3

Methods to Reveal theStructure of Lignin

Prof. Dr. Gösta BrunowDepartment of Chemistry, Laboratory of Organic Chemistry, University of Helsinki,P. O. Box 55, 00014 University of Finland, Finland; Tel: �358-919140361;Fax: �358-919140366, E-mail : gosta.brunow@helsinki.fi

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

2 Historical Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

3 Lignin Preparations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913.2 Milled Wood Lignin (MWL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

4 Determination ofTotal Lignin in Lignocellulosics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

5 Methods to Determine the Molecular Mass of Lignin . . . . . . . . . . . . . . . 935.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935.2 Gel Permeation Chromatography . . . . . . . . . . . . . . . . . . . . . . . . . . 935.3 Vapor Pressure Osmometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945.4 Light Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945.5 Ultrafiltration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

6 Degradative Methods of Lignin Analysis . . . . . . . . . . . . . . . . . . . . . . 946.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 946.2 Acidolysis and Thioacidolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 956.3 Permanganate Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 966.4 Nitrobenzene and Cupric Oxide Oxidation . . . . . . . . . . . . . . . . . . . . 966.5 Ozonolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 966.6 Reductive Cleavage after Derivatization (DFRC) . . . . . . . . . . . . . . . . . 986.7 Functional Group Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 986.7.1 Methoxyl Group Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 986.7.2 Phenolic Hydroxyl Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

89

6.7.3 Carbonyl Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006.7.4 Quinoid Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

7 Nondegradative Methods of Lignin Analysis . . . . . . . . . . . . . . . . . . . . 1007.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007.2 Functional Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007.3 Aromatic Nuclei and Side Chain Structures . . . . . . . . . . . . . . . . . . . . 101

8 Structural Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028.2 Softwood Lignin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028.3 Hardwood Lignin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1038.4 Non-Wood Lignin (Straw, Grass) . . . . . . . . . . . . . . . . . . . . . . . . . . 103

9 Patents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

CEL cellulolytic enzyme ligninDHP dehydrogenation polymerGPC gel permeation chromatographyHPSEC high-pressure size exclusion chromatographyIR infraredMWL milled wood ligninNMR nuclear magnetic resonanceSEC size exclusion chromatographyUV/Vis ultraviolet/visibleVPO vapor pressure osmometry

1

Introduction

Lignin in the cell walls of vascular plants isintimately mixed with the carbohydrate com-ponents. The structure of the polymer iscomplex and irregular, and preparation ofpure samples of unchanged lignin is not easy.This makes structure determination of ligninmore challenging than that of other biopoly-mers. In the following chapter, the mostusefulmethods for the isolationandstructureelucidation of lignin are described, with asummary description of the structural pat-tern emerging from these studies.

2

Historical Outline

Long after Anselme Payen (1839) first de-scribed the `encrusting material' in wood,researchers were unclear about the nature ofthisveryabundantmaterial.Althoughithadahigher carbon content than the carbohy-drates, its chemical nature remained obscurefor a long time; indeed, it was not until about20 years later that the term `lignin' becameaccepted for this material (Schulze, 1857).Aromatic products formed on alkali fusion(Bente, 1868 and references cited therein) ledto the conclusion that the noncellulosicconstituent of wood, or lignin, was aromaticin nature. That the methoxyl group was

3 Methods to Reveal the Structure of Lignin90

typical of lignin and lacking in cellulose wasshown in 1890 (Benedikt and Bamberger,1890). The methods to reveal the structure oflignin evolved very slowly, and as recently as1960 it might be read in a review that ªthelignin building stone has a phenylpropanestructure that may be regarded as proven, buthow the stones are linked together in proto-lignin is still a mysteryº (Brauns, 1960). Nodimeric degradationproduct, thatmight haverevealed the nature of the bonding betweentheunits,hadbeen isolatedat that time. Itwasin fact not degradation that led to a break-through in the understanding of ligninchemistry, but an idea that usually is tracedback to P. Klason, who postulated that ligninwas an oxidation product of coniferyl alcohol(Adler, 1977). The fruitfulness of this ideawas demonstrated by Freudenberg (1968)when he succeeded in producing polymersby oxidative dehydrogenation of coniferylalcohol. These dehydrogenation polymers(DHPs) were found to contain the samestructural units as native lignin, and havebeen used extensively as models for lignin. (Itwas in fact the advent of new reducing agentsduring the early 1950s that made it possible tosynthesize coniferyl alcohol in the largeamounts necessary for such a study.) To thisday, all analytical results in lignin chemistryare interpreted in the light of the dehydrogen-ation hypothesis. Most structural units inlignin can be regarded as oxidation productsof coniferyl alcohol, and the fact that there aresome structural units that do not fit into thisscheme, has not diminished its usefulness.

3

Lignin Preparations

3.1

Introduction

A critical review of methods for the isolationof lignin was published by Lai and Sarkanen

(1971). The most commonly used methodinvolves thorough milling of the plant mate-rial, followed by extraction with dioxan±water(Björkman, 1957; Lundquist, 1992a); thismaterial is referred to as milled wood lignin(MWL). The yields are usually low, and thepossibility of chemical changes occurringduring the isolation process must always betaken into account.

3.2

Milled Wood Lignin (MWL)

When choosing a sample of plant tissue forthe isolation of MWL, consideration shouldbe taken into the possible variations in lignincomposition in different parts of the plant. Toprepare a lignin sample representative of acertain species of wood, it is customary tochoose sapwood, which is free of reactionwood. The milling is carried out either in anonswelling medium such as toluene, or inthe dry state. Treatment of the finely groundwood meal with cellulolytic enzymes prior tosolvent extraction removes part of the poly-saccharides and increases the yield of lignin.Such preparations are referred to as cellulo-lytic enzyme lignins (CEL) (Chang et al.,1975). After extraction with dioxane±water,the crude extract still contains carbohydrates.In the original Björkman procedure, CEL ispurified by precipitation into water from asolution in acetic acid. An alternative methodhas been proposed that is based on liquid±liquid extraction (Lundquist, 1992a), thisyielding lignin preparations with less carbo-hydrate contamination. The remaining car-bohydrate contaminants are difficult to re-move, as some may be covalently bound to thelignin. Hardwood MWLsare foundto containmore carbohydrates than softwood MWLs.MWL usually constitutes about 25% of thelignin in wood, and the question of itsmorphological origin has been discussed(Lai and Sarkanen, 1971; Lee et al., 1981;

3 Lignin Preparations 91

Whiting and Goring, 1981; Eom et al., 1987).The milling probably causes chemicalchanges (Itoh et al. , 1995), but the extent ofthese changes is largely unknown. In struc-tural studies (see below) it has not beenpossible to find any significant differences inthe chemical structure of MWL and lignin insitu. CEL is probably more representative ofthe total lignin, with higher yields and lesschemical change (Chang et al. , 1975), but ithas other drawbacks, such as high carbohy-drate content and its preparation procedure ismore tedious (Lapierre et al. , 1985).

4

Determination of Total Ligninin Lignocellulosics

There is no generally applicable method forthe quantitative determination of total ligninin lignocellulosics. The oldest and mostcommon method is based on gravimetry. Inthe Klason lignin determination which hasbeen standardized by the Technical Associa-tion of the Pulp and Paper Industry (Dence,1992), the sample is first treated with 72%sulfuric acid and subsequently heated withdilute acid to hydrolyze the polysaccharides tosoluble fragments. The solid residue iswashed, dried and weighed. The method isreliable provided that standard conditions arestrictly applied, but also has serious limita-tions that must be taken into consideration.Correct values are obtained for softwoods, buthardwoods contain variable amounts of `acid-soluble lignin' which must be estimated byUV spectrophotometry. The method is alsonot suitable for herbaceous and annualplants. Such material tends to contain varia-ble amounts of proteins and siliceous materi-al that interfere with the determination ofKlason lignin. For such plant material there isa modification of the Klason determination,developed by Ellis (1949) and adapted as an

official method by the Association of OfficialAgricultural Chemists (Dence, 1992). Thismethod involves pretreatment with a proteo-lytic enzyme, but since there is no assurancethat such pretreatment effects completeremoval of protein, it is often be necessaryto correct the values for acid-insoluble ligninwith calculated values based on nitrogenanalysis. In most cases, acid-soluble ligninhas also to be determined and the valueapplied as a correction to arrive at an accuratevalue for total lignin.

There are also methods that do not involveisolation of the lignin. UV microspectropho-tometry has been used to estimate lignin indifferent morphological regions (Fergus andGoring, 1970). Other spectrometric tech-niques include infrared spectroscopy (Schultzet al., 1985) and solid-state NMR (13C CP/MAS/NMR; Leary and Newman, 1992).

Some spectral techniques involve dissolv-ing the whole plant material in a suitablesolvent and measuring the UV-absorbance ata wavelength characteristic for the lignin inquestion, usually 280 nm. Acetyl bromide inacetic acid (Iiyama and Wallis, 1988; Hatfieldet al. , 1999) seems to have earned widespreadacceptance as a rapid and simple method,adaptable to samples of small size.

Special methods, which are used exclusive-ly for the analysis of unbleached pulps, arebased on the consumption of oxidant. Ligninconcentration determined by such proce-dures is usually expressed as the amount ofoxidant per unit weight of pulp (for instancethe Roe chlorine number, or the kappanumber). These numbers may be convertedto Klason lignin or other lignin values byapplication of empirically determined con-version factors. The two oxidants most com-monly used are chlorine and potassiumpermanganate. It is assumed that the oxidantreadily oxidizes lignin, whereas the carbohy-drate constituents are relatively unreactive(Dence, 1992).

3 Methods to Reveal the Structure of Lignin92

5

Methods to Determine the Molecular Massof Lignin

5.1

Introduction

An important limitation of the study of woodpolymers is that properties such as molecularmass, molecular shape, crystallinity, density,etc., have been investigated almost exclusive-ly with isolated samples, and these character-istics of the wood polymers in situ can bededuced only by inference (Goring, 1971).Considering the large number of differentlinkages between the phenylpropane units, ithas been assumed that lignin constitutesa three-dimensional network polymer. Tobreak this down to soluble fragments, bondsmust be broken, and random scission ofbonds in such a network will lead to a widerange of molecular sizes. It is typical of ligninthat isolated samples are very polydisperse,and the measured molecular size range isvery much dependent on the isolation proce-dure. The polydispersity is accompanied byrather small changes in chemical properties.Soluble lignin preparations tend to be com-posed of molecules of similar constitution,but differing widely in size. Most methods todetermine themolecularmassof ligninshavebeen developed for the study of industrialspent liquors. For MWLs, the most importantmethods aregelpermeationchromatography(GPC), light scattering and vapor pressureosmometry. There are also reports on theapplication of ultrafiltration (Lin, 1992b) andmass spectrometry (Metzger et al. , 1992;Evtuguin et al., 1999) to lignins.

5.2

Gel Permeation Chromatography

A commonly used method of measuring thesizeofmacromolecules isbasedonmolecular

sieving in a chromatographic column. Forsynthetic polymers in organic solvents thetechniques of GPC or `size exclusion chro-matography' (SEC) (Gellerstedt, 1992) havebeen used; when applied to water-solublebiopolymers, the technique is known as `gelfiltration'.

The procedure involves passing a solutionof macromolecules through a column of asolvent-filled porous gel. Depending on theirsize, the macromolecules can diffuse invarying proportions into the porous gel.Molecules with a small hydrodynamic radiuscan penetrate into the gel, while largemolecules are excluded from the gel and passdirectly through the column. The elutionvolume of any particular fraction is a functionof thedimensionsof themacromoleculesandthe sizes of the pores in the gel. In order toconvert the elution volumes into molecularweight, the relationship between molecularsize and mass of the macromolecular soluteshould be known. The simplest solution is torun monodisperse fractions of known mo-lecular weight through the column and thusestablish the relationship between molecularmass and elution volume (Månsson, 1981).In earlier work, cross-linked polydextran gelshaving varying pore sizes were employed asstationary phases. The gels were swollen inwater and used for fractionating aqueoussolutions. Subsequently, modified polydex-trans have become available that permitfractionation in organic solvents (Connors,1978). Recently, cross-linked rigid poly-styrene gels have made it possible to per-form GPC in high-pressure SEC systems(HPSEC), where the time of analysis can begreatly reduced and the resolution enhancedconsiderably when compared to normal GPC(Yau et al. , 1979). For milled wood lignins,elution with tetrahydrofuran has been foundto give reliable results (Månsson, 1981;Chum et al. , 1987). For added solubility, thelignin samples are often derivatized (by

5 Methods to Determine the Molecular Mass of Lignin 93

methylation or acetylation) before analysis.For calibration, polystyrene fractions or poly-peptides (Brunow et al., 1993) of knownmolecular weight and lignin model com-pounds have been used.

5.3

Vapor Pressure Osmometry

Vapor pressure osmometry (VPO) is com-monly used to determine number±averagemolecular masses in the range of 100 to10,000. Beyond the upper limit the sensitivitybecomes unsatisfactory. In spite of that, VPOnow appears to be the preferred methodfor determining number-average molecularweights of lignins and lignin products. Allcolligativemethodsareuncertain intherangeof 10,000 to 25,000 and average molecularmasses within this range are difficult todetermine accurately (Pla, 1992a). The prin-ciple of the method is based on the measure-ment (at a given temperature) of the vaporpressure depression of the solvent for dilutepolymer solutions. The instrument must becalibrated with a substance of known molec-ular mass.

5.4

Light Scattering

For measuring the number-average molec-ular mass of soluble lignins, low-angle laserlight-scattering photometers have been used(Pla, 1992b), and are reported to overcomesome of the sources of error associated withobtaining reproducible results from light-scattering measurements. The presence ofaggregates in the solution, and the absorb-ance or fluorescence of lignin in solution,may influence the measurements.

5.5

Ultrafiltration

Membrane separation is a relatively newtechnology that can be used for separatinglignins on the basis of molecular size.Although not a standard method, it can beused for the determination of molecular sizedistribution of lignin samples. Membranesare available for a wide range of molecularsizes, from 1000 to 300,000. The solution oflignin is filtered through a succession ofmembranes and the yields of fractions aredetermined by weighing the lignin fraction,or by ultraviolet absorption (Lin, 1992a). Thechief advantage of the method is the insensi-tivity to impurities. Lignin permeatesthrough ultrafiltration membranes almostunhinderedbysugarsor inorganicsalts.Suchimpurities render classical physical tech-niques, such as VPO and light scattering,unusable.

6

Degradative Methods of Lignin Analysis

6.1

Introduction

It is not possible to degrade lignin intomonomeric fragments like most other bio-polymers. In addition to the uncertaintycaused by the difficulty of preparing repre-sentative lignin samples that have not under-gone any chemical change, all known meth-ods of chemical degradation yield identifiableproductsofsmallmolecularweight inmodestyields only. Under such circumstances thetask of devising a satisfactory structuralpicture of lignin macromolecules has beenlikened (Sarkanen and Ludwig, 1971) to anattempt to compose a picture-puzzle with anincomplete number of pieces. What is con-tained in the missing pieces can only be

3 Methods to Reveal the Structure of Lignin94

guessed at, and there may be more than oneway to assemble the available pieces. Thedegradative methods described below are allaimed at some particular aspect of thestructure of lignin, and an extensive struc-tural analysis implies the use of more thanone method and a combination with aspectroscopic method.

6.2

Acidolysis and Thioacidolysis

Acid-catalyzed degradation aims at cleavingthe most important ether bonds in lignins,thearylglycerol-b-arylethers.Acidolysis is theterm used for the method where the sample isheated at 100 8C in 0.2 M HCl in dioxane±water (9:1, v/v; Lundquist, 1976, 1992b). Thiscauses selective cleavage of arylglycerol-b-

ethers and some other types of labile etherlinkages.

Monomeric and dimeric acidolysis prod-ucts that contain a phenylpropane skeleton,can be analyzed by gas chromatography aftersilylation. The results are then interpretedwith the aid of low molecular weight `model'compounds, that have undergone the sametreatment. The structural elements detectedwith the aid of acidolysis studies include b-O-4, b-5, b-b, b-1, glyceraldehyde-2-aryl ether,2-aryloxypropiophenone, cinnamaldehyde,cinnamic acid, benzaldehyde, benzoic acidand quinoid types (see Figure 1). Some ofthese elements have also been estimatedquantitatively. The same structural units aredetected by thioacidolysis, where the sampleis treated with boron trifluoride in dioxan±ethanethiol solution (Rolando et al. , 1992).Monomeric products substituted with thio-

6 Degradative Methods of Lignin Analysis 95

Scheme 1 Acidolysis (1) and thioacidolysis (2).

ethyl groups are formed, and these can beanalyzed by gas chromatography after silyla-tion. Dimeric products can be analyzed afterremoval of the sulfur substituents by reduc-tion with Raney-nickel (Lapierre et al., 1991).The yields of thioacidolysis products aretypically higher than in acidolysis, and thereare some noteworthy differences in thestructural information obtained. One suchdifference is the occurrence of 5,5 or 5-O-4bonded dimers (Figure 2); these are found inthioacidolysis but are absent in acidolysismixtures.Also, theratiosofunitsof type1and2 (Figure 3) are different in the productmixtures from the two methods (Lapierreet al. , 1985). An important limitation, as faras the structure of lignin is concerned, is thatboth acidolysis and thioacidolysis only candetect structural units bound by arylglycerol-b-ether bonds (Scheme 1).

6.3

Permanganate Oxidation

Inpermanganateoxidation, thesidechains inthe lignin are degraded to carboxyl groupsattached to the aromatic ring. The structuresof the products thus reveal the pattern ofsubstitution on the aromatic rings. Prior tothe oxidation, the sample is alkylated, usuallywith dimethyl or diethyl sulfate (Freuden-berg, 1968; Erickson et al., 1973a; Bose et al.,1998). This protects the phenolic aromaticrings from degradation, while all otheraromatic rings are degraded in the perman-ganate oxidation. Some biphenyls and di-phenyl ethers also survive the oxidativedegradation. Themixture ofacids is esterifiedand analyzed by gas chromatography. Thestructures provide information about thesubstitution pattern of the phenolic structur-al units in the lignin. Monocarboxylic acidsare formed from uncondensed phenolicgroups, and dicarboxylic acids are formedfrom condensed units with an additional

carbon substituent in the 5 or 6 position. Thepresence of catechol units can be detected ifthe alkylation is carried out with diethylsulfate. The structural information obtainedwith permanganate oxidation is mostly qual-itative. The generally low yields of degrada-tion acids make it difficult to estimate thequantitative distribution of structural unitsfrom permanganate oxidation data (cf. Erick-sonet al., 1973b;Bose et al., 1998) (Scheme 2).

6.4

Nitrobenzene and Cupric Oxide Oxidation

Nitrobenzene oxidation and cupric oxideoxidation give very similar degradation prod-ucts, and are also more simple to performthan permanganate oxidations. They areprimarily carried out in order to classify thelignin in terms of the proportions of phenyl-propane units of types 1±3 (Figure 3). It isthus possible to characterize lignins of differ-ent botanical origin and in different morpho-logical region of a plant. The reactions arecarried out in alkaline solutions at high tem-perature, though the reaction mechanismsare still not well understood. The structuralinformation from these oxidations is limitedand they are usually combined with spectro-scopic studies (Chen, 1992a) (Scheme 3).

6.5

Ozonolysis

Degradationof ligninwithozoneachieves theopposite of permanganate oxidation. Whilepermanganate degrades the side chains,ozone destroys double bonds and the aro-matic rings, leaving the side chains intact inthe form of carboxylic acids. The maininterest in ozonolysis lies in the fact that thestereochemical configuration of the sidechain carbons in lignin is retained in theoxidation products. Thus, the erythro form ofguaiacylglycerol-b-aryl ethers yield erythron-

3 Methods to Reveal the Structure of Lignin96

6 Degradative Methods of Lignin Analysis 97

Scheme 2 Permanganate oxidation.

Scheme 3 Nitrobenzene/cupric oxide.

Scheme 4 Ozonolysis.

ic acid and threonic acid is formed from thecorresponding threo isomer (Sarkanen et al.,1992) (Scheme 4).

6.6

Reductive Cleavage after Derivatization (DFRC)

Acidolysis and thioacidolysis are specificreactions that cleave the b-aryl ether bondsin lignins. The strongly acidic treatmentsinvolved cause extensive acid-catalyzed reac-tions that complicate the analysis of theproducts of degradation. In an effort todevelop a simpler method with cleanercleavage reactions of b-aryl ether bonds, aprocedure has been introduced that involvesderivatization of lignin with acetyl bromideand reductive cleavage of the resulting benzylbromides with zinc dust. This sequence ofreactions cleaves the b-aryl ether bonds, andthe liberated phenylpropane units can beanalyzed as cinnamyl alcohol derivatives. Theproduct mixtures in the derivatization fol-lowed by reductive cleavage (DFRC) (Lu andRalph, 1997) tend to be simpler than thoseobtained in acidolysis and thioacidolysis.Apart from monomeric products, a numberof dimers and trimers have been identified.These include representatives of all the com-mon interunit linkages in softwood ligninexcept the b-O-4which is efficiently cleaved inthe degradation reaction (b-1, b-b, 5 ± 5, b-5,and 5-O-4) (Peng et al. , 1998). Among thetrimers, a novel isochroman structure hasbeen found (Peng et al. , 1999), and it issuggested that such structures may exist

in unchanged lignins (Ralph et al., 1998)(Scheme 5).

6.7

Functional Group Analysis

6.7.1

Methoxyl Group AnalysisThe methoxyl content is a widely used basisfor characterization of lignins. Methoxylgroups are present in guaiacylpropane units(1) and syringylpropane units (2). The meth-oxyl content of a lignin sample reflects thedistribution of units 1 ± 3 (Figure 3). Themethoxyl content may give a measure of thepurity and the plant origin of lignin samples.The method most widely used is a modifica-tion of one originally introduced by Zeisel(1885), and is based on the formation ofmethyl iodidewhenthesample is treatedwithhydriodic acid at reflux temperature (Chen,1992b). The methyl iodide is then treatedwith bromine, which liberates the iodine andconverts it to iodic acid. The iodic acid istreated with potassium iodide, and the liber-ated iodine is titrated with standard thiosul-fate solution. In this sequence of reactions sixatoms of iodine are produced for eachmethoxyl group, which allows determina-tions with a high degree of accuracy, even on amicro scale (Scheme 6).

6.7.2

Phenolic Hydroxyl GroupsCommon instrumental methods are poten-tiometric and conductometric titration, ion-ization difference UV spectroscopy (see be-

3 Methods to Reveal the Structure of Lignin98

Scheme 5 Reductive cleavage after derivatization (DFRC).

low), and NMR spectroscopy (Lai, 1992).Oxidation of guaiacyl and syringyl structureswith aqueous sodium periodate to o-qui-nones, with the release of 1 mol of methanolfrom each phenolic units, has been used toestimate phenolic hydroxyl groups in lignins.The amount of methanol is measured by gaschromatography. The good reproducibilityand comparative simplicity makes it usefulfor routine purposes (Francis et al., 1991).The aminolysis method is based on the

finding that phenolic acetates are cleavedwith pyrrolidine much faster than aliphaticacetates (Månsson, 1983). It is a morelaborious procedure, and the accuracy iscritically dependent on the extent of acetyla-tion and the selectivity of the deacetylation.The inherent difficulties of obtaining reliabledata for phenolic hydroxyl contents in ligninsis reflected in the large variation in values(from 18 to 33%) obtained for spruce MWLwith different methods (Lai, 1992).

6 Degradative Methods of Lignin Analysis 99

Scheme 6 Methoxyl group analysis.

Scheme 7 Analysis of phenolic hydroxylgroups; (1) oxidation with periodate, (2)Aminolysis, (3) quinoid structures.