Membrane filtration of kraft lignin: Structural charactristics and antioxidant activity of the low-molecular-weight fraction

Industrial Crops & Products 112 (2018) 200–209

Contents lists available at ScienceDirect

Industrial Crops & Products journal homepage: www.elsevier.com/locate/indcrop

Membrane filtration of kraft lignin: Structural charactristics and antioxidant activity of the low-molecular-weight fraction

T

Selda Aminzadeha, Maris Laubertsb, Galina Dobeleb, Jevgenija Ponomarenkob, Tuve Mattssond, ⁎ Mikael E. Lindströma, Olena Sevastyanovaa,c, a

Wallenberg Wood Science Center (WWSC), Department of Fibre and Polymer Technology, KTH Royal Institute of Technology, Teknikringen 56-58, SE 100 44, Stockholm, Sweden b Latvian State Institute of Wood Chemistry, Laboratory of Lignin Chemistry, 27 Dzerbenes Str., LV 1006, Riga, Latvia c Department of Fibre and Polymer Technology, KTH Royal Institute of Technology, Teknikringen 56-58, SE 100 44, Stockholm, Sweden d Wallenberg Wood Science Center (WWSC), Forest and Chemical Engineering Department, Chalmers University of Technology, SE 412 96, Gothenburg, Sweden

A R T I C L E I N F O

A B S T R A C T

Keywords: Antioxidant activity LignoBoost lignin Low-MW lignin Structural analysis Ultrafiltration,

Lignin, which is the second most abundant biomass component and has carbon-rich phenolic content, is a promising renewable raw material for multiple applications, such as carbon fibers, adhesives, and emulsifiers. To use lignin efficiently, it is important to ensure its purity and homogeneity. As a result, the separation of lignin into fractions with high purity and narrow molecular-weight distributions is likely a prerequisite for several applications. Ultrafiltration using ceramic membranes has many advantages, including enabling direct lignin extraction from Kraft pulp cooking liquors without pH and temperature adjustment. One challenge with membrane filtration using such a system is the potential for reduced membrane performance over time, which is associated with fouling. In this study, LignoBoost Kraft lignin was fractionated using a ceramic membrane with a molecular weight cut-off of 1 kDa. The separation behavior during ultrafiltration fractionation was investigated and the antioxidant properties of the recovered low-molecular-weight (low-MW) lignin samples were evaluated. Using this model system, the permeate fluxes were unstable during the 100 h of membrane operation. However, a decrease in the average MW in the permeate over time was observed. The shift in MW was most pronounced for virgin membranes, while a more stable MW distribution was evident for membranes subjected to multiple cleaning cycles. According to 2D NMR analysis, low-MW lignin that was recovered after 100 h of operation, consisted of smaller lignin fragments, such as dimers and oligomers, with a high content of methoxygroups. This was confirmed using the size exclusion chromatography method, which indicated an weigh average molecular weight in the range of 450–500 Da. 31P NMR spectroscopy showed that, despite the lower total content of phenolic OH groups, the low-MW sample had a higher proportion of non-condensed phenolic OH groups. The results of the antioxidant tests demonstrated the strong potential of lignin and its low-MW fraction as a natural antioxidant, particularly for lipid-containing systems. The low-MW lignin fraction showed better antioxidant activity than the non-fractionated LignoBoost lignin in the kinetic oxygen radical absorbance capacity (ORAC) test and demonstrated three-fold stronger inhibition of the substrate (fluorescein) than the reference antioxidant Trolox (a water-soluble derivative of vitamin E).

1. Introduction Replacing fossil raw materials with renewable materials has become a primary focus of many research efforts. As a result, there is growing interest in lignocellulosic material, an abundant and renewable resource for the production of biofuel and high value bioproducts, as a feasible alternative (Luterbacher et al., 2014; Ragauskas et al., 2014).

Lignocellulose biomass is composed of three main components: cellulose, hemicelluloses and lignin. Lignin is the most abundant biopolymer after cellulose, corresponding to 15–30% of the wood weight of softwood and approximately 20% of that of hardwood. Its structure consists of a phenyl propane (C6–C3) units joined together through various interunit linkages, e.g., ether linkages (CeOeC) or carbon-carbon bonds (CeC). These interunit linkages impart heterogeneous complex

⁎ Corresponding author at: Wallenberg Wood Science Center (WWSC), Department of Fibre and Polymer Technology, KTH Royal Institute of Technology, Teknikringen 56-58, SE 100 44 Stockholm, Sweden. E-mail address: [email protected] (O. Sevastyanova).

https://doi.org/10.1016/j.indcrop.2017.11.042 Received 30 June 2017; Received in revised form 27 October 2017; Accepted 22 November 2017 0926-6690/ © 2017 Elsevier B.V. All rights reserved.

Industrial Crops & Products 112 (2018) 200–209

S. Aminzadeh et al.

et al., 2007), an emulsifier (Boeriu et al., 2004) an adsorbent (Mohan et al., 2006; Podkościelna et al., 2015), a carbon fiber precursor (Kadla et al., 2002; Kubo and Kadla, 2005), and an antioxidant (EspinozaAcosta et al., 2016); and co-reagent in phenol-formaldehyde resins (Jönsson et al., 2008; Mansouri and Salvadó, 2006; Sarkar and Adhikari, 2000; Tejado et al., 2007; Villar et al., 2001) and thermoplastics synthesis (Wang et al., 2016). As a polyphenol, lignin has a strong potential as an antioxidant to prevent oxidation reactions in biofuels, animal feeds and polymeric composite materials (Gulcin, 2012; Salem et al., 2014; Sindhi et al., 2013). The complex structure of lignin, composed of aromatic rings with hydroxy and methoxy functional groups, is responsible for this antioxidant potential due primarily to the termination of the oxidation propagation reaction through hydrogen donation (Jamshidian et al., 2012; Lu et al., 1998; Malík and Kröhnke, 2006). As a result of recent observations of possible cytotoxic and carcinogenic effects of synthetic antioxidants, i.e., butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), tert-butylhydroquinone (TBHQ) and propyl gallate (PG) at high dosages (Carocho and Ferreira, 2013; Gulcin, 2012), an attention towards the use of naturally occurring antioxidants (Gulcin, 2012), including lignin, is growing. The antioxidant activity of technical lignin is well established (Boeriu et al., 2004; Dizhbite et al., 2004; KoŠíková et al., 2006; Lu et al., 1998; Pan et al., 2006; Pouteau et al., 2003), however, the application of technical lignins as a natural antioxidants has not been implemented in the industrial sector mainly due to the high non-homogenous complex structure and high polydispersity of lignin (García et al., 2009; Hussin et al., 2015). Typically, purification and fractionation steps are necessary to improve the compatibility of lignin material with various substrates and enhance its stabilizing effect (García et al., 2010; García et al., 2009; Hussin et al., 2015; Lauberts et al., 2017; Ponomarenko et al., 2014; Ponomarenko et al., 2015b). Although, to predict the antioxidant properties of lignin material, several different tests should be applied, mainly due to the different reaction kinetics of lignin material in different systems. The aim of this study was to investigate the separation behavior during fractionation of dissolved LignoBoost Kraft lignin using a ceramic membrane with molecular weight cut-off of 1 kDa and to evaluate the structural characteristics and properties of the fractions obtained (low-MW fractions), including their antioxidant activity in different systems. The changes in MW and composition of lignin in the permeate were monitored over four days of membrane operation. A detailed structural characterization of lignin material was performed by nuclear magnetic resonance spectroscopy (NMR) analysis (2D-NMR and 31P NMR), analytical pyrolysis (pyrolysis-gas chromatography-mass spectrometry (PyGC/MS) and size-exclusion chromatography (SEC). To evaluate the antioxidant activity of the low-MW Kraft lignin fraction and original LignoBoost lignin, the radical scavenging activity towards DPPH% and ABTS% radicals and the oxygen radical absorbance capacity (ORAC) were tested. In addition, the Oxipres test, which evaluates the inhibitory properties of lignin in the oxidation of rich-in-lipids substrates, e.g., vegetable oils, bio-diesel, and cosmetic creams, was performed. These characteristics are expected to provide guidance on the optimization of the lignin fractionation using ultrafiltration and on possible applications of lignin as a natural antioxidant.

structures to the lignin (Adler, 1957; Crestini et al., 2011; Lin and Dence, 1992; Sarkanen and Ludwig, 1971). Technical lignin is a byproduct of the pulp and paper and biorefinery industries, where the Kraft and sulfite processes generates approximately 70 million tons of lignin annually (Lora and Glasser, 2002; Tomani, 2010). During the Kraft pulping process, lignin is degraded and solubilized in a strong alkaline aqueous solution, forming (together with degraded hemicelluloses and spent cooking chemicals) a black liquor (BL) (Marton, 1971; Sjöström, 1981). As a result, the chemical structure of Kraft lignin in terms of molecular weight (Mw), content of functional eOH and eCOOH groups and cross-linking density differ significantly from that of the lignin structure in the original plants, depending to a large extent on the conditions during the pulping process (Gellerstedt and Lindfors, 1984; Robert et al., 1984). The generated Kraft lignin is today used almost exclusively for energy production to cover the energy demand of the Kraft pulp process, however increasing energy efficiency of the Kraft pulp process enables the extraction of excess Kraft lignin for uses in value-added applications (Loutfi et al., 1991; Wallberg et al., 2005; Wallberg and Jönson, 2003). To use lignin efficiently, an important step is the separation of lignin. Three methods for extracting lignin can be found in the literature. One method is precipitation of the lignin from the BL by acid (Alen et al., 1979; Passinen, 1968). In 1910, Hough patented a method in which lignin can be precipitated by acidification of spend liquors from the alkaline pulping process using sulfuric acid. To improve dead-end filtration and prevent any possibility of colloid formation (which would interfere with the downstream separation steps), the precipitation step was performed at higher temperature (Hough, 1910). A recent development is the LignoBoost process, where through displacement washing, a lignin with a higher purity can be produced. In this method, after the precipitation of lignin, the material is filtered, and the filtrate cake is re-dispersed at a low pH (2–4). The newly formed suspension can easily be filtered and washed by displacement washing (Tomani, 2010; Öhman and Theliander, 2006; Öhman et al., 2007). A third technique that has been suggested is membrane separation by ultrafiltration (Colyar et al., 2008; Helander et al., 2013; Jönsson et al., 2008; Sevastyanova et al., 2014; Toledano et al., 2010; Wallberg and Jönsson, 2003). Membrane separation is a versatile operation that has found application in a wide range of sectors, e.g., the food industry, water purification and the pharmaceutical industry. For lignin separation using membrane filtration, several studies can be found in the literature, e.g., concentration or recovery of byproducts from BL (Brodin et al., 2009; Dafinov et al., 2005; Kevlich et al., 2017; Keyoumu et al., 2004; Tanistra and Bodzek, 1998; Wallberg et al., 2003). The cost of lignin separation by applying ultrafiltration to BL streams in pulp mills has also been estimated (Jönsson and Wallberg, 2009; Wallberg et al., 2005). Relative to other separation methods, membrane filtration has several advantages, e.g., it is possible to utilize ultrafiltration for lignin separation directly from a BL stream without temperature or pH adjustment (Wallberg and Jönsson, 2006). Such partial lignin separation will decrease the organic load on the recovery boiler system, which is often a bottleneck in the pulp mill (Hill et al., 1988; Wallberg and Jönsson, 2006). Additionally, fractionation using membranes with different molecular cut-offs produces lignin with more well-defined molecular mass, lower polydispersity and tailored properties (Sevastyanova et al., 2014). Thus, ultrafiltration of Kraft lignin can provide a pathway for upgrading lignin, opening new routes for the production of high-value materials and chemicals from this lignin source. Lignin as a source of phenolic units, including various aromatic and aliphatic hydroxyl groups, is one of the best candidates for modification and various applications (Espinoza-Acosta et al., 2016), e.g., in the production of vanillin (Tarabanko et al., 1995) and phenols; as a dispersant in cement and gypsum blends (Matsushita et al., 2008; Yang

2. Experimental 2.1. Materials The LignoBoost Kraft lignin powder used in this study was produced from Nordic softwood and kindly supplied from a plant in northern Europe. Sodium hydroxide (Sigma Aldrich, Germany) was used to dissolve lignin, and sulfuric acid (> 95%) (Fisher Scientific, UK) was used for the precipitation of the fractionated products. Sodium chloride (Sigma 201

Industrial Crops & Products 112 (2018) 200–209

S. Aminzadeh et al.

Ultrafiltration fractionation with lower ionic strength solutions (1 LV-A and 1 LV-B) was performed with virgin membranes. Following the filtration experiments, the entire system (including the membrane) was washed with Ultrasil 10 for 1 h at 50 °C, followed by water washing to neutrality. The experiments with higher ionic strength feed solutions were performed with these washed membranes, referred to as reused membranes (i.e., the membrane from 1 LV-A was used for 2HR-A and 1 LV-B for 2HR-B). To investigate the impact of the washing procedure on the membrane performance, an additional experiment (Experiment 3 LRS) was performed, reusing a stabilized membrane after experiments 1 LV-A and 2HR-A. A final experiment, using a virgin membrane washed twice with Ultrasil 10 (Experiment 4 VWS) was also performed. The experimental details are shown in Table 1. After collection of the permeate, samples were stored in a cold room (4 °C). The lignin was recovered by precipitation via dropwise addition of 6 M sulfuric acid while stirring to pH 2. During this procedure, the beaker containing the permeate solution was placed in an ice bath. The recovered precipitates were freeze-dried.

Aldrich, Germany) was used to increase the ionic strength. Ultrasil 10, used for washing the membrane system, was purchased from Ecolab AB (Alvsjö, Stockholm, Sweden) 2.2. Methods 2.2.1. Ultrafiltration experiments Membrane cross-flow filtration experiments were performed using bench-scale membrane equipment. The membrane unit was a Kerasept™ unit purchased from Novasep, Pompay, France. For this investigation, a single-channel ceramic membrane (TAMI Industries, model MSKTB 0251001, Nyons, France) with a cutoff of 1000 Da, internal diameter of 0.006 m and total surface area of 0.0047 m2 was used. A gear pump (Micropump IDEX Corporation, Vancouver, Canada) was used to feed the lignin solution from a 10 L reservoir tank into the membrane unit. The crossflow velocity was measured using a flow meter (Floab Flödesprodukter AB, Model MS501, ML4-F1), and the pressure directly before and after the membrane module was measured using two manometers. The trans- membrane pressure (TMP), was estimated as the average overpressure. The retentate was recirculated back to the feed tank (except when recovering samples), and the total amount withdrawn as permeate samples from the system was less than 1% of the total volume of the system mode. Fig. 1 shows a schematic representation of the apparatus. Two sets of Kraft lignin solutions were used in this study: 10% (w/ w) Kraft lignin solution with 5% (w/w) NaOH (experiments 1 LV-A, 1 LV-B, 3 LRS and 4 LVWS) and 10% (w/w) Kraft lignin solution with 5% (w/w) NaOH, with higher ionic strength adjusted by adding NaCl to the total amount of 6 kg of prepared solution (experiments 2HR-A and 2HRB) to achieve the concentration of 1.2 mol/L. A detailed description of each experiment can be found in Table 1. All solutions were prepared from Kraft lignin powder one day prior to the filtration experiment, and the entire solution was filtered with a Munktell filter paper, grade 3 to remove undissolved particles. The filtration experiments were performed at a TMP of 3 bar, ambient temperature (24 °C) and a cross flow velocity of 3.9 m/s (corresponding to a turbulent flow regime). For each set of experiments, 6 kg of solution was placed in the 10 L feed tank as an initial feed and circulated through the system for one night (12 h) with the permeate valve closed. The feed solution reached a steady-state temperature (24 °C) during this time. All membrane filtrations were performed for 100 h, and permeate samples were withdrawn at 0, 2, 4, 7, 24, 27, 32, 49, 57, 75, and 100 h and collected for 30 min per sample. Between sample collection events, permeate was circulated back to the feed tank (Fig. 1). The permeate flow was estimated by weighing the sample.

2.2.2. Evaluation of antioxidant properties The DPPH% (2,2-diphenyl-1-picrylhydrazyl) radical scavenging test was performed as described elsewhere (Dizhbite et al., 2004). The freeradical scavenging activity was expressed as the IC50 (concentration required for 50% inhibition of free radicals). A lower IC50 value indicates higher radical scavenging activity of the investigated sample. The ABTS%+ radical was produced according to earlier published method (Baltrušaitytė et al., 2007). An antioxidant activity was measured by spectrophotometrically using a PerkinElmer Lambda 650 UV/ VIS spectrometer (PerkinElmer Instruments, Shelton, Connecticut, USA). The free-radical scavenging activity was also expressed as the IC50. The ORAC assay was performed according to a method described previously (Prior et al., 2005) using a BioTek Synergy HT microplate reader (BioTek Instruments, Winooski, Vermont, USA). The lignin radical absorbance capacity was expressed as the Trolox equivalent, TE, per gram of lignin sample. A higher TE value indicated higher antioxidant capacity. The protection activity of non-fractionated lignin and its low-Mw fraction in the process of lipid oxidation was determined in mayonnaise containing 65% (w/w) fats (control and with synthetic antioxidant TBHQ) in an Oxipres apparatus (Mikrolab Aarhus, Højbjerg, Denmark) at 120 °C according to the method suggested earlier (Lauberts et al., 2017). The induction period (IP, hours) of oxidation was recorded when the pressure begun decrease rapidly. The antioxidant activity of additives was estimated by their protection factor (PF): PF = IPx/IPk

(1)

where IPx is the oxidation IP with an additive and IPk is the IP for the sample without additive. All values measured are shown as an average with a confidence interval (significance level α = 0.05). Each measurement was performed at least in triplicate. Statistics were calculated in MS Excel 2013 and IBM SPSS Statistics 21.0. 2.3. Analytical characterization of lignin samples The total solid content of the lignin powder and the fractionated samples were measured by heating the sample in the oven at 105 °C for 24 h and recording the weight of dry residue. Alternatively, samples were dried using a Mettler PM460 with a Mettler LP16 infrared radiation dryer device (Mettler Toledo, Greifensee, Switzerland). Both methods gave approximately similar results. Sample preparation was based on the dry weight measured in the oven. The organic content of the fractionated samples were measured by burning samples in a furnace at 525 °C overnight (TAPPI T 211

Fig. 1. Experimental configuration for the ultrafiltration experiment.

202

Industrial Crops & Products 112 (2018) 200–209

S. Aminzadeh et al.

Table 1 Description of solutions and membranes used for individual ultrafiltration experiments (L = low ionic strength, H = high ionic strength, V = virgin membrane, R = reused membrane, S = stabilized with NaOH, W = washed with Ultrasil 10). Experiments

Content of each solution

1LV-A 1LV-B 2HR-A 2HR-B 3LRS 4LVWS

10% 10% 10% 10% 10% 10%

lignin lignin lignin lignin lignin lignin

dissolved dissolved dissolved dissolved dissolved dissolved

in in in in in in

5% 5% 5% 5% 5% 5%

NaOH NaOH NaOH NaOH NaOH NaOH

solution solution solution solution solution solution

w/w (low ionic strength)/virgin membrane w/w/(low ionic strength)/virgin membrane (replicate of 1 LV-A) w/w and 1.2 M NaCl (high ionic strength)/reused membrane w/w and 1.2 M NaCl (high ionic strength), reused membrane (replicate of 2HR-A) w/w (low ionic strength)/reused membrane/stabilized with NaOH prior to experiment w/w (low ionic strength)/virgin membrane, washed with Ultrasil 10, stabilized with NaOH

ionization detection (Py-GC/MS(FID) was performed as described earlier (Lauberts et al., 2017) by using a Frontier Lab (Fukushima, Japan) Micro Double-Shot Pyrolyzer Py-2020iD. Approximately 1–2 mg of the sample was used for analysis (residual moisture content < 1%). The oven program was: held at 60 °C for 1 min, ramped at 6 °C min−1 to 270 °C, and held at 270 °C for 10 min. Identification of the individual compounds was performed based on GC/MS data in the NIST 147.LI13 MS library (Frontier Lab Fukushima, Japan). The relative areas of the peaks of individual compounds were calculated from the GC/FID data using Shimadzu software. The summed molar areas of the relevant peaks were normalized to 100%; the data for five repetitive pyrolysis experiments were averaged. Contents of lignin (lignin derivatives, L), carbohydrates (C), sulfur containing compounds (S-containing) and other identified compounds were calculated as the ratio of the sum of the areas of the peaks from corresponding component divided by the sum of the area of all used peaks multiplied by 100% (Alves et al., 2006). The area originated from CO2, an overlapping peaks from carbohydrates and lignin and minor products was not used for the calculations. The retention time used for quantification was from 6 to 53 min.

Standard and ISO 1762) and recording the weight loss gravimetrically. The molecular weight properties of lignin samples was determined by gel permeation chromatography (GPC) using a Waters instrument system (Waters Sverige AB, Sollentuna, Sweden) consisting of a 515 HPLC-pump, 2707 autosampler and 2998 photodiode array detector (operated at 254 and 280 nm). Two to five mg of dried lignin sample was acetobrominated prior to analysis (Guerra et al., 2006). A total volume of 1 mL of acetyl bromide and glacial acetic acid (1/9 v/v) was added to the sample, which was then stirred for 2 h at room temperature. The solution was evaporated by nitrogen gas to dryness. The residue was dissolved in HPLC-grade tetrahydrofuran (THF), and the solution was filtered through a 5 μm syringe filter. HPLC-grade THF, filtered through a 2 μm PTFE membrane filter and degassed, was used as the mobile phase. The flow rate of the mobile phase was set to 0.3 mL/min. Separation was achieved on Waters Ultrastyragel HR4, HR2 and HR0.5 4.6 × 300 mm solvent-efficient columns connected in series and operated at 35 °C. A sample volume of 20 μL was injected into the GPC instrument and UV signal was recorded at different wavelengths, 254 and 280 nm. Calibration was performed at 254 nm using polystyrene standards with nominal MW’s ranging from 480 to 176,000 Da. The final analysis was performed using the intensity of the UV signal at 280 nm with a help of Waters Empower 3 build 3471 software. The content of functional groups in lignin samples was measured by 31 P NMR (Argyropoulos, 1994). Approximately 20–30 mg of lignin sample was weighed and dissolved in 100 μL of DMF and 100 μL of pyridine. Endo-N-hydroxy-5-norbornene-2,3-dicarboximide (e-HNDI) (Sigma Aldrich, 40 mg/mL) and chromium (Ш) acetylacetonate (Aldrich, 5 mg/mL) were used as an internal standard and relaxation reagent, respectively. In addition, 2-chloro-4,4,5,5-tetramethyl-1,3,2-dioxaphospholane was used as a phosphorylating reagent, and the derivatized sample was dissolved in CDCl3 prior to analysis. The 31P NMR experiment was performed with a 90° pulse angle, inverse gated proton decoupling and a delay time of 10 s. For analysis, 256 scans with a time delay of 6 s and a total runtime of 34 min were collected. Measurements were performed in duplicates. The structural characteristics of lignin samples were also studied using a 2D HSQC method. Prior to the analysis, approximately 100 mg of sample was acetylated by the addition of 1 mL of anhydrous pyridine and 1 mL acetic anhydride and was stirred overnight. Excess acetic anhydride was decomposed by addition of methanol and cooling the system in an ice bath. After evaporation of the mixture, the removal of pyridine continued by addition of toluene and evaporation (repeated three times). Finally, toluene was removed by addition of methanol and evaporation (Gellerstedt, 1992). The residue was dissolved in 700 μL of DMSO-d6. The 2D HSQC NMR spectrum was acquired using the Bruker pulse program ‘hsqcetgpsi’ a relaxation delay of 1.7 s, a coupling constant of 145 Hz, an INEPT transfer delay time of 1.72 ms (d4 = 1/4J), a spectral window of 10.5 ppm in F2 and 166 ppm in F1 with 1024 × 512 increments, 240 scans per increment, and a spectral center set at 90.0 ppm in F1 and 5.3 ppm in F2. The 2D NMR data set was processed with 2 K × 1 K data points using a π/2-shifted sine-bell window function in both dimensions. Pyrolysis-gas chromatography-mass spectrometry with flame

3. Results and discussion 3.1. Membrane fractionation The membrane performance was investigated over 100 h for a number of experiments with Kraft lignin solutions of varying ionic strength and 1 kDa cut-off membranes with various pretreatments. The pretreatments included various types of washing of virgin and reused membranes (see Table 1). Liquid permeate samples were collected, and the permeate flux is reported based on either the weight of the liquid sample or the weight of the organic content in the sample. In Fig. 2, the permeate flux is reported as the total liquid weight for

Fig. 2. Permeate flux reported as total liquid weight (kg/h·m2) for samples with low (1 LV experiments) and high ionic strength (2HR experiments). Permeate flux for additional experiments with low ionic strength and reused membrane stabilized with NaOH (3 LRS) and with lower ionic strength and virgin membrane washed twice with Ultrasil 10 stabilized with NaOH (4 LVWS) is also included in the figure.

203

Industrial Crops & Products 112 (2018) 200–209

S. Aminzadeh et al.

the cross-flow ultrafiltration experiments of the two prepared LignoBoost Kraft lignin solutions. As shown in the figure, poor reproducibility between individual experiments was observed: 1LV-A vs 1LV-B and 2HR-A vs 2HR-B. The permeate fluxes were also unstable during the period of membrane operation (100 h), particularly for experiments performed at higher ionic strength with membranes that had been used previously and cleaned with Ultrasil 10 (experiments 2HR-A and 2HRB, see Table 1). The permeate flux was also typically higher for these experiments compared to 1LV-A and 1LV-B. Since reused, cleaned membranes were used for both experiments with a higher ionic strength feed solution, it is difficult, from these experiments, to separate the two potential effects on permeate flux, i.e., the increase in flux could be due to either the cleaning procedure with Ultrasil 10 performed after the first filtration experiment and/or the increase in ionic strength. Therefore, two additional experiments were performed with lowerionic-strength feed solutions to investigate how the cleaning procedure affects permeate flux. Experiment 3 LRS was performed by reusing the same membrane used in 1 LV-A, subjected to two cleaning cycles with Ultrasil 10. Before starting the filtration, the membrane was stabilized by circulating a 1 M NaOH solution through the system for 24 h to ensure alkaline conditions at the start of the filtration experiment. For the 4 LVWS experiment, a virgin membrane cleaned two times with Ultrasil 10 was used; the system was also stabilized as described above with a 1 M NaOH solution. The permeate flux for these experiments is included in Fig. 2. Fig. 2 shows that the flux significantly increased after earlier use and cleaning of the membranes. This finding indicates that the chemical conditions and cleaning chemicals affected the membranes. The permeate flux, reported as organic weight, is shown for three selected experiments in Fig. 3. The organic flux generally follows the trend of the total liquid flux, with some outliers: At the first sampling points for experiment 1 LV-B and 2HR-B, the organic flux is considerably higher than that at later sampling times, while the corresponding trend for the total liquid flux (Fig. 2) is constant or increasing. Furthermore, at 24 h in experiment 2HR-B, a relatively low organic flux was measured, considering the substantial increase in total flux. These data points are challenging to evaluate due to the relatively high experimental error, as relatively small amounts of organic material in the permeate (14–100 mg) were collected and used for gravimetric analysis. Table 2 summarizes the organic content found in the permeate fractions recovered at 100 h for experiments 1 LV-B, 2HR-B and 4 LVWS based on gravimetrical analysis. The organic content of the permeate was approximately 1 wt% in the two experiments performed at lower ionic strength, while it was found to be lower in the experiment performed at higher ionic strength. The difference in organic content between these two conditions could indicate an effect of ionic strength on the production of low-Mw lignin under the investigated conditions. However, the errors in these estimates are large, as stated above. For these experiments (1 LV-B, 2HR-B and 4 LVWS), the Mw distribution was also monitored for lignin samples isolated from the

Table 2 Organic content in the permeate (g/kg) collected after 100 h of operation (final point). Organic content at the last point (g/kg) 1 LV-B 9.4

2HR-B 6.0

4 LVWS 9.9

permeate fractions collected at various times (Fig. 4). The average Mw of each lignin fraction, collected during 100 h of filtration, was calculated as described in Section 2.3 (see Fig. 5). As shown in Figs. 4 and 5, there was a gradual shift in the MW with increasing sampling time towards lower MW values: fractions collected later displayed a higher retention time (Fig. 4). This trend of decreasing MW was more pronounced for virgin membranes (1LV-B), and a more stable MW distribution emerged as the number of cleaning cycles increased (4LVWS). Ultrafiltration with a virgin membrane without initial stabilization by NaOH (1 LV-B) produced lignin fractions with Mw of 1600–1100 Da during the first hours of ultrafiltration, with a subsequent decrease towards 500 Da at the end of the operation. The experiment with the twice-washed and the stabilized membrane, 4 LVWS, produced a lignin fraction of 600–500 Da over the entire 100 h of operation. This process is illustrated in Fig. 4c, where MW distribution curves with similar shapes are shown for 26–35 min in the 4 LVWS experiment, while in 1 LV-B and 2HR-B ultrafiltration experiments, the distribution curves are located at 23–35 min and 24–35 min, respectively, and have broader shapes. Despite these differences, all three experiments produced permeate with a similar Mw of 400–500 Da at the final sample time at 100 h No significant difference in molecular weight averages was seen for the experiment preformed at higher ionic strength at this sample time. The final permeate fractions were selected for further analysis and property investigation. 3.2. Structure of LignoBoost lignin and its low-Mw fractions The chemical composition and structure of LignoBoost lignin and selected low-Mw fractions (1LV-B and 2HR-B) obtained after ultrafiltration at a sampling time of 100 h, were studied and compared using NMR techniques, namely, 31P NMR and 2D NMR and by the Py-GC/MS (FID) method described in Section 2.3. The main purpose of the 2D NMR analysis was to observe the structural changes in lignin samples as a result of ultrafiltration through a 1 kDa membrane. All peak assignments were done based on the previous works (Capanema et al., 2004; del Río et al., 2012; Kim and Ralph, 2010; Zhang and Gellerstedt, 2007). For the original LignoBoost lignin, signals from the β-aryl ether (β-O-4), pinoresinol (β-β), phenylcoumarin (β-5) and coniferyl alcohol structures and the methoxy groups could be clearly detected in the 2D NMR spectra (Fig. 6a). In the spectra of the low-MW fraction, some low intensities signals are still present in the area of oxygenated aliphatics, however clear signals from a typical lignin interunit linkages, such as β-O-4, β-β, β-5 could not be detected anymore (Fig. 6b). This observation in combination with SEC results, indicating the average molecular size for the low-MW fraction is in a range of 400–500 Da, suggests that the highly modified and oxidized lignin fragments, likely in a form of monomers, dimers and oligomers, have accumulated in the permeate during the prolonged ultrafiltration. The larger content of carboxylic groups in the low-MW sample detected by 31P NMR serves an additional support for this assumption (see Table 3) (Berlin and Balakshin, 2014; Crestini et al., 2017). 31 P NMR was applied to both lignin samples (initial LignoBoost lignin and 2HR-B fractionated sample) to investigate the changes in the contents of various functional groups, as they are predominantly responsible for the reactivity of lignin, including the radical scavenging ability.

Fig. 3. Permeate flux reported as the organic flux (g/h·m2) for three sets of experiments.

204

Industrial Crops & Products 112 (2018) 200–209

S. Aminzadeh et al.

Fig. 4. Changes in the molecular weight distributions of lignin with sampling time for the experiments: a) 1 LV-B, b) 2HR-B, and c) 4 LVWS.

as a result of cleavage of the β-O-4 linkages present in the native lignin (Tsutsumi et al., 1995). As reported in previous work on the fractionation of Kraft lignin—including fractionation with ceramic

Table 3, shows that the total content of phenolic groups is actually higher in the original LignoBoost lignin. It is well known that significant amounts of free phenolic groups form during the Kraft pulping of wood 205

Industrial Crops & Products 112 (2018) 200–209

S. Aminzadeh et al.

composition and structure of complex and insoluble natural materials, including lignocelluloses and various types of lignin (Alves et al., 2006; Faix et al., 1992; Gutierrez et al., 1996; Meier and Faix 1992; Ohra-aho et al., 2005; Rodriges et al., 2001). The list of compounds identified from the chromatogram is present in Table 4 and in Table SI-1. The presence of carbohydrates in technical lignins, subjected to the thorough washing procedures, is usually associated with the presence of socalled lignin-carbohydrate complexes, LCCs (Balakshin et al., 2011; Lawoko et al., 2005). During fractionation, carbohydrates typically accumulate in the higher-Mw fraction (Sevastyanova et al., 2014). However, in the current study, low-MW fractions, collected after 100 h of ultrafiltration, had similar relative carbohydrate content as the original lignin: 2.9 and 2.7% compared to 2.7% (based on the content of identified peaks from the Py-GC–MS/FID chromatograms) (Table 4). This indicates the accumulation of some low MW carbohydrates-derived degradation products in the permeate sample. The sulfur content was higher for the low-Mw fractions, in agreement with previous studies (Sevastyanova et al., 2014). Notably, the content of lignin derivatives (L) was lower for the low-MW fractions, particularly for the sample obtained using a low-ionic-strength solution (1LV-B), while a portion of the degradation products (methanol, water and carbon dioxide) was highest for the corresponding sample. For the lignin sample obtained in the 2HR-B experiment, the content of aliphatic, aromatic and cyclic monomers, which were stable during pyrolysis and avoided degradation to water, carbon dioxide and methanol, was 3.7 times higher than that in the sample in the 1LV-B experiment. This together with the difference in the estimated organic content in the permeate fraction (Table 2) could indicate a difference in lignin separation behavior. It should however be noted that the average MW weights for the

Fig. 5. Comparison of changes in weight average molecular weight for different experiments, measured by SEC.

membranes low-Mw fractions are usually enriched phenolic content (Sevastyanova et al., 2014). In the present work, ultrafiltration was performed for almost 100 h prior to collection of the lignin sample, producing samples with MW much lower than reported earlier. However, the ratio of non-condensed to condensed phenolic groups is higher for fractionated samples (see Table 3), indicating possible higher reactivity of available phenolic OH groups. In addition, the amount of carboxylic group is higher for the fractionated sample, due to the presence of highly oxidized lignin units. Lignin samples were characterized by Py-GC/MS method. Analytical pyrolysis is a useful technique for comparative investigation as it is a rapid and sensitive quantitative method for assessing the chemical

Fig. 6. Interunit analysis by 2D-NMR for the (a) initial LignoBoost lignin and (b) fractionated low-MW sample (2HR-B).

206

Industrial Crops & Products 112 (2018) 200–209

S. Aminzadeh et al.

Table 3 Content of functional groups determined by Sample

Lignoboost lignin Low-MW fraction

31

P NMR.

Phenolic OH (mmol/g) condensed

Non-condensed

Total phenolic

2 ± 0.04 0.95 ± 0.05

2.27 ± 0.0 2.85 ± 0.21

4.27 ± 0.04 3.8 ± 0.27

Non-Condensed/Condensed Ph-OH

Aliphatic OH (mmol/g)

Total OH (mmol/g)

- COOH (mmol/g)

1.13 3

1.97 ± 0.07 0.43 ± 0.02

6.24 ± 0.11 4.24 ± 0.29

0.45 ± 0.01 1 ± 0.08

Table 4 Composition of LignoBoost lignin and two low-Mw fractions identified based on Py-GC/ MS/FID analysis%.

Table 5 Antioxidant properties of non-fractionated LignoBoost lignin and its low-Mw fraction. Radicals

Compounds identified from chromatogram, %

Lignin (L) - Phenyl and benzyl derivatives - Guaiacyl derivatives Carbohydrates (C) Carbon dioxide, water, methanol (summary) Sulphur-containing (S-containing) Aliphatic, aromatic and cyclic monomers Phenols with various substitutions, % from Summarized content of methoxylated phenols (G derivatives) Saturated side chain Double bond in C1 (C1 = C2) Oxygen in side chain (ArC1 + ArC2)/ArC3 CH2 in the Cα

LignoBoost lignin

Lignin samples

Low-Mw fractions 1 LV-B

2HR-B

50.2 8.0 42.2 2.7 42.6

37.5 6.7 30.8 2.9 55.9

41.8 5.1 36.7 2.7 46.8

0.5 1.8 L 84.1

0.8 1.5

1.1 5.5

82.1

87.6

31.5 21.6 11.5 2.51 38.9

32.3 14.3 13.4 4.2 38.5

24.3 15.6 22.4 3.3 32.8

−1

DPPH%, IC50 (mg L ) ABTS%+, IC50 (mg L−1) ORAC assay (TE mmol g−1)

a

LignoBoost lignin

Low-Mw fraction

23.7 ± 0.9 5.5 ± 0.1 9.09 ± 0.1

32.8 ± 1.0 3.9 ± 0.2 14.9 ± 0.2

Inhibitory properties of lignin samples in mayonnaise oxidation (Oxipres test) Additive concentration, mg g−1

Induction period (IP), h

0 2.5b 5.0

1.9 1.93 2.73 Protection factor 1 1.02 1.44

0 2.5b 5.0

1.9 2.27 3.39 1 1.19 1.78

a Antioxidant capacity calculated based on Trolox; 1 g of Trolox corresponds to 3.9 TE mmol g−1. b Addition of 1 mg g−1 (0.1%) TBHQ to a mayonnaise sample resulted in an IP of 3.48 h and a protection factor of 1.83.

two different fractions were very similar. As demonstrated previously (Lauberts et al., 2017; Ponomarenko et al., 2015a), there is a correlation between certain structural features and the antioxidant activity of polyphenolic compounds, particularly lignin. Several important structural descriptors in addition to OHphen content, which regulate antioxidant activity of lignin samples were characterized during this work. As shown in Table 4 (and Table 1 in SI), the ratio (ArC1 + ArC2)/ (ArC3) for the low-Mw samples are higher than that of the initial nonfractionated lignin, imparting enhanced antioxidant activity. The content of quaiacyl (G) phenols (OCH3 groups) are higher and phenols with CH2 in the Cα position is lower for the low-Mw sample obtained in the 2HR-B experiment. For this sample, high methoxy group content was also shown using 2D HSQC NMR. However, for both low-Mw fractions, a high content of oxygen in the side chain is observed, which may have a negative impact on the antioxidant activity in various tests.

commercial antioxidant Trolox. Such a discrepancy between lignin sample efficiency in the DPPH• test, and the two other assays can be explained by the difference in the mechanisms of the radical scavenging reactions in the tests applied (Ponomarenko et al., 2014). This could have been due to the different mechanisms occurring in the various tests: a PCET mechanism based on the hemolytic OeH bond cleavage in test with DPPH% and a SPLET mechanism based on the heterolytic dissociation of some phenolic OH groups in the test with ABTS%+ and the ORAC assay; The low-Mw fraction, which demonstrated high results in the ABTS %+ and ORAC tests, revealed itself as a potential protector of substrates with high lipid content from oxidation. The decomposition of lipid components in foods, especially in those having a high content of polyunsaturated fatty acids, shortens their shelf-life dramatically. Thus, antioxidants (compounds that are delaying lipid oxidation) are often added to fat-containing food. According to European regulations (EC 2004), the concentration of synthetic antioxidants in foods should be kept below 0.2 mg g−1 of the oil or fat content in such foods. Higher content of synthetic additives is associated with an increased risk of pathological enzyme or lipid alterations, as well as with carcinogenic effects. There are no regulations, on the other hand, concerning concentrations of added antioxidants of natural origin, which explains the growing interest in the use of these type of compounds (Li et al., 2015). The antioxidant activity of LignoBoost lignin and its low-Mw fraction in an oxygen atmosphere was determined in commercial additivefree mayonnaise. The changes in the induction periods (IP, in hours) of oxidation for the additive-free mayonnaise sample and the sample with various concentrations of lignin are shown in Fig. 7. The low-Mw fraction was an efficient antioxidant even at relatively low concentration. Generally, low-Mw lignin demonstrated higher antioxidant stability than LignoBoost lignin. For instance, 2.5 mg and 5 mg of the lowMw lignin fraction per 1 g of mayonnaise increased its oxidation IP

3.3. Antioxidant properties of the low-Mw lignin To get more objective results, the antioxidant activity of the samples under study was characterized, using different methods, taking into account the mechanisms of the reactions with free radicals, kinetic parameters and other factors as described elsewhere (Ponomarenko et al., 2014; Trouillas et al., 2008). The results of the ABTS%+ and DPPH% tests are shown (Table 5) in terms of the IC50 (mg L−1). A lower IC50 values for the investigated antioxidant sample indicate higher radical scavenging activity, as the IC50 corresponds to the concentration of the sample required for 50% inhibition of the radical species. For ABTS%+ scavenging, the low-Mw fraction exhibited better antioxidant activity than the unfractionated LignoBoost lignin. However, the radical scavenging activity in the test with DPPH• free radical was higher for the non-fractionated lignin. The results of the ORAC assay (Table 5) conform to those observed in the ABTS%+ test. Moreover, the antioxidant activity of low-Mw is three times higher than that for the 207

Industrial Crops & Products 112 (2018) 200–209

S. Aminzadeh et al.

References Öhman, F., Theliander, H., 2006. Washing lignin precipitated from kraft black liquor. Pap. Puu 88, 287–292. Öhman, F., Wallmo, H., Theliander, H., 2007. A novel method for washing lignin precipitated from kraft black liquor −Laboratory trials. Nordic Pulp Pap. Res. J. 22, 9–16. Adler, E., 1957. Structural elements of lignin. Ind. Eng. Chem. 49 (9), 1377–1383. Alen, R., Patja, P., Sjöström, E., 1979. Carbon dioxide precipitation of lignin from pine kraft black liquor. Tappi J. 62 (11), 108–110. Alves, A., Schwanninger, M., Pereira, H., Rodrigues, J., 2006. Analytical pyrolysis as a direct method to determine the lignin content in wood. Part 1 Comparison of pyrolysis lignin with Klason lignin. J. Anal. Appl. Pyrol. 76, 209–213. Argyropoulos, D.S., 1994. Quantitative phosphorus-31 NMR analysis of lignins, a new tool for the lignin chemist. J. Wood Chem. Technol. 14, 45–63. Balakshin, M., Capanema, E., Gracz, H., Chang, H.M., Jameel, H., 2011. Quantification of lignin-carbohydrate linkages with high-resolution NMR spectroscopy. Planta 233, 1097–1110. Baltrušaitytė, V., Venskutonis, P.R., Čeksterytė, V., 2007. Radical scavenging activity of different floral origin honey and beebread phenolic extracts. Food Chem. 101, 502–514. Berlin, A., Balakshin, M., 2014. Industrial Lignins: analysis, Properties, and Applications. Bioenergy Research: Advances and Applications 18, 315–335 Chapter 18. Boeriu, C.G., Bravo, D., Gosselink, R.J.A., van Dam, J.E.G., 2004. Characterisation of structure-dependent functional properties of lignin with infrared spectroscopy. Ind. Crops Prod. 20, 205–218. Brodin, I., Sjöholm, E., Gellerstedt, G., 2009. Kraft lignin as feedstock for chemical products: the effects of membrane filtration. Holzforschung 63, 290–297. Capanema, E.A., Balakshin, M.Y., Kadla, J.F., 2004. A comprehensive approach for quantitative lignin characterization by NMR spectroscopy. J. Agric. Food Chem. 52, 1850–1860. Carocho, M., Ferreira, I.C.F.R., 2013. A review on antioxidants, prooxidants and related controversy Natural and synthetic compounds, screening and analysis methodologies and future perspectives. Food Chem. Toxicol. 51, 15–25. Colyar, K.R., Pellegrino, J., Kadam, K., 2008. Fractionation of pre-Hydrolysis products from lignocellulosic biomass by an ultrafiltration ceramic tubular membrane. Sep. Sci. Technol. 43, 447–476. Crestini, C., Melone, F., Sette, M., Saladino, R., 2011. Milled wood lignin. Linear Oligomer Biomacromol. 12, 3928–3935. Crestini, C., Lange, H., Sette, M., Argyropoulos, D.S., 2017. On the structure of softwood kraft lignin. Green Chem. 19, 4104–4121. Dafinov, A., Font, J., Garcia-Valls, R., 2005. Processing of black liquors by UF/NF ceramic membranes. Desalination 173, 83–90. Dizhbite, T., Telysheva, G., Jurkjane, V., Viesturs, U., 2004. Characterization of the radical scavenging activity of lignins–natural antioxidants. Bioresour. Technol. 95, 309–317. Espinoza-Acosta, J.L., Torres-Chávez, P.I., Ramírez-Wong, B., López-Saiz, C.M., MontañoLeyva, B., 2016. Antioxidant, antimicrobial, and antimutagenic properties of technical lignins and their applications. BioResources 11, 5452–5481. Faix, O., Bremer, J., Meier, D., Fortmann, I., Scheijen, M.A., Boon, J.J., 1992. Characterization of tobacco lignin by analytical pyrolysis and Fourier transform-infrared spectroscopy. J. Anal. Appl. Pyrol. 22, 239–259. García, A., Toledano, A., Serrano, L., Egüés, I., González, M., Marín, F., Labidi, J., 2009. Characterization of lignins obtained by selective precipitation. Sep. Purif. Technol. 68, 193–198. García, A., Toledano, A., Andrés, M.Á., Labidi, J., 2010. Study of the antioxidant capacity of Miscanthus sinensis lignins. Process Biochem. 45, 935–940. Gellerstedt, G., Lindfors, E.-L., 1984. Structural changes in lignin during kraft pulping. Holzforschung 38, 151–158. Gellerstedt, G., 1992. In: Lin, S.Y., Dence, C.W. (Eds.), Methods in Lignin Chemistry. Springer-Verlag, Berlin, pp. 491. Guerra, A., Filpponen, I., Lucia, L.A., Argyropoulos, D.S., 2006. Comparative evaluation of three lignin isolation protocols for various wood species. J. Agric. Food Chem. 54, 9696–9705. Gulcin, I., 2012. Antioxidant activity of food constituents: an overview. Arch. Toxicol. 86, 345–391. Gutierrez, A., Bocchini, P., Galletti, G.C., Martinez, A.T., 1996. Analysis of ligninpolysaccharide; complexes formed during grass lignin degradation by cultures of Pleurotus species. Appl. Environment. Microbiol. 62, 1928–1934. Helander, M., Theliander, H., Lawoko, M., Henriksson, G., Zhang, L., Lindström, M.E., 2013. Fractionation of technical lignin: molecular mass and pH effects. BioResources 8, 2270–2282. Hill, M.K., Violette, D.A., Woerner, D.L., 1988. Lowering kraft black liquor viscosity by ultrafiltration. Sep. Sci. Technol. 23, 1789–1798. Hough, W.J., 1910. Process of treating spent alkaline pulping liquors, United States Patent, 949-324. Hussin, M.H., Rahim, A.A., Mohamad Ibrahim, M.N., Perrin, D., Brosse, N., 2015. Enhanced properties of oil palm fronds (OPF) lignin fractions produced via tangential ultrafiltration technique. Ind. Crops Prod. 66, 1–10. Jönsson, A.-S., Wallberg, O., 2009. Cost estimates of kraft lignin recovery by ultrafiltration. Desalination 237, 254–267. Jönsson, A.-S., Nordin, A.-K., Wallberg, O., 2008. Concentration and purification of lignin in hardwood kraft pulping liquor by ultrafiltration and nanofiltration. Chem. Eng. Res. Des. 86, 1271–1280. Jamshidian, M., Tehrany, E.A., Desobry, S., 2012. Release of synthetic phenolic

Fig. 7. The impact of the lignin samples (at various concentrations) on the oxidative stability of lipid-based sample (commercial mayonnaise) at 120 °C.

from 1.9 h to 2.27 and 3.39 h, respectively (Fig. 7 and Table 5). For the non-fractionated lignin, the stabilization effect was less pronounced. The addition of the LignoBoost lignin in low concentrations (2.5 mg per 1 g of mayonnaise) did not have any noticeable effect on the oxidation (induction period). The addition of 5.0 mg g−1 of LignoBoost lignin, on the other hand, had an obvious protection effect (induction period increased from 1.9 to 2.73 h). The addition of 5 mg g−1 of the low-Mw fraction was similar to the stabilization effect of the addition of 1 mg g−1 of the commercial antioxidant TBHQ (Table 5). 4. Conclusions

• For

• • • •

the model system used, the permeate fluxes were unstable during the 100 h membrane operation, but a decrease in the average Mw of the permeate during the operation time was observed. This shift was more pronounced for the virgin membrane, and a more stable Mw distribution was displayed as the number of cleaning cycles increased; The permeate fractions collected after 100 h had a similar average Mw regardless of the two investigated ionic strength levels and membrane pretreatments; Low-Mw fractions (in the range of 500 Da) had a lower total content of phenolic OH groups relative to the initial LignoBoost lignin, but the proportion of non-condensed phenolic groups was higher in these samples; The low-Mw lignin fraction demonstrated high antioxidant activity in the tests with ABTS•+ radical and the ORAC assay (positively charged free radicals), while the tests with DPPH% radical gave better results for the original lignin; The low-Mw lignin fraction demonstrated favorable stabilizing activity in the test with the lipid-based sample and thus could be considered a promising natural antioxidant in lipid-containing products.

Acknowledgments This work was supported by grants from the Knut and Alice Wallenberg Foundation in connection with the Wallenberg Wood Science Centre Program (WWSC). The cooperation of the Latvian Research grant LigProBK and State program VPP Nr.6 ResProd project Nr.3 is gratefully acknowledged. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.indcrop.2017.11.042. 208

Industrial Crops & Products 112 (2018) 200–209

S. Aminzadeh et al. antioxidants from extruded poly lactic acid (PLA) film. Food Control 28, 445–455. Kadla, J.F., Kubo, S., Venditti, R.A., Gilbert, R.D., Compere, A.L., Griffith, W., 2002. Lignin-based carbon fibers for composite fiber applications. Carbon 40, 2913–2920. Kevlich, N.S., Shofner, M.L., Nair, S., 2017. Membranes for Kraft black liquor concentration and chemical recovery Current progress, challenges, and opportunities. Sep. Sci. Technol. 52, 1070–1094. Keyoumu, A., Sjödahl, R., Henriksson, G., Ek, M., Gellerstedt, G., Lindström, M.E., 2004. Continuous nano- and ultra-filtration of kraft pulping black liquor with ceramic filters: a method for lowering the load on the recovery boiler while generating valuable side-products. Ind. Crops Prod. 20, 143–150. Kim, H., Ralph, J., 2010. Solution-state 2D NMR of ball-milled plant cell wall gels in DMSO-d(6)/pyridine-d(5). Org. Biomol. Chem. 8, 576–591. KoŠíková, B., Lábaj, J., Gregorová, A., Slameňová, D., 2006. Lignin antioxidants for preventing oxidation damage of DNA and for stabilizing polymeric composites. Holzforschung 60, 166. Kubo, S., Kadla, J.F., 2005. Lignin-based carbon fibers: effect of synthetic polymer blending on fiber properties. J. Polym. Environ. 13, 97–105. Lauberts, M., Sevastyanova, O., Ponomarenko, J., Dizhbite, T., Dobele, G., Volperts, A., Lauberte, L., Telysheva, G., 2017. Fractionation of technical lignin with ionic liquids as a method for improving purity and antioxidant activity. Ind. Crops Prod. 95, 512–520. Lawoko, M., Henriksson, G., Gellerstedt, G., 2005. Structural differences between the lignin-Carbohydrate complexes present in wood and in chemical pulps. Biomacromolecules 6, 3467–3473. Li, J., Bi, Y., Liu, W., Sun, S., 2015. Simultaneous analysis of tertiary butylhydroquinone and 2-tert-Butyl-1,4-benzoquinone in edible oils by normal-Phase high-Performance liquid chromatography. J. Agric. Food Chem. 63, 8584–8591. Lin, S.Y., Dence, C.W., 1992. Methods in Lignin Chemistry. Springer, Berlin/Verlag. Lora, J.H., Glasser, W.G., 2002. Recent industrial applications of lignin: a sustainable alternative to nonrenewable materials. J. Polym. Environ. 10, 39–48. Loutfi, H., Blackwell, B., Uloth, V., 1991. Lignin recovery from Kraft black liquor: preliminary process design. Tappi J. 74, 203–210. Lu, F.J., Chu, L.H., Gau, R.J., 1998. Free radical-scavenging properties of lignin. Nutr. Cancer 30, 31–38. Luterbacher, J., Alonso, D.M., Dumesic, J., 2014. Targeted chemical upgrading of lignocellulosic biomass to platform molecules. Green Chem. 16, 4816–4838. Malík, J., Kröhnke, C., 2006. Polymer stabilization: present status and possible future trends. C.R. Chim. 9, 1330–1337. Mansouri, N.-E.E., Salvadó, J., 2006. Structural characterization of technical lignins for the production of adhesives Application to lignosulfonate, kraft, soda-anthraquinone, organosolv and ethanol process lignins. Ind. Crops Prod. 24, 8–16. Marton, J., 1971. Reaction in alkaline pulping. In: Sarkanen, K.V., Ludwig,C, H. (Eds.), Lignin: Occurence, Formation, Structure and reactions. John Wiley & Sons, USA, pp. 639–694. Matsushita, Y., Imai, M., Iwatsuki, A., Fukushima, K., 2008. The relationship between surface tension and the industrial performance of water-soluble polymers prepared from acid hydrolysis lignin, a saccharification by-product from woody materials. Biores. Technol. 99, 3024–3028. Meier, D., Faix, O., 1992. In: Lin, S.Y., Dence, C.W. (Eds.), Methods in Lignin Chemistry. Springer-Verlag, Berlin, pp. 177. Mohan, D., Pittman Jr., C.U., Steele, P.H., 2006. Single, binary and multi-component adsorption of copper and cadmium from aqueous solutions on Kraft lignin–a biosorbent. J. Colloid Interface Sci. 297, 489–504. Ohra-aho, T., Tenkanen, M., Tamminen, T., 2005. Direct analysis of lignin and lignin-like components from softwood kraft pulp by Py-GC/MS techniques. J. Anal. Appl. Pyrol. 74, 123–128. Pan, X., Kadla, J.F., Ehara, K., Gilkes, N., Saddler, J.N., 2006. Organosolv ethanol lignin from hybrid poplar as a radical Scavenger: Relationship between lignin structure, extraction conditions, and antioxidant activity. J. Agric. Food Chem. 54, 5806–5813. Passinen, K., 1968. Chemical Composition of Spent Liquors, Proceeding of IUPAC EUCEPA Symposium on Recovery of Pulping Chemicals. pp. 183–209 (Helsinki, Finland). Podkościelna, B., Sobiesiak, M., Zhao, Y., Gawdzik, B., Sevastyanova, O., 2015. Preparation of lignin-containing porous microspheres through the copolymerization of lignin acrylate derivatives with styrene and divinylbenzene. Holzforschung 69, 769. Ponomarenko, J., Dizhbite, T., Lauberts, M., Viksna, A., Dobele, G., Bikovens, O., Telysheva, G., 2014. Characterization of softwood and hardwood LignoBoost kraft lignins with emphasis on their antioxidant activity. BioResources 9, 2051–2068. Ponomarenko, J., Dizhbite, T., Lauberts, M., Volperts, A., Dobele, G., Telysheva, G., 2015a. Analytical pyrolysis −A tool for revealing of lignin structure-antioxidant activity relationship. J. Anal. Appl. Pyrol. 113, 360–369. Ponomarenko, J., Lauberts, M., Dizhbite, T., Lauberte, L., Jurkjane, V., Telysheva, G.,

2015b. Antioxidant activity of various lignins and lignin-related phenylpropanoid units with high and low molecular weight. Holzforschung 69, 795. Pouteau, C., Dole, P., Cathala, B., Averous, L., Boquillon, N., 2003. Antioxidant properties of lignin in polypropylene. Polym. Degrad. Stab. 81, 9–18. Prior, R.L., Wu, X., Schaich, K., 2005. Standardized methods for the determination of antioxidant capacity and phenolics in foods and dietary supplements. J. Agric. Food Chem. 53, 4290–4302. Ragauskas, A.J., Beckham, G.T., Biddy, M.J., Chandra, R., Chen, F., Davis, M.F., Davison, B.H., Dixon, R.A., Gilna, P., Keller, M., Langan, P., Naskar, A.K., Saddler, J.N., Tschaplinski, T.J., Tuskan, G.A., Wyman, C.E., 2014. Lignin valorization: improving lignin processing in the biorefinery. Science 344 (1246843), 1–10. Robert, D.R., Bardet, M., Gellerstedt, G., Lindfors, E.L., 1984. Structural changes in lignin during kraft cooking part 3: on the structure of dissolved lignins. J. Wood Chem. Technol. 4, 239–263. Rodriges, J., Graca, J., Pereira, H., 2001. Influence of tree eccentric growth on syringyl/ guaiacyl ratio in Eucalyptus globulus wood lignin assessed by analytical pyrolysis. J. Anal. Appl. Pyrol. 58, 481–489. Salem, M., Abdel-Megeed, A., Ali, M.H., 2014. Stem wood and bark extracts of delonix regia (Boj. ex. hook): chemical analysis and antibacterial antifungal, and antioxidant properties. BioResources 9, 2382–2392. Sarkanen, K.V., Ludwig, C.H., 1971. Lignins: Occurrence, Formation, Structure and Reactions. Wiley-Interscience. Sarkar, S., Adhikari, B., 2000. Lignin-modified phenolic resin: synthesis optimization, adhesive strength, and thermal stability. J. Adhes. Sci. Technol. 14, 1179–1193. Sevastyanova, O., Helander, M., Chowdhury, S., Lange, H., Wedin, H., Zhang, L., Ek, M., Kadla, J.F., Crestini, C., Lindström, M.E., 2014. Tailoring the molecular and thermo–mechanical properties of kraft lignin by ultrafiltration. J. Appl. Polym. Sci. 131 (40799), 1–11. Sindhi, V., Gupta, V., Sharma, K., Bhatnagar, S., Kumari, R., Dhaka, N., 2013. Potential applications of antioxidants −A review. J. Pharm. Res. 7, 828–835. Sjöström, E., 1981. Wood Chemistry: Fundamentals and Applications. Academic Press Inc., Orlando, pp. 124–139 (Chapter 7). Tanistra, I., Bodzek, M., 1998. Preparation of high-purity sulphate lignin from spent black liquor using ultrafiltration and diafiltration processes. Desalination 115, 111–120. Tarabanko, V.E., Fomova, N.A., Kuznetsov, B.N., Ivanchenko, N.M., Kudryashev, A.V., 1995. On the mechanism of vanillin formation in the catalytic oxidation of lignin with oxygen. React. Kinet. Catal. Lett. 55, 161–170. Tejado, A., Peña, C., Labidi, J., Echeverria, J.M., Mondragon, I., 2007. Physico-chemical characterization of lignins from different sources for use in phenol–formaldehyde resin synthesis. Biores. Technol. 98, 1655–1663. Toledano, A., García, A., Mondragon, I., Labidi, J., 2010. Lignin separation and fractionation by ultrafiltration. Sep. Purif. Technol. 71, 38–43. Tomani, P., 2010. The LignoBoost process. Cell. Chem. Technol. 53–58. Trouillas, P., Marsal, P., Svobodova, A., Vostalova, J., Gazak, R., Hrbac, J., Sedmera, P., Kren, V., Lazzaroni, R., Duroux, J.L., Walterová, D., 2008. Mechanism of the antioxidant action of silybin and 2,3-dehydrosilybin flavonolignans: a joint experimental and theoretical study. J. Phys. Chem. A 112, 1054–1063. Tsutsumi, Y., Kondo, R., Sakai, K., Imamura, H., 1995. The difference of reactivity between syringyl lignin and guaiacyl lignin in alkaline systems. Holzforschung 49, 423–428. Villar, C.J., Caperos, A., García-Ochoa, F., 2001. Oxidation of hardwood kraft-lignin to phenolic derivatives with oxygen as oxidant. Wood Sci. Technol. 35, 245–255. Wallberg, O., Jönsson, A.S., 2003. Separation ProcessesInfluence of the membrane cut-off during ultrafiltration of kraft black liquor with ceramic membranes. Chem. Eng. Res. Des. 81, 1379–1384. Wallberg, O., Jönsson, A.-S., 2006. Separation of lignin in kraft cooking liquor from a continuous digester by ultrafiltration at temperatures above 100 °C. Desalination 195, 187–200. Wallberg, O., Jönsson, A.-S., Wimmerstedt, R., 2003. Fractionation and concentration of kraft black liquor lignin with ultrafiltration. Desalination 154, 187–199. Wallberg, O., Holmqvist, A., Jönsson, A.-S., 2005. Ultrafiltration of kraft cooking liquors from a continuous cooking process. Desalination 180, 109–118. Wang, C., Kelley, S.S., Venditti, R.A., 2016. Lignin-Based thermoplastic materials. ChemSusChem 9, 770–783. Yang, D., Qiu, X., Zhou, M., Lou, H., 2007. Properties of sodium lignosulfonate as dispersant of coal water slurry. Energ. Convers. Manage. 48, 2433–2438. Zhang, L., Gellerstedt, G., 2007. Quantitative 2D HSQC NMR determination of polymer structures by selecting suitable internal standard references. Magn. Reson. Chem. 45, 37–45. del Río, J.C., Rencoret, J., Prinsen, P., Martínez, Á.T., Ralph, J., Gutiérrez, A., 2012. Structural characterization of wheat straw lignin as revealed by analytical pyrolysis, 2D-NMR, and reductive cleavage methods. J. Agric. Food Chem. 60, 5922–5935.

209