Jump to ContentJump to Main Navigation
Show Summary Details
More options …
Wood Research and Technology


Cellulose – Hemicelluloses – Lignin – Wood Extractives

Editor-in-Chief: Faix, Oskar

Editorial Board: Daniel, Geoffrey / Militz, Holger / Rosenau, Thomas / Salmen, Lennart / Sixta, Herbert / Vuorinen, Tapani / Argyropoulos, Dimitris S. / Balakshin, Yu / Barnett, J. R. / Burgert, Ingo / Rio, Jose C. / Evans, Robert / Evtuguin, Dmitry V. / Frazier, Charles E. / Fukushima, Kazuhiko / Gindl-Altmutter, Wolfgang / Glasser, W. G. / Holmbom, Bjarne / Isogai, Akira / Kadla, John F. / Koch, Gerald / Lachenal, Dominique / Laine, Christiane / Mansfield, Shawn D. / Morrell, J.J. / Niemz, Peter / Potthast, Antje / Ragauskas, Arthur J. / Ralph, John / Rice, Robert W. / Salin, Jarl-Gunnar / Schmitt, Uwe / Schultz, Tor P. / Sipilä, Jussi / Takano, Toshiyuki / Tamminen, Tarja / Theliander, Hans / Welling, Johannes / Willför, Stefan / Yoshihara, Hiroshi

12 Issues per year

IMPACT FACTOR 2016: 1.868
5-year IMPACT FACTOR: 1.875

CiteScore 2016: 1.83

SCImago Journal Rank (SJR) 2015: 0.817
Source Normalized Impact per Paper (SNIP) 2015: 0.954

See all formats and pricing
More options …
Volume 69, Issue 3


Characterization of dissolved lignins from acetic acid Lignofibre (LGF) organosolv pulping and discussion of its delignification mechanisms

Heli Kangas / Tiina Liitiä / Stella Rovio / Taina Ohra-aho / Harri Heikkinen / Tarja Tamminen / Kristiina Poppius-Levlin
Published Online: 2014-09-27 | DOI: https://doi.org/10.1515/hf-2014-0070


Birch chips were cooked by means of the Lignofibre (LGF) organosolv process in acetic acid (AA) and phosphinic acid (H3PO2) at 150°C. The delignification rate and structure of the dissolved lignin was followed as a function of time. The degree of delignification increased steadily up to 88% during the 120 min treatment time. The dissolved lignins were precipitated from the spent liquor (SL) by water addition, washed, and purified for the analyses. Elemental analysis, 31P nuclear magnetic resonance (NMR), heteronuclear single-quantum coherence (HSQC) NMR, pyrolysis-gas chromatography/mass spectrometry (GC/MS), and gel permeation chromatography (GPC) were applied for the structural elucidation. It was found that the cleavage of the β-aryl ether linkages is the main reaction leading to delignification, accompanied by the formation of free phenolic hydroxyl groups and reduction in the content of aliphatic hydroxyl groups. The structure of the dissolved lignin remained the same after the drastic changes at the early stages of cooking (up to 30 min cooking time), indicating that secondary reactions (e.g., condensation) do not take place to a significant extent. H3PO2 probably enhances the acidolysis reaction via an ester derivative that both boosts the cleavage reaction and prevents the formation of the carbocation intermediate that induces condensation. Homolytic cleavage reactions may take place parallel to the acidolytic reactions.

Keywords: acetic acid; degradation mechanism; delignification; lignin; Lignofibre (LGF) process; organosolv pulping; phosphinic acid


Replacing oil-based products with materials based on renewable raw materials is the driving force for searching novel process technologies for biomass utilization. The so-called biorefinery processes aim at the utilization of the whole biomass for producing fibers, energy, value-added chemicals, and polymers. Organosolv pulping processes can be considered as one of the key methods for the realization of biorefinery concepts.

The term “organosolv pulping” encompasses a wide variety of methods in which lignocellulosic materials are treated in organic solvents for the efficient separation of lignin, cellulose, and other carbohydrates (Hergert 1998). In the past, organosolv processes were considered as alternative methods for the traditional chemical cooking processes. Their permanent industrial application was not successful due to problems related to the fiber quality, commercialization of the byproducts (mainly lignin), process cost-efficiency, and the recovery of chemicals, just to mention a few. However, the interest toward various organosolv processes within the scope of biorefinery processes is still vivid. The advantages are the selective separation of lignocellulosic biomasses into high-value fractions, the suitability for nonwood biomasses with high silicate content, and the absence of sulfur in the process. Thus, the products are sulfur free, there are no sulfur emissions, and the chemical recovery is relatively simple (Aziz and Sarkanen 1989; Hergert 1998).

The Lignofibre (LGF) organosolv process has been developed at VTT (Mikkonen et al. 2009). It is based on organic solvents such as ethanol or acetic acid (AA) and phosphinic acid (H3PO2), which is a strong reducing agent (Mehrotra 2013). With the LGF process, it is possible to fractionate the major lignocellulosic biomass components, that is, cellulose, lignin, and hemicelluloses. The process produces fibrous material with high cellulose content, dissolved lignin (which can be recovered easily by precipitation in water), as well as dissolved hemicelluloses and their reaction products such as furfural (Hakala et al. 2010; Liitiä et al. 2011, 2013; Tamminen et al. 2011; Kangas et al. 2013).

In acidic organosolv pulping, delignification is promoted by the cleavage of β-aryl ether bonds via either acidolysis or homolytic cleavage. The two reaction mechanisms typically take place simultaneously, while the prevailing route is dependent on acidity and temperature (Li et al. 2000).

The acidolytic reactions are favored by high acidity. In the first phase of the acidolysis, benzylic carbocation is formed (Li et al. 2000; Lai 2001). The degradation can then progress to produce the so-called “Hibbert ketones”, or homovanillin derivatives, via cleavage of the γ carbon (Adler et al. 1957; Lundquist and Lundgren 1972; Karlsson et al. 1988; Karlsson and Lundquist 1992). These reactions are, however, competed by nucleophilic addition reactions to the carbocation, typically resulting in the formation of β-aryl-type condensation products (Li et al. 2007). The condensation of lignin via this mechanism can be prevented by the addition of small-molecular-weight phenolic nucleophiles, such as naphthol (Lora and Wayman 1980; Li et al. 2007).

Homolytic cleavage is expected to dominate in organic solvents at neutral or moderately acidic conditions at high temperatures (Li et al. 2000; Lai 2001). In this case, the quinone methide intermediate is first formed followed by the homolytic cleavage of the β-aryl ether bond. The formed radicals may further participate in radical coupling reactions, leading to lignin condensation (Li et al. 2000). Alternatively, they may be stabilized via radical exchange reactions ending up in lignin degradation.

The aim of the present work is to follow the various degradation and condensation reactions expected to take place under the acidic high-temperature conditions of LGF cooking. Special attention is paid to the role of H3PO2 in these reactions.

Materials and methods

European silver and/or white birch (Betula pendula Roth and/or Betula pubescens Ehrh.) obtained from Espoo, southern Finland, was cooked. The stems were debarked manually and chipped at Teollisuuden Hake Oy (Kuusankoski, Finland). The dry matter content of the chips was 54.8%; composition: 22.2% lignin, 44% cellulose, and 31.6% hemicelluloses (consisting of 26% glucuronoxylan and 3.2% glucomannan) (Kangas et al. 2013). The cooking trial was performed with chip batch size of 300 g in an air bath digester built in-house and fitted with four autoclaves (V=2.5 L). AA (99.85% from Algol Chemicals, Espoo, Finland. Diluted to 80% by the process water) was used as the cooking solvent and 3.5% H3PO2 (based on wood; Sigma-Aldrich, Helsinki, Finland) was added. The process temperature was 150°C and the liquid-to-wood ratio 5:1. The temperature was raised to the final cooking temperature in 60 min. During the cooking process, liquor samples (50 ml) were taken from the autoclave through a valve at the bottom at time intervals of 30, 60, 90, and 120 min, corresponding to samples LGF30, LGF60, LGF90, and LGF120, respectively.

The lignin in the spent liquor (SL) samples was precipitated by the addition of water (5:1). The precipitate was separated from the liquid by centrifugation (Eppendorf 5804R, Hamburg, Germany; 7–27 min at 20°C and 9000 rpm, corresponding to 10 300 G) and sequentially washed with water until pH around 5. The samples were freeze dried.

The lignin samples were further purified for analytical purposes. Then, 0.5 g lignin was dissolved in 35 ml of 0.1 M NaOH (Merck, Darmstadt, Germany) for 30 min at room temperature. The solution was filtered (Whatman no. 4) and the filter was further washed with 15 ml of 0.1 M NaOH (Merck, Darmstadt, Germany). The solution was then acidified with 1 M HCl (Sigma-Aldrich, Helsinki, Finland) to pH around 2.4 and centrifuged at 9000 rpm and 4°C for 25 min (Eppendorf 5804R, Hamburg, Germany). The precipitate was washed twice with acidified Milli-Q water (pH 2.4) and separated by centrifugation at 10 000 rpm and 4°C for 30 min (Eppendorf 5804R, Hamburg, Germany). The sample was freeze dried and Soxhlet extracted with hexane (Rathburn Chemicals Ltd., Walkenburn, Scotland) for 8 h.

Birch milled wood lignin (MWLbirch) as a reference was isolated from the same birch species, from which the LGF lignins were prepared, but was isolated previously from another raw material batch. The MWLbirch was isolated by a slightly modified Björkman (1956) method, including an ultrasonic extraction step (90 min at 15°C) after the ball milling.

The Klason lignin content in the cooking liquor was analyzed gravimetrically by precipitation with H2SO4 (4%, w/v; Sigma-Aldrich, Helsinki, Finland) in an autoclave (Getinge A-96082, Göteborg, Sweden) at 121°C. For the analysis, 1.8 ml H2SO4 (Sigma-Aldrich, Helsinki, Finland) and 13.2 ml Milli-Q water were added to 10 ml of the sample. The precipitate was filtered and the acid-soluble lignin was detected in the clear filtrate at 215 and 280 nm according to Goldsmith (1971) and the acid soluble lignin concentration cL (g l-1) was calculated by Equation (1): cL=(4.53 a215–a280)/300, where a215 and a280 are the ultraviolet absorbance values of the filtrate.

For the determination of lignin and residual carbohydrate content of the precipitated lignins, the freeze-dried samples were hydrolyzed with sulfuric acid (4%, w/v ; Sigma-Aldrich, Helsinki, Finland) in an autoclave and the resulting monosaccharides were determined by HPAEC, Dionex ICS 3000 (Dionex, Sunnyvale, CA, USA) equipped with CarboPac PA1 column with pulse amperometric detection (PAD) (Hausalo 1995; Willför et al. 2009). The Klason lignin content (i.e., the insoluble residue from the hydrolysis) was determined gravimetrically from the same hydrolysate according to the NREL method (Sluiter et al. 2008). The acid-soluble lignin in the hydrolysate was detected at 215 and 280 nm according to Goldsmith (1971).

The elemental analysis C, O, H, N, P, and content of OCH3 was performed at the Analytical Laboratories (Lindlar, Germany). The C, O, H, and N contents were determined according to standard ASTM D 5291 and the phosphorus analyses were performed after digestion with sulfuric and nitric acids by the photometric procedure described in ASTM D 515 Method A. The methoxy (OMe) content was determined by the Zeisel method followed by iodometric titration as described by Pregl and Roth (1958). The P content of the unpurified LGF120 sample was determined as a reference at Labtium Oy (Espoo, Finland). The analysis was performed after wet combustion in hydrogen peroxide and nitric acid in a microwave oven and the P content was determined from the solution by inductively coupled plasma-optical emission spectrometry (Iris Itrepid, ICP-OES from Thermo Elemental, Maryland, USA).

The empirical formula CaHbOcNd(OPH2O)e(OCH3)f(CH3COO)g (OHaliph)h(OHphen)i of LGF lignins per C900 units was calculated according to Vazquez et al. (1997) with some modifications to include acetyl and hydroxyl groups and H3PO2 esters in the formula: a=(%C)/ MC-f-2g, b=(%H)/MH-2e-3f-3g-h-i, c=(%O)/MO-2e-f-2g-h-i, d=(%N)/MN, e=(%P)/MP, and f=(%OCH3)/MOCH3.

The content of acetoxy (OAc) groups was analyzed by capillary electrophoresis (P/ACE, Beckman-Coulter, Fullerton, CA, USA) after alkaline hydrolysis. The sample was weighted (4–9 mg) into 1 ml solvent (100 mM NaOH; stock solution 1 M NaOH, FF-Chemicals, Haukipudas, Finland). The solution was first mixed in a vortex mixer (Scientific Industries Inc., Bohemia, USA) and then with ultrasound (model M12, FinnSonic Oy, Lahti, Finland) (30+30 min). The mixture was centrifuged (model EBA21, Hettich Zentrifugen, Tuttlingen, Germany) at 14 000 rpm for 5 min and the supernatant was taken to CE analysis. A reference treatment was performed with ultrapure water. The CE analysis of detached OAc groups was performed at pH 8.1 with the chemical and separation parameters as described by Rovio (2010).

For 31P nuclear magnetic resonance (NMR) analyses performed with Bruker Avance III 500 Mhz NMR instrument (Bruker Corp., Billerica, MA, USA) equipped with a double-resonance z-gradient BBFO probe head, the lignin samples were accurately weighed (25 mg) and dissolved in 150 μl N,N-dimethylformamide (Sigma-Aldrich, St. Louis, USA) in 10 ml vial. After total dissolution, 100 μl pyridine (Sigma-Aldrich, St. Louis, USA), 200 μl internal standard solution (ISTD) of endo-N-hydroxy-5-norbornene-2,3-dicarboximide (e-HNDI; 0.005 mmol, Sigma-Aldrich, St. Louis, USA) in pyridine/CDCl3 (1.6:1, v/v), and 50 μl of the Cr(acac)3 solution (11.4 mg ml-1; Sigma-Aldrich, St. Louis, USA) in pyridine/CDCl3 (Sigma-Aldrich, St. Loius, USA) (1.6:1, v/v) were added. Then, 150 μl of the phosphitylation reagent 2-chloro-4,4,5,5-tetramethyl-1,3,2-dioxaphopholane [P.R.(II), Sigma-Aldrich, St. Louis, USA] were added dropwise. Finally, 300 μl CDCl3 was added to the solution and clear brown to black solution was achieved. A freshly prepared sample was measured with 31P NMR immediately after preparation at room temperature. The chemical shifts are reported relative to the sharp signal (132.2 ppm) originating from the reaction between water and P.R.(II). NMR parameter: scans=512, pulse delay=5 s, 90° pulse, and line broadening=2 Hz. 31P NMR measurement is based on the method developed by Granata and Argyropoulos (1995).

Pyrolysis-gas chromatography/mass spectrometry (GC/MS) analyses were performed with a filament pulse pyrolyzer (Pyrola2000, Pyrol AB, Lund, Sweden) connected to a GC/MS instrument (Varian 3800 GC/2000 MS, Walnut Creek, USA; column J&W, DB-1701, Folsom, USA, 30 m×0.25 mm, film 1 μm) in combination with an ion-trap mass spectrometer (EI 70 eV) detection. Then, 100 μg of the sample was weighed accurately and pyrolyzed at 580°C for 2 s. The peak molar areas of lignin pyrolysis degradation products were calculated and normalized to 100%. Each sample was analyzed at least twice. A more detailed description of the method can be found in Ohra-aho et al. (2013).

1H-13C heteronuclear single-quantum coherence (HSQC) spectroscopy NMR experiments were carried out at 303 K on a Bruker Avance III 500 MHz instrument (Bruker Biospin GmbH, Rheinstetten, Germany) equipped with a z-gradient double-resonance probe. The samples were prepared by dissolving approximately 50 mg of the material in 0.7 ml d6-DMSO (Sigma-Aldrich, St. Louis, USA). The solvent signal was taken as an internal reference (δ1H 2.49 ppm and δ13C 39.5 ppm). HSQC NMR experiments were acquired with 3 s pulse delays and chromium(III)acetylacetonate (0.15 M, Sigma-Aldrich, St. Louis, USA) for a complete relaxation of all nuclei. The spectral widths were 5000 and 20 000 Hz for the 1H and 13C dimensions. The number of transients was 32 with 256 time increments on the 13C dimension. The 1JCH was 145 Hz. All NMR data were processed with Bruker Topspin-NMR software (Bruker Biospin Gmbh, Rheinstetten, Germany).

The molar mass measurements were performed by SEC; instrument: HPLC (Waters, Milford, MA, USA) in 0.1 M NaOH (stock solution 1 M NaOH from Merck Titrisol, Darmstadt, Germany; diluted 1:10 and filtered) eluent and MCX 1000 and 100 000 Å columns (PPS Polymer Standards Service, Mainz, Germany) with ultraviolet (UV) detection at 280 nm (Waters 2998 Photodiode Array Detector, Milford, MA, USA). The average molar masses (Mn, Mw) and the molar mass distributions were calculated relative to Na-polystyrene sulfonate standards (Na-PSS, 3420–148 500 g mol-1, American Polymer Standards Corporation, Mentor, OH, USA) with Empower 3 software (Waters, Milford, MA, USA). For the analysis, about 4 mg of lignin were dissolved overnight in 4 ml analytical NaOH (0.1 M, Merck Titrisol, Darmstadt, Germany) and filtered with 0.45 μm PTFE membrane syringe filters (VWR International, Radnor, PA, USA).

Results and discussion

Structural considerations

The amount of lignin dissolved during the LGF process was estimated based on the lignin content in the SL (Table 1). To avoid overlap in the UV determination due to the presence of furanoic carbohydrate degradation products, the lignin content in the SLs was analyzed as the sum of precipitated Klason lignin and acid-soluble lignin obtained after acid hydrolysis. The amount of lignin in SL increased with increasing process time, as expected. After 120 min, almost 90% of the lignin in the wood chips was dissolved.

Table 1

Lignin (L) contents in the SLs of the LGF process and lignin recovery yields after precipitation and purification such as composition of the purified LGF lignins.

The lignins dissolved during the LGF cooking were separated from the SL by water precipitation and analyzed. The material recovered by precipitation corresponds to 66.3%–77.5% of the lignin found in the SL (Table 1). A large part of the water-soluble lignin could not be not precipitated resulting in yield losses. The decreased yield in the LGF120 sample is most probably due to an experimental error. The yield of the purification step was around 84%. The chemical compositions of the water precipitated and purified LGF lignin samples are listed in Table 1. Only xylose was detectable in the hydrolysates after acid hydrolysis. The elemental compositions and contents of OMe and OAc groups of LGF lignins are presented in Table 2. Very little time-dependent differences were observed in the samples. LGF lignins contain more carbon, less oxygen, and slightly less OMe groups compared to MWLbirch reported in the literature (59.7% C, 34.2% O, and 21.4% OMe). The hydrogen contents around 6.1% are similar (Fengel and Wegener 1989). The higher carbon and lower oxygen contents of LGF lignins indicate that lignin condensation (formation of new C-C linkages) and/or reduction had taken place. A significant demethylation of lignin during LGF cooking did not occur, as the OMe contents of LGF lignins were only slightly lower than that in MWLbirch. Some phosphor remained attached to the lignin from the H3PO2 in the cooking liquor. The remaining phosphor corresponds to 2–3 P per C900 units. The P content in the purified and unpurified LGF120 samples was at the same level (0.4%), confirming that phosphor had not been removed by hydrolysis of H3PO2 lignin esters present in LGF lignins. Some OAc groups remained in the samples despite purification that includes dissolution in alkali. The unpurified sample contains 4.3% OAc groups (13 OAc per C900). Accordingly, the LGF lignins are heavily acetylated and the purification removes a large part of the OAc groups.

Table 2

Elemental composition and contents of OMe and OAc groups in LGF lignins and the empirical formulas based on 100 phenylpropane units.

The OH group analysis by 31P NMR (Table 3) shows that the content of OHaliph groups decreased significantly compared to the MWL, while the content of OHphen groups increased. Only minor changes in OH group contents were observed as a function of LGF process time, indicating that the most significant effects in this context had already taken place before the first 30 min measurement time. The practically constant OHtotal groups during the cooking time allow us to interpret that the cleavage of β-aryl ether linkages, leading to the formation of new OHphen groups, is accompanied by the elimination of a similar amount of OHaliph groups.

Table 3

Amount of different hydroxyl groups in LGF lignins and MWLbirch for comparison.

In pyrolysis-GC/MS analysis, lignin is degraded to a mixture of monomeric phenols that retain the aromatic ring substitution pattern in the original lignin. It is thus possible to identify products from p-hydroxyphenyl (H), guaiacyl (G), and syringyl (S) lignin units and calculate the S/G/H ratio (Table 4). The S/G ratios of LGF lignins (1.8) are somewhat lower than that of MWLbirch (2.1). The S/G ratios do not change essentially as a function of cooking time. As both G- and S-type pyrolysis degradation products have similar side chain structures, it is easy to present their structural changes with increasing cooking time (Figure 1). The amount of coniferyl and syringyl alcohols and aldehydes, typical to native lignin, is more abundant in MWLbirch than in LGF lignins. These units decrease in the LGF lignins in the beginning of the process, but after 60 min no further changes are observable, indicating that the cleavage of β-aryl ether linkages takes place mainly during the first 60 min cooking.

Table 4

Distribution of S, G, and H lignin units (after normalization to 100%), the amount of main interunit linkages, and molecular mass data. The possible elevated S data are not corrected.

Relative molar abundances of the pyrolysis degradation products of the guaiacyl and syringyl units of lignin.
Figure 1

Relative molar abundances of the pyrolysis degradation products of the guaiacyl and syringyl units of lignin.

The content of benzenediols (catechol and methoxybenzenediol) among the pyrolysis products is higher in LGF lignins than in MWLbirch with increasing tendency with longer cooking times. However, these changes are so minor that demethylation cannot be considered as relevant reaction during LGF cooking. The minor changes in OMe contents in the course of cooking support this view (Table 2).

2D NMR spectra of MWLbirch and LGF lignin at 120 min were recorded (Figure 2a and b, respectively). The intermonomeric linkages of MWLbirch and LGF lignins, as analyzed by integration of signals in 2D HSQC NMR spectra, are presented in Table 4. Accordingly, the content of β-O-4 bonds in LGF lignin are essentially lower compared to MWLbirch and some changes still took place between 60 and 90 min cooking time. The contents of β-5- and β-β-type C-C bonds did not increase simultaneously, indicating that the condensation of lignin via recoupling of the radical intermediates did not occur. Stilbene structures, known to be formed via β-5 structures, were detectable in the HSQC NMR spectra (marked with K in Figures 2b and 3). The main lignin structures detected in the 2D NMR spectra are presented in Table 5. In LGF lignins, several new 1H-13C correlation signals appeared in the 2D spectra (marked grey in Figure 2b), indicating relevant structural changes already after 30 min cooking. Especially, the signals at δCH 127.58/5.308 and 129.42/5.318 ppm – most likely originating from vinylic structures based on their chemical shifts – are clearly detectable (Figure 2b). Besides the new correlation signals detected in the aromatic region (δCH 105.66/6.426 and 112.38/6.685 ppm), an intense signal at δCH 59.26/3.636 ppm was observed. This overlapping signal is partly due to OMe and partly to γ-correlations of the interunit linkages. Another new signal was detected in the interunit linkage region resonating at δCH 81.18/4.76 ppm.

Contour plot 1H-13C HSQC NMR spectra (δC/δH 40–143/ 2.5–7.7): (a) MWLbirch and (b) LGF120 lignin.
Figure 2

Contour plot 1H-13C HSQC NMR spectra (δCH 40–143/ 2.5–7.7): (a) MWLbirch and (b) LGF120 lignin.

Main lignin structures detected in 2D 1H-13C HSQC NMR spectra.
Figure 3

Main lignin structures detected in 2D 1H-13C HSQC NMR spectra.

Table 5

Main lignin structures in MWLbirch and LGF lignin detected from the 2D 1H-13C HSQC NMR spectrum.

The cleavage of β-aryl ether linkages in the very early stage of LGF cooking is also evident based on molar mass measurements. The Mw values (Table 4) are in a similar range, but the molar mass distribution is narrower in the case of LGF lignins compared to that of MWLbirch (Figure 4a), clearly showing the cleavage of aryl ether linkages. During the LGF cooking, the molar mass of lignin remained rather constant. No clear differences were detected in the shape of the molar mass distributions either. This result also supports the conclusion that only minor changes took place after the very fast initial reactions. A minor shift to higher molecular weights with cooking times longer than 30 min can be interpreted to be due to some condensation reactions occurring with prolonged cooking times (Figure 4b). This low degree of lignin condensation may be one reason for the good bleachability of LGF pulps (Kangas et al. 2013).

(a) Molar mass distributions of MWLbirch and LGF120 showing the effect of LGF cooking and (b) molar mass distributions of LGF lignins as a function of cooking time.
Figure 4

(a) Molar mass distributions of MWLbirch and LGF120 showing the effect of LGF cooking and (b) molar mass distributions of LGF lignins as a function of cooking time.

Delignification mechanisms

According to the literature, the β ether cleavage under AA cooking at high temperature may take place via ionic acidolysis and/or homolytic reactions (Li et al. 2000). The cleavage reactions are competing with re-condensation reactions. Kangas et al. (2013) found that AA cooking in the presence of H3PO2 results in an improved delignification compared to other AA processes, in which mineral acids serve as catalysts. It was also demonstrated that only minor delignification takes place without H3PO2, leaving the wood chips unfiberized (unpublished results). Thus, one aim of the study was to reveal the boosting mechanism of H3PO2.

Acidolysis has been proposed to cleave β-aryl ether bonds via aryl-enol ethers as intermediates, partly including the cleavage of the γ-C via retro-Diels-Alder reaction. In the presence of H3PO2, an alternative reaction path is possible. Under the reaction conditions, the OH groups of lignin are expected to be in equilibrium with the corresponding esterified forms. Both H3PO2 and AA may participate in these equilibria. The α-C-OH group esterified by H3PO2 is postulated to be the key intermediate leading to the cleavage of β-O-4 bonds via a mechanism depicted in Figure 5. The concerted mechanism route necessitates the presence of H3PO2 esters in the α-position. This way, the same end-products as derived from the acidolysis reaction would be formed in a straightforward reaction, without the formation of the intermediate aryl-enol ethers. The relative abundance of the β-carbonyl structures among the pyrolysis degradation products (R=CH2 -CHO, CH2-CO-CH3) (Figure 1) supports the presence of β-carbonyl-type substructures in the LGF lignins. As the end-products in H3PO2-boosted and unboosted acidolysis are the same, the effect of H3PO2 cannot be detected on that basis. The intermediate enol-ether structures of the acidolysis mechanism were not detected by 2D NMR, which supports the presence of the alternative catalytic route but does not prove it, because the intermediates may have reacted further.

Proposed mechanism for acidolysis in the presence of H3PO2.
Figure 5

Proposed mechanism for acidolysis in the presence of H3PO2.

The carbocations in the α-position are formed as the first step of the normal acidolysis reaction. They are prone to condensation side-reactions, leading to the formation of α-aryl structures. The H3PO2-catalyzed acidolysis reaction does not include carbocation intermediates, which is expected to decrease the condensation tendency and promote delignification also this way.

H3PO2 was added to the LGF process in approximately 1:4 molar ratio to the C9 subunits. This high charge is probably necessary to make sure that the cleavage reaction happens so fast that the competing condensation reaction cannot occur. In addition, in delignification reactions involving H3PO2, the ester formed in the α-position protects lignin toward condensation reactions. H3PO2 is partly regenerated in the cleavage reaction, being a catalyst rather than a reagent in that sense. However, a part of H3PO2 is “lost” in the course of oxidation-reduction reactions and some remain attached to the lignin as shown by 2–3 P per C900 units. The reduction of lignin during LGF cooking is visible by the lower oxygen content compared to MWLbirch.

A homolytic cleavage may also occur under the cooking conditions applied and the radical recoupling reactions form various new C-C bonds. However, the radical intermediates may also stabilize via other mechanisms. H3PO2 is a reducing agent and thus could stabilize the lignin radicals by direct reduction and block the condensation via the homolytic route. No new C-C linkages were detectable in the LGF lignins (Table 4), suggesting that condensation via the homolytic route is not a major reaction. However, it is not yet clear whether the efficient acidolysis or the reductive stabilization of the intermediates by H3PO2 is more relevant.


The LGF dissolved lignin was found to be structurally different from native lignin. The majority of β-O-4 linkages were cleaved, only 9 β-O-4 per 100 C9 units were left after 120 min LGF process time, and new OHphen groups were generated. The number of OHaliph was lessened. These changes took place mainly during the early stages of the process, and after 60 min reaction time, no further structural changes were observable. However, lignin dissolution continued during the whole reaction time (120 min). The delignification reaction is fast, and only minor secondary reactions take place once the lignin has dissolved.

The main delignification reaction is probably acidolysis, catalyzed by H3PO2 esterification. In addition to boosting the aryl ether cleavage, H3PO2 may also protect lignin against the typical condensation reactions under acidolytic conditions. In parallel, homolytic cleavage may also occur, but its significance is not yet clear. H3PO2 is expected to stabilize lignin also against condensation via radical coupling, based on its reductive nature. Potentially, softwoods could benefit more from the LGF process, because their G units are more prone to condensation reactions than the S units in hardwoods.


The financial support from the Academy of Finland under grant no. 252564 is gratefully acknowledged. The authors would like to thank Jari Leino, Juha Haakana, Eila Turunen, Yukho Sok-Sar, Nina Vihersola, and Tuija Kössö for their skilful technical assistance.


  • Adler, E., Pepper, J.M., Eriksoo, E. (1957) Action of mineral acid on lignin and model substances of guaiacylglycerol-beta-aryl ether type. Ind. Eng. Chem. 49:1391–1392.CrossrefGoogle Scholar

  • Aziz, S., Sarkanen, K. (1989) Organosolv pulping – a review. Tappi J. 72:169–175.Google Scholar

  • Björkman, A. (1956) Studies on finely divided wood. Part I. Extractions of lignin with neutral solvents. Svensk Papperstidn. 60:477–485.Google Scholar

  • del Rio, J.C., Rencoret, J., Prinsen, P., Martinez, A.T., Ralph, J., Gutierrez, 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.Google Scholar

  • Fengel, D., Wegener, G. Wood Chemistry, Ultrastructure, Reactions. Walter de Gruyter, Berlin/New York, 1989. p. 152.Google Scholar

  • Goldsmith, O. (1971) Ultraviolet spectra. In: Lignins: Occurrence, Formation, Structure and Reactions. Eds. Sarkanen, K.V., Ludwig, C.H. John Wiley & Sons, New York. pp. 241–298.Google Scholar

  • Granata, A., Argyropoulos, D.S. (1995) 2-Chloro-4,4,5,5-tetramethyl-1,3,2-dioxaphospholane, a reagent for the accurate determination of the uncondensed and condensed phenolic moieties in lignins. J. Agric. Food Chem. 43:1538–1544.Google Scholar

  • Hakala, T.K., Hyvärinen, S., Liitiä, T., Suurnäkki, A. (2010) Enzyme-assisted extraction of hemicelluloses and lignin from organosolv pulps. 11th European Workshop on Lignocellulosics and Pulp. August 16–19, 2010, Hamburg, Germany. pp. 153–157.Google Scholar

  • Hausalo, T. (1995) Analysis of wood and pulp carbohydrates by anion exchange chromatography with pulsed amperometric detection. 8th International Symposium on Wood and Pulping Chemistry (ISWPC). June 6–9, 1995, Helsinki, Finland. pp. 131–136.Google Scholar

  • Hergert, H. (1998) Developments in organosolv pulping – an overview. In: Environmentally Friendly Technologies for the Pulp and Paper Industry. Eds. Young, R.A., Akhtar, M. John Wiley & Sons, Inc., New York. pp. 5–67.Google Scholar

  • Kangas, H., Tamminen, T., Liitiä, T., Asikainen, S., Poppius-Levlin, K. (2013) Lignofibre (LGF) organosolv process for producing dissolving pulps. 17th International Symposium on Wood, Fibre and Pulping Chemistry (ISWFPC). June 12–14, 2013, Vancouver, BC, Canada.Google Scholar

  • Karlsson, O., Lundquist, K. (1992) Acid reactions of hydrobenzoins: effects of catalyst and reaction medium on product composition. Acta Chem. Scand. 46:283–289.CrossrefGoogle Scholar

  • Karlsson, O., Lundquist, K., Meuller, S., Westlid, K. (1988) On the acidolytic cleavage of arylglycerol B-aryl ethers. Acta Chem. Scand. 42B:48–51.Google Scholar

  • Lai, Y.-Z. (2001) Chemical degradation. In: Wood and Cellulose Chemistry. 2nd ed., revised and expanded. Eds. Hon, D.N-S., Shiraishi, N. Marcel Dekker, Inc., New York, pp. 443–512.Google Scholar

  • Li, S., Lundquist, K., Westermark, U. (2000) Cleavage of arylglycerol B-aryl ethers under neutral and acidic conditions. Nord. Pulp Paper Res. J. 15:292–299.CrossrefGoogle Scholar

  • Li, J., Henriksson, G., Gellerstedt, G. (2007) Lignin depolymerization/repolymerization and its critical role for delignification of aspen wood by steam explosion. Biores. Technol. 98: 3061–3068.CrossrefGoogle Scholar

  • Liitiä, T., Maunu, S.L., Hortling, B., Toikka, M., Kilpeläinen, I. (2003) Analysis of technical lignins by two- and three-dimensional NMR spectroscopy. J. Agric. Food Chem. 51: 2136–2143.CrossrefGoogle Scholar

  • Liitiä, T., Mikkonen, H., Paananen, M. Uusitalo, J., Tamminen, T., Colodette, J.L. (2011) Organosolv pretreatment of lignocellulosics for bioethanol production. 5th ICEP, May 9–12, 2011, Porto Seguro, Bahia, Brazil. 8 pp.Google Scholar

  • Liitiä, T., Barth, D., Tamminen, T., Colodette, J.L. (2013) Organosolv pulping as a pretreatment for bioethanol production from Eucalyptus and elephant grass. 17th International Symposium on Wood, Fibre and Pulping Chemistry. June 12–14, 2013, Vancouver, BC, Canada.Google Scholar

  • Lora, J.H., Wayman, M. (1980) Simulated autohydrolysis of aspen milled wood lignin in the presence of aromatic wood additives. Changes in molecular weight distribution. J. Appl. Polym. Sci. 25:589–596.CrossrefGoogle Scholar

  • Lundquist, K., Lundgren, R. (1972) Acid degradation of lignin. Part VII. The cleavage of ether bonds. Acta Chem. Scand. 26:2005–2023.CrossrefGoogle Scholar

  • Mehrotra, R.N. (2013) Three oxyacids of phosphorous: tautomerism and oxidation mechanisms. Eur. Chem. Bull. 2: 758–776.Google Scholar

  • Mikkonen, H., Peltonen, S., Kallioinen, A., Suurnäkki, A., Kunnari, V., Malm, T. (2009) Process for defibering a fibrous material. Patent WO2007/066007.Google Scholar

  • Ohra-aho, T., Gomes, F., Colodette, J., Tamminen, T. (2013) S/G ratio and lignin structure among Eucalyptus hybrids determined by Py-GC/MS and nitrobenzene oxidation. J. Anal. Appl. Pyrol. 101:166–171.CrossrefGoogle Scholar

  • Pregl, F., Roth, H. (1958) Quantitative Organische Mikroanalyse, 7. Auflage. Springer-Verlag, Wien. 245 pp.Google Scholar

  • Ralph, J., Landucci, L.L. (2010) NMR of lignins. In: Lignin and Lignans; Advances in Chemistry. Eds. Heitner, C., Dimmel, D.R., Schmidt, J.A. CRC Press (Taylor & Francis Group), Boca Raton, FL. pp. 137–234.Google Scholar

  • Rovio, S. (2010) Determination of the carboxylic acids in acidic and basic process samples by capillary zone electrophoresis. J. Chromatogr. A 1217:1407–1413.Google Scholar

  • Sette, M., Wechselberger, R., Crestini, C. (2011) Elucidation of lignin structure by quantitative 2D NMR. Chem. Eur. J. 17:9529–9535.CrossrefGoogle Scholar

  • Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D. (2008) NREL Standard, Determination of Structural Carbohydrates and Lignins in Biomass. Laboratory Analytical Procedure (LAP).Google Scholar

  • Tamminen, T., Liitiä, T., Hyvärinen, S., Mikkonen, H., Hakala, T.K., Ohra-aho, T., Rovio, S., Suurnäkki, A., Colodette, J.L. (2011) Characterisation of xylans and lignins isolated from various Eucalyptus species. 5th ICEP, May 9–12, Porto Seguro, Bahia, Brazil. 8 pp.Google Scholar

  • Vazquez, G., Antorrena, G., Gonzalez, J., Freire, S. (1997) FTIR, 1H and 13C NMR characterization of acetosolv-solubilized pine and Eucalyptus lignins. Holzforschung 51:158–166.Google Scholar

  • Willför, S., Pranovich, A., Tamminen, T., Puls, J., Laine, C., Suurnäkki, A., Saake, B., Uotila, K., Simolin, H., Hemming, J., Holmbom, B. (2009) Carbohydrate analysis of plant materials with uronic acid-containing polysaccharides – a comparison between different hydrolysis and subsequent chromatographic analytical techniques. Ind. Crops Prod. 29:571–580.CrossrefWeb of ScienceGoogle Scholar

About the article

Corresponding author: Heli Kangas, VTT Technical Research Centre of Finland, P.O. Box 1000, FI-02044 VTT, Finland, e-mail:

Received: 2014-03-08

Accepted: 2014-08-22

Published Online: 2014-09-27

Published in Print: 2015-04-01

Citation Information: Holzforschung, Volume 69, Issue 3, Pages 247–256, ISSN (Online) 1437-434X, ISSN (Print) 0018-3830, DOI: https://doi.org/10.1515/hf-2014-0070.

Export Citation

©2015 by De Gruyter. Copyright Clearance Center

Citing Articles

Here you can find all Crossref-listed publications in which this article is cited. If you would like to receive automatic email messages as soon as this article is cited in other publications, simply activate the “Citation Alert” on the top of this page.

L. Lagerquist, A. Pranovich, A. Smeds, S. von Schoultz, L. Vähäsalo, J. Rahkila, I. Kilpeläinen, T. Tamminen, S. Willför, and P. Eklund
Industrial Crops and Products, 2018, Volume 111, Page 306
Jingming Tao, Omid Hosseinaei, Lukas Delbeck, Pyoungchung Kim, David P. Harper, Joseph J. Bozell, Timothy G. Rials, and Nicole Labbé
RSC Adv., 2016, Volume 6, Number 82, Page 79228

Comments (0)

Please log in or register to comment.
Log in