CELLULOSE CHEMISTRY AND TECHNOLOGY

ANALYTICAL METHODS FOR LIGNIN CHARACTERIZATION. II. SPECTROSCOPIC STUDIES CARMEN-MIHAELA POPESCU, CORNELIA VASILE, MARIA-CRISTINA POPESCU, GH. SINGUREL*, V. I. POPA** and B. S. MUNTEANU* Romanian Academy, “P.Poni” Institute of Macromolecular Chemistry, Iasi, ROMANIA * “Al. I. Cuza” University, Optics and Spectroscopy Department, Iaşi, ROMANIA ** Gh. Asachi Technical University, Iasi, ROMANIA Received October 5, 2006 Lignin’ characterisation is a very difficult task, if considering its diversity with respect to both origin and method of separation. The heterogeneity of lignin is caused by variations in polymer’s composition, size, crosslinking, functional groups, and linkage type between the phenyl propane monomers (p – hydroxyl phenyl, guaiacyl and syringyl units). The elaboration of well-defined analytical methods for lignin characterization is very important for its industrial applications as a raw material. Two groups of lignin from woody species and annual fibre crops have been studied by FT-IR, UV, fluorescence and 13C NMR spectroscopy. The relative content of the different functional groups (p - hydroxy phenyl, guaiacyl and syringyl units) was appreciated by normalised intensities and deconvolution of the spectral bands. Correlation of the results provided by these methods permits differentiation between the structural characteristics of two lignin groups from the viewpoint of their particularities.

Keywords: lignin, guaiacyl, syringyl, p – hydroxyl phenyl, IR spectroscopy, UV spectroscopy, 13C NMR spectroscopy, fluorescence spectroscopy

INTRODUCTION Lignin, one of the main structural polymers in the cell wall of any higher plant, is the second, after cellulose, the most abundant macromolecular compound of phytomass.1 It plays a major role in plant tissues, as a mechanical support and stress protection agent, providing support against gravity,2 as well as in water transport regulation, alongwith other important functions.3 Even if lignin has been the object of intense interest for chemists and biologists since over 150 years, its three dimensional structure, as well as the role it plays in plant cell metabolism, is still not completely elucidated.4 It was by now established that lignin is a three-dimensional heteropolymer

biosynthesized by linking together some precursors hydroxycinnamyl alcohol, coniferryl alcohol and sinapyl alcohol,5,6 to give p–hydroxyphenyl, guaiacyl and syringyl lignin monomeric units (Schemes 1 and 2). Several inter-unit linkages, usually involving C – O or C – C bonds, are present, the arylglyceryl - β - aryl ether linkages (β O – 4 linkages, labelled “a” in Scheme 2) being the most common ones. Unlike cellulose, which has a unique well-defined structure, there is a large variety of lignin types,7 its heterogeneity being caused by variations in the composition, size of morphological entities, crosslinking and functional groups. The proportion of the different monomeric units varies among the

Cellulose Chem. Technol., 40 (8), 597-622 (2006)

Carmen –Mihaela Popescu et al. different types of lignin, 6,8,9 alongwith differences existing in molecular architecture, due to the types of linkages between the phenyl propane structural units: p-hydroxyphenyl (H), guaiacyl (G) and syringyl (S) units.10 The major chemical functional groups in lignin include hydroxyl, methoxyl, carbonyl and carboxyl groups, in various amounts and proportions, as depending on their genetic origin and extraction processes11 applied. Lignins composition is different not only among plants of different genetic origin, but also among different tissues of the same plant.

Scheme 1: Monomeric units in lignin: (a) trans-pcoumaryl alcohol, (b) trans-coniferyl alcohol (guaiacyl unit) and (c) trans-sinapyl alcohol (syringyl unit)

Scheme 2: Fragment of lignin structure and inset of a phenylpropanoid subunit (in this case, coniferyl alcohol). Different types of linkages are shown: a, β – O – 4; b, conjugated phenyl in position 5 – 5; c, γ – O – α. Figure is adapted from Hammel6

598

Lignin characterization As known, in softwood lignin, the structural elements (over 95%) are predominantly derived from coniferyl alcohol (2 – methoxy – 4 – propyl – phenol – guaiacol, especially). In hardwoods and dicotyl crops like flax and hemp lignins, various ratios of coniferyl / sinapyl have been determined (a mixture of 2 – methoxy – 4 – propyl – phenol – guaiacol and 1, 5 – dimethoxy – 4 – propyl – phenol – syringol), whereas in the lignin isolated from cereal straws and grasses, the presence of coumaryl alcohol is typical.10 In the plant cell walls, lignin is intertwined and covalently bounded with macromolecules of cellulose and hemicelluloses, thus forming supramolecular structures, such as micro fibrils and lamellae as constituents of the cell plant wall membranes. The linkage types (more than 10) include ether bonds between aryl carbons in the lignin and the carbohydrate, ester bonds between aryl carbons, and also uronic acid residues and lignin-glycosidic bonds. The resulting lignin-hemicelluloses matrix encrusts and protects the cellulose component of the cell wall from microbial attacks.9 The macromolecular structure and supramacromolecular assembly of lignin into a cell wall structure is still poorly understood,12,13 most of the information on lignin structures coming from spectroscopic and microscopic studies.14 In recent years, extensive research efforts have been taken towards the elucidation of lignin structure, characterisation its chemical reactivity and functional properties and development of new applications. Thus, it has been shown that lignin represents a versatile molecule possessing multiple properties, providing numerous possibilities for its modification.15 The lignin from woods and annual fibre crops with different chemical composition and properties can be obtained by several extraction methods. Commercial chemical pulping processes (sulphite and Kraft process) produce lignosulfonates and Kraft lignins as by-products. Recently, the commercialized alkaline pulping – precipitation process has supplied sulphurfree, free-flowing lignins. Other delignification technologies use an organic

solvent or a steam high-pressure treatment. However, it is practically impossible to quantitatively isolate pure lignin from the cell walls in an intact native state. The lignins isolated by the known methods (physical, chemical or enzymatic treatments) represent a mixture of degraded or solubilised lignin from various non-identified morphological regions of the raw material.15 To enhance the industrial use of lignins, there is need for a continuous supply of lignin products with constant quality related to purity, chemical composition and functional properties. These requirements need a very detailed characterisation, for their possible subsequent standardization. In such cases, the spectroscopical methods appeared as very useful tools in the investigation of plant cell wall polymers, such as polysaccharides and lignins. Boeriu et al.15 studied a set of samples containing various non-wooden, hardwood and softwood lignins isolated by different processing technologies. They stated that FTIR spectra represent an important tool for a quick qualitative and quantitative characterisation of the chemical structure and functional properties of lignins. Principal Component Analysis (PCA) allows classification of lignin materials with respect to botanical origin, pulp processing and modification treatments. The partial least squares (PLS) models developed allow an accurate determination of the concentration of lignin polymers and phenolic hydroxyl groups. The results presented in the abovementioned paper suggest that Fourier Transform Infrared spectroscopy (FT-IR), combined with chemometrics, can be used as a fast and reliable non-destructive technique for the characterisation and quality control of lignin-based materials.17 Tejado et al. assert that FT-IR spectroscopy reveals that Kraft lignins from softwood are mainly composed by guaiacyl units with nonetherified hydroxyl groups, while alkaline flax lignin has mainly syringyl units and less free hydroxyls.16 Gosselink et al.11 used five types of lignins for their structural characterisation, mentioning that a comparison between FT – IR spectroscopy and wet chemical methods can indicate that the acetylation reaction

599

Carmen –Mihaela Popescu et al. applied for the determination of hydroxyl groups under the experimental conditions is incomplete, resulting in an unreliable total hydroxyl determination. UV spectrophotometry is one of the most convenient and useful methods for quantitative and qualitative analyses of lignin in solution. Due to its aromatic nature, lignin strongly absorbs ultraviolet light, exhibiting characteristic maxima in the ultraviolet light region. The position and intensity of the absorption maxima depend on the type of lignin, on its chemical modifications, as well as on the solvent used for photometric measurements.17 UV spectra were used to effectively estimate the amounts of certain functional groups, especially the phenolic hydroxyls of lignin. Hence, this method was used to establish the carbonyl groups content through the reduction of various aldehydes and ketones in guaiacyl-propane structures with sodium borohydride, in an alkaline solution.18 Fluorescence spectroscopy of lignin has been investigated by several researchers.19–24 It is well-known, e.g., that the luminescence observable in lignin (in wood as such or in liquid solutions) at ambient temperature is due to fluorescence.22,24 Fluorescence spectroscopy can be used as a sensitive method for the photochemistry of wood fibres and paper, and also for the analysis of lignin constituents in wastewaters from pulp mills.23 Experiments on lignin model compounds showed that structural elements of cinnamyl alcohol type or phenyl coumarone type are conceivable candidates for “energy sink”. End groups of cinnamyl alcohol type are present in lignin, while phenyl coumarone structures have not been detected in untreated lignin; acid treatment of lignin results in the conversion of its structural phenylcoumaran type units into structures of phenyl coumarone type and stilbene structures, detected by fluorescence spectroscopy. Albinsson et al.23 studied spruce lignin samples, examination of the lignin samples and model compounds suggesting that the small amounts of phenylcoumarone structures in lignin act as a conceivable acceptor. Such structures are formed from

600

phenylcoumarane structural elements 23 existing in lignin. 13 C NMR spectroscopy has been shown to be a method with a significant potential in providing detailed structural information for lignin. In particular, the development of multidimensional NMR techniques has considerably extended the prospect of lignin structural analysis. In addition, 13C NMR is indispensable in the quantitative determination of the amounts of different structural units in lignin,23–34 while the broad proton NMR signals occurring over a narrow frequency range render limited quantitative information; 13C NMR spectroscopy provides an elegant alternative method, mainly due to its significantly larger chemical shift dispersion. Current practices in the use of quantitative 13C NMR spectroscopy for the study of lignin are confined to using the aromatic and methoxyl signals as internal standards in expressing the various functional groups per C9 or per methoxyl unit, respectively.33,34 Such a practice, although applicable to native lignin, may be a source of serious errors during analyses performed on technical lignins, once they contain degraded side chains and fragmented aromatic rings.35 Quantitative analysis was achieved by carefully applying the chosen internal standards that displayed clear, unoverlapped signals in the middle of the 13C NMR spectra. Consequently, the various environments containing the carbon skeleton of lignins have been reliably quantified.35 Despite of its random and highly heterogeneous nature, the structural details of complex lignin polymer can be revealed by 13 C NMR. That is why; this method was established as a useful tool for a detailed structural characterization, supported by an almost precise agreement between carbons’ chemical shifts in low-molecular-mass model compounds and in the polymer.36 In a previous paper,37 the correlation between the thermal characteristics of woody and annual fibre crops lignins and their content in functional groups, as determined by chemical methods,37 was discussed. The present review analyzes our results46-52 on the spectral characterization of the same groups of lignins, to bring new insights in lignin

Lignin characterization structural characterization, on emphasizing the differences between lignins of different origin (such as those from woody plants and fibre crops). In each analysed group some spectral characteristics are also evidenced, alongwith their relationships with other properties. EXPERIMENTAL Materials The lignin samples under study were classified into two groups: lignin from woody plants and lignin from annual fibre crops. The lignins from wheat straw, sisal, abaca, hemp, jute and flax were obtained from Granit SA

No.

Source

(Lausanne, Switzerland), while Alcell organosolv lignin from mixed hardwoods (maple, birch and poplar) was obtained from Repap Technologies Inc., (Valley Forge, PA, USA). Lignosulfonate from softwood (Borresperse 3A and WAFEX-P) and Kraft lignin from softwood (Curan 100) were provided by Lignotech (Sweden). Sulphur-free lignins from softwood were obtained from Kiram AB, (Sweden). The chemical characterization of lignins was made by Boeriu et al.15 by determining the lignin content, the phenolic and carboxyl groups and also the sugar content (Table 1), as well as by FT-IR spectroscopy, to assess their antioxidant activity.

Table 1 Characteristics of the lignins under study15 Lignin COOH Pulp process (%) (mmol/g)

PhenolicOH (mmol/g)

Total sugars

Lignins from woody plants 1

Mixed hardwood organosolv

2

Softwood sulphur free HpH

3

Softwood Sulphur free LpH

4

Softwood Curan 100

5

Softwood Kraft 1

6

Alcell lignin

7

Softwood Pine

8

Aspen wood lignin

9 Softwood lignosulfonate 10 Softwood lignosulfonate white Lignins from annual fibre crops 1 Straw 2 Hemp 3 Jute 4 Sisal 5 Abaca 6 Flax 7

Flax

8

Bagasse

Organosolv EtOH Sulphur-free (HpH) Sulphur-free (LpH) Sulphate (Kraft) Sulphate (Kraft) Sulphate (Kraft) Steam Explosion Sulphite (LS) Sulphite (LS) Sulphur–free Sulphur–free Sulphur–free Sulphur–free Sulphur–free Sulphur–free Sulphur–free oxidized Sulphur–free

One may observe that, in all samples, the lignin content is higher than 80%, with several exceptions, therefore the efficiency of separation procedures may be appreciated as good. The content of residual sugars is of 0.3-7.7%, except that of the lignin resulted by a sulphite procedure, where the mere determination of these

96.5

0.78

2.4

0.32

64.7

Nd

Nd

1.77

99.2

1.4

1.0

0.65

96.8

2.0

2.5

0.71

88.6

0.9

2.1

2.26

Ndl

0.8

2.2

Ndl

90.0

2.5

1.8

2.06

Ndl

Ndl

Ndl

Ndl

Ndl Ndl

3.5 1.2

1.1 1.1

1.3 24.5

Nd Nd Nd Nd Nd 87.8

1.7 2 1.8 1.2 1.15 1.8

2.6 1.9 2.4 2.3 2.7 1.6

Nd 2.4 Nd 7.7 5.5 1.7

92.1

1.6

1.5

1.6

Nd

Nd

Nd

Nd

components is difficult. Sample’s origin is one of the decisive factors determining both its structure and spectral characteristics. Boeriu et. al 15 evidenced some deviations for the softwood samples, mainly LS and HpH (65%), which have a low content of lignin and a higher one of residual sugars, the efficiency of the extraction

601

Carmen –Mihaela Popescu et al. procedure being thus unsatisfactory. The characteristics of the samples obtained by sulphite procedures are different from those of the samples obtained by sulphate procedures. On such a basis, a classification of the samples, with respect to their origin and separation methods, may be made (Table 1). Their spectroscopic behaviour has been discussed versus both their origin and chemical analysis. Methods FTIR spectra of the lignin samples were recorded on KBr pellets with a Bomem MB-104 FT-IR spectrophotometer with a spectral resolution of 4 cm-1. Samples’ concentration in the pellets was constant, i.e. 3 mg samples/ 500 mg KBr. Part of the samples was heated at a rate of 1 ºC min-1 up to 105 oC. The IR spectra have been recorded at every 3 min, during both heating and cooling. An IR spectrum of the cooled pellets was recorded, at room temperature, in the second day, to check the reversibility of some processes like the rebuilt of the H-bonds by cooling. Deconvolution of the overlapped bands was carried out with a Grams/32 program [Galactic Industry Corporation]. 13 C NMR spectra were recorded on a Bruker Avance device (400 MHz), in a DMSO solution, at a temperature of 50 oC and concentration of the sample in solution of 10% w/w. Tetramethylsilane was used as an internal standard. UV spectra of the lignin samples were recorded in a NaOH 2M solution, on a UV/Vis

Lambda 3 Perkin – Elmer Spectrophotometer. The concentration of the sample in solution was constant: 1.5 mg sample/5 mL NaOH 2 M solution (0.03%). The Fluorescence spectra of the lignin samples were recorded in a NaOH 2M solution, on a Fluorescence Spectrophotometer SLM 8000. The concentration of the sample in solution was constant: 1.5 mg sample/5 mL NaOH 2 M solution (0.03%). Recording of the emission spectra was made over a wavelength range of 420 – 650 nm, at an excitation radiation of 390 nm.

RESULTS AND DISCUSSION FT-IR spectroscopy. As lignin is a complex system, obviously, spectra overlapping of the bands of some characteristic absorptions of functional groups is frequent over its IR. The FT – IR spectra of the two groups of lignin, from woody plants and annual fibre crops, were plotted, bands’ assignment according to literature data38–43 being given in Table 2. The FT – IR spectra are particular for each kind of lignin, differing only by the intensities of the bands (Table 3); in the spectra of several samples some bands are missing, other are very intense or very weak in the same region, which provides a solid basis for a good characterization and interpretation of such experimental results.

Table 2 Assignment of the absorption bands in the FT-IR spectra of the lignin samples Wavenumber (cm-1) Bands assignment 3550 Weakly absorbed water 3410 – 3460 Hydroxyl groups in phenolic and aliphatic structures CH stretching in aromatic methoxyl groups and in aliphatic methyl 2938 – 2920 and methylene groups of side chains CH stretching in aromatic methoxyl groups and in methyl and 2840 – 2835 methylene groups of side chains C-O stretch in unconjugated ketones, carbonyls and in ester groups 1740-1720 (frequently of carbohydrate origin); 1715 - 1675 C=O stretching in conjugated p-substituted aryl ketones; 1650 - 1640 Protein impurity and water associated with lignin 1610 – 1595 C=C stetching of the aromatic ring (S), CH deformation 1515 – 1505 C=C stretching of the aromatic ring (G)CH deformation 1470 – 1460 C-H asymmetric deformation in CH2 and CH3 1460 – 1370 Asymmetric C-H bending from methoxyl groups 1430 – 1422 C-H asymmetric deformation in –OCH3 symmetric C-H bending from methoxyl group, O-H and C-O of 1370 – 1365 phenol and tertiary alcohol 1330 – 1320 C1-O vibrations in S derivatives Aromatic C-O stretching vibrations (C-O stretching are those of the 1300 – 1200 methoxyl and phenol groups)

602

Lignin characterization Guaiacyl ring breathing, C-O stretch in lignin, C-O linkage in guaiacyl aromatic methoxyl groups Syringyl ring breathing with C-O stretching C-O bonds C-H in plane deformation of G ring Aromatic C-H in plane deformation; typical for G units, whereby G condensed > G etherified Aromatic C-H in plane deformation (typical for S units) plus secondary alcohols plus C=O stretch C-O deformation in secondary alcohols and aliphatic ethers Calkyl – O ether vibrations methoxyl and – O – 4 in guaiacol C-O valence vibration =CH out of plane deformation (trans) Aromatic C-H out of plane deformations Ch out of plane vibrations in position 2, 5 and 6 of guaiacyl units C-H out of plane in position 2 and 6 of S units

1268 1235 - 1230 1180 1160 1140 1128 – 1115 1086 1047 – 1004 996 – 985 970 925 – 915 858 – 853 843-835

1220cm-1

1030cm-1

866cm-1

1275cm-1

1330cm-1

1425cm-1

1510cm-1

1658cm-1

1710cm-1

Spectral region Lignins from woody plants Mixed 1 hardwood w vw organosolv Softwood 2 w vvw sulphur free HpH Softwood 3 w vw Sulphur free LpH Softwood 4 w vvw Curan 100 Softwood 5 w vw Kraft 6 Alcell lignin w w Softwood 7 m w Pine Aspen wood 8 m w lignin Softwood w vvw 9 lignosulfonate Softwood 10 lignosulfonate s vw white Lignins from annual fibre crops 1 Straw w vw 2 Hemp w w 3 Jute w w 4 Sisal m w 5 Abaca m w 6 Flax m s 7 Flax ox. w vs 8 Bagasse w w

2850cm-1

2920cm-1

No

3415 cm-1

Table 3 FT-IR results for lignins of various origins from woody plants and annual fibre crops Samples

vvw

__

vvw

w

w

w

vw

w

vw

vvw

vvw

__

vw

vw

vw

vvw

vw

vw

vw

__

vvw

__

vvw

w

vvw

__

w

w

W

vvw

vvw

__

vw

w

vw

__

w

vw

vw

Vvw

vvw

__

vw

w

vw

__

w

w

w

Vvw

vw

__

__

m

w

w

w

m

w

Vw

vvw

vvw

w

s

w

vw

m

m

m

Vw

vw

__

__

m

w

w

w

m

w

__

vvw

vvw

vw

vw

vw

__

vvw

w

w

__

vvw

vvw

w

w

w

w

w

w

w

Vvw

vvw vw w w vw w m vw

W W W W W M W W

vvw vw vvw w vw vw vw vw

w w m m s m m m

vw w w w w w w w

vw vw w m m w w w

vw w w w w m m w

w w m m s m m m

w vw w w w w w w

__ Vvw Vvw Vvw W Vw Vvw Vvw

603

Carmen –Mihaela Popescu et al. vs - very strong, s - strong, m - medium, w - weak, vw - very weak,

Lignins spectra evidence two important regions, such as: the bands in the 3700 – 2750 cm-1 region, assigned to the –OH groups of the physically absorbed water or bonded in aliphatic and aromatic groups, as well as to the methyl groups, and the bands in the 1800 – 900 cm-1 region (“fingerprint region”), assigned mainly to guaiacyl and syringyl structural units, but also to other functional groups.

(A)

These regions are enlarged (Figs. 1 and 8), to clearly evidence the differences between samples, and analysed in detail. The assignments of the IR bands are presented in Table 2. In the 3700 – 2700 cm-1 region, the absorption band centred at about 3415 cm-1, assigned to the valence vibration of the –OH groups has approximately the same shape as for the lignins obtained from annual fibre crops, while the peak position varies within close limits - between 3412 and 3417 cm-1 (Fig. 1B). For woody plants lignin, this band takes various shapes, with two maxima, as for Pine wood, or one shoulder, as for Softwoods lignin, its intensity being different for each sample; the highest absorbance values are recorded for Softwood lignosulfonate (white) and Aspen wood (Fig. 1A) lignin. For woody plants lignin, the peak position is shifted towards higher wavenumbers, comparatively with that of annual fibre crops, which is an indication of the stronger interactions between the absorbed water and the lignin molecule, or between the functional groups. By hydrogen bonding, the energy of the hydrogen bonds was evaluated with the following equation:44

EH =

(B) Figure 1: The 3750 – 2700 cm-1 region of the FT – IR spectra for different lignin types: woody plants (A) and annual fibre crops (B) lignins (the numbers correspond to those in Table 1)

604

1  (ν 0 − ν )   λ  ν0 

(1)

where ν0 is the standard frequency corresponding to the free –OH groups (3650 cm-1) and ν is the frequency of the bonded – OH (3550 cm-1 assigned to the –OH groups in weakly absorbed water, 3415 cm-1 assigned to the –OH groups in phenolic/aliphatic structures) groups, while λ is a constant equal to 6.72*10-2 kJ-1. Generally, the energies of the hydrogen bonds are higher for the lignins from annual fibre crops than those of the samples from woody plants, however all of them are distributed around the same straight line, when plotted versus the content of functional groups (Figs. 2), as phenolic -OH (Fig. 2A) and carboxyl groups, respectively (Fig. 2B).

Lignin characterization

No. 1 2 3 4 5 6 7 8 9 10

Lignins from woody plants

Table 4 Energy of the hydrogen bonds Energy of the hydrogen Lignins from No. bonds annual fibre crops 3550 cm-1 3415 cm-1

Mixed hardwood organosolv Softwood sulphur free HpH Softwood Sulphur free LpH Softwood Curan 100 Softwood Kraft 1 Alcell lignin Softwood Pine Aspen wood lignin Softwood lignosulfonate Softwood lignosulfonate white

2.035

3.867

1

Straw

1.696

3.850

1.831

3.799

2

Hemp

1.713

3.935

1.543

3.833

3

Jute

2.001

3.850

1.882

3.816

4

Sisal

1.916

3.901

1.628

3.596

5

Abaca

1.645

3.545

1.713

3.816

6

Flax

2.205

3.629

2.086

4.104

7

Flax ox.

2.103

3.870

1.730

4.087

8

Bagasse

1.832

3.922

1.357

3.765

0.933

3.969

The energy of the hydrogen bonds increases with the phenolic – OH content and decreases with the – COOH one, respectively. For a good assessment of the variations between the bands corresponding to the different functional groups of the different lignins, the normalised intensity for the main bands of the spectra of both types of lignins was calculated with the formulae:39

I NIλ = λ Cm

Energy of the hydrogen bonds 3550 cm-1 3415 cm-1

(2)

(A)

where: NIλ → normalised intensity at λ cm-1; Iλ → integral absorption at λ cm-1; Cm → sample concentration and

Cm =

pm *100 p KBr

where: pm → sample mass pKBr → KBr mass

(3) (B) Figure 2: Variation in the energy of the hydrogen bonds with the content of functional groups: phenolic -OH (A) and carboxyl (B)

605

Carmen –Mihaela Popescu et al. Variation of the normalised intensity of the 3415 cm-1 band is plotted in Figure 3.

(A)

Figure 3: Variation of the normalised intensities of the bands at 3415 cm-1 of the FT-IR spectra of lignin samples (the numbers correspond to those in Table 1)

One can observe that the strongest absorption bands of the hydrogen bonds in woody plant lignins are present in Lignosulfonate (white) (0.64), Aspen (0.53), and also in Pine (0.39), while in the case of the annual fibre crops lignins, the strongest bands (at 3415 cm-1) are present in Abaca (0.51) and Sisal (0.49) samples. To evidence the presence of the physically-bonded water, the lignins were conditioned. The samples were kept for 1 hour in a conditioning atmosphere in which the specimens were exposed to circulating air at 50.0% ±2.0% relative humidity (RH) and a temperature of 23.0 ±1.0 oC. Further on, the samples were oven-dried for 2 hours at 105 ± 3 oC, cooled in a desiccator and weighed, the operation being repeated until reaching a constant mass. The spectra were recorded both before and after these operations and the differences between them were determined, for evidencing the influence of the absorbed water. The spectra recorded before and after conditioning/drying are plotted in Figure 4, which also shows the differences recorded between them for two representative lignins, namely: aspen wood lignin (Figs. 4A and 4B) and abaca lignin (Figs. 4C and 4D). For woody plant lignins, the differences in absorbance are higher than those observed for annual fibre crops samples, meaning that water desorption is faster in the first case.

606

(B)

(C)

(D) Figure 4: The 3750 – 2700 cm-1 region of the FTIR spectra: A – aspen lignin spectra and C – abaca lignin spectra, before and after conditioning/drying; B – differences in the aspen lignin spectra and D – differences in the abaca lignin spectra

Lignin characterization However, after water desorption, this band remains important, as due to the – OH groups of the lignin macromolecules which participate to intramolecular hydrogen bonding.

and 2850 cm-1 are much stronger in lignins from annual fibre crops than in those from woody plants (Figs. 6A and 6B).

(A) (A)

(B)

(B) Figure 5: FT-IR spectra of the lignin from hardwood (a) and from straw (b) heated up to 105 o C. Each 3 minutes, a spectrum was recorded, in figures plotting the spectrum recorded each 15 minutes

Following water desorption by recording spectra at various times of heating, some differences between two groups of lignins may be also observed (Fig. 5). The variations between spectra are higher for the first group (Fig. 5A – woody plant lignins) versus the second one (Fig. 5B – annual fibre crops lignins), which means that the energetics of the processes are different in these groups, depending on the content of functional groups in the lignin structure Other differences between these two groups of lignins can be observed in the 3000-2750 cm-1 region. The bands from 2920

Figure 6: Variation of the normalised intensities of the bands at 2920 cm-1 (a) and 2850 cm-1 (b) of the FT-IR spectra of the lignin samples (the numbers correspond to those in Table 1)

The strongest absorbance values of the bands from 2920 and 2850 cm-1 belong to lignins obtained from flax sulphur free and flax oxidized sulphur free. According to bands’ assignments (Table 2), this might mean that the lignins obtained from annual fibre crops have a higher number of –CH2 or –CH3 groups participating to asymmetric or symmetric valence vibrations, comparatively with the lignins from woody plants. A higher number of methyl and methylene groups can be correlated with a large number of syringyl units, which also agrees with the intensity of the band at 1460 cm-1, assigned to C – H deformations, in syringyl derivatives. This band was observed to be strongest in lignins from annual fibre crops and hardwoods, versus those from softwoods samples.

607

Carmen –Mihaela Popescu et al.

Figure 7: Ratio of aliphatic-to-aromatic signals (I2930/I 1510) of the intensities of the FT-IR spectra of the lignin samples (the numbers correspond to those in Table 1)

This indicates that, after their extraction, the lignins from woody plants, such as: softwood sulphur free LpH, aspen wood and lignosulfonate, have more aliphatic CHx groups with respect to the aromatic CH groups. In annual fibre crops lignins, the aliphatic-to-aromatic CHx groups ratio is of approximatively 1:1 (the aliphatic content of CHx is a little bit higher than of the aromatic ones). The 1800 – 900 cm-1 “fingerprint region” (Figs. 8A and 8B) is very complex, containing numerous bands, peaked, for example, at the following wavenumbers: 1710 cm-1; 1680 cm-1, 1650 cm-1, 1595 cm-1, 1510 cm-1, 1460 cm-1; 1425 cm-1; 1370 cm-1, 1325 cm-1, 1275 cm-1, 1220 cm-1, 1086 cm-1 and 1030 cm-1 (for bands’ assignment, see Table 2), many of them evidencing shoulders and peculiarities for each lignin under study.

(A)

(B) Figure 8: FT – IR spectra for different types of lignins from the 1850 – 900 cm-1 region: woody plants lignins (A) and fibre crops lignins (B) (the numbers correspond to those in Table 1)

In the carbonyl / carboxyl region, one may observe the band from 1710 cm-1 assigned to the valence vibration for these groups. This band is very weak (possibly evidenced only by a second order derivative and by spectra deconvolution) in most of the woody plants lignins, such as: softwood sulphur free (LpH), curan wood, softwood Kraft, pine wood, aspen wood lignins and softwood lignosulfonate, being much more visible in hardwood, softwood sulphur free

608

Lignin characterization LpH, Alcell lignin and aspen wood lignins. In the annual fibre crops lignins, this band is better evidenced in this region. The second order derivative of the IR spectra can obviously enhance the apparent resolution and amplify the tiny differences in the IR spectrum. The second order derivative of the FT-IR spectra were obtained with the Savitsky-Golay method (second order polynomial with twenty five data points, using Grams 32 (Galactic Industries Corp.). The second order derivative spectra of lignins of different sources for the 1800-900 cm-1 spectral region are plotted in Figures 9A and 9B.

(A)

(B) Figure 9: Secondary derivatives of the FT – IR spectra for different types of lignins from the 1850 – 900 cm-1 region: woody plants lignins (A) and fibre crops lignins (B) (the numbers correspond to those in Table 1)

As seen, bands’ intensities vary among the samples from the two series and also within the same series. In the second order derivative spectra of lignins from woody plants, there are some differences, such as: the intensity of the band at 1598 cm-1 (assigned to C=C stretching of the aromatic ring –syringyl and CH deformation) is very weak in softwood HpH, and in both softwood lignosulfonates samples. Also, the intensity of the band at 1330 cm-1 (C1-O vibrations in syringyl derivatives) in the lignins from softwood HpH, softwood LpH, curan 100, softwood Kraft, pine wood and softwood samples is very weak, hardly observable. Instead, the intensity of the band at 1115 cm-1, assigned to the aromatic C-H in-plane deformation (typical for syringyl units), and also to the secondary alcohols or C=O stretch of the hardwood, aspen wood and alcell lignin, is much stronger than the intensity of the same band of other lignins from the woody plants. In the second-order derivative spectra of the lignins from annual fibre crops differences appear in the intensity of the band at 1650 cm-1 (assigned to protein impurity and water associated), seen as very weak in hemp lignin, appearing only as a shoulder in jute and abaca lignins and being the strongest in straw lignin. The intensities of the bands at 1330 cm-1 (C1-O vibrations in syringyl derivatives), 1220 cm-1 (syringyl ring breathing with C-O stretching) and 1115 cm-1 (aromatic C-H in plane deformation of syringyl units, secondary alcohols and C=O stretch) are stronger in abaca, bagasse and jute lignins; in straw and other lignins, these bands are overlapped. In annual fibre crops lignins, the band from 1650 cm-1 is easily observable, whereas in the lignins from woody plants this band can be seen only as a shoulder or it is even missing. Also, the second order derivative spectra of hardwood, alcell and aspen wood lignins are similar to those from annual fibre crops. The variations in intensity may be explained by the fact that the bands are, of course, closely related to the compositional differences between lignins. One may observe that the bands of the conjugated

609

Carmen –Mihaela Popescu et al. ketones are not present in softwood lignins, those of the aromatic structures are very big in hardwoods, while annual fibre crops lignins from abaca, bagasse and jute show higher intensities of the bands corresponding to the syringyl units. For better differentiating among the bands, deconvolution of the “fingerprint” region was carried out. Figure 10 plots the deconvolutions of the 1800 – 1550 cm-1 region for four types of lignins.

(D) Figure 10: Deconvolution in the 1850-1550 cm-1 region of the FT-IR spectra for: Softwood lignosulfonate(white) (a), Aspen wood steam explosion (b), Sisal sulfur free (c) and Straw sulfur free (d)

(A)

After deconvolution of this region, one can observe that the band at 1710 cm-1, assigned to C=O stretch in carboxyl and conjugated ketones groups, is strongest in lignins from annual fibre crops while, in the other group, this band is very weak (0.01 – 0.03 a. u.) (Fig. 11). A band peaked around 1740 cm-1 also appears, yet with very small intensity values in all spectra. This band could be assigned to the C=O stretch in unconjugated ketones, and therefore both type of ketone groups are present in lignins’ structure. Variation of the normalised intensities of the 1710 cm-1 band for the studied lignins is plotted in Figure 11.

(B)

(C) 610

Figure 11: Variation of the normalised intensities of the bands at 1710 cm-1 in the FT-IR spectra of the lignin samples (the numbers correspond to those in Table 1)

Lignin characterization This variation shows that carboxyl groups are present in high amounts in lignins from annual fibre crops, while in those from woody plants their number is more reduced. Exceptions are registered in the case of lignins from hardwoods (hardwood mixed, aspen wood and alcell lignin), in which the intensity of the band is comparable with the intensities recorded in annual fibre crops lignins.

groups at 1460 cm-1 are evidenced in both lignins series under study (Figs. 13A and 13B); here, it is only the absorbance intensities of the bands that differ. However, the most intense bands from 1595 and 1460 cm-1 are present in lignins from annual fibre crops (between 0.25 and 0.31 a. u.) while, in the case of woody plants samples, most of the lignins are characterized by less intense absorbance values of these bands (between 0.10 and 0.20 a. u.). Exceptions may be observed for the hardwood, aspen wood and alcell wood lignins, and the values of which being comparable with the band intensities of the lignins separated from annual fibre crops.

Figure 12: Ratio of carbonyl-to-aromatic signals (I1710/I1510) of the intensities in the FT-IR spectra vs. the content of the (chemically determined) COOH groups of the lignin samples

The ratio of carbonyl to aromatic signals (I1710/I1510) increases in the lignin series with increasing the –COOH groups content (chemically determined) (Fig. 12). The data points agree well, since they are situated between the upper and lower confidence bands of linear fitting. The band at 1650 cm-1, assigned to the – OH of absorbed water in lignins can be observed in all spectra (not shown). The intensities of this band are almost two times higher in lignins from annual fibre crops than in the samples from woody plants, agreeing, with only minor exceptions, with the intensities of the bands from around 3530 cm-1 , assigned, too, to the –OH vibrations of weakly bounded water; the intensity values of this band are almost two times higher in lignins from hardwoods and annual fibre crops than those from softwoods, thus agreeing with results on the thermal behaviour of the same lignins.37 The vibrations of the aromatic ring of the syringyl groups at 1595 cm-1 and the asymmetric C-H bending of the methoxyl

(A)

(B) Figure 13: Variation of the normalised intensities of the bands at 1598 cm-1 (a), 1460 cm-1 (b) in the FT-IR spectra of the lignin samples (the numbers correspond to those in Table 1)

The bands assigned to the valence vibrations of the aromatic ring in guaiacyl groups at 1510 cm-1 and C-H asymmetric deformation in –OCH3 at 1425 cm-1 are the most representative ones for lignins,39,45–52

611

Carmen –Mihaela Popescu et al. their intensities being different in the lignins from both woody plants and annual fibre crops (Figs. 14 A and 14B).

The normalised intensity values evidence that the bands assigned to the valence vibrations of the aromatic ring in syringyl groups (Fig. 13) show higher absorbance intensities (between 0.15 and 0.30 a. u.) than those assigned to the valence vibration of the aromatic ring in the guaiacyl groups (Fig. 14) (between 0.04 and 0.12 a. u.), which means that the lignins contain a high amount of syringyl units in their structure.

(Fig. 13). The intensities of the band at 1460 cm-1 (Fig. 13B) are more intense (from 0.10 to 0.32 a. u.) than those of the one at 1425 cm-1 (from 0.04 to 0.14 a. u.) (Fig. 14B), being stronger in hardwoods and annual fibre crops lignins, as compared to softwoods samples. The spectral region beyond 1400 cm-1 is more difficult to analyse, as most of the bands are complex, with more than one vibration type contribution; that is why, deconvolution of this region was necessary (Fig. 15).

(A) (A)

(B) Figure 14: Variation of the normalised intensities of the bands at 1510 cm-1 (a) and 1425 cm-1 (b) in the FT-IR spectra of the lignin samples (the numbers correspond to those in Table 1)

Generally, the absorption intensities of the syringyl units of hardwoods lignins are approximately equal with those of the annual fibre crops, but they are higher than those of the softwoods samples 612

(B) Figure 15: Deconvolution in the 1400-900 cm-1 region of the FT-IR spectra for: Hardwood mixed organosolv (a) and Sisal sulfur free lignins (b)

Following deconvolution, the bands are better evidenced at 1370 cm-1, being assigned to –OH and C-O of phenols and tertiary alcohol and aliphatic C-H stretch – visible as a shoulder; at 1270 cm-1 being assigned to guaiacyl ring breathing, C-O stretch in lignin, C-O linkage in guaiacyl aromatic methoxyl groups – visible as a shoulder; at

Lignin characterization 1180 cm-1 C-O bonds – almost invisible in normal FT – IR spectra; at 1150 cm-1 assigned to aromatic C-H in plane deformation; typical for guaiacyl units, whereby guaiacyl condensed > guaiacyl etherified – visible as a shoulder; at 1086 cm1 assigned to C-O deformation in secondary alcohols and aliphatic ethers – visible as a very weak shoulder. As expected, the bands corresponding to phenolic – OH groups at ~ 1370 cm-1 are present in all spectra, yet their position is shifted to 1365 cm-1 for the lignins from annual fibre crops, which indicates the presence of some interactions between the functional groups and the absorbed water different for each group of lignins from woody plants and, respectively, annual fibre crops.

(A)

intense than those registered for softwood samples (between 0.01 and 0.04 a. u.) (Fig. 16 A). The band at 1330 cm-1, assigned to the C1 – O vibrations in syringyl derivatives, shows a strong intensity in hardwoods lignins (such as: hardwoods mixed, alcell lignin and aspen wood) and in lignins from annual fibre crops while, in softwoods samples, this band appears as very weak, after deconvolution (in softwood HpH, curan 100, softwood Kraft, softwood, pine and lignosulfonate). This is due to the fact that the lignins extracted from annual fibre crops present a higher amount of syringyl units, which agrees with the data recorded for the bands evidencing maxima at 1598 and 1460 cm-1, also assigned to syringyl groups vibrations. The ratio representing the content of phenolic –OH groups (I1370 vs. I1510) correlates well with the content of chemically-determined phenolic –OH groups (Fig. 17A), since all data points are situated between the upper and lower confidence bands of linear fitting. For woody plants and annual fibre crops lignins, the intensity of 1370 cm-1 in all spectra also present an increasing order with the increasing content of the (chemically determined) – OH phenolic groups (Fig. 17B). In this case, all data points are enclosed between confidence bands of linear fitting for both lignin groups. The differentiation between the two types of lignin can be made by normalised intensity of 1370 cm-1 bonds plotted versus the phenolic – OH content.

(B) Figure 16: Variation in the normalised intensities of the bands at 1370 cm-1 (A) and 1330 cm-1 (B) of the FT-IR spectra of lignin samples (the numbers correspond to those in Table 1)

The intensities of this band, for hardwoods (hardwood mixed, aspen wood and alcell wood) and annual fibre crops lignins (between 0.06 and 0.1 a. u.), are more

(A)

613

Carmen –Mihaela Popescu et al.

(B) Figure 17: Content of phenolic OH groups (I1370/I1510) (A) and the band at 1370 cm-1 intensity of the FT-IR spectra vs. the content of (chemically determined) phenolic –OH groups (B) of the lignin samples

All spectra evidence vibrations characteristics for guaiacyl units, others than those discussed before, such as: 1270 cm-1, 1150 cm-1 and 855 cm-1; and vibrations characteristic for the syringyl units at 1220, 1120 and 830 cm-1. The absorbance intensity varies among the samples. The bands at 1270 cm-1, assigned to guaiacyl ring breathing, C-O stretch in lignin, C-O linkage in guaiacyl aromatic methoxyl groups and 1220 cm-1 assigned to syringyl ring breathing with C-O stretching, are much stronger or, generally, are characterized by the same intensity in lignins from woody plants (from 0.10 to 0.3 a. u.), than in those from annual fibre crops (from 0.15 to 0.20 a. u. and from 0.15 to 0.40). The bands at 1270 cm-1 and 1220 cm-1 in the FT-IR spectra of annual fibre crops lignins vary within large limits, which can be explained by the fact that guaiacyl and syringyl unit’s content is different for each lignin type, and in each annual fibre crops, respectively. For example, the intensity of the band at 1270 cm-1 for straw lignin is of 0.13 a.u. while, for the same series, the flax lignin is of 0.47 a.u. The intensities of this band vary within wide limits from 0.04 a.u. in softwood lignosulfonate to 0.54 a.u. in pine wood lignin. With some exceptions, the intensities of the band at 1220 cm-1 are modified, contrary to the other band from 1270 cm-1.

614

Figure 18: Ratio of syringyl-to-guaicyl signals (I1330/I1270) of the intensities of the FT-IR spectra of the lignin samples (the numbers correspond to those in Table 1)

Softwoods lignins have a low syringyl content (some softwood lignins may contain up to 95 % guaiacyl units). Therefore, S/G (I1330/I1270) signal ratio shows very low values in comparison with hardwoods and annual fibre crops lignins (in which the content of syringyl and guaiacyl units are present within different limits) (Fig. 18). In the later case, the ratio values of S/G are higher than in the former.

Figure 19: Content of the C–O groups (I1086/I1510) from the intensities of the FT-IR spectra of the lignin samples

The ratio of signals (I1086/I1510) representing the ratio of ether contents base on aromatic structures increases with increasing the content of the chemically determined – COOH groups, a good

Lignin characterization agreement being observed, since almost all data points are situated between the upper and lower confidence bands of linear fitting (confidence value – 98 %). After conditioning and drying or heating of lignin samples in the “fingerprint” region, no differences were observed in the characteristic absorption bands between the initial spectrum and the spectra recorded after these operations. Therefore, one may say that, after conditioning and heating, the structure of lignins remains unchanged.

13

C NMR spectroscopy. The 13C NMR spectra for hardwood, softwoods (curan 100 and lignosulfonate) lignins and for sisal and bagasse lignins are plotted in Figure 20. In the 13C NMR spectra of the studied lignins, several chemical shift ranges can be observed.53-54

Figure 20: 13C NMR spectra of some of the studied lignins

The 150 – 100 ppm range, assigned to the aromatic carbons, can be divided into tertiary carbons range (125 – 110 ppm), namely: 122 – 117 ppm carbon 6 in guaiacyl; 117 – 114 ppm carbons 3 and 5 in coumaryl – OH and coumaryl – O –; 114 – 108 ppm carbon 2 in guaiacyl; 109 – 101 ppm carbons 2 and 6 in syringyl, and quaternary carbons range (160-125 ppm), namely: 153-151 ppm carbons 3 and 5 in syringyl – O – or syringyl – OMe; 150 – 145 carbons 3 and 5 in syringyl – OH, carbon 3 in guaiacyl – O –, 3 guaiacyl – OH or 3 guaiacyl – OMe, carbon 4 in guaiacyl – OMe or 4 guaiacyl – O. The signals corresponding to the aliphatic carbons and corresponding to the side chain carbons (90 – 50 ppm) and methoxyl carbons (- OCH3) (60 – 50 ppm) appear in the same range. Based on these assignments, the composition of the studied lignins was estimated (Table 5). Comparing the shape of the 13C NMR spectra (Fig. 20) and estimating the contents of syringyl, guaiacyl, coumaryl and side chain carbons (Table 5), differences

can be found between mixed hardwood and softwoods curan, as well as between sisal and bagasse lignins. Thus, bagasse and sisal lignins contain more side chain carbons and syringyl units, and less coumaryl and guaiacyl units than the mixed hardwood and softwood Curan 100 samples. As expected, the content of syringyl units in softwood is lower than in hardwood and annual fibre crops, while the content of guaiacyl units is higher in softwood lignin than in other cases. The content of coumaryl is the lowest in annual fibre crops lignins and highest in softwood lignin, while the side chain atoms are in lowest amounts in hardwood lignin and in highest amounts, respectively, in annual fibre crops lignins. The content of methoxyl groups is almost the same in all lignins, while the content of other types of carbons is lowest in Curan lignin and highest in bagasse lignin. The composition of lingnosulfonate differs to a high extent from that of the other studied lignins.

615

Carmen –Mihaela Popescu et al.

Samples

Aromatics and aliphatics α=β (a) 60

Table 5 Composition of the studied lignins Syringyl Guaiacyl Coumaryl (b) (b) (b, c)

Α+β+γ (side chain atoms) (a) 6.3

OMe (a)

Other types of carbons 15.7

Mixed 3.5 4.3 2.2 18 hardwood organosolv Softwood 62.4 2.0 4.0 4.4 8.9 16 12.7 Curan 100 Softwood 39 2.3 4 0.2 21.9 25.2 13.9 lignosulfonate Sisal 54 5.7 2.7 0.6 14.8 16 15.2 Bagasse 55.8 6.0 2.9 0.4 13.2 14 17 (a) percents of the total number of carbon atoms (b) percents of the number of aromatic units (c) A/6 – S – G where A, S and G stands for aromatics + aliphatics (α=β ), syringyl and guaiacyl groups respectively.

UV spectroscopy. Due to their aromatic nature, lignins show strong absorption bands in the UV spectra. Figures 21A and 21B show the UV spectra of the lignins for the two studied groups: woody plants and annual fibre crops. The basic UV spectra of typical lignins exhibit the following absorption maxima: 220 nm, 240 nm, 298 nm, 330 nm and 360 nm. The absorption bands of all studied lignins are broad and indistinct, because of the interactions between electronic transition and vibrational and rotational ones, and to the overlapping of the electronic transition bands originating from the absorbing units in the polymeric lignin structure. The shapes of the lignins spectra are particular for every series of lignins. For annual fibre crops lignins, the bands are better evidenced than in woody plants samples, they being more intense and sharp. For a good assignment, the complex shape of the UV spectra of the lignin samples required their deconvolution. Following UV spectra deconvolution, the position of the maxima and the integral area of the bands are modified in the studied lignin samples (Table 6).

616

(A)

(B) Figure 21: UV spectra of woody plants (A) and annual fibre crops (B) lignins (the numbers correspond to those in Table 1)

Lignin characterization Representative deconvolutions for the two groups of lignins are presented in Figure 22.

(C)

(A)

(D) Figure 22: Deconvolutions in the UV spectra for: jute lignin (A), straw lignin (B), curan 100 lignin (C) and softwood Kraft lignin (D) (B) Table 6 Maxima of the UV absorption bands from different lignin samples No 1 2

Sample

Mixed hardwood organosolv Softwood sulphur free HpH 3 Softwood Sulphur free LpH 4 Softwood Curan 100 5 Softwood Kraft 1 6 Alcell lignin 7 Softwood Pine 8 Aspen wood lignin 9 Softwood lignosulfonate 10 Softwood lignosulfonate white Lignins from annual fibre crops 1 Straw 2 Hemp 3 Jute 4 Sisal 5 Abaca 6 Flax 8 Bagasse

First peak Second peak maximum maximum Lignins from woody plants 219.6 243.1 220.9 239.5 221.2 236.0 219.2 237.0 220.7 237.8 219.3 240.5 219.5 237.4 220.0 243.4 226.4 246.7 223.4 238.2 218.3 218.8 219.3 219.4 218.9 218.6 216.6

240.5 242.4 245.8 247.1 241.9 243.4 245.8

Third peak maximum

Forth peak maximum

Fifth peak maximum

288.2 297.5 298.3 296.2 298.5 289.2 296.9 287.6 299.3 295.8

331.8 333.0 329.9 331.4 330.4 330.1 329.6 337.3 331.2 330.9

364.9 357.1 362.7 353.5 358.4 366.2 355.2 365.4 344.8 329.6

289.6 292.4 291.2 292.7 290.6 290.7 288.4

330.1 329.9 330.8 328.4 331.3 328.3 326.2

359.3 353.8 358.2 360.7 361.9 351.5 356.0

617

Carmen –Mihaela Popescu et al. Based on the interpretation of lignin’s UV spectra, a 220 nm wavelength is assigned to the methoxylated phenol ring. It was also shown that free and etherified hydroxyl groups contribute significantly to the characteristic absorption maximum around 298 nm. Comparatively with the simple model compounds, the maxima of the biphenyl derivatives present the absorption band around 240 nm, while the aromatic carboxylic acids and the α- carbonyl groups are characterized by absorption at 320 nm. The bands with maxima at 365 nm were assigned to carbonyl groups or double-bondconjugated phenols. Dissociated aromatic carboxylic acids are also absorbed at 365 nm.55-60 Chemical pulping processes because significant changes in the structure of lignins, as reflected in their light absorption properties. The woody plants lignins differ not only in their origin but also in the extraction procedure and, consequently, the maxima of the UV bands vary within larger limits than in the case of annual fibre crops. The maximum at around 298 nm, originating from the non–conjugated phenolic groups (aromatic ring) in lignins is characteristic only to guaiacyl lignin, while the absorbance at lower wavelength, around 288 nm, is characteristic to guaiacyl – syringyl lignin. Consequently, shifting of the band from 298 nm to lower wavelength (to 288 nm) in annual fibre crops lignins (which present both guaiacyl and syringyl units in various ratios) can be assigned to the higher number of syringyl units present in this group of lignins. The main difference among softwood, hardwood and annual fibre crops lignins is the shift of the 298 nm maximum to shorter wavenumbers (289–290 nm), explained by a higher content of the syringyl groups. The differences in the 360 nm region probably indicate variations in the number of carbonyl groups created during lignins’ extraction from the raw material. The intensities of the bands’ maxima are modified in the following order: for 220 nm, the intensities of the band vary within close limits between these two groups of lignins; at 240 nm, the intensities are stronger in woody

618

plant lignins than in the annual fibre crops samples; at 298 nm, the same dependence as in the last case may be observed.

Figure 23: Variations in the absorbance values of the 298 nm band from UV spectra with (chemically-determined) –OH group, for woody plant and annual fibre crops lignins

Figure 23 plots the intensity of the maximum of the 298 nm band versus the – OH groups’ content (chemically determined) of all lignins. All data are situated inside of confidence interval of linear fitting. Fluorescence spectroscopy. Fluorescent emission in lignin has been attributed to aromatic structures such as conjugated carbonyl, biphenyl, phenylcoumarone and stilbene groups. The fluorescence excitation and emission spectra for the two series of lignins samples studied in the NaOH solution 2M are presented in Figure 24.

(A)

Lignin characterization

(B)

(C)

The shapes of the emission spectra (Figs. 24C and 24D) are almost the same, only bands intensities being different. The spectra of the woody plants lignins present higher emission intensities (almost twice) than those observed in the spectra of the annual fibre crops samples. This is due to the presence, in woody plants lignin, of a small amount of carbonyl groups which, according to FT-IR spectra results, are essentially nonfluorescent. In contrast, the excitation spectra (Figs. 24A and 24B) show different shapes, as depending on the lignin sample. The most different is the spectrum of the lignosulfonate and lignosulfonate white, presenting more than one maxima (Fig. 24A). The other lignin spectra show a maximum at 395 nm and a shoulder at around 350 nm. Therefore, these fluorescence spectral characteristics can be useful in the identification of different lignins. The excitation wavelength dependence and the broad nature of the fluorescence spectra with shoulders at longer wavelength suggest the presence of many different fluorophoric species (different “active” species) present in the lignin polymer.61 The proximity between the potential fluorophores and the different functional groups present in the lignins should also induce variations in the behaviour of those fluorophores. Lignins exhibit fluorescence emission spectra maxima after excitation with a 390 nm wavelength radiation, in the 487.9–501 nm region, which may be explained by the non-radiative energy transfer from lignin flourophores, which are excited in this wavelength range, to an acceptor emitting fluorescent light at 390 nm. The maxima of the emission spectra differ between the lignin samples, larger variations occurring in woody plants lignin, which is the result of the different extraction methods applied and of its different origin.

(D) Figure 24: Excitation fluorescence spectra of woody (A) and annual fibre crops (B) lignins and emission fluorescence spectra of woody (C) and annual fibre crops (D) lignins (the numbers correspond to those in Table 1)

619

Carmen –Mihaela Popescu et al.

Table 7 Fluorescence spectroscopic data for lignin samples

No

1 2 3 4 5 6 7 8 9 10

Sample

Mixed hardwood organosolv Softwood sulphur free HpH Softwood Sulphur free LpH Softwood Curan 100 Softwood Kraft 1 Alcell lignin Softwood Pine Aspen wood lignin Softwood lignosulfonate Softwood lignosulfonate white

Emission wavelengt h [nm]

Emission intensity [a.u.]

Excitation intensity [a.u.]

No

Sample

Emission wavelength [nm]

Emission intensity [a.u.]

Excitation intensity [a.u.]

496.4

81.61

81.44

1

Straw

495

112.6

119.77

491.5

65.98

67.56

2

Hemp

493

201.4

207.74

494.3

366.44

383.25

3

Jute

496

56.0

56.83

490.3

241.99

247.24

4

Sisal

501

48.8

45.88

494

150.23

153.08

5

Abaca

493

74.3

79.90

494.3 487.9

61.31 570.47

62.33 574.98

6 8

Flax Bagasse

494 493

382.5 193.2

421.60 211.78

493.5

70.75

72.79

498.4

56.64

58.92

495.5

651.48

716.45

In annual fibre crops lignins, the range of variations is quite reduced, examinations of the samples suggesting that only a small amount of lignin’s phenylcoumarone structures acts as an acceptor. Similarly with stilbene, such structures are created by modifying lignin’s structural units, which is accompanied by the formation of phenylcoumarane during the various treatments applied for lignins’ separation. CONCLUSIONS The differences observed among hardwoods, softwoods and annual fibre crops lignins have been evaluated by spectroscopical analyses. The different origin and extraction methods applied on the lignin samples have had a major influence on the shape of FT-IR, 13 C NMR, UV and fluorescence spectra. All spectroscopic methods permit a quantitative and qualitative evidencing of the differences

occurring among the lignin samples isolated form woody plants and annual fibre crops. Quantitative analysis of FT-IR and 13C NMR spectra shows that the lignins from hardwoods and annual fibre crops present different characteristics compared with those of softwood lignins. As expected; the lignins from softwoods present a high amount of guaiacyl units than syringyl units. The most representative seems to be the S/G (I1330/I1270) ratio, clearly evidencing this difference. The relationships of some spectral characteristics such as I2920/I1510, I1710/I1510, I1370/I1510, I1330/I1270, I1086/I1510 with the content of the (chemically determined) functional groups was established, which can be useful for a rapid structural characterization of the various lignin samples. ACKNOWLEDGEMENTS: This paper contains results obtained under the contract GIRT-CT-2002-05088 from the European project EUROLIGNIN founded by EC.

620

Lignin characterization REFERENCES 1

N. G Lewis, Ann. Rev. Plant. Phys., 41, 455 (1990). 2 M. Micic, K. Radotic, I. Benitez; M. Ruano, M. Jeremic, V. Moy, M. Mabrouki and R. M. Leblanc, Biophys. Chem., 94, 257 (2001). 3 R. Laschimke, Thermochim. Acta, 151, 35 (1989). 4 N.G. Lewis, Curr. Opin. Plant Biol., 2, 153 (1999). 5 T. Dizhbite, G. Zakis, A. Kizima, E. Lazareva, G. Rossinskaya, V. Jurkjane, G. Telysheva and U. Viesturs, Bioresource Technol., 67, 221 (1999). 6 D. W. Hopkins, E. A. Webster, J. A. Chudek and C. Halpin, Soil Biol. Biochem., 33, 1455 (2001). 7 M. N. Belgacem, A. Blayo and A. Gandini, Ind. Crop. Prod., 18, 145 (2003). 8 J. S. Sperry, Int. J. Plant Sci., 164, 115 (2003). 9 B. Monties, Anim. Feed Sci. Tech., 32, 159 (1991). 10 C. W. Dence and S. Y. Lin, in “Methods in Lignin Chemistry”, S. Y. Lin and C. W. Dence (Eds.), Springer Verlag, pp. 3 – 6 (1992). 11 R. J. A. Gosselink, A. Abächerli, H. Semke, R. Malherbe, P. Käuper, A. Nadif and J .E. G. van Dam, Ind. Crop. Prod., 19, 271 (2004). 12 W. Boerjan, J. Ralph and M. Baucher, Ann. Rev. Plant Biol., 54, 519 (2003). 13 M. Micic, K. Radotic, M. Jeremic, D. Djikanovic and S. B. Kämmer, Colloid Surface B: Bio interfaces, 34, 33 (2004). 14 A. M. Rouhi, Chem. Eng. News, 13, 29 (2000). 15 C. G. Boeriu, D. Bravo, R. J. A. Gosselink and J. E. G. van Dam, Ind. Crop. Prod., 20, 205 (2004). 16 J. Ralph, J. M. Marita, S. A. Ralph, R. D. Hatfield, F. Lu, R. M. Ede, J. Peng, L. L. Landucci, in “Advances in Lignocellulosics Characterization”, D. Argyropoulos (Ed.), Tappi Press: Atlanta, GA, pp. 55 - 108 (1999). 17 S.Y. Lin, in “Methods in Lignin Chemistry”, S.Y. Lin and C.W. Dence (Eds.), Springer Verlag, pp 217 – 232 (1992). 18 K. V. Sarkanen, C. H. Ludwig and H. L. Hergert, in “Lignins: Occurrence, Formation, Structure and Reactions: Infrared Spectra”, Wiley-Interscience, New York, pp 267 - 299 (1971). 19 K. Lundquist, B. Josefsson and G. Nyquist, Holzforschung, 32, 27 (1978). 20 K. Lundquist, I. Egyed, B. Josefsson and G. Chem. Technol., 15, 699 Nyquist, Cellulose (1981). 21 R. S. Davidson, L. Dumn, A. Castellan and A. Nourmamode, J. Photochem. Photobiol., A 58, 349 (1991). 22 A. Castellan and R.S. Davidson,. J. Photochem. Photobiol., A 78, 275 (1994). 23 B. Albinsson, S. Li, K. Lundquist and R. Stomberg, J. Mol. Struct., 508, 19 (1999).

24

A. E. H. Machado, D. E. Nicodem, R. Ruggiero, S. Perez and A. Castellan, J. Photoch. Photobio. A: Chemistry, 138, 253 (2001). 25 G. Gellerstedt and D. Robert, Acta Chem. Scand. Ser. B, 41, 541 (1987). 26 L. Landucci, Holzforschung, 39, 355 (1985). 27 L. Landucci, Holzforschung, 45, 559 (1991). 28 C. Lapierre, B. Monties, E. Guittet and J.Y. Lallemand, Holzforschung, 38, 333 (1984). 29 C. Lapierre, B. Monties, E. Guittet and J. Y. Lallemand, Holzforschung, 39, 367 (1985). 30 J. R. Obst and L. Landucci, Holzforschung, 40, 87 (1986). 31 D. Robert and G. Brunow, Holzforschung, 38, 85 (1984). 32 D. Robert, G. Gellerstedt and M. Bardet, Nord. Pulp Pap. Res. J., 1, 18 (1986). 33 D. Robert and C. L. Chen, Holzforschung, 43, 323 (1989). 34 D. Robert, in “Methods in Lignin Chemistry”, S. Y. Lin and C. W. Dence (Eds.), Springer Verlag, pp 250 – 273 (1992). 35 J. Gierer, J. Wood Sci. Technol., 19, 289 (1985). 36 H.D. Ludemann and H. Nimz, Makromol. Chem., 175, 2409 (1974). 37 C. Vasile, C.-M. Popescu, A. Stoleriu and R. Gosselink, in “New Trends in Natural and Synthetic Polymer Science”, C. Vasile and G. E. Zaikov (Eds.), Nova Science Publishers, pp 135 - 163 (2006). 38 O. Faix, in “Methods in Lignin Chemistry”, S. Y. Lin and C. W. Dence (Eds.), Springer Verlag, pp 83 - 110 (1992). 39 A. Bermello, M. Del Valle, U. Orea and L. R. Carballo, Int. J. Polym. Mat., 51, 557 (2002). 40 E. Saliba, N. M. Rodriguez, S. A. L. Morais and D.P. Veloso, Ciencia Rural, 31, 917 (2001). 41 M. P. Turker, Q. A. Nguyen, F. P. Eddy, K. L. Kadam, L. M. Gedvilas and J. D. Webb, Appl. Biochem. Biotech., 91 – 93, 51 (2001). 42 M. Likon and A. Perdih, Acta Chim. Slov., 46, 87 (1999). 43 S. A. L. Morais, E. A. Nascimento, C. Queiroz, D. Pilo-Veloso and M. G. Drumond, J. Braz. Chem. Soc., 10, 447 (1999). 44 H. Struszczyk, J. Macromol. Sci. A, 23, 973 (1986). 45 J. Rodrigues, O. Faix and H. Pereira, Holzforschung, 52, 46 (1998). 46 G. Cazacu, C.-M. Popescu, V. I. Popa, Gh. Singurel and C. Vasile, Studia Univ. Babes-Bolyai Physica, Special Issue II, pp. 434-437 (2003). 47 C.-M. Popescu, M.-C. Popescu, V. I. Popa, Gh. Singurel and C. Vasile, Analele Universitatii de Vest din Timisoara, Physics, 44, 69 (2003). 48 C.-M. Popescu, V. I. Popa, Gh. Singurel, G. Cazacu and C. Vasile, Workshop on Fundamental

621

Carmen –Mihaela Popescu et al. and Applied Research in Physics, Iasi, November, pp. 58 (2003). 49 C.-M. Popescu, G. Cazacu, Gh. Singurel and C. Vasile, Romanian J. Phys., 51, 259 (2006). 50 C. Vasile, C.-M. Popescu, B. S. Munteanu and G. Cazacu, Proceedings. Eight European Workshop on Lignocellulosics and Pulp, Utilization of lignocellulosics and by-products of pulping, Riga, Latvia, August 22-25, pp. 267 – 270 (2004). 51 C.-M. Popescu, B. S. Munteanu, M.-C. Popescu and C. Vasile, Procs. 7th ILI Forum, Barcelona, April 27 – 28, pp. 133 – 136 (2005). 52 C.-M. Popescu, V. Pohoata, C. Vasile and Gh. Singurel, Procs. 3rd Italian Meeting on Lignin Chemistry, Wood Derivatives and Agroindustrial Waste Valorisation, L’Aquila, Italia, June 23-24, pp 126 – 129 (2005). 53 L. L. Landucci, S. A. Ralph and K. E. Hammel, Holzforschung, 52, 160 (1998).

54

A. R. Gonçalves, U. Schuchardt, M. L. Bianchi and A. A. S. Curvelo, J. Braz. Chem. Soc., 11, 491 (2000). 55 X.-J Pan and Y. Sano, Holzforschung, 53, 511 (1999). 56 G. Koch, B. Rose, R. Patt and O. Kordsachia, Holzforschung, 57, 611 (2003). 57 T. Bikova and A. Treimanis, Carbohyd. Res., 55, 315 (2004). 58 R. C. Sun, J. Tomkinson, X. F. Sun and N. J. Wang, Polymer, 41, 8409 (2000). 59 T. Bikova, A. Treimanis, G. Rossinska and G. Telysheva, Holzforschung, 58, 489 (2004). 60 A. Scalbert, B. Monties, E. Guittet and J. Y. Lallemand, Holzforschung, 40, 119 (1986). 61 K. K. Pandey, Polym. Degrad. Stabil., 90, 9 (2005).

622