J.; Gavach, C. Electrotransport of alanine through ion-exchange membranes. J. Membr. Sci. 1992, 69, 273–281. [CrossRef] 24. Eliseeva, T.V.; Shaposhnik, V.A.; Krisilova, E.V.; Bukhovets, A.E. Transport of basic amino acids through the ion-exchange membranes and their recovery by electrodialysis. Desalination 2009, 241, 86–90. [CrossRef] 25. Sandeaux, J.; Fares, A.; Sandeaux, R.; Gavach, C. Transport properties of electrodialysis membranes in the presence of arginine I. Equilibrium properties of a cation exchange membrane in an aqueous solution of arginine chlorhydrate and sodium chloride. J. Membr. Sci. 1994, 89, 73–81. [CrossRef] 26. Teng, Y.; Sanders, J.P.M.; Scott, E.L.; Van Zeeland, A.N.T. The use of l-lysine decarboxylase as a means to separate amino acids by electrodialysis. Green Chem. 2011, 13, 624–630. [CrossRef] 27. Fares, A.; Sandeaux, J.; Sandeaux, R.; Gavach, C. Transport properties of electrodialysis membranes in the presence of arginine II. Competition between the electrotransport of organic and inorganic ions through a cation exchange membrane in an aqueous solution of arginine chlorhydrate and sodium chloride. J. Membr. Sci. 1994, 89, 83–91. [CrossRef] 28. Firdaous, L.; Dhulster, P.; Amiot, J.; Gaudreau, A.; Lecouturier, D.; Kapel, R.; Lutin, F.; Vézina, L.-P.; Bazinet, L. Concentration and selective separation of bioactive peptides from an alfalfa white protein hydrolysate by electrodialysis with ultrafiltration membranes. J. Membr. Sci. 2009, 329, 60–67. [CrossRef] 29. Lentsch, S.; Aimar, P.; Orozco, J.L. Enhanced separation of albumin-poly(ethylene glycol) by combination of ultrafiltration and electrophoresis. J. Membr. Sci. 1993, 80, 221–232. [CrossRef] 30. Bargeman, G.; Houwing, J.; Recio, I.; Koops, G.-H.; Van Der Horst, C. Electro-membrane filtration for the selective isolation of bioactive peptides from an ?s2-casein hydrolysate. Biotechnol. Bioeng. 2002, 80, 599–609. [CrossRef] 31. Daufin, G.; Kerhervé, F.L.; Aimar, P.; Mollé, D.; Léonil, J.; Nau, F. Electrofiltration of solutions of amino acids or peptides. Le Lait 1995, 75, 105–115. [CrossRef] 32. Bargeman, G.; Dohmen-Speelmans, M.; Recio, I.; Timmer, M.; Van Der Horst, C. Selective isolation of cationic amino acids and peptides by electro-membrane filtration. Le Lait 2000, 80, 175–185. [CrossRef] Membranes 2021, 11, 429 23 of 24

33. Bargeman, G.; Koops, G.-H.; Houwing, J.; Breebaart, I.; Van Der Horst, H.; Wessling, M. The development of electro-membrane filtration for the isolation of bioactive peptides: The effect of membrane selection and operating parameters on the transport rate. Desalination 2002, 149, 369–374. [CrossRef] 34. Kimura, S.; Tamano, A. Separation of Aminoacids by Charged Ultrafiltration Membranes. In Membranes and Membrane Processes; Drioli, E., Nakagaki, M., Eds.; Springer Science and Business Media LLC: Berlin, Germany, 1986; pp. 191–197. 35. Higuchi, A.; Hara, M.; Horiuchi, T.; Nakagawa, T. Optical resolution of amino acids by ultrafiltration membranes containing serum albumin. J. Membr. Sci. 1994, 93, 157–164. [CrossRef] 36. Tone, S.; Masawaki, T.; Hamada, T. The optical resolution of amino acids by ultrafiltration membranes fixed with plasma polymerized l-menthol. J. Membr. Sci. 1995, 103, 57–63. [CrossRef] 37. Nakao, S.; Osada, H.; Kurata, H.; Tsuru, T.; Kimura, S. Separation of proteins by charged ultrafiltration membranes. Desalination 1988, 70, 191–205. [CrossRef] 38. Gotoh, T.; Iguchi, H.; Kikuchi, K.-I. Separation of glutathione and its related amino acids by nanofiltration. Biochem. Eng. J. 2004, 19, 165–170. [CrossRef] 39. Timmer JM, K.; Speelmans MP, J.; Van Der Horst, H.C. Separation of amino acids by nanofiltration and ultrafiltration membranes. Sep. Purif. Technol. 1998, 14, 133–144. [CrossRef] 40. Garem, A.; Daufin, G.; Maubois, J.L.; Leonil, J. Selective separation of amino acids with a charged inorganic nanofiltration membrane: Effect of physicochemical parameters on selectivity. Biotechnol. Bioeng. 1996, 54, 291–302. [CrossRef] 41. Hong, S.U.; Bruening, M.L. Separation of amino acid mixtures using multilayer polyelectrolyte nanofiltration membranes. J. Membr. Sci. 2006, 280, 1–5. [CrossRef] 42. Vyas, B.; Ray, P. Preparation of nanofiltration membranes and relating surface chemistry with potential and topography: Application in separation and desalting of amino acids. Desalination 2015, 362, 104–116. [CrossRef] 43. Wang, X.-L.; Ying, A.-L.; Wang, W.-N. Nanofiltration of l-phenylalanine and l-aspartic acid aqueous solutions. J. Membr. Sci. 2002, 196, 59–67. [CrossRef] 44. Yoshikawa, M.; Izumi, J.-I.; Kitao, A.T.; Sakamoto, S. Molecularly imprinted polymeric membranes containing DIDE derivatives for optical resolution of amino acids. Macromolecules 1996, 29, 8197–8203. [CrossRef] 45. Lin, S.-H.; Chen, C.-N.; Juang, R.-S. Extraction equilibria and separation of phenylalanine and aspartic acid from water with di(2-ethylhexyl)phosphoric acid. J. Chem. Technol. Biotechnol. 2006, 81, 406–412. [CrossRef] 46. Capela, E.V.; Quental, M.V.; Domingues, P.; Coutinho, J.A.P.; Freire, M.G. Effective separation of aromatic and aliphatic amino acid mixtures using ionic-liquid-based aqueous biphasic systems. Green Chem. 2017, 19, 1850–1854. [CrossRef][PubMed] 47. Fukumoto, K.; Yoshizawa, M.; Ohno, H. Room temperature ionic liquids from 20 natural amino acids. J. Am. Chem. Soc. 2005, 127, 2398–2399. [CrossRef] 48. Esfahani, M.R.; Aktij, S.A.; Dabaghian, Z.; Firouzjaei, M.D.; Rahimpour, A.; Eke, J.; Escobar, I.C.; Abolhassani, M.; Greenlee, L.F.; Esfahani, A.R.; et al. Nanocomposite membranes for water separation and purification: Fabrication, modification, and applications. Sep. Purif. Technol. 2019, 213, 465–499. [CrossRef] 49. Yue, X.; Li, Z.; Zhang, T.; Yang, D.; Qiu, F. Design and fabrication of superwetting fiber-based membranes for oil/water separation applications. Chem. Eng. J. 2019, 364, 292–309. [CrossRef] 50. Wang, G.; Zhang, C.; Sun, M.; Zhang, X.; Wu, C.; Wu, Y. Separation of mixed amino acids by BMED process using porous SPES and SPSf cation exchange membranes. Sep. Purif. Technol. 2017, 188, 539–547. [CrossRef] 51. Chen, S.; Lv, C.; Hao, K.; Jin, L.; Xie, Y.; Zhao, W.; Sun, S.; Zhang, X.; Zhao, C. Multifunctional negatively-charged poly (ether sulfone) nanofibrous membrane for water remediation. J. Colloid Interface Sci. 2019, 538, 648–659. [CrossRef] 52. Ingole, P.G.; Bajaj, H.C.; Singh, K. Membrane separation processes: Optical resolution of lysine and asparagine amino acids. Desalination 2014, 343, 75–81. [CrossRef] 53. Ji, Y.-L.; Qian, W.-J.; An, Q.-F.; Lee, K.-R.; Gao, C.-J. Polyelectrolyte nanoparticles based thin-film nanocomposite (TFN) membranes for amino acids separation. J. Ind. Eng. Chem. 2018, 66, 209–220. [CrossRef] 54. Stepniak, P.; Lainer, B.; Chmurski, K.; Jurczak, J. pH-Controlled recognition of amino acids by urea derivatives of β-cyclodextrin. RSC Adv. 2017, 7, 15742–15746. [CrossRef] 55. Liu, L.; Yang, L.-L.; Jin, K.-Y.; Xu, D.-Q.; Gao, C.-J. Recovery of l-tryptophan from crystallization wastewater by combined membrane process. Sep. Purif. Technol. 2009, 66, 443–449. [CrossRef] 56. Bukhovets, A.; Savel’Eva, A.; Eliseeva, T. Separation of amino acids mixtures containing tyrosine in electromembrane system. Desalination 2009, 241, 68–74. [CrossRef] 57. Lammens, T.M.; De Biase, D.; Franssen, M.C.R.; Scott, E.L.; Sanders, J.P.M. The application of glutamic acid α-decarboxylase for the valorization of glutamic acid. Green Chem. 2009, 11, 1562–1567. [CrossRef] 58. Kyne, G.M.; Light, M.E.; Hursthouse, M.B.; De Mendoza, J.; Kilburn, J.D. Enantioselective amino acid recognition using acyclic thiourea receptors. J. Chem. Soc. Perkin Trans. 1 2001, 1, 1258–1263. [CrossRef] 59. Mutihac, L.; Lee, J.H.; Kim, J.S.; Vicens, J. Recognition of amino acids by functionalized calixarenes. Chem. Soc. Rev. 2011, 40, 2777–2796. [CrossRef][PubMed] 60. Kumar, R.; Vikramachakravarthi, D.; Pal, P. Production and purification of glutamic acid: A critical review towards process intensification. Chem. Eng. Process. Process. Intensif. 2014, 81, 59–71. [CrossRef] Membranes 2021, 11, 429 24 of 24

61. Fähnrich, A.; Mavrov, V.; Chmiel, H. Membrane processes for water reuse in the food industry. Desalination 1998, 119, 213–216. [CrossRef] 62. Drioli, E.; Romano, M. Progress and new perspectives on integrated membrane operations for sustainable industrial growth. Ind. Eng. Chem. Res. 2001, 40, 1277–1300. [CrossRef] 63. Mavrov, V.; Chmiel, H.; Bélières, E. Spent process water desalination and organic removal by membranes for water reuse in the food industry. Desalination 2001, 138, 65–74. [CrossRef] 64. Ferreira, A.R.V.; Alves, V.D.; Coelhoso, I.M. Polysaccharide-Based Membranes in Food Packaging Applications. Membranes 2016, 6, 22. [CrossRef] 65. Ghaseminezhad, S.M.; Barikani, M.; Salehirad, M. Development of graphene oxide-cellulose acetate nanocomposite reverse osmosis membrane for seawater desalination. Compos. Part B Eng. 2019, 161, 320–327. [CrossRef] 66. Ghimpusan, M.; Nechifor, G.; Din, I.S.; Nechifor, A.C.; Passeri, P. Application of hollow ﬁbre membrane bioreactor instead of granular activated carbon ﬁltration for treatment of wastewater from car dismantler activity. Mat. Plast. 2016, 53, 578–584.

67. Nechifor, A.; Cotorcea, S.; Bungău, C.; Albu, P.; Pas, cu, D.; Oprea, O.; Grosu, A.; Pîrt, ac, A.; Nechifor, G. Removing of the sulfur compounds by impregnated polypropylene fibers with silver nanoparticles-cellulose derivatives for air odor correction. Membranes 2021, 11, 256. [CrossRef][PubMed] 68. Din, I.S.; Cimbru, A.M.; Rikabi, A.A.K.K.; Tanczos, S.K.; Cotorcea, S.T.; Nechifor, G. Iono-molecular separation with composite membranes VI. Nitro-phenol separation through sulfonated polyether ether ketone on capillary polypropylene membranes. Rev. Chim. 2018, 69, 1603–1607. [CrossRef] 69. Nafliu, I.M.; Al Ani HN, A.; GROSU, A.R.; Tanczos, S.K.; Maior, I.; Nechifor, A.C. Iono-molecular separation with composite membranes. VII. nitrophenols pertraction on capilary polypropylene S-EPDM composite membranes. Mater. Plast. 2018, 55, 511–516. [CrossRef] 70. Ghimpusan, M.; Nechifor, G.; Nechifor, A.-C.; Dima, S.-O.; Passeri, P. Case studies on the physical-chemical parameters’ variation during three different purification approaches destined to treat wastewaters from food industry. J. Environ. Manag. 2017, 203, 811–816. [CrossRef] 71. Grosu, A.R.; Nafliu, I.M.; Din, I.S.; Cimbru, A.M.; Nechifor, G. Neutralization with simultaneous separation of aluminum and copper ions from condensed water through capillary polypropylene and cellulose. UPB Sci. Bull. Ser. B Chem. Mater. Sci. 2020, 82, 25–34. 72. Subroto, E.; Lembong, E.; Filianty, F.; Indiarto, R.; Primalia, G.; Putri, M.S.K.Z.; Theodora, H.C.; Junar, S. The Analysis Tech-niques Of Amino Acid And Protein In Food And Agricultural Products. Int. J. Sci. Technol. Res. 2020, 9, 29–36. 73. Croxton, R.; Kent, T.; Littlewood, A.; Smith, M. An evaluation of inkjet printed amino acid fingerprint test targets for ninhydrin process monitoring − and some observations. Forensic Sci. Int. 2021, 321.[CrossRef] 74. Chen, L.; Chen, Q.; Zhang, Z.; Wan, X. A novel colorimetric determination of free amino acids content in tea infusions with 2,4-dinitrofluorobenzene. J. Food Compos. Anal. 2009, 22, 137–141. [CrossRef] 75. Koter, S.; Szczepański,P.; Mateescu, M.; Nechifor, G.; Badalau, L.; Koter, I. Modeling of the cadmium transport through a bulk liquid membrane. Sep. Purif. Technol. 2013, 107, 135–143. [CrossRef] 76. Szczepański,P.; Diaconu, I. Transport ofp-nitrophenol through an agitated bulk liquid membrane. Sep. Sci. Technol. 2012, 47, 1725–1732. [CrossRef] 77. Kowalski, M.; Salerno-Kochan, R. An effect of washing on durability of hydrophobic finishes of membrane laminates. Autex Res. J. 2018, 18, 137–142. [CrossRef]