On the antioxidant activity of lignin and the structural impacts Master’s thesis in Biotechnology Paria Asadzadehkhaneghah DEPARTMENT OF CHEMISTRY AND CHEMICAL ENGINEERING CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2025 www.chalmers.se http://www.chalmers.se/ MASTER’S THESIS 2025 On the antioxidant activity of lignin and the structural impacts Paria Asadzadehkhaneghah Forest Products and Chemical Engineering Unit Department of Chemistry and Chemical Engineering Chalmers University of Technology Gothenburg, SE-41296 Sweden 2025 iv On the antioxidant activity of lignin and the structural impacts Paria Asadzadehkhaneghah © Paria Asadzadehkhaneghah, 2025. Supervisor: Liyang Liu, Chalmers Ahilan Manisekaran, Chalmers Examiner: Liyang Liu, Chalmers Master’s Thesis 2025 Forest Products and Chemical Engineering Unit Department of Chemistry and Chemical Engineering Chalmers University of Technology SE-412 96 Gothenburg Telephone +46 31 772 1000 Cover: Lignin is an abundant natural resource existing in plants, including wood and grass. It is produced in large amounts (> 50 million tons) as a byproduct of the pulp and paper industry annually.[1] With its unique structure, lignin offers powerful antioxidant properties, making it a plentiful and sustainable antioxidant that nature has gifted us.[2] Photo taken by Felix Blomfelt. Typeset in Microsoft word Printed by Chalmers Reproservice Gothenburg, Sweden 2025 v On the antioxidant activity of lignin and the structural impacts Paria Asadzadehkhaneghah Department of Chemistry and Chemical Engineering Chalmers University of Technology Abstract The development of nature-based antioxidants is a crucial step toward a sustainable society, including lignin, an organic complex biopolymer in plants, offering versatile properties, particularly antioxidant characteristics. To effectively use lignin as a commercial antioxidant, it is essential to identify the specific structural features responsible for its antioxidant activity. The polyphenolic structure of lignin is particularly relevant, since these functional groups are found in many antioxidants. (e.g., vitamin E). This study investigates the correlation between phenolic hydroxyl group content (phenolic -OH) and antioxidant activity in seven types of lignin, including original and modified lignins. To quantify the phenolic-OH groups, conductometric titration and quantitative Phosphorus-31 nuclear magnetic resonance spectroscopy (31P NMR) were performed. Antioxidant activity was assessed using DPPH(2,2-diphenyl-1-picrylhydrazyl) and ABTS (2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)) radical scavenging methodologies. Furthermore, linear regression, Pearson, Spearman, and Kendall’s Tau correlation analyses were performed to evaluate the correlation between the number of phenolic -OH groups and antioxidant activity. Lastly, the results from Fourier transform infrared spectroscopy (FT-IR) were interpreted to investigate the presence of other lignin structures/functional groups that might influence antioxidant activity. A strong linear correlation between titration and 31P NMR results (R² = 0.97) reveals that conductometric titration can be a suitable quantitative method for measuring phenolic-OH content. Correlation results (R² = 0.51 for phenolic content vs DPPH; R²= 0.66 for phenolic content vs ABTS) indicate that the phenolic-OH content has a moderate effect on antioxidant activity, but it is not always the dominant factor. Other correlation analyses, including Kendall’s Tau, Pearson, and Spearman, have been performed in the study and have shown similar trends. For instance, hydroxyethyl softwood kraft lignin (the modified lignin without any phenol groups) shows comparable antioxidant properties to the original lignin. This finding ultimately suggests that other structural factors, such as chromophores (conjugation and carbonyl groups), may play a crucial role in lignin antioxidant activity. Keywords: Lignin, phenolic content characterization, antioxidant activity, titration, 31P NMR, FT- IR, DPPH, ABTS, chromophore, conjugated carbonyl. vi Acknowledgements I would like to express my deepest gratitude to my supervisors, Liyang Liu and Ahilan Manisekaran, for their invaluable guidance, unwavering support, and constant encouragement throughout this research. Their wisdom and expertise have been a source of inspiration, and I am truly fortunate to have had their mentorship. I am also profoundly grateful to Yuge Yao and Elahe Sharifi for their generous support, insightful guidance, and kindness during my thesis journey. Your patience and willingness to help have made this experience both enriching and rewarding. I would also like to thank Emilia Rózsa, without whom I wouldn’t be where I am today. Most importantly, I would like to thank my family for their unconditional love and support. Their belief in me has been my greatest strength. This journey has been one of growth, challenges, and discoveries, and I could not have navigated it without such incredible mentors and friends. Thank you for your belief in me, your time, and your unwavering support. I will always cherish the knowledge and experiences gained along the way. Paria Asadzadehkhaneghah, Gothenburg, April 2025 vii List of Acronyms Below is the list of acronyms that have been used throughout this thesis listed in alphabetical order: ABTS 2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) A Absorbance DPPH 2,2-diphenyl-1-picrylhydrazyl FT-IR Fourier transform infrared spectroscopy 31P NMR Phosphorous nuclear magnetic resonance spectroscopy RSA Radical scavenging activity BDE Bond dissociation energy wt. Weight Mw Weight average molecular weight DS Degree of substitution Mn Number average molecular weight OSHL Organoslov hardwood lignin SKL1 Softwood kraft lignin 1 SKL2 Softwood kraft lignin 2 SKL3 Softwood kraft lignin 3 HEOSHL Hydroxyethyl organosolv hardwood lignin HESKL1 Hydroxyethyl softwood kraft lignin Amine SKL3 Aminated softwood kraft lignin BHA Butylated hydroxyanisole BHT Butylated hydroxytoluene PG Propyl gallate TBHQ Tert-butyl hydroquinone EHL Enzymatic hydrolysis Lignin HAT Hydrogen atom transfer SET Single electron transfer SPLET Sequential proton loss electron transfer PCET Proton coupled electron transfer IS Internal standard R2 Coefficient of determination SST Total sum of squares SSR Regression sum of squares SSE Error sums of squares ρ̂(r) Sample Pearson correlation coefficient between u and v viii Nomenclature Below is the nomenclature of indices, sets, parameters, and variables that have been used throughout this thesis. Sets Symbol Description Unit {𝑦𝑖} Set of all experimental data points - {𝑦𝑖̂} Set of all model-predicted values - {𝑢𝑖} Sets of observed 𝑢 data points - {𝑣𝑖} Sets of observed 𝑣 data points - {𝑞𝑖} Sets of all ranks for variable 𝑞 - {𝑟𝑖} Sets of all ranks for variable 𝑟 - Parameters Symbol Description Unit 𝑏0 Intercept of the regression line (constant term) - 𝑏1 Slope of the regression line - 𝑛 Total number of samples - 𝜏̂ Kendall rank correlation coefficient - 𝑛𝑡 The total number of concordant and discordant pairs in the sample - 𝑛𝑐 The number of concordant pairs - 𝑛𝑑 The number of discordant pairs - 𝐶𝐻𝐶𝑙 Concentration of hydrochloric acid M M(mol/L) 𝑚𝑙𝑖𝑔𝑛𝑖𝑛 Mass of lignin sample g 𝑀𝑁𝐻𝑁𝐷 NHND molar mass g/mol Variables Symbol Description Unit 𝑥 Independent variable (predictor) - 𝑦̂ Estimate mean value of the response variable - 𝑦𝑖 The experimental data points/observations - 𝑦̅ The mean value of the observations - 𝑦𝑖̂ The data points predicted by the model - 𝑢𝑖 Measured value of variable 𝑢 - 𝑣𝑖 Measured value of variable 𝑣 - 𝑢̅. Mean of 𝑢 values - 𝑣̅. Mean of 𝑣 values - 𝑞𝑖 Rank of the 𝑖-th observation in variable 𝑞 - 𝑟𝑖 Rank of the 𝑖-th observation in variable 𝑟 - 𝑞̅. Mean rank of 𝑞 values - 𝑟̅. Mean rank of 𝑟 values - 𝑁[𝐶6𝐻5𝑂𝐻] Number of phenol groups per gram of lignin. mmol/g 𝑉𝐵,𝐻𝐶𝑙 Total volume of HCl added at point B. mL 𝑉𝐴,𝐻𝐶𝑙 Total volume of HCl added at point A. mL 𝑁[𝑅 − 𝑂𝐻] 𝑔 𝑜𝑓 𝑙𝑖𝑔𝑛𝑖𝑛 The number of OH groups per gram of lignin mmol/g 𝐼𝑂𝐻 Area under the NMR peak corresponding to phenolic OH - ix 𝐼𝑁𝐻𝑁𝐷 Area under the NMR peak of the internal standard (NHND) - 𝑅 Integration ratio of the spectral region of interest (IOH) over the internal standard region (INHND) - 𝑚𝑁𝐻𝑁𝐷 [𝑔] Mass of NHND used in the NMR sample g 𝑅𝑆𝐴𝐷𝑃𝑃𝐻 Radical scavenging activity against DPPH radicals % 𝑅𝑆𝐴𝐴𝐵𝑇𝑆 Radical scavenging activity against ABTS radicals % 𝐴𝐵𝑆𝑡=0 Initial solution absorbance at time 0 min 𝐴𝐵𝑆𝑡=12 𝑚𝑖𝑛 Solution absorbance after 12 minutes incubation min 𝐴𝐵𝑆𝑡=30 𝑚𝑖𝑛 Solution absorbance after 30 minutes incubation min x Contents Introduction ------------------------------------------------------------------ 1 1.1 Aim ----------------------------------------------------------------------------------------------- 2 Theory ------------------------------------------------------------------------- 3 2.1 What is lignin ----------------------------------------------------------------------------------- 3 An introduction to the lignin structure ------------------------------------------------------------ 3 2.1.1 Lignin --------------------------------------------------------------------------------------- 3 2.1.2 Structure ----------------------------------------------------------------------------------- 3 2.1.3 Species-based variations in lignin structure ----------------------------------------- 5 2.2 Lignin types ------------------------------------------------------------------------------------- 5 Industrial-based variations in the lignin structure, processing techniques, and common modifications ----------------------------------------------------------------------------------------- 5 2.2.1 Lignin’s current place in industry and potential ------------------------------------ 5 2.2.2 Industrial-based variations in lignin structure -------------------------------------- 5 2.2.3 Lignin modifications --------------------------------------------------------------------- 7 2.2.4 Lignins studied in this project ---------------------------------------------------------- 9 2.3 Antioxidant activity of lignin ---------------------------------------------------------------10 An insight into lignin's antioxidant activity, structural impact, and limitations -------------10 2.3.1 Antioxidant activity ---------------------------------------------------------------------10 2.3.2 Lignin as an antioxidant ----------------------------------------------------------------11 2.3.3 Antioxidant activity measurement ----------------------------------------------------12 2.3.4 Limitations and challenges -------------------------------------------------------------13 2.4 Correlation statistical analysis -------------------------------------------------------------14 2.4.1 Linear regression analysis --------------------------------------------------------------14 2.4.2 Pearson analysis --------------------------------------------------------------------------15 2.4.3 Spearman analysis -----------------------------------------------------------------------16 2.4.4 Kendall’s Tau analysis ------------------------------------------------------------------16 Materials and Methods ----------------------------------------------------21 3.1 Lignin, materials, and chemicals -----------------------------------------------------------21 xi 3.1.1 Lignin resources -------------------------------------------------------------------------21 3.1.2 Chemicals ---------------------------------------------------------------------------------22 3.2 Phenol group characterization -------------------------------------------------------------22 3.2.1 Conductometric titration analysis ----------------------------------------------------22 3.2.2 31P NMR spectroscopy ------------------------------------------------------------------23 3.3 FT-IR analysis ---------------------------------------------------------------------------------24 3.4 Antioxidant activity measurement ---------------------------------------------------------24 3.4.2 ABTS+ radical scavenging assay ------------------------------------------------------25 Results ------------------------------------------------------------------------26 4.1 Characterization of phenolic groups in lignin -------------------------------------------26 4.1.1 Data analysis ------------------------------------------------------------------------------22 4.1.2 Linear regression analysis --------------------------------------------------------------22 4.1.3 Kendall’s Tau, Spearman’s, and Pearson’s correlation analysis ----------------23 4.2 Antioxidant characterization of lignin ----------------------------------------------------23 4.2.1 Data analysis ------------------------------------------------------------------------------23 4.2.2 Linear regression analysis --------------------------------------------------------------24 4.2.3 Kendall’s Tau, Spearman’s, and Pearson’s correlation analysis ----------------25 4.3 Correlation analysis between phenol group quantity and antioxidant activity ----25 4.3.1 31P NMR Vs radical scavenging activity ---------------------------------------------25 4.4 Fourier Transform Infrared Spectroscopy (FT-IR) analysis --------------------------27 Discussion --------------------------------------------------------------------36 5.1 Conductometric titration analysis vs 31P NMR spectroscopy -------------------------36 5.1.1 Improvements ----------------------------------------------------------------------------36 5.2 DPPH vs ABTS+ radical scavenging assay -----------------------------------------------36 5.2.1 Improvements ----------------------------------------------------------------------------37 5.3 Phenol group vs antioxidant activity ------------------------------------------------------38 5.3.1 Improvement -----------------------------------------------------------------------------38 5.4 FT-IR analysis and identification of conjugated carbonyl groups -------------------38 5.4.1 Improvement -----------------------------------------------------------------------------39 Conclusions ------------------------------------------------------------------36 xii List of Figures Figure 1. The complex and heterogeneous polymer structure of lignin within the secondary cell walls of plants. Figure reproduced from Balk et al. (2023) [13] ................................... 3 Figure 2. (A) Chemical structures of three key cinnamic alcohols in the lignin structure: p- coumaryl alcohol (4-hydroxycinnamyl), coniferyl (3-methoxy-4-hydroxycinnamyl), and Sinapyl (3,5-dimethoxy-4-hydroxycinnamyl). (B) Structure of Monolignols (A) and softwood Kraft lignin (B), reproduced with permission from Crestini, C, et al [15] courtsy of Ahilan Manisekaran. ............................................................................................................ 4 Figure 3. A picture of the Kraft process. Adopted with permission from Ref.[21] ................ 6 Figure 4. A picture of a) LignoBoost®; and b) LignoForce® lignin recovering techniques. Adapted with permission from Ref [21] ........................................................................ 7 Figure 5. Reaction scheme of green Hydroxy-ethylating process using ethylene carbonate. Adopted from Ref.[26] ............................................................................................... 8 Figure 6. Reaction scheme of green animation process using 2-oxazolidinone (OZD). Adopted from Ref.[25] ............................................................................................... 8 Figure 7. Intracellular imbalance between ROS (reactive oxygen species) and antioxidant concentrations and its consequences. Adopted from Ref.[28].......................................... 10 Figure 8. Reaction mechanism of (A) DPPH(2,2-diphenyl-1-picrylhydrazyl) and (B) ABTS (2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid)) assays. Adapted from ref.[32] ..... 13 Figure 9. The picture of the titration curve (a) and its derivative (b). ............................... 22 Figure 10. Results from phenol group characterization using conductometric titration and 31P NMR for OSHL, SKL1, SKL2, and SKL3 lignin. 31P NMR results are based on one sample per analysis. i.e., no error bars. .................................................................................. 26 Figure 11. The linear regression analysis between the data obtained from conductometric titration, visualized on the x axis, and 31P NMR spectroscopy, visualized on the y axis. ...... 22 Figure 12. The Kendall’s Tau, Spearman’s, and Pearson’s correlation coefficient between the data obtained from conductometric titration and 31P NMR spectroscopy. .......................... 23 Figure 13. The radical scavenging activity (RSA) measured by DPPH (a) and ABTS (b) assays. ................................................................................................................... 23 Figure 14. The linear regression analysis between the data obtained from DPPH radical scavenging activity, visualized on the x axis and ABTS radical scavenging activity, visualized on the y axis. .......................................................................................................... 24 Figure 15. The Kendall’s Tau, Spearman’s, and Pearson’s correlation coefficient between the data obtained from DPPH and ABTS analysis. ............................................................. 25 Figure 16. The linear regression correlation analysis between 31P NMR results and Radical scavenging activity: DPPH (a), and ABTS (b). ............................................................. 25 Figure 17. The Spearman’s, Pearson’s, and Kendall’s Tau correlation coefficient between the data obtained from (a) 31P NMR spectroscopy and DPPH analysis, and (b) 31P NMR spectroscopy and ABTS analysis. ............................................................................... 26 xiii Figure 18. FT-IR spectra of OSHL, SKL1, SKL2, SKL3, and Amine SKL3. .................... 27 Figure 19. Experiment set up for conductometric titration ............................................. 39 Figure 20. Implied instruments in the experiment set up for conductometric titration. ........ 39 Figure 21. The picture of the titration curve (a) and its derivative (b). ............................. 40 Figure 22. The example titration curve (a) and the derivate of the titration curve (b) ......... 41 Figure 23. Set up for dissolving the Amine SKL3 lignin in sodium hydroxide solution (0.1M) by applying heat (~50 °C) ......................................................................................... 43 xiv List of Tables Table 1. Presents the abbreviations used for the investigated lignin in this study. ................ 9 Table 2. Interpretation of model fits based on the coefficient of determination (R2). Adapted from ref.[37] ........................................................................................................... 15 Table 3. Overview of the key characteristics, relevant information, and corresponding sources of the different lignin types analyzed. .............................................................. 21 Table 4. An overview of the methodology characteristic of conductometric titration and 31P NMR Spectroscopy.................................................................................................. 24 Table 5. Summarizes the identified chemical structures and functional groups in peak cluster 3 for OSHL, SKL1, SKL2, SKL3, and Amine SKL. The points are identified by [25] and the reference chart from Thermo Fisher.[44] ..................................................................... 28 Table 6. Qualitative analysis of chromophore content in all lignin types by comparing the ration between peaks 4 and 8a. .................................................................................. 29 Table 7. The phenolic -OH content (N[-OH]), and the antioxidant activity (RSA) and their ranking. ................................................................................................................. 38 Table 8. Data from conductometric titration analysis of SKL2. ....................................... 41 Table 9. Determined points of Phenolic -OH group’s reacting interval (A and B), and carboxyl group’s reaction area (C and D) .................................................................... 41 iv 1 1 Introduction Lignin is one of the most generous gifts of nature, offering remarkable potential in various applications, including packaging[3], biofuels[4], and the food/biomedical industry[3] as an antioxidant. Lignin’s benefits stem from its unique structural features. With strategic investment in studying lignin’s structure and properties, this abundant byproduct can provide distinctive characteristics on an industrial scale.[5] Lignin is a complex polymer and a key component of plant cell walls, playing a crucial role in plant strength, structure, and overall viability. It provides essential protection for plants against fungi, bacteria[6], UV light[7], and moisture[6]. These diverse properties arise from a complex chemical structure composed mainly of three cinnamic alcohols, which form a phenylpropanoid backbone.[6] These couplings involve various C–C and C–O bonds, such as β-aryl ether (β-O-4), with differing bond strengths.[6] Lignin is rich in functional groups, such as carboxyl, hydroxyl, carbonyl, and phenol groups, which can contribute to its unique properties, particularly its antioxidant activity.[8] However, the polyphenolic structure of lignin is widely regarded as the key contributor to its antioxidant activity since this structural characteristic is also observed in various antioxidants, such as vitamin E.[8], [9] Antioxidants are crucial for protecting the body against oxidative stress, a condition that can arise from an imbalance between free radicals and the body's ability to neutralize them. Free radicals are highly reactive molecules with unpaired electrons that play a significant role in physiological processes. However, when their concentration becomes unbalanced due to environmental factors like UV radiation and heat, free radicals can trigger oxidative stress, increasing the risk of cardiovascular disease, cancer, and other health problems. These health conditions occur due to the reaction of the free radical with DNA bases, amino acid side chains in proteins, and double bonds in unsaturated fatty acids.[8], [10] Antioxidants can react with free radicals by exchanging electrons and protecting against oxidative stress. They can either completely neutralize the free radicals or convert them to less harmful forms. The intake of antioxidants is significant for enhancing the body's natural defense system, which includes enzymatic activity, metal chelation, and scavenging of free radicals.[10] Therefore, studying natural antioxidants like lignin and utilizing their radical scavenging properties is of great importance. However, fully exploiting the potential of lignin, particularly as an antioxidant, presents several challenges. Effective characterization of lignin, especially technical lignin obtained from industrial processes, remains a significant challenge. The monomeric construction of lignin varies in different plant species. Additionally, different extraction and purification processes lead to further variations in the lignin’s structure.[6] Therefore, establishing a reliable structure-property relationship is essential for identifying the specific structural features that influence lignin’s antioxidant properties. Conventional characterization methodologies only provide partial insight into the lignin’s complex structure. For instance, phosphorous nuclear magnetic resonance spectroscopy, 31P NMR, can selectively identify hydroxyl groups on the lignin surface. However, factors such as high cost, complexity of the 1. Introduction 2 methodology, toxicity of the reagents used, and concerns about stability and accuracy are limiting the effectiveness of this method.[11] To overcome these limitations, innovative and more efficient characterization techniques are needed. It is crucial to identify functional groups and other elements in the lignin structure that are responsible for its antioxidant properties and determine the relative magnitude of their influence. Several studies indicated that lignin's antioxidant activity is due to its polyphenolic structure. Therefore, evaluating the impact of phenolic functional groups on lignin's antioxidant properties is of great importance. 1.1 Aim This study aims to investigate whether the number of phenolic -OH groups is the dominant factor of antioxidant activity. To accomplish this objective, the project comprises three steps. (1) To quantify phenolic-OH groups in various types of lignin using conductometric titration and ³¹P NMR spectroscopy. It also allows for a comparison between these two methods in terms of accuracy, suitability, and precision. (2) To measure the antioxidant activity in various types of lignin using DPPH (2,2-diphenyl-1- picrylhydrazyl) and ABTS (2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid)) radical scavenging assays, and analyze whether there is a clear relation between antioxidant activity and the number of phenol-OH content. (3) To look into the lignin structure and check for other structural properties that might influence antioxidant activity. The results from these steps can provide data to investigate whether there is a direct correlation between the phenolic-OH group quantity and the antioxidant activity in the lignin. 3 2 Theory 2.1 What is lignin An introduction to the lignin structure 2.1.1 Lignin Lignin is an abundant biopolymer in plant cell walls, providing strength and hydrophobicity for the lignocellulosic matrix.[6] Figure 1 illustrates lignin within lignocellulosic biomass. Lignin accounts for 15–30% of the mass fraction of lignocellulosic biomass.[3] It contributes to maintaining strength by protecting cellulose and hemicellulose and reducing biodegradation.[6] Importantly, lignin functions as a natural antioxidant protecting plant cells against oxidative stress.[5] 2.1.2 Structure Lignin has an aromatic structure composed of various functional groups and building blocks. Its structure is mainly composed of phenylpropane units, including three key cinnamic alcohols: sinapyl (3,5-dimethoxy-4-hydroxycinnamyl), coniferyl (3-methoxy-4-hydroxycinnamyl), and p-coumaryl alcohol (4-hydroxycinnamyl). These alcohols, also known as S-units, G-units, and H-units, form the Figure 1. The complex and heterogeneous polymer structure of lignin within the secondary cell walls of plants. Figure reproduced from Balk et al. (2023) [13] 2. Theory 4 fundamental structures and build up phenylpropanoid groups, carbonyl groups (e.g., aldehyde, ketone, and carboxyl groups), and phenolic hydroxyl groups(phenolic-OH).[3] The radical coupling between these phenylpropane units occurs by various C-C and C-O linkages with different bond dissociation energies (BDE). Some of these crosslinking molecules are β-aryl ether (β-O-4), α-aryl ether (α-O-4), 1,2-diaryl propone (β-1), phenyl coumaran (β-5), biphenyl (5–5, α-1), and diphenyl ether (5-O-4). Figure 2 presents the chemical structure of softwood kraft lignin. The β-O-4 bond exhibits the lowest BDE of all lignin linkages, making it the most susceptible to cleavage. This cleavage generates new aromatic hydroxyl groups, including phenolic hydroxyls, which can potentially enhance antioxidant activity by increasing the phenolic content.[3] Chromophores are regions in lignin's structure that are associated with UV absorption properties and color variations.[7], [14] The chromophore components in lignin are mainly structures related to quinoids, catechols, aromatic ketones, stilbenes, and conjugated carbonyls with phenolics. Also, additional chromophores might be generated through oxidation or during the pulping process.[7] These structures are rich in conjugated functional groups. The conjugated double bonds can stabilize the phenolic radical formed after free-radical neutralization, so-called phenoxyl radical (Ar–O•), by resonance delocalization and stereoelectronic effects.[14] Figure 2. (A) Chemical structures of three key cinnamic alcohols in the lignin structure: p-coumaryl alcohol (4- hydroxycinnamyl), coniferyl (3-methoxy-4-hydroxycinnamyl), and Sinapyl (3,5-dimethoxy-4-hydroxycinnamyl). (B) Structure of Monolignols (A) and softwood Kraft lignin (B), reproduced with permission from Crestini, C, et al [15] courtsy of Ahilan Manisekaran. https://www.sciencedirect.com/topics/chemistry/quinonoid-zwitterion https://www.sciencedirect.com/topics/chemistry/catechol https://www.sciencedirect.com/topics/chemistry/aromatic-ketone https://www.sciencedirect.com/topics/chemistry/stilbene 2. Theory 5 2.1.3 Species-based variations in lignin structure The structure of the lignin varies between plant species. Softwood plants, such as fir, pine, and spruce, are gymnosperms and originate from coniferous trees. Hardwood plants, such as oaks, maples, and birches, are angiosperms and originate from deciduous trees.[6] Based on this classification, the lignocellulosic biomass also divides into softwood and hardwood. Additionally, a third category of lignocellulosic biomass originates from annual grasses.[6], [8] The weight percentage of lignin content varies in different biomass categories. The lignin content constitutes 30 wt.% in softwood, 20-25 wt.% in hardwood, and 10–15 wt.% in grass-based wood biomass.[8] Additionally, the monomeric structure of lignin varies in different plant species. In hardwoods, Lignin mainly contains S-units (50–75%) and G-units (25–50%), with a low number of H- units. In softwoods, almost all the lignin consists of G-units (90–95%), with very few S-units (0–1%) and a small percentage of H-units (0.5–3.4%). Grasses have a mix, with 25–50% G-units, 25–50% S- units, and 10–25% H-units.[5] 2.2 Lignin types Industrial-based variations in the lignin structure, processing techniques, and common modifications 2.2.1 Lignin’s current place in industry and potential The original form of lignin, which exists in lignocellulosic biomass, is not directly available, as it undergoes structural modifications during the extraction process. Lignin is produced in large quantities as a by-product in pulp and paper industries, referred to as technical lignin.[16] Lignin holds significant potential across a range of applications due to its unique characteristics and its availability. The application of lignin in food packaging has been shown to enhance UV protection, anti-oxidation, and protection against oxygen.[17] The strong and stable structure of lignin can improve the mechanical strength and thermostability of polymer materials.[7] Introducing plastic materials with lignin-based macromolecules (e.g., as a reinforcement/blend) can reduce their environmental carbon footprint.[18] Due to its higher energy and carbon content per mass compared to other bio polymers such as cellulose, lignin is a good candidate as a biofuel source.[19] Due to its UV-absorbing and oxidation-resistant properties, Lignin’s application as a natural ingredient in sunscreen products is under investigation.[20] Several techniques have been developed for lignin separation and recovery from the pulping products for various applications. 2.2.2 Industrial-based variations in lignin structure Lignin is a heterogeneous polymer that exists in various types. These differences arise from variations in its molecular structure and size, which are influenced not only by the plant species but also by various industrial processes. Lignin undergoes slight structural changes during different processing steps and techniques in the paper industry. Lignin is isolated from wood through various pulping methods, including the Kraft, Soda, and Organosolv processes.[6] 2.2.2.1 Kraft pulping Currently, Kraft pulping (Figure 3) accounts for 90% of the produced chemical pulp worldwide. It is responsible for a significant portion of the lignin produced, known as kraft lignin. Kraft pulping induces fragmentation and dissolution of lignin through a series of chemical reactions and leads to the cleavage of the β-O-4’ alkyl-aryl ether bonds in its structure. In some cases, the monomeric or oligomeric lignin 2. Theory 6 fragments released from this process can undergo radical redox reactions, leading to recondensation processes and fragmentation, and reduction processes. Nevertheless, the Kraft process offers valuable advantages, including a high quantity of the initial product and relatively low production cost.[21] After the pulping process, lignin can be extracted from the pulp mill using several techniques. It can either be recovered as crude black liquor or extracted by precipitation. LignoForce® and LignoBoost® are some examples of industrial extraction technologies to obtain high-purity lignin from black liquor. These methods extract lignin by decreasing its solubility in the liquor through acidification, followed by lignin separation via filtration.[21] The LignoBoost® extraction process consists of two main procedures: precipitation and filtration. The precipitation step occurs by acidifying black liquor (with about 40% dry substance) using carbon dioxide. During this step, lignin's phenolic compounds will be protonated and precipitated as solid particles. The resulting slurry is filtered by a filter press, yielding a solid composed of moist lignin with high ash content. The ash content is separated by acidic washing with sulfuric acid (pH 2-3)[22], resulting in the production of lignin with high purity and less than 1% of ash. A final filtration step separates the final lignin product, which has a 35% moisture content. Figure 4a illustrates the LignoBoost® process. This technique has been applied in two commercial plant companies so far: Stora Enso's Sunila mill in Finland, with a demonstration plant also operating in Bäckhammar, Sweden, and Domtar's Plymouth mill in North Carolina, USA.[21] The LignoForce® process is like the LignoBoost® process with some minor modifications (Figure 4b). This process is relatively simple and can reduce toxic H2S emissions. Briefly, the LignoForce® process starts with the oxidation of black liquor, which alters the particle-forming behavior of lignin. This alteration eliminates the additional filtration steps performed in the LignoBoost® process. In this extraction, the filtration occurs during a single filter press, which also includes washing and protonation of lignin with sulfuric acid. Despite this, the final lignin will have similar characteristics to LignoBoost® lignin.[21] According to Maria Juliane Suota et al.[6], softwood LignoForce® lignin has Figure 3. A picture of the Kraft process. Adopted with permission from Ref.[21] 2. Theory 7 higher total OH content, lower β-O-4′, and higher molecular weight (Mw) than hardwood lignin. While hardwood Lignin is less condensed, more soluble, and less stable than softwood Lignin.[6] 2.2.2.2 Organosolv pulping Organosolv processing uses organic solvents to extract lignin, the extracted lignin is known as Organosolv lignin. Organosolv extraction techniques can be an alternative to the kraft process. Similar to Kraft pulping, the breakage of ether bonds leads to the formation of new phenol groups.[23], [24] The lignin product from the Organosolv process is highly phenolic, relatively hydrophobic, and less contaminated. Compared to Kraft lignin, it has a lower molecular weight and higher purity (less ash content, and carbohydrate content).[24] This process is also free of sulfur. However, this process is not cost-effective and thus not suitable for industrial scale.[24] 2.2.3 Lignin modifications Although lignin has a lot of potential, the practical utilization of this material has limitations due to its complicated chemical structure, thermal instability, and low reactivity.[25] Modification of the lignin is typically performed with the aim of improving or adding specific characteristics to this component, which in turn enables its use in various applications. These improvements can include reactivity, solubility, blending, and composite making. Lignin modifications can be performed through different mechanisms. Some of the examples are alkylation, esterification, etherification, phenolation, and urethanization.[12] In this study, we used aminated and hydroxy-ethylated lignin, produced by Liyang Liu et.al. Figure 4. A picture of a) LignoBoost®; and b) LignoForce® lignin recovering techniques. Adapted with permission from Ref [21] 2. Theory 8 2.2.3.1 Etherified lignin derivatives (HEOSHL & HESKL1) Hydroxy-ethylated (HE) softwood kraft lignin (SKL1) and organosolv hardwood lignin (OSHL) were produced through a green process by Liu et al.[26] An excess amount of ethylene carbonate (8 g), abbreviated as EC, was mixed with 5 g dried lignin powders with a molar ratio equal to EC/(ArOH + COOH) = 6.6 . The reaction was performed in a 50 ml round-bottom flask at 80 –120 °C for 0–6 h. The flask was sealed with a rubber-septa, Teflon film, and parafilm to avoid leakage. The CO2 gas was collected using a self-designed analytical equipment and recorded as a function of time.[26] This modification replaces phenolic -OH groups with aliphatic -OH groups in OSHL and SKL, forming of HESKL1 and HEOSHL. For the HESKL1, the weight-average molecular weight (Mw) was 278.5 kDa, corresponding to 86 mL/g CO₂ released during the reaction, and the degree of substitution (DS) were 88%.[26] These numbers were 14.9 kDa Respective 91% for the HEOSHL lignin.[27] Figure 5 briefly presents the reaction mechanism of HE-modification. 2.3.3.2 Aminated lignin (Amine SKL3) Liu et al.’s aminated lignin building blocks are produced from the softwood kraft Lignin. The resulting lignin demonstrated enhanced reactivity and altered solubility. During this process, the 2-oxazolidinone (OZD) was used as a solvent and reagent, and the modification was performed for 2-4 hours at 150 °C in the presence of an alkaline catalyst (NaOH) (Figure 6). During the reaction, the deprotonated ArOH and COOH groups in lignin were selectively modified by the OZD. This process leads to the formation of primary amine groups from aminoethylation or secondary amine groups arising from urea-linked hydroxyethyl groups. (DS= 66% and Molecular Weight (Mn) = 9.0 kDa) More specifically, this process leads to the replacement of phenolic -OH groups with amine groups, resulting in the creation of Aminated building blocks from softwood kraft lignin.[25] Figure 6. Reaction scheme of green animation process using 2-oxazolidinone (OZD). Adopted from Ref.[25] Figure 5. Reaction scheme of green Hydroxy-ethylating process using ethylene carbonate. Adopted from Ref.[26] https://www.sciencedirect.com/topics/engineering/lignin-powder https://www.sciencedirect.com/topics/engineering/molar-ratio https://www.sciencedirect.com/topics/engineering/teflon-film 2. Theory 9 2.2.4 Lignins studied in this project Seven lignin variants were investigated in this study. The selected lignins represent diverse categories based on species origin (e.g., hardwood, softwood), processing methods (e.g., LignoBoost®, LignoForce®), and chemical modifications (i.e., native and modified forms). A brief description of each lignin variant is provided below. Organosolv hardwood lignin (OSHL) is kindly donated by our collaborators at Suzano (Brazil). It is extracted from Eucalyptus trees using ethanol/water with acidic catalysis. Softwood kraft lignin 1 (SKL1) is provided by Domtar Corp. (US). This lignin is isolated from black liquor generated in the Kraft pulping process via the LignoBoost® method. This lignin usually has the lowest ash content compared with other lignin resources. Softwood kraft lignin 2 (SKL2) is another kind of alkali lignin from Ingevity Corporation (US, South Carolina). This lignin was also extracted from black liquor produced during Kraft pulping process. The raw wood materials are softwood, but the purification process is slightly different from the LignoBoost® process. Softwood kraft lignin 3 (SKL3) is isolated from black liquor via the LignoForce® process. This lignin is kindly donated by WestFraser (Canada, Alberta). Unlike SKL1 and SKL2, this lignin contains a significant amount of acidity, requiring an additional washing procedure before characterization. Hydroxyethyl softwood kraft lignin (HESKL1) is acquired by modifying SKL1 using ethylene carbonate. Previous studies have described the modification methods.[26] This ethylene carbonate can selectively functionalize the phenol and carboxylic acid groups on the lignin surface, leaving primary aliphatic OH groups. After the modification, lignin is quite difficult to dissolve in organic solvents and aqueous alkali, except DMSO. Hydroxyethyl hardwood organosolv lignin (HEOSHL) is obtained by a similar modification to the HESKL1 methods with slightly different conditions. The organosolv hardwood lignin is used as a raw material, resulting in products that usually have good solubility in organic solvents than HESKL1. Aminated softwood kraft lignin (Amine SKL3) is acquired by modifying SKL3 using OZD.[25] The OZD can selectively convert phenol groups on the lignin surface into amine groups, associated with an increase in molar mass and changes in solubility parameters. As such, the aminated lignin has difficulty dissolving organic solvents and aqueous alkali, except in DMSO. Table 1. Presents the abbreviations used for the investigated lignin in this study. Hardwood organosolv lignin OSHL Softwood kraft lignin 1 SKL1 Softwood kraft lignin 2 SKL2 Softwood kraft lignin 3 SKL3 Hydroxyethyl hardwood organosolv lignin HEOSHL Hydroxyethyl softwood kraft lignin HESKL1 Aminated softwood kraft lignin Amine SKL3 2. Theory 10 2.3 Antioxidant activity of lignin An insight into lignin's antioxidant activity, structural impact, and limitations 2.3.1 Antioxidant activity The antioxidant characteristic refers to the ability of a substance to protect cells from damage caused by free radicals.[9] This protection occurs by preventing the oxidation of molecules through scavenging free radicals. Free radicals are molecules with one or more unpaired electrons, which prompts them to stabilize themselves by reacting with other molecules. Factors such as high temperature, oxidants, or ionizing radiation caused by UV light can lead to the formation of free radicals.[8], [9] Figure 7 provides a brief overview of intracellular oxidative stress, its consequences, and the role of antioxidants in preventing it. Free radicals accumulate excessively when the production of reactive oxygen species (ROS) is imbalanced due to the above-mentioned factors. ROS refers to a group of oxygen-containing molecules generated during physiological processes. An imbalance in the production rate of ROS can result in oxidative stress, leading to the formation of various free radicals, including peroxyl (ROO•) and hydroxyl(•OH) radicals. These molecules are highly reactive and unstable at room temperature. Excessive ROS can react with cell membranes, DNA, and other biomolecules, leading to protein degeneration, lipid peroxidation, and cell death.[8] These unwanted reactions can lead to rapid aging and cancer.[8], [9] The efficacy of antioxidants depends on various factors, such as chemical structure, concentration, and the antioxidant reactivity with ROS.[29] Antioxidants can be natural or synthetic.[8] Natural antioxidants refer to the antioxidants that exist in fruits and vegetables, including vitamin C, tocopherols, carotenoids, and flavonoids. Synthetic antioxidants are developed in laboratories. Figure 7. Intracellular imbalance between ROS (reactive oxygen species) and antioxidant concentrations and its consequences. Adopted from Ref.[28] 2. Theory 11 Examples of synthetic antioxidants include butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), propyl gallate (PG), and tert-butyl hydroquinone (TBHQ).[8] Although synthetic antioxidants have optimal antioxidant properties, they have some limitations and drawbacks. The production of synthetic antioxidants is costly. Some synthetic antioxidants, such as BHA and BHT, are cytotoxic and carcinogenic.[8] Additionally, they can lead to greater side effects compared to natural antioxidants.[8] Reports about PG demonstrate that it can induce mutagenesis, which in turn can lead to DNA damage and also accelerate aging. Natural antioxidants, on the other hand, are more suitable for human consumption and can be derived from existing natural resources.[8] Lignin, a byproduct already produced in excess in the paper industry, is an excellent example of a nature-derived antioxidant. 2.3.2 Lignin as an antioxidant Lignin offers antioxidant properties and can scavenge free radicals.[29] During plant growth, the massive polymeric structure of lignin protects plants against any biological, chemical, and mechanical stresses. Neutralizing these harmful free radicals reduces cellular damage.[8] A study by Rumpf et al. on lignins from Paulownia and Silphium showed that these lignins possess great antioxidant activity and have potential to be used as additives in food packaging or for biomedical applications.[29] Azadfar et al. demonstrates that the antioxidant capability of lignin is comparable to commercial antioxidants, guaiacol, and butylated hydroxytoluene.[30] There is a general belief that the antioxidant properties of lignin are primarily due to the presence of excess phenolic compounds and other oxygen-containing functional groups in its structure.[12] They include aliphatic hydroxyl, carbonyl, and carboxyl groups. Particularly, the phenolics can form quinomethides, a common characteristic shared with many other antioxidants such as polyphenols, phenolic acids, and vitamins.[8], [9] The polyphenolic structure of lignin reportedly prevents oxidative stress and excess ROS in several ways. The phenolic hydroxyl groups (Ar-OH) in lignin allow it to scavenge the free radicals (R •) through hydrogen atom transfer (HAT) and single electron transfer (SET) reactions (Scheme 1).[12] They can protect molecules against factors that provoke ROS production, such as light exposure and radiation. Additionally, the phenolic groups can generate less reactive, partially oxidized free radicals instead of highly reactive and harmful ones.[8] According to reports, lignin's antioxidant activity varies with the availability of phenolic hydroxyl groups and their stability.[8] ArO–H + R • → ArO • + R–H [12] Scheme 1. Radical scavenging mechanism of Phenol hydroxyl groups The stability of phenoxy radicals arises from structural features. An example is ortho-substituents (methoxy groups) in lignin, which stabilize (ArO•) through resonance and improve antioxidant activity. In addition, the π-π system in lignin increases the stability of the generated free radicals by enabling the delocalization of unpaired electrons. Conjugated double bonds also stabilize phenoxy radicals through extended delocalization, and conjugated carbonyl groups have little effect on antioxidant activity.[8] In addition to functional groups and structural characteristics, other factors such as molecular weight, polydispersity, biomass source, extraction method, and post-treatments can influence antioxidant.[12] For instance, Kaur et al. demonstrated that unmodified lignin from sugarcane bagasse had https://www.sciencedirect.com/topics/agricultural-and-biological-sciences/sugarcane-bagasse 2. Theory 12 higher antioxidant activity than its chemically modified version by acetylation and epoxidation.[12] Wang et al. proposed that the antioxidant activity of enzymatic hydrolysis lignin fractions increases as their molecular weight decreases.[31] Various assays can be employed to measure lignin's antioxidant activity, including chemical assays with specific reagents and cell-based assays.[29] 2.3.3 Antioxidant activity measurement Antioxidant assays can be categorized based on the reaction mechanism: Hydrogen atom transfer (HAT) or single electron transfer (SET).[8], [29] SET mechanism refers to the transfer of a single electron from antioxidants to free radicals, protecting other compounds from reacting with them.[8] HAT, on the other hand, refers to the donation of hydrogen atoms from antioxidants and the transfer of one bonding electron between protons to neutralize free radicals. However, in practice, the antioxidant activity can occur through more complicated mechanisms, such as mixed HAT/SET, stepwise electron transfer- proton transfer, concerted electron–proton transfer, or sequential proton loss electron transfer (SPLET).[8], [29] It is reported that the antioxidant mechanisms in lignin are the HAT, SPLET, and proton-coupled electron transfer (PCET). PCET mechanism refers to the transfer of both protons and multiple electrons. In the HAT mechanism, electron transfer occurs to a single radical orbital, while in PCET, multiple molecular orbitals are involved.[8] The DPPH (2,2-diphenyl-1-picrylhydrazyl) and ABTS (2,2′-azino-bis (3-ethylbenzothiazoline-6- sulfonic acid)) assays are widely used chemical-based analytical methods for quantifying radical- scavenging activity. DPPH and ABTS molecules act like a stable free radical with a reactive electron after activation with a solvent. The effectiveness of lignin neutralizing these stable free radicals (DPPH, ABTS) can be represented as radical scavenging activity (RSA). RSA is the amount of free radicals (DPPH and ABTS) neutralized by lignin, which is expressed as a percentage. Both assays follow HAT and SET mechanisms. It is important to note that DPPH is suitable for hydrophobic antioxidants, while ABTS is suitable for both hydrophilic and hydrophobic antioxidants.[8] In the DPPH assay, the DPPH free radical solution has a purple color due to the delocalization of excess electrons on the molecule, which has a distinct absorption peak at 515-520 nm in UV-Vis spectrophotometer measurements. Addition of an antioxidant leads to the neutralization of DPPH since the antioxidant’s hydrogen ions react with DPPH. Depending on the strength of the antioxidant, the intensity of the absorption peak decreases and the solution color changes toward yellow (Figure 8). The antioxidant capacity of an antioxidant can be determined by measuring the initial absorbance of the DPPH solution and the final absorbance of the mixture after the addition of the antioxidant. The difference between these two absorptions divided by the initial absorbance, in percent, shows the radical scavenging activity. In other words, the RSA shows the percentage of neutralized DPPH radicals. The formula used to calculate RSA_DPPH is provided in the Materials and Methods section (Section 3.4).[8] The DPPH assay is simple, rapid and inexpensive and is suitable for solid, liquid and biological samples. However, the DPPH molecule’s steric hindrance, stability, visible light exposure, concentration, and reaction time may influence the RSA measurement.[8] https://www.sciencedirect.com/topics/chemistry/antioxidant-capacity https://www.sciencedirect.com/topics/agricultural-and-biological-sciences/acetylation https://www.sciencedirect.com/topics/chemistry/epoxidation 2. Theory 13 In the ABTS assay, the ABTS+ free radical solution is produced by reacting with potassium persulfate (K2S2O8), forming dark blue-green solution through oxidation reaction. This solution has a distinct absorption at three points: 660, 734, and 820 nm. Among these peaks, the absorbance at 734 nm is generally considered and measured to avoid other interferences. After the addition of the antioxidant and neutralization of ABTS+ free radicals with hydrogen ions, the peak intensity at 734 will be decreased, and the solution’s color becomes lighter (Figure 8). The RSA in this methodology is determined similar to the DPPH assay. The formula used to calculate RSA_ABTS is provided in the Materials and Methods section (Section 3.5).[8]An advantage of the ABTS assay is the longer lifetime of the radical compared to hydroxyl radical, and oxygen ion radical. This method doesn’t require high temperature to produce free radicals and can study antioxidant activity over a wide pH range. This can avoid interference from endogenous peroxidase activity and make the determination of the antioxidant activity of amphiphilic substances more accurate.[8] The chemical structure of DPPH and ABTS and the reaction mechanism with antioxidants are presented in Figure 8. 2.3.4 Limitations and challenges Although lignin has considerable antioxidant properties, its practical application is limited by several challenges, both in antioxidant packaging materials and in biomedicine. Studying antioxidant activity is complicated in general, since no single methodology can fully capture how a possible antioxidant would function in vivo.[29] When it comes to lignin, its variety and complex structure further complicate the determination of specific chemical structures that lead to the antioxidant Figure 8. Reaction mechanism of (A) DPPH(2,2-diphenyl-1-picrylhydrazyl) and (B) ABTS (2,2′-azino-bis (3- ethylbenzothiazoline-6-sulfonic acid)) assays. Adapted from ref.[32] 2. Theory 14 activity, compared to other polyphenols such as flavonoids and tannins.[8] The heterogeneity of lignin in its natural form is also challenging for developing and applying this polymer for biomedical purposes.[12] There is insufficient understanding of the mechanism of lignin breakdown by the human body.[12] In vivo effects of lignin as an anti-oxidants should be investigated to bring it to actual application. In polymer applications, technical lignin have poor miscibility with numerous polymer matrices. Thus, chemically modified lignins were developed.[33] The antioxidant properties remain intact in lignin- based copolymers. An example is lignin–poly (ε-caprolactone-co-lactide), produced by Dan Kai et. al.[34], reported to exhibit good antioxidant activity. In conclusion, application of lignin as an antioxidant requires a directed study of its structure and its correlation with antioxidant activity. 2.4 Correlation statistical analysis Correlation analyses are useful to investigate the influence of a specific quantitative variable on another variable. More specifically, when the value of one variable increases, the value of another tends to increase or decrease, can be defined as how correlated the two variables are. Additionally, the intensity and direction of this correlation can be determined through various correlation tests.[35] This research aims to investigate the correlation status between antioxidant activity and the number of phenol groups in lignin. We mainly performed linear regression analysis. Additionally, Pearson analysis, Spearman analysis and Kendall’s Tau analysis were performed to validate the results from linear regression. 2.4.1 Linear regression analysis A simple approach is to plot the data in a scatter plot and fit a regression linear trendline to the data points. The equation for the linear trendline is as follows: 𝑦̂ = 𝑏0 + 𝑏1𝑥 (1) In equation 1, ŷ is an estimate mean value of the response variable, b0 is the y-intercept, and b1 is the slope, and x is the predictor variable. After fitting a trendline to the data points, the next step is to determine how well the data is fit to the trendline, and thereby how well the two variables are related. This can be determined by the analysis of variance and calculating the Coefficient of determination (R2). To achieve this, the total variation in the data is divided into two parts: the variation explained by the regression line and the variation that is left unexplained as random error. These are calculated using sums of squares. As a result, the total variability in the sample measurements (SST) equals the variability explained by the regression line (SST) plus the sums of squares for error (SSE). Equation 3 shows how these three sum of squares are calculated. In this equation the 𝑦𝑖 is the experimental data points (e.g. phenolic content/the antioxidant activity of a lignin sample), The 𝑦̅ is the mean value of the observations (mean phenolic content or mean antioxidant activity across all samples), and 𝑦𝑖̂ is the data points predicted by the model. In other words, 𝑆𝑆𝑇 = 𝑆𝑆𝑅 + 𝑆𝑆𝐸 (2) ∑(𝑦𝑖 − 𝑦̅)2 = ∑(𝑦𝑖̂ − 𝑦̅)2 + ∑(𝑦𝑖 − 𝑦̂)2 (3) 2. Theory 15 the SSR is the explained variation of the data points by the model, and the SSE is the unexplained variation of the model. From these sums of squares, the R2 can be calculated as follows. 𝑅2 = 𝐸𝑥𝑝𝑙𝑎𝑖𝑛𝑒𝑑 𝑣𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑇𝑜𝑡𝑎𝑙 𝑣𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 = 𝑆𝑆𝑅 𝑆𝑆𝑇 = 1 − 𝑆𝑆𝐸 𝑆𝑆𝑇 (4) The Coefficient of Determination measures the percentage of variation in the response variable (y) explained by the model. Values range from 0 to 1, with values close to 0 will worse fit and values close to 1, indicating very good fit.[36] Table 2 visualizes the model fit for different R2 values. Table 2. Interpretation of model fits based on the coefficient of determination (R2). Adapted from ref.[37] Linear regression analyses were performed in Google Colab using the linregress function, following the general command: slope, intercept, r_value, p_value, std_err = linregress (x, y) r_squared = r_value ** 2 Where x and y represent the independent and dependent variables, respectively. The resulting outputs include the slope, intercept, correlation coefficient (r_value), significance level (p_value), and standard error (std_err). R2 is calculated as the square of the correlation coefficient. A detailed example of the full analysis code and the plotting is provided in Appendix A. 2.4.2 Pearson analysis Pearson’s product-moment correlation coefficient, or Pearson’s r, is the most widely used technique for correlation analysis. This technique requires that both variables are normally distributed and there is a linear relationship between them. When these requirements are not fulfilled, it can lead to errors in the conclusion.[35] For a total amount of n samples and two measured continuous variables, 𝑢𝑖 and 𝑣𝑖 within the sample (1≤i≤n), the Pearson’s r(𝜌̂) is calculated as follows.[38] 𝜌̂ = ∑ (𝑢𝑖 − 𝑢̅.)(𝑣𝑖 − 𝑣̅.) 𝑛 𝑗=1 √∑ (𝑢𝑖 − 𝑢̅.) 2𝑛 𝑖=1 √∑ (𝑣𝑖 − 𝑣̅.) 2𝑛 𝑖=1 (5) 𝑢̅.= 1 𝑛 ∑ 𝑢𝑖 𝑛 𝑖=1 , 𝑣̅.= 1 𝑛 ∑ 𝑣𝑖 𝑛 𝑖=1 (6) R2 Interpretation 0.00-0.199 Very weak 0.20-0.399 Weak 0.40-0.599 Medium 0.60-0.799 Strong 0.80-1.00 Very strong 2. Theory 16 where 𝑢̅. and 𝑣̅. are the means of 𝑢𝑖respective 𝑣𝑖data points. The Pearson correlation 𝜌̂ ranges between -1 and 1, in which 1 shows a perfect positive correlation, -1 shows a perfect negative correlation, and 0 shows no correlation between the variables.[38] Pearson correlation analysis was performed in Excel using the general form of the commands =CORREL(array1; array2) or =PEARSON(array1; array2), where array1 and array2 represent the ranges of cells containing the two variables of interest. 2.4.3 Spearman analysis For cases in which the normal distribution is not fulfilled, the Spearman rank order correlation coefficient (ρ) substitutes the original data for their ordered ranks. It doesn’t require linear relationship between variables, as long as they exhibit monotonic behavior. In other words, they must exhibit a gradual relationship in the same direction (rising or falling) for the whole domain of the data studied.[35] Let 𝑞𝑖 (𝑟𝑖) denote the rankings of 𝑢𝑖 (𝑣𝑖), (1 ≤ 𝑖 ≤n). Spearman's ρ is defined as: 𝜌̂ = ∑ (𝑞𝑖 − 𝑞̅.)(𝑟𝑖 − 𝑟̅.) 𝑛 𝑗=1 √∑ (𝑞𝑖 − 𝑞̅.) 2𝑛 𝑖=1 √∑ (𝑟𝑖 − 𝑟̅.) 2𝑛 𝑖=1 (7) 𝑞̅.= 1 𝑛 ∑ 𝑞𝑖 𝑛 𝑖=1 , 𝑟̅.= 1 𝑛 ∑ 𝑟𝑖 𝑛 𝑖=1 (8) Whereas 𝑞𝑖and 𝑟𝑖 are the rankings of the original variables 𝑢𝑖and 𝑣𝑖. The ranking refers to the order of the variables. What is 𝑞̅. and 𝑟̅..The Spearman 𝜌̂ ranges between -1 and 1, in which 1 shows a perfect positive correlation, -1 shows a perfect negative correlation, and 0 shows no correlation between the variables 𝑢𝑖 and 𝑣𝑖.[38] In other words, 𝜌̂ equal to 1 means that both variables have the same ranking (𝑞𝑖 = 𝑟𝑖), and thereby. 𝑢𝑖 < 𝑢𝑗, 𝑣𝑖 < 𝑣𝑗 𝑜𝑟 𝑢𝑖 > 𝑢𝑗, 𝑣𝑖 > 𝑣𝑗 𝑓𝑜𝑟 𝑎𝑙𝑙 1 ≤ 𝑖 < 𝑗 ≤ 𝑛 It means that whenever one observation has a higher (or lower) rank in u, it also has a similar higher (or lower) rank in 𝑣. Any two sets of variables that follow the relationship above are concordant. In contrast, 𝜌̂ equal to -1 means that variables have a perfect negative correlation ( 𝑞𝑖 = 𝑛 − 𝑟𝑖 + 1), which means that whenever one observation has a higher rank in u, it has lower rank in 𝑣 and vice versa. 𝑢𝑖 < 𝑢𝑗, 𝑣𝑖 > 𝑣𝑗 𝑜𝑟 𝑢𝑖 > 𝑢𝑗, 𝑣𝑖 < 𝑣𝑗 𝑓𝑜𝑟 𝑎𝑙𝑙 1 ≤ 𝑖 < 𝑗 ≤ 𝑛 Any two sets of variables that follow the relationship above are discordant.[38] Spearman correlation analysis was performed in Excel using the general form of the commands =CORREL(array1; array2) applied to the rank of the data. In other words, array1 and array2 in this case correspond to the ranges of cells containing the ranked values of the two variables. 2.4.4 Kendall’s Tau analysis Another alternative for the not-normally distributed data, is the Kendall’s rank correlation coefficient (Tau-b), which substitutes the original data for their ordered ranks. It doesn’t require linear relationship 2. Theory 17 between variables, as long as they exhibit monotonic behavior. The Kendall’s Tau-b coefficient (τ or tb) is robust to extreme data (outliers), giving it a greater capacity for populational inference and a smaller estimation error.[35] Similar to Spearman's ρ, Kendall's Tau also implies the concepts of concordance and discordance to access the correlation between two variables. However, this methodology uses the notion of concordant and discordant pairs directly in the definition of this correlation measure. Specifically, Kendall's τ (sample version) is calculated through Equation 9. 𝜏̂ = 𝑛𝑐 − 𝑛𝑑 𝑛𝑡 (9) 𝑛𝑡 = 1 2 𝑛(𝑛 − 1) = 𝑛𝑐 + 𝑛𝑑 (10) 𝑛𝑐 = 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑐𝑜𝑛𝑐𝑜𝑟𝑑𝑎𝑛𝑡 𝑝𝑎𝑖𝑟𝑠 𝑛𝑑 = 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑑𝑖𝑠𝑐𝑜𝑟𝑑𝑎𝑛𝑡 𝑝𝑎𝑖𝑟𝑠 Where 𝑛𝑡 is the total number of concordant and discordant pairs in the sample. 𝜏̂ varies between -1 and 1, with -1 corresponding to the perfect discordance, 0 indicating no association between the variables, and 1 corresponding to the perfect concordance.[38] Kendall’s Tau correlation analysis was performed using Equations 9 and 10 (see Appendix B for a detailed example of the calculation). 21 3 Materials and Methods To investigate potential correlations between lignin structure and antioxidant activity, a range of analytical techniques was applied to seven different lignin types. However, incomplete dissolution of some samples in the selected interfered with the accuracy of the measurements. As a result, the modified lignins—Amine SKL3, HEOSHL, and HESKL1—were excluded from the titration analysis. In addition, the presence of amine side groups in Amine SKL3 was expected to increase conductivity, potentially affecting the reliability of pH-meter readings. This chapter is organized into four main sections: Lignin, materials and chemicals, characterization of phenolic groups, FT-IR analysis, and quantification of antioxidant activity. 3.1 Lignin, materials, and chemicals 3.1.1 Lignin resources Seven lignin types were investigated in this study. These include hardwood organosolv lignin (OSHL); softwood kraft lignin 1 (SKL1); softwood kraft lignin 2 (SKL2); and softwood kraft lignin 3 (SKL3). In addition, modified versions of OSHL, SKL1, and SKL3 were prepared according to the previously described procedure.[25], [26] These modified lignins include hydroxyethyl organosolv hardwood lignin (HEOSHL), hydroxyethyl softwood kraft lignin (HESKL1), and Aminated softwood kraft lignin (Amine SKL3). Table 3 summarizes the key information and corresponding sources for each lignin type. Table 3. Overview of the key characteristics, relevant information, and corresponding sources of the different lignin types analyzed. Lignin Wood category Pulping method Purification method Mn [KDa] Mw [KDa] Description OSHL Hardwood Organosolv - 2.07 [27] 5.84 [27] Suzano corporation, Brazil. SKL1 Softwood Kraft LignoBoost® 1.15 [27] 4.12 [27] Domtar Corporation (USA) SKL2 Softwood Kraft Indulin AT 1.2~1.34 [21] 2.99~3.4 [21] provided by Ingevity Corporation (South Carolina, USA) SKL3 Softwood Kraft LignoForce® 6 [39] 6~12.5[40] donated by West Fraser (Alberta, Canada) HEOSHL Hardwood Kraft - 2.72 [27] 14.9 [27] Liyang L., et al [27] HESKL1 Softwood kraft LignoBoost® 0.92 [27] 8.80 [27] Liyang L., et al [27] Amine SKL3 Softwood kraft LignoForce® 9.0 [25] 43.3 [25] Liyang L., et al [25] 3. Materials and Methods 22 3.1.2 Chemicals Sodium hydroxide pellets (NaOH, 40 g/mol), aqueous hydrochloric acid (HCl, 1 M), absolute ethanol (99.7 %), and dimethyl sulfoxide (DMSO, 99.7 %) were purchased from Sigma-Aldrich (Darmstadt, Germany). Deuterated chloroform (CDCl₃), anhydrous pyridine, chromium (III) 2,4-pentanedionate [Cr(acac)₃], N-hydroxy-5-norbornene-2,3-dicarboximide (NHND, 97%), ABTS(2,2′-Azino-bis(3- ethylbenzothiazoline-6-sulfonicacid)) diammonium salt, and potassium persulfate (K₂S₂O₈) were also from Sigma-Aldrich. DPPH (2,2-diphenyl-1-picrylhydrazyl) and Trolox (6-hydroxy-2,5,7,8- tetramethylchroman-2-carboxylic acid) were purchased from Sigma-Aldrich. 3.2 Phenol group characterization Conductometric titration and Phosphorus-31 nuclear magnetic resonance spectroscopy (³¹P NMR) were performed to determine the number of phenol groups in lignin structure. 3.2.1 Conductometric titration analysis Conductometric titration determines the number of phenol groups based on inflection points on the titration curve, obtained from constant conductivity measurements during the procedure.[25] Oven-dried lignin (0.1 g) was dissolved in aqueous sodium hydroxide (0.1 M, 40 mL). The solution was titrated by adding HCl (0.1 M, ~60 mL) stepwise, under magnetic stirring, and at ambient temperature. Meanwhile, the conductivity of the solution was continuously recorded by a conductivity meter (CO301, VWR, Germany). (see Appendix C for experimental setup). The titrant volume was added to the lignin–NaOH solution following a flow rate schedule (mL/time). This addition rate was designed to enhance time efficiency and clearly identify the conductivity inflection points. Briefly, 5 mL of HCl was added every 5 minutes up to a total volume of 20 mL, followed by 1 mL per minute up to 25 mL, then 0.25 mL every 30 seconds up to 40 mL, and finally 1 mL every 30 seconds, corresponding to a total titrant volume of 60 mL.[25] The titration curve was obtained by plotting conductivity (mS cm⁻¹) against titrant volume (mL) to visualize the inflection points. The derivative plot of the titration curve was generated, leading to the identification of four inflection points (A, B, C, D), illustrated in Figure 9 (a and b). The first inflection point, A, indicates the neutralization of the solution. Data points from A to B correspond to the volume interval in which the protonation of phenol groups occurs. The area between B and C corresponds to an Figure 9. The picture of the titration curve (a) and its derivative (b). 3. Materials and Methods 23 equilibrium state between protonation of phenol groups and carboxyl groups. Data points from C to D represent the protonation of carboxyl groups. The calculation of the number of phenolic-OH groups in the lignin structure was performed using Equation 11. (See Appendix D for additional calculations used to derive Equation 11 and Appendix E for a detailed example calculation.) N[C6H5OH] is the number of phenol groups, expressed in mmol per gram of lignin. V B, HCl is the total added HCl volume at point B, and VA, HCl is the total added HCl volume at point A. The HCl concentration (C HCl) is equal to 0.1 M, and the lignin mass (m Lignin) is 0.1 g.[25] 3.2.2 31P NMR spectroscopy The 31P NMR analysis was performed using a modified version of a protocol presented by Meng X et al.[11] Before analysis, three solutions were prepared; a solvent mixture (deuterated chloroform and anhydrous pyridine 1:1.6 (vol/vol)), a relaxation agent mixture (5.0 mg/mL solution of chromium (III) 2,4-pentanedionate (Cr(acac)3) in solvent mixture), and an internal standard (18.0 mg/mL of N- hydroxy-5-norbornene-2,3-dicarboximide (NDHD) in solvent mixture), which were prepared and stored in sealed containers. Oven-dried lignin samples (20 mg) were mixed with 400μl of the solvent mixture, 100μl of internal standard, and 50μl of relaxation agent. The mixtures were vortexed carefully before adding 100 mL of the phosphorate agent (2-chloro-4,4,5,5-tetramethyl-1,3-2-dioxaphospholane (TMDP)) to the lignin solutions. The mixtures were vortexed thoroughly and transferred into a 5-mm NMR tube. The NMR tubes were placed in the measurement queue of the NMR spectrometer (600 MHz, Oxford magnet equipped with Bruker NEO console and 5mm QCIP 31P cold probe) at the Swedish NMR Centre in Gothenburg.[11] The results were subsequently processed using TopSpin software (version 3.7.0, Bruker BioSpin GmbH). The calculation of the phenolic -OH group was performed using Equation 12. 𝑵[𝑹 − 𝑶𝑯] 𝒈 𝒐𝒇 𝒍𝒊𝒈𝒏𝒊𝒏 = 𝑹 × 𝒎𝒎𝒐𝒍 𝒐𝒇 𝑵𝑯𝑵𝑫 𝒊𝒏 𝑵𝑴𝑹[𝒎𝒎𝒐𝒍] 𝑫𝒓𝒚 𝒘𝒆𝒊𝒈𝒕𝒉 𝒐𝒇 𝒍𝒊𝒈𝒏𝒊𝒏 (12) Where N[R−OH] 𝑔 𝑜𝑓 𝑙𝑖𝑔𝑛𝑖𝑛 is the number of OH groups of the lignin’s weight. R is the ratio of the integration of the spectral region of interest (IOH) over the internal standard (IS) region (INHND), which is already calculated by the program, using Equation 13. 𝑹 = 𝑰𝑶𝑯 𝑰𝑵𝑯𝑵𝑫 = 𝑰𝒏𝒕𝒆𝒈𝒓𝒂𝒕𝒊𝒐𝒏 𝒐𝒇 𝒔𝒑𝒆𝒄𝒕𝒓𝒂𝒍 𝒓𝒆𝒈𝒊𝒐𝒏 𝒐𝒇 𝒊𝒏𝒕𝒆𝒓𝒆𝒔𝒕 𝑰𝒏𝒕𝒆𝒓𝒈𝒓𝒂𝒕𝒊𝒐𝒏 𝒐𝒇 𝑵𝑯𝑵𝑫 𝒓𝒆𝒈𝒊𝒐𝒏 (13) 𝐍[𝑪𝟔𝑯𝟓𝐎𝐇] = (𝑽𝑩,𝑯𝑪𝒍 − 𝑽𝑨,𝑯𝑪𝒍) × 𝑪𝑯𝑪𝒍 𝒎𝒍𝒊𝒈𝒏𝒊𝒏 (11) 3. Materials and Methods 24 The mole quantity of the internal standard, which is mentioned as mmol of N-hydroxy-5-norbornene- 2,3-dicarboximide (NHND-97%) in NMR in the equation above, is calculated using Equation 14: 𝒎𝒎𝒐𝒍 𝒐𝒇 𝑵𝑯𝑵𝑫 𝒊𝒏 𝑵𝑴𝑹[𝒎𝒎𝒐𝒍] = 𝒎𝑵𝑯𝑵𝑫 [𝒈] 𝑴𝑵𝑯𝑵𝑫 [ 𝒈 𝒎𝒐𝒍 ] × 𝟗𝟕% × 𝟏𝟎𝟎𝟎 (14) Where m NHND is the mass of NHND in grams, MNHND is the molecular weight of the NHND, which is 179.17 grams per mole.[11] The utilized methods for phenol group characterization are compared in Table 4. Table 4. An overview of the methodology characteristic of conductometric titration and 31P NMR Spectroscopy 3.3 FT-IR analysis The Fourier Transform Infrared Spectroscopy (FT-IR) results were included and interpreted to obtain an overall view of lignin structure, by searching for other functional groups or structural characteristics that may promote antioxidant activity. 3.4 Antioxidant activity measurement DPPH (2,2-diphenyl-1-picrylhydrazyl) radical scavenging activity and ABTS (2,2′-Azino-bis(3- ethylbenzothiazoline-6-sulfonicacid) radical scavenging activity are performed to measure the antioxidant activity in different types of lignin. 3.4.1 DPPH free radical scavenging capacity Criteria Conductometric titration ³¹P NMR spectroscopy Reagents Aqueousacids/bases (e.g., NaOH, HCl [11] Derivatization reagents (e.g. TMDP), organic solvents. (e.g. Pyridine, CDCl₃)[11] Toxicity Low High (organic solvents, phosphitylating agents) Ease of use Simple, accessible to most labs, can be performed by both manual and automatic Complex, requires trained personnel and specialized equipment Equipments pH-meter, pipette, stirrer etc. (alternatively, auto titrator) NMR spectrometer Calculation Basic stoichiometric calculations Requires spectral integration, calibration, and TopSpin software.[11] Result consistency High reproducibility and reliability Results can be affected due to the stability of the used reagents. Time consumption Quick (~36 minutes to 1 hour) Time-consuming sample preparation and analysis (at least 24 hours inclusive queuing) Cost Cheap (Total price of the analysis~ 497 - 1680 SEK/sample[41]) Expensive including instruments, time, reagents, maintenance, etc. ( Total price of the analysis ~528.00 US$[42]) 3. Materials and Methods 25 DPPH analysis was performed using a modified version of the method described by Duan et al.[43] A DPPH stock solution (C = 0.6 mmol L⁻¹) was prepared with absolute ethanol. This solution was diluted (dilution factor = 1:10, C = 0.06 mmol L⁻¹) to achieve an absorbance below 1.000. Lignin samples were subsequently dissolved in DMSO (C = 0.05 mg mL⁻¹). Each lignin sample (0.13 mL) was mixed with 2.87 mL of diluted DPPH solution. A standard solution was prepared by mixing Trolox-DMSO solution (0.13 mL, C = 0.05 mg mL⁻¹) with 2.87 mL of diluted DPPH solution. Similarly, a control sample (2.87 mL DPPH solution and 0.13 mL DMSO) and a blank sample (2.87 mL pure ethanol and 0.13 mL DMSO) were prepared. All mixtures were incubated in the dark for 30 minutes, and the DPPH concentrations were monitored over 300-700 nm using UV-Vis spectroscopy (SPECORD® PC 205, Analytik Jena, Germany). The absorbance at the peak maximum around 517 nm was used to calculate the radical scavenging activity of the samples using Equation 15. Measurements were performed in triplicate. 𝑹𝑺𝑨𝑫𝑷𝑷𝑯[%] = ( 𝑨𝑩𝑺𝒕=𝟎 − 𝑨𝑩𝑺𝒕=𝟑𝟎 𝒎𝒊𝒏 𝑨𝑩𝑺𝒕=𝟎 ) × 𝟏𝟎𝟎 (15) Where, RSA DPPH stands for radical scavenging activity, and the ABSt=30 min, is the absorbance of the antioxidant + DPPH sample after 30 minutes of incubation. ABSt=0 is the absorbance of the solution before any reaction occurred at time point 0, which consequently corresponds to the absorbance of the control sample.[29] 3.4.2 ABTS+ radical scavenging assay ABTS analysis was performed using a modified version of the method described by Rumpf et al.[29] An ABTS stock solution (10 mL, C = 7 mM) and a K₂S₂O₈ solution (10 mL, C = 2.25 mM) were prepared separately in distilled water. The solutions were mixed (1:1 v/v) and incubated in the dark for 16 hours at ambient temperature to generate the ABTS•⁺ radical cation solution. The ABTS•⁺ solution was diluted with distilled water at a 1:10 ratio to yield an absorbance of 0.700-1.000 at 734 nm. The appropriate ABTS•⁺ concentration was calculated using the Beer-Lambert law (see Appendix F for detailed calculations). The calculated concentration of ABTS•⁺ was 0.080 mM, which corresponds to an absorbance of 1.000 at 734 nm. Lignin samples were dissolved in DMSO at a concentration of 0.5 mg mL⁻¹. Each sample solution (0.05 mL) was mixed with 2.5 mL of ABTS•⁺ solution. A control sample (0.05 mL DMSO and 2.5 mL ABTS•⁺ solution) and a blank sample (0.05 mL DMSO and 2.5 mL distilled water) were also prepared. The samples were incubated for 12 minutes, and the absorbance was measured against the blank over the wavelength range of 500–900 nm using UV-Vis spectroscopy (SPECORD® 205 PC, Analytik Jena, Germany). The absorbance at the peak maximum around 734 nm was used to calculate the radical scavenging activity (RSA) using Equation 17. All measurements were performed in triplicate. 𝑹𝑺𝑨𝑨𝑩𝑻𝑺 = 𝑨𝒕=𝟎 − 𝑨𝒕=𝟏𝟐𝒎𝒊𝒏 𝑨𝒕=𝟎 (17) Where the RSAABTS stands for radical scavenging activity of antioxidant toward ABTS, and the At=12min is the absorbance of antioxidant + ABTS+ sample after 12 minutes of incubation. At=0 is the absorbance of the solution before any reaction occurred at time point 0, which consequently corresponds to the absorbance of the control sample.[29] 26 4 Results The study results are presented in four key sections. Section 4.1 presents the quantification of the phenolic-OH group and lignin structure. Section 4.2 presents antioxidant activity measurement. Section 4.3 investigates any possible correlations between phenolic-OH group characterization and antioxidant activity. Lastly, Section 4.4 interprets the Fourier transform infrared (FT-IR) spectroscopy results. 4.1 Characterization of phenolic groups in lignin This section presents the correlation analysis between different characterization methodologies for the phenolic -OH group. 31P NMR spectroscopy analysis was performed on all lignin types introduced in the methods section. For conductometric titration, however, the modified lignins were excluded from analysis. The reasons were incomplete dissolution of lignin in the solvent and the probable conductivity of the added amine group in Amine SKL3, which affected the pH-meter readings. Therefore, the correlation analysis between conductometric titration and 31P NMR spectroscopy was performed only on the original, unmodified lignin types, as presented in Figure 10. Figure 10. Results from phenol group characterization using conductometric titration and 31P NMR for OSHL, SKL1, SKL2, and SKL3 lignin. 31P NMR results are based on one sample per analysis. i.e., no error bars. 4. Results 22 4.1.1 Data analysis The comparison in Figure 10 is performed on OSHL, SKL1, SKL2, and SKL3, showing that the reported phenolic-OH quantity for each lignin varies between two analyses, except for OSHL. This variation may be due to the different methods used to measure OH groups and the distinct reagents employed. At the same time, both analyses report the same sequence arrangement from lowest to highest number of phenolic -OH groups: SKL1 < SKL2 < OSHL < SKL3. The observed standard deviation in the conductometric titration analysis is small for all lignin types except SKL3, indicating the homogeneity and repeatability of the technique. Nevertheless, higher variability observed for SKL3 may be attributed to its lower solubility during the experimental procedure. 4.1.2 Linear regression analysis A linear regression analysis was performed between the results from conductometric titration and ³¹P NMR, as presented in Figure 11. According to the analysis, a strong linear correlation was observed, with a slope of 1.85162 and an intercept of -3.70047. The obtained R2 is equal to 0.97393, indicating that 97% of the variation in titration results can be explained by the corresponding 31P NMR values, meaning that results from both methodologies are linearly correlated. Figure 11. The linear regression analysis between the data obtained from conductometric titration, visualized on the x axis, and 31P NMR spectroscopy, visualized on the y axis. y = 1.86 x – 3.70 R2= 0.97 4. Results 23 4.1.3 Kendall’s Tau, Spearman’s, and Pearson’s correlation analysis The Spearman, Pearson, and Kendall’s Tau correlation results are reported in Figure 12, reporting coefficients close to 1 in all analyses and indicating a strong positive correlation between the reported results from conductometric titration and ³¹P NMR. 4.2 Antioxidant characterization of lignin This section presents the results of antioxidant activity, determined by DPPH and ABTS radical scavenging analyses, and followed by a correlation analysis between these two methodologies. The investigation compared mean values from both assays and assessed their correlation using four methods: linear regression, Pearson’s correlation, Spearman’s correlation, and Kendall’s Tau. Both radical scavenging assays were performed on seven lignin types: OSHL, SKL1, SKL2, SKL3, HESKL1, HEOSHL, and Amine SKL3. 4.2.1 Data analysis Figure 12. The Kendall’s Tau, Spearman’s, and Pearson’s correlation coefficient between the data obtained from conductometric titration and 31P NMR spectroscopy. Figure 13. The radical scavenging activity (RSA) measured by DPPH (a) and ABTS (b) assays. a b 4. Results 24 Figure 13 presents a comparative summary of the DPPH and ABTS radical scavenging activities (RSA) for each lignin type. According to Figure 13, the DPPH analysis suggested the following sequence of lignin from highest to lowest RSA_DPPH: OSHL>SKL1> SKL3> SKL2> HESKL1> HEOSHL >Amine SKL3. At the same time, ABTS analysis suggested the following sequence of lignin from highest to lowest RSA_ABTS: SKL1> SKL3> OSHL >SKL2>HESKL1> Amine SKL3> HEOSHL. In both analyses, the three least RSA belonged to the modified lignins: HEOSHL, HESKL1, and Amine SKL3. Among them, HESKL1 showed the best RSA. In contrast, the unmodified lignin types, OSHL, SKL1, SKL2, and SKL3, had the four highest RSA in both methodologies. 4.2.2 Linear regression analysis A linear regression analysis was performed to evaluate the correlation between DPPH and ABTS analyses, presented in Figure 14. The linear regression analysis generated a trendline with a slope of 3.09450, an intercept of -5.50416, and an R2 value of 0.84814. This indicates that 85% of the variation in RSA_DPPH values can be explained by the corresponding RSA_ABTS values, demonstrating a strong linear correlation between the two methodologies. Figure 14. The linear regression analysis between the data obtained from DPPH radical scavenging activity, visualized on the x axis and ABTS radical scavenging activity, visualized on the y axis. y = 3.09 x – 5.50 R2= 0.85 4. Results 25 4.2.3 Kendall’s Tau, Spearman’s, and Pearson’s correlation analysis Spearman, Pearson, and Kendall’s Tau calculated coefficients were close to 1, as presented in Figure 15, demonstrating a strong positive correlation between the results of the DPPH and ABTS analyses. 4.3 Correlation analysis between phenol group quantity and antioxidant activity This section introduces the correlation analysis results between phenolic-OH content and ABTS/DPPH- RSA, followed by linear regression, Pearson’s, Spearman’s, and Kendall’s Tau analysis between RSA and 31P NMR, as presented in Section 4.3.1. 4.3.1 31P NMR Vs radical scavenging activity Figure 16 presents the linear correlation between 31P NMR results and radical scavenging activity. y = 8.03 x + 10.15 R2= 0.66 y = 2.09 x + 6.34 R2= 0.51 a b Figure 16. The linear regression correlation analysis between 31P NMR results and Radical scavenging activity: DPPH (a), and ABTS (b). Figure 15. The Kendall’s Tau, Spearman’s, and Pearson’s correlation coefficient between the data obtained from DPPH and ABTS analysis. 4. Results 26 The linear regression analysis presented in Figure 16a demonstrates a moderate linear correlation between 31P NMR-based phenolic-OH content and RSA_DPPH results. The regression equation has a slope of 2.09296 and an intercept of 6.34337. The R² value is 0.51, indicating that 51% of the variation in RSA_DPPH is explained by the obtained linear trend line with the ³¹P NMR-based phenolic-OH content. According to the 31P NMR and ABTS correlation analysis (Figure 16b), the regression equation has a slope of 8.03317 and an intercept of 10.15183. The obtained R2 is 0.66185, indicating that 66% of the variation in RSA_ABTS is explained by the obtained linear relationship with the 31P NMR-based phenolic content. Kendall’s, Spearman’s, and Pearson’s correlation analyses are performed between the number of phenol groups obtained by 31P NMR, and the RSA-analyses obtained by DPPH and ABTS, presented in Figure 17. The results show a positive correlation across all three analyses, although the strength of the correlations is moderate. a b Figure 17. The Spearman’s, Pearson’s, and Kendall’s Tau correlation coefficient between the data obtained from (a) 31P NMR spectroscopy and DPPH analysis, and (b) 31P NMR spectroscopy and ABTS analysis. 4. Results 27 4.4 Fourier Transform Infrared Spectroscopy (FT-IR) analysis This section presents and interprets the results from Fourier Transform Infrared Spectroscopy (FT-IR) from a previous study by L. Liu et al[25]. Figure 18 introduces the result from the FI-TR analysis. . -0,02 0 0,02 0,04 0,06 050010001500200025003000350040004500 SKL1 -0,2 0 0,2 0,4 0,6 050010001500200025003000350040004500 Amine SKL3 -1 0 1 2 3 050010001500200025003000350040004500 SKL3 -0,5 0 0,5 1 1,5 050010001500200025003000350040004500 SKL2 -0,5 0 0,5 1 1,5 050010001500200025003000350040004500 OSHL A b so rb an ce Wavenumber (cm-1) Figure 18. FT-IR spectra of OSHL, SKL1, SKL2, SKL3, and Amine SKL3. 2 3 4 6 7 1 5 8 9 10 11 12 13 4. Results 28 Chemical structures and functional groups identified in all five lignin types, along with their positions, are summarized in Table 5. Among the observed positions, the wavelength interval [1685-1700] cm-1 included several peaks for different types of lignin. The number and location of peaks in this specific area were unique for some lignin types. This region contains peaks corresponding to aromatics, conjugated groups, and carbonyl groups, which might be able to stabilize the phenolic radical formed after free-radical neutralization due to the resonance delocalization and stereoelectronic effects. These structures are primarily associated with lignin chromophores, such as quinones, stilbenes, catechols, aromatic ketones, and conjugated carbonyls with phenolics.[7] These findings suggest the hypothesis that these chromophores may also contribute to the antioxidant activity of lignin. Table 5. Summarizes the identified chemical structures and functional groups in peak cluster 3 for OSHL, SKL1, SKL2, SKL3, and Amine SKL. The points are identified by [25] and the reference chart from Thermo Fisher.[44] To obtain a qualitative measurement of the chromophore’s peak in FT-IR analysis, a ratio between the height of peaks 4 and 8a is calculated for each lignin type. In this ratio, the height of peak 4 is a representative of carbonyl groups, while the height of peak 8a corresponds to the aromatic structures in lignin. A higher ratio between peak 4 and 8a indicates a higher chromophore content. Peak Peak position (cm-1) Description of the corresponding structure 1 ~3445 O-H stretching from alcohol/phenol group 2 ~3384 N-H stretching from aliphatic primary amine 3 ~2937 C-H stretching from alkane 4 ~1707 carbonyl groups (C=O stretching from amid group or carboxylic acid or ester or aldehyde) 5 ~1651 C=O stretching from urea/δ-lactam/conjugated ketone/ conjugated alkene 6 ~1597 C-C stretching from lignin aromatic 7 ~1565 N-H bending from amine 8 a) ~1514 C-C stretching from the aromatic rings in lignin. b) ~1368 c) ~1423 9 ~1371 O-H bending from phenol 10 ~1270 C-O stretching from alkyl-aryl ether and methoxy groups 11 ~1214 C-O (H) + C-O (C) stretching from Aromatic OH +ether 12 ~1145 C-N stretching from amine group or C-H bending from Guaiacyl unit 13 ~1031 C-O (H) + C-O (C) stretching from Aliphatic OH + ether 4. Results 29 Table 6. Qualitative analysis of chromophore content in all lignin types by comparing the ration between peaks 4 and 8a. Lignin type Peak 4 Peak 8a Ratio Peak (4/8a) Ratio Percentage [%] Ranking of chromophore content OSHL 0.40658 0.66002 0.616 62 1 SKL1 0.01132 0.0381 0.297 30 4 SKL2 0.26095 0.99457 0.262 26 5 SKL3 0.58226 1.58131 0.368 37 3 AMINE SKL3 0.27078 0.47046 0.575 58 2 The ratio analysis suggested that the chromophore content from highest to lowest is ranked as follows: OSHL >AMINE SKL3>SKL3> SKL1>SKL2. OSHL has significantly higher chromophore contents than SKL 1-3. All SKLs show similar chromophore contents (30 ± 7). Amine SKL3 also shows similar chromophore contents to OSHL. 36 5 Discussion This chapter is structured into four parts. Section 5.1 presents and discusses the results from the phenolic-OH group characterization, while Section 5.2 examines the results from the radical scavenging activity analysis. Section 5.3 explores the correlation between phenolic group characterization and radical scavenging activity. Finally, Section 5.4 introduces and discusses the results from the FT-IR analysis. 5.1 Conductometric titration analysis vs 31P NMR spectroscopy Figure 10 displays the comparison of the conductometric titration and 31P NMR spectroscopy mean values of phenol groups. Although the methodologies report different OH values, both suggest the same ranking of lignin types: SKL1< SKL2< OSHL< SKL3 from lowest to highest phenolic-OH content. The standard deviation of conductometric titration analysis is also presented in Figure 10. The reported numbers are relatively small for OSHL and SKL2 and almost zero for SKL1, which indicates the results are reliable and reproducible. Several correlation analyses were performed to evaluate the degree of correlation between the two assays, illustrated in Figures 11 and 12. The obtained R2 is 0.97 for linear regression analysis. The reported values from Kendall’s Tau, Spearman, and Pearson are 1 or very close to 1, indicating a perfect positive correlation between the number of phenolic hydroxyl groups obtained by titration and 31P NMR. These results show that, with optimizations, conductometric titration can replace 31P NMR spectroscopy as a simple, less toxic method for quantifying phenolic-OH groups. 5.1.1 Improvements A limitation of these analyses was the small sample size. Titration analysis was performed in triplicate, and the mean values of the measurements are presented in Figure 10. ³¹P NMR spectroscopy was performed as a single replicate, leading to no reported standard deviation for the corresponding data. Conducting replicate measurements for the ³¹P NMR spectroscopy analysis represents an important area for methodological improvement. The modified lignin types were excluded from the conductometric titration analysis due to conductivity and solubility challenges. Consequently, only four technical lignins were analyzed and compared for phenolic-OH characterization in this assay. This challenge limits the feasibility of robust statistical analyses due to insufficient sample size. It can be addressed by increasing the number of replicates and expanding the range of lignin types examined in future studies. The modification of Amine SKL3 lignin involves the replacement of phenolic –OH groups with amine (–NH₂) groups, whereas in HEOSHL and HESKL1, the phenolic groups are substituted with aliphatic –OH groups. The amine and ester functionalities appear to interfere with conductivity measurements 5. Discussion 36 during the titration analysis. A potential improvement to address this challenge would be to quantify the amine/ester group content and their respective conductivities on the modified lignin surfaces and incorporate these parameters into the calculation of phenolic content. Another challenge encountered during the titration analysis was the limited solubility of certain lignins in the solvent system. Lignin is an amphiphilic macromolecule with complex structural characteristics, exhibiting variable solubility across different solvent systems. During the analysis, Amine SKL3 lignin did not fully dissolve in the aqueous sodium hydroxide solution, with a small fraction (<5%) remaining undissolved even after heating to approximately 50 °C (see Appendix G). Identifying a more suitable titrant solvent, applying elevated temperatures to enhance lignin dissolution, or incorporating an appropriate cosolvent to improve lignin solubility would be critical steps toward optimizing the titration protocol for future applications. 5.2 DPPH vs ABTS+ radical scavenging assay This section will review and discuss the results from radical scavenging activity, presented in Figures 13, 14, and 15. The initial definition of antioxidant ability is to react with and neutralize free radicals. Consequently, radical scavenging activity (RSA) assays were chosen for a direct measurement of antioxidant activity. Due to the amphiphilic nature of lignin, both DPPH, which is suitable for hydrophobic components, and ABTS, which is suitable for both hydrophobic and hydrophilic components, are performed in this study. As shown in Figure 13, the ABTS and DPPH assays don't show identical rankings of antioxidant activity. At the same time, in both analyses, the three modified lignin types, Amine SKL3, HESKL1, and HEOSHL have relatively lower RSA. While the four original lignins, OSHL, SKL1, SKL2, and SKL3, have relatively higher RSA. This separation is significant. All modified lignin types have a fewer phenolic-OH groups. Both Amine SKL3 and HESKL1 are dephenolized derivatives of SKL3 and SKL1 and display lower antioxidant activity compared to their unmodified counterparts. However, the reduction in antioxidant activity was less pronounced in HESKL1 than in Amine SKL3. Likewise, HEOSHL also showed a drastic decrease in antioxidant activity compared to its original version, OSHL. The results demonstrated that the phenolic content significantly influenced antioxidant acti