Analytical methods for determining functional groups in various technical lignins

Industrial Crops and Products 26 (2007) 116–124

Analytical methods for determining functional groups in various technical lignins Nour-Eddine El Mansouri 1 , Joan Salvad´o ∗ Rovira i Virgili University, Department of Chemical Engineering, Avinguda dels Pa¨ısos Catalans 26, 43007 Tarragona (Catalunya), Spain Received 8 March 2006; accepted 7 February 2007

Abstract In this paper we compare various analytical methods for determining functional groups in technical lignins of five different origins: kraft, sulfite, soda/anthraquinone, organosolv and ethanol process lignins. These lignins were characterized in terms of methoxyl, phenolic and aliphatic hydroxyl, carbonyl, carboxyl and sulfonate groups. The analytical methods used were: gas chromatography, aminolysis, UV-spectroscopy, 1 H and 13 C NMR spectroscopy, the oximating method, FTIR spectroscopy, acid number determination, and non-aqueous and aqueous potentiometry. The statistical comparison of the various analytical methods for hydroxyl groups (phenolic and aliphatic) shows that the results obtained are not fully comparable. Aminolysis and non-aqueous potentiometry are assumed to be the most reliable for phenolic hydroxyl. We observed the same trend for the methods for carbonyl groups and selected the oximating method as reliable for determining total carbonyl. The results for the methods used for carboxylic groups showed correspondence at a significance level of 0.05. We selected aqueous and non-aqueous titration as being reliable for the lignins studied. We also compare all the commercial lignins in terms of functional groups. Finally, by completely characterizing the functional groups of various technical lignins, we have established the most complete representative expanded formula C9 for each lignin under study. © 2007 Elsevier B.V. All rights reserved. Keywords: Characterization; Technical lignins; Analytical methods; Functional groups; Expanded molecular formula C9

1. Introduction With the exception of cellulose, no other renewable natural resource is more abundant than lignin. Lignin is a highly-branched, three dimensional polymer with a wide variety of functional groups providing active centers for chemical and biological interactions. In wood, the ∗ Corresponding author. Tel.: +34 977 559 641; fax: +34 977 558 544. E-mail addresses: [email protected] (N.-E. El Mansouri), [email protected] (J. Salvad´o). 1 Tel.: +34 977 558 656; fax: +34 977 558 544.

0926-6690/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.indcrop.2007.02.006

lignin content generally ranges from 19 to 35% (Dence and Lin, 1992). It is extracted by several pulping techniques and ethanol production process as a by-product available inexpensively in large quantities. Technical lignins are divided into two categories (Gosselink et al., 2004b). The first one comprises sulfur-containing commercial lignins, including lignosulfonates and kraft lignins, which are produced in huge quantities. The second one comprises lignins without sulfur in their composition, such as organosolv, soda/anthraquinone lignin and lignin from the ethanol process production. The potential of lignins is clearly not valued because almost all are burned to generate energy and recover

N.-E. El Mansouri, J. Salvad´o / Industrial Crops and Products 26 (2007) 116–124

chemicals. Only a limited quantity has been used for applications such as biomaterials, fuels, biocides and biostabilisers, animal feed, health products and crops cultivations (Lora and Glasser, 2002). However, industrial applications are only possible if lignin’s added value is enhanced, which is only possible if industrial and scientific research can be intensified to find better applications. Current research faces several problems that could be avoided. These problems are the low purity, heterogeneity, odour, colour of lignin-based products and the absence of reliable analytical methods (Gosselink et al., 2004a). Thus, the availability of the analytical methods for chemical and physical properties adopted by both suppliers and users can allow any laboratory to reproduce the results and analyze any various existing types of lignins. Using these methods lignin can be properly characterized and its behavior with regard to several potential uses can be determined (Gosselink et al., 2004c). Several studies have established new methods or compared existing methods to characterize lignins (Gosselink et al., 2004a; Milne et al., 1992; Faix et al., 1998). Much interest has focused on functional groups analyses. The main chemical functional groups in lignin are the hydroxyl, methoxyl, carbonyl and carboxylic groups. The proportion of these groups depends on the genetic origin and isolation processes applied. Functional group analysis can be used to determine the lignin structure. However, the increasing interest in using analytical methods to determine the functional groups is mainly due to the following reasons: (i) the appearance of new technical lignin generated from new and more environmentally friendly cellulose-production methods. To understand the reaction mechanisms during delignification and to predict and develop different uses for byproducts of the pulping process, we therefore need to study their functional properties; (ii) lignin is currently of interest to the specialist in various fields of science and industry searching for new practical applications. Functional group analysis is therefore an indispensable research method. The only way to achieve these aims is to compare the various analytical methods. In this paper, we review the main analytical methods in the field of lignin chemistry, especially for functional groups analysis, and select 11 analytical methods. We selected five technical lignins for the structural characterization, focusing on different functional groups, with these analytical methods. These characterization permit a critical comparison between these methods, and choose the most adequate in each case, and the comparison between these lignins in term of functional groups. Finally, we established the most representative formula

117

C9 , which contains the important information about the structure of each lignin. 2. Materials and methods 2.1. Raw materials Kraft lignin (KL) and lignosulfonate (LS) derived from softwood were purchased from Ligno-Tech Iberica. Soda/anthraquinone lignin (SAL) from a mixture of long fiber plants, was supplied by CELESA “Celulosa de Levante S.A.” of Tortosa, Catalonia, Spain. Organosolv lignin (ORS) obtained from Miscanthus sinensis, was of the formasolv lignin type, which was supplied by the University of Santiago de Compostela (Galicia-Spain). Ethanol process lignin (EPL) was supplied by CIEMAT (Centro de Investigaci´on Energ´eticas, Medioambientales y Tecnol´ogicas) of Madrid, Spain, from Populus wood pretreated by steam explosion and the simultaneous saccharification and fermentation process (SSF). These lignins were purified and analyzed for chemical composition in a previous study (El Mansouri and Salvad´o, 2006). The characteristics of these lignins are: total lignin content of over 94% (except lignosulfonate) and a sugar content of close to 2% (except ethanol process lignin). All lignins were air-dried at room temperature to equilibrium moisture content and stored in plastic bottles for characterization. The technical lignins were analyzed in this study by the following methods. 2.2. Analytical methods 2.2.1. Elemental analysis Carbon, hydrogen, sulfur and nitrogen contents were determined using a Perkin Elmer 640-C Analyzer. After correction for ash content, the percentage of oxygen was calculated by difference. 2.2.2. FTIR spectroscopy for unacetylated lignins The FTIR spectra of the unacetylated lignin samples embedded in KBr disk were obtained with a BRUKER spectrometer using a resolution of 4 cm−1 and 32 co addition scans in a frequency range of 400–4600 cm−1 . The spectra were analyzed by Nicolet software to compare the absorbance corresponding to each functional group. The absorption bands were assigned as suggested by Faix (1992). 2.2.3. Methoxyl groups Methoxyl group was determined as suggested by V´azquez et al. (1997). The lignin (0.15 g) was treated with refluxing concentrated sulfuric acid (10 ml) for

118

N.-E. El Mansouri, J. Salvad´o / Industrial Crops and Products 26 (2007) 116–124

Fig. 1. Types of phenolic structure determined in different lignin samples.

10 min. The reaction mixture was cooled, 70 ml of distilled water was added, and the methanol produced in the reaction was distilled off under vacuum and quantified by gas chromatography. 2.2.4. Acetylation A weighted amount of each lignin except lignosulfonate was acetylated for 48 h with a mixture of purified pyridine-acetic anhydride (1:1, v/v). Methanol was used to quench the remaining acetic anhydride. Finally, a flow of nitrogen was applied to evaporate the solvents and the samples were dried under vacuum (Chum et al., 1985). 2.2.5. Hydroxyl groups: aliphatic and phenolic Phenolic hydroxyl groups were determined by three wet chemical methods (aminolysis, ultravioletspectroscopy and non-aqueous potentiometry) and two spectroscopy methods (1 H NMR and 13 C NMR). The two spectroscopy methods enabled aliphatic hydroxyl quantification. These methods are described below. 2.2.5.1. Aminolysis. The procedure described by Lai was used to determine free phenolic hydroxyl groups in lignin (Lai, 1992). The acetylated lignin, dissolved in 1.0 ml of dioxane containing 5 mg of 1-methylnaphtalene, was treated with 1.0 ml of dioxanepyrrolidine (1:1, v/v) solution, which initiated the aminolysis reaction. After the addition of pyrrolidine, samples were taken from the reaction mixture at different times (total reaction time was approximately 120 min) and analyzed by gas chromatography. The amount of 1acetylpyrrolidine formed (equivalent to the amount of hydroxyl groups) was recorded as a function of time. The content of phenolic hydroxyl groups was calculated by extrapolation of the curve at zero time. 2.2.5.2. Phenolic hydroxyl groups by ultravioletspectroscopy (ε method). The content of various phenolic units in lignin samples was determined by UV spectroscopy as described by Zakis (1994). This method is based on the difference in absorption at 300

and 360 nm between phenolic units in neutral and alkaline solutions. The content of ionizing phenol hydroxyl groups can be quantitatively evaluated by comparing the ε values of substances studied at certain wavelengths to the values of ε of the respective model compounds (I, II, III, IV types) (Fig. 1). 2.2.5.3. Proton nuclear magnetic resonance spectroscopy (1 H NMR). We used proton nuclear magnetic resonance to analyze all acetylated technical lignins under study. 1 H NMR spectra of 10 mg acetylated lignin samples dissolved in 0.5 ml of CDCl3 were recorded on a VARIAN GEMINI 300 Hz apparatus using tetramethylsilane as internal standard under the same condition as those described by Lundquist (1992). Proton signals were integrated from the baseline and referred to the integrated signal of the methoxyl protons for proton quantification of aliphatic and phenolic hydroxyl. 2.2.5.4. Carbon nuclear magnetic resonance spectroscopy 13 C NMR. 13 C nuclear magnetic resonance is the most suitable method for determining benzylic alcohol groups in lignins. For all acetylated lignins, the 13 C NMR spectra were recorded in acetone-d under 6 the same conditions as those described by Robert and Br¨unow (1984). The quantitative estimation of different hydroxyl groups (located at 170.8 and 170 ppm of primary and secondary aliphatic hydroxyl groups, respectively, and 168.9 ppm for the phenolic hydroxyl group) were achieved by expanding ten times, before integration, the signal areas corresponding to each functional group and combining these results with those of elemental analysis and methoxyl groups. 2.2.6. Carbonyl groups Carbonyl groups for all lignins were determined by two wet chemical methods: the Modified Oximating method and differential UV-spectroscopy. The Modified Oximating method was described by Faix et al. (1998) that present a correction technique, which is necessary for lignins containing carboxyl groups. Differential UV-

N.-E. El Mansouri, J. Salvad´o / Industrial Crops and Products 26 (2007) 116–124

119

Fig. 2. Types of carbonyl structures determined in various lignins.

spectroscopy was developed by Alder and Marton in 1966 and reported by Zakis (1994). It involves differential absorption measurements that take place when carbonyl groups are reduced at the benzylic alcohol corresponding with sodium borohydride. This method determines some carbonyl lignin structures such as aldehydes and ketones structures described in Fig. 2. 2.2.7. Carboxyl groups We analyzed carboxyl groups using three methods: acid number determination and aqueous and nonaqueous titration methods. These methods are described below. 2.2.7.1. Acid number determination. Carboxylic groups were determined as described by Gosselink et al. (2004a). The pH of 100 ml of 95% ethanol in water was adjusted to 9.0 using 0.1 mol/l sodium hydroxide in water. After adding 1 g of dried lignin, the mixture was stirred for 4 h and subsequently titrated back to pH 9.0 with 0.1 mol/l sodium hydroxide solution. 2.2.7.2. Aqueous titration method. This method was used by Gosselink et al. (2004a). A weight of lignin sample (1 g) was suspended in 100 ml of alkaline aqueous solution. After stirring for 3 h, the pH was adjusted to 12 with sodium hydroxide. After stirring again, the solution was potentiometrically titrated with 0.1 mol/l aqueous hydrochloride acid. 2.2.7.3. Non-aqueous potentiometry method. This procedure, reported by Dence, involves a non-aqueous potentiometric titration of lignin with tetra-n-butylammonium hydroxide in the presence of an internal standard, which is p-hydroxybenzoic acid (Dence, 1992; Gosselink et al., 2004a). The advantage of this method is

that it determines not only the carboxyl groups in lignin but it concurrently determines the weakly acidic phenolic hydroxyl groups. When combined with an ion-exchange treatment, the aforementioned titrimetric procedure was also used to determine the strongly acidic groups (sulfonates groups) in lignosulfonate. 2.2.8. Sulfonate groups Sulfonate groups were determined by non-aqueous potentiometry, as described above (Dence, 1992). 2.2.9. Expanded C9 formulae The expanded formulae C9 contain complete information about the lignin structure. They are obtained by combining the results from elementary analysis and functional groups analysis. 2.2.10. Statistical analysis We compared the methods for determining the functional groups in lignins by applying paired two-sided t-tests at a 95% confidence level for mean values and combining the two methods. The results are presented as averages and their standard deviation. 3. Results and discussions 3.1. Structural characterization with FTIR spectroscopy The IR absorption spectra of the five technical lignins studied were recorded in the 400–4000 cm−1 region (see Fig. 3). These spectra show that there were clear differences between these lignins. The band at 3400 cm−1 , which is attributed to OH groups in lignins, had a lower absorption intensity for KL and SAL than for ORS, EPL and LS. This is attributed to the high oxidation and degra-

120

N.-E. El Mansouri, J. Salvad´o / Industrial Crops and Products 26 (2007) 116–124

Fig. 3. FTIR spectra of unacetylated lignin samples.

dation power of soda during the two pulping processes. The 3000–2800 cm−1 region of the C H stretch in the methyl and methylene groups was present in different quantities. These bands, which were mainly attributed to methoxyl groups, were substantially higher for SAL, EPL and ORS and presented relatively lower absorbance bands for KL and LS. The carbonyl stretching vibration at 1720 cm−1 appeared in the IR spectra of KL, SAL and ORS but was absent in the spectra of EPL and LS. At 1600 and 1500 cm−1 , aromatic skeletal vibration bands were observed for all lignins. Between 1300 and 1000 cm−1 , the bands and peak ratios were very different due to various vibrations modes such as C O, C H and C O. The distinct band appearing at 620 cm−1 in the spectra of LS was assigned to the sulphonic groups (S O stretching vibration) formed from the reaction of sodium sulphite with the secondary OH of the aliphatic side chain of lignin. FTIR spectroscopy showed that the lignins studied were clearly structurally different. The structural differences between other lignins analyzed

by FTIR spectroscopy were reported by Carmen et al. (2004). This will be analyzed in further detail in this study. 3.2. Hydroxyl groups: phenolic and aliphatic hydroxyl The phenolic hydroxyl groups of all lignin samples were determined by several methods: aminolysis, UVspectroscopy, 1 H NMR, 13 C NMR and non-aqueous potentiometric titration (Table 1). Aliphatic hydroxyl groups were determined by 1 H NMR and 13 C NMR spectroscopy (Table 2). The amounts of the various phenolic structures present in lignin as determined by UV-spectroscopy are shown in Table 3. Comparison of the methods used for phenolic hydroxyl quantification by statistical analysis (paired t-test) as listed in Table 4 shows that aminolysis/13 C NMR, UV-spectroscopy/13 C NMR and non-aqueouspotentiometry/1 H NMR show a poor correspondence

Table 1 Phenolic hydroxyl content in various technical lignins determined by different methods (%, w/w)

KL SAL ORS EPL LS

Aminolysis

Non-aqueous potentiometry

1H

4.60 (0.04) 4.90 (0.07) 2.80 (0.10) 2.55 (0.08) NA

4.54 (0.15) 5.10 (0.23) 3.56 (0.12) 2.92 (0.18) 2.55 (0.31)

4.10 4.50 3.33 2.65 NA

NA: Not acetylated; ( ) standard deviation. a Data from El Mansouri and Salvad´ o (2006).

NMRa

13 C

NMR

4.99 5.31 3.23 2.70 NA

UV-spectroscopya 4.50 (0.32) 4.40 (0.30) 2.66 (0.32) 2.30 (0.36) 2.00 (0.16)

N.-E. El Mansouri, J. Salvad´o / Industrial Crops and Products 26 (2007) 116–124 Table 2 Aliphatic hydroxyl content of various technical lignins determined by NMR spectroscopy methods (%, w/w) 1H

KL SAL ORS EPL

13 C

NMR

10.09 3.10 3.50 4.73

NMR

9.80 2.45 3.20 4.55

at a significance level of 0.05. In contrast, the other paired analyses show a correspondence at a significance level of 0.05. This variability in results is attributed to an incomplete acetylation in the case of methods based on lignin acetylation, such as 1 H NMR, 13 C NMR and aminolysis. This incomplete acetylation was confirmed by Gosselink et al. for sulphur-free lignin and model compounds that may be attributed to steric hindrance by the methoxyl groups present in lignins (Gosselink et al., 2004a). Moreover, NMR-spectroscopy is characterized by an overlapping signal that lowers the accuracy of these techniques. Also, UV-spectroscopy determines only some phenolic structures, so the phenolic groups might be underestimated. For non-aqueous potentiometry it is difficult to observe the inflection point with some lignins. From the results obtained, we can see that the methods used are not fully comparable. The standard deviations for each lignin analysis lead us to assume that aminolysis and non-aqueous potentiometry are the most reliable for the determination of phenolic hydroxyl. These results are in agreement with those of Milne et al. (1992) and Gosselink et al. (2004a). The two selected methods provide quantitative data on the frequency with which the

121

phenolic OH occurs in lignin, but they do not reveal the structural environment in which it occurs. This information about the lignin structure can be obtained by the spectral techniques. The UV spectroscopy is an easy method to quickly estimate some phenolic hydroxyl structures. The 1 H NMR showed a poor correspondence in the results with 13 C NMR at a significance level of 0.05 for aliphatic hydroxyl determination (p-value is 0.04 < 0.05). This is attributed to the overlapping signals that can easily introduce significance errors and to the well-known incomplete acetylation of lignin with NMR spectroscopy. A similar discrepancy was observed by Gosselink et al. (2004a) when estimating the ratio of phenolic/aliphatic hydroxyl by methods such as 1 H NMR and 13 C NMR spectroscopy. 1 H NMR and 13 C NMR are therefore not comparable for aliphatic hydroxyl quantification. These results show the phenolic hydroxyl contents were highest for kraft and soda/anthraquinone lignins, high for organosolv lignin and relatively low for ethanol process lignin and lignosulfonate. The aliphatic hydroxyl content was highest for the kraft lignin and relatively low for the other samples.

3.3. Carbonyl groups Table 5 shows the quantitative determination of carbonyl groups by differential UV-spectroscopy and modified oximating method with and without the correction technique. Table 6 shows the amount of different carbonyl structures as determined by differential UVspectroscopy.

Table 3 Relative abundance of different phenolic structures in lignins determined by UV-spectroscopy (%, w/w)

[OH]I [OH]III [OH]II [OH]IV

Non-conjugated phenolic structures (I + III) Conjugated phenolic structures (II + IV)

KL

LS

SAL

ORS

EPL

2.63 0.49 1.30 0.08

1.34 0.48 0.14 0.03

2.74 0.57 1.10 0.02

0.89 0.44 1.31 0.02

1.43 0.68 0.14 0.05

Table 4 Comparison of methods for the determination of phenolic hydroxyl content by paired t-test (two-sided p-values) Method

Aminolysis

UV-spectroscopy

1H

Aminolysis UV-spectroscopy 1 H NMR 13 C NMR Non-aqueous potentiometry

– – – – –

0.07 – – – –

0.79 0.48 – – –

NMR

13 C

NMR

0.01 0.013 0.20 – –

Non-aqueous potentiometry 0.16 0.05 0.02 0.88 –

122

N.-E. El Mansouri, J. Salvad´o / Industrial Crops and Products 26 (2007) 116–124

Table 5 Content of carbonyl groups in samples from various analytical methods (%, w/w) Lignin types Oximating method

UV-spectroscopy

Without correction With correction KL SAL LS ORS EPL

3.13 (0.05) 2.62 (0.10) 5.30 (0.10) 4.05 (0.10) 6.48 (0.11)

2.91 (0.05) 2.13 (0.10) 4.50 (0.10) 3.94 (0.09) 5.73 (0.11)

2.35 (0.32) 1.94 (0.25) 4.70 (0.27) 2.90 (0.19) 5.20 (0.23)

Ethanol process lignin and lignosulfonate showed higher contents of carbonyl groups than other lignins. Values for kraft and organosolv lignins were within the range found by Faix et al. (1998) when analyzing alcell-organosolv from yellow poplar (4.40%) and kraft indultin AT (3.32%). The higher carbonyl contents of technical lignins than of ball milled enzyme lignin 2.2% are plausible because technical lignins underwent oxidation during the treatment process (Faix et al., 1998).

( ) Standard deviation.

3.4. Carboxyl groups Table 7 indicate that the correspondence between the results for the carbonyl groups determined by oximating method without correction and for the oximating method with correction and differential UVspectroscopy method were poor at a significance level of 0.05. These differences in the results are attributed to a correction method introduced in order to subtract CO from carboxylic origin in the oximating method and to the existence of other forms of carbonyl groups underestimated by differential UV-spectroscopy for example quinone forms, which exist in highly oxidized lignins such as those in this study. The results from UV-spectroscopy and the oximating method with the correction technique corresponded at a 0.05 significance level. These results show that the methods are not completely comparable. From the standard deviations of each lignin analysis, we concluded that the oximating method with the correction technique is reliable for total carbonyl quantification, which was confirmed by Faix et al. (1998). Differential UV-spectroscopy enables some carbonyls, such as aldehydes and ketones structures, to be determined.

Table 8 lists the carboxyl content for lignins determined by acid number and aqueous and non-aqueous titration methods, as described above. Statistical comparison of these methods shows that there were no significant differences at a 95% confidence level (Table 9). However, the carboxylic contents of lignins were different for the three titration methods. These differences were due to the solubility of the lignins in the selected solvents. The same trend was observed by Gosselink when analyzing soda lignins with the same methods (Gosselink et al., 2004a). The accessibility of the carboxylic groups is therefore higher when DMF is used as solvent for non-aqueous titration and when the agitation time is longer in the alkaline medium for aqueous titration. From the standard deviations for each lignin analysis, we concluded that non-aqueous titration and aqueous titration, in this order, provide reliable results for the determination of carboxyl groups. The acid number method cannot be used for lignosulfonate because this lignin is insoluble in 95% ethanol. With this method the solubility of the other lignins is also poor, which is reflected in the low values for the carboxylic groups.

Table 6 Relative abundance of some aldehydes and ketones types in samples obtained by differential UV-spectroscopy (%, w/w) KL

LS

SAL

ORS

EPL

Coniferyl aldehyde structures (I + II)

[CO]I [CO]II

0.38 1.09

0.98 1.80

0.31 0.56

1.03 1.14

1.50 1.53

Ketones structures (III + IV)

[CO]III [CO]IV

0.51 0.37

0.90 1.02

0.73 0.34

0.66 0.07

1.28 0.89

Table 7 Comparison of methods for the determination of carbonyl content by paired t-test (two-sided p-values) Method

Oximating without correction

Oximating with correction

UV-spectroscopy

Oximating without correction Oximating with correction UV-spectroscopy

– – –

0.03 – –

0.01 0.11 –

N.-E. El Mansouri, J. Salvad´o / Industrial Crops and Products 26 (2007) 116–124

123

Table 8 Contents of carboxylic and sulfonate obtained by the analytical methods (%, w/w) Lignin type

Non-aqueous potent

Acid number

Aqueous titration

Sulfonate

KL SAL ORS EPL LS

7.06 (0.15) 6.91 (0.22) 3.15 (0.20) 2.02 (0.27) 4.63 (0.21)

5.97 (0.50) 5.42 (0.56) 2.79 (0.60) 1.82 (0.69)

7.10 (0.31) 6.90 (0.37) 2.86 (0.45) 2.17 (0.41) 4.30 (0.42)

– – – – 12.23 (0.39)

a

( ) Standard deviation. a Sample not completely dissolved. Table 9 Comparison of methods for the determination of carboxyl content by paired t-test (two-sided p-values) Methods

Non-aqueous potent

Acid number

Aqueous titration

Non-aqueous potent Acid number Aqueous titration

– – –

0.08 – –

0.40 0.10 –

The contents of carboxylic groups for kraft and soda/anthraquinone were higher than for the other lignins. This indicates that the two lignins were highly degraded during the kraft and soda/anthraquinone pulping. Ethanol process lignin seemed to be less degraded. 3.5. Sulfonate groups Table 8 shows the sulfonate group contents of lignosulfonate. These groups ensure ready water solubility in the presence of a suitable counter ion (Na, Ca, Mg, NH4 , etc.). These results are in agreement with those in the literature. The results from non-aqueous potentiometric titration and elementary analysis show that not all the sulfur content in lignosulfonate is in the form of sulfonate. 3.6. Expanded molecular formulae Table 10 lists the expanded molecular formulae for the various technical lignins under study. The expanded

Table 11 Elemental composition of different lignins studied (El Mansouri and Salvad´o, 2006)

KL SAL LS ORS EPL

%C

%H

%N

%S

%O

65.00 65.00 44.84 63.51 58.34

5.41 6.12 5.15 5.55 6.01

0.05 0.17 0.02 0.02 1.26

1.25 0.00 5.85 0.00 0.00

28.24 28.64 44.14 30.92 34.40

C9 formulae were obtained from elemental analysis (Table 11) and functional groups analysis, which provides a number for each functional group per expanded formula C9 . Each expanded formula C9 summarizes all the information about the structure of these technical lignins. 4. Conclusions We have conducted a comparative study of the different analytical methods for the functional groups in various technical lignins. Statistical comparison shows that the methods used for phenolic OH are not fully equivalent. Each method has some disadvantage or other: incomplete acetylation with techniques based on acetylation, an overlapping signal in nuclear magnetic resonance, the difficulty of showing the inflection point in non-aqueous titration, and the underestimation of phenolic hydroxyl content with UV-spectroscopy. Despite

Table 10 Expanded molecular formulae for the technical lignins studied Lignins

Expanded formulae C9

KL SAL LS ORS EPL

C9 H6,010 O0,269 N0,006 S0,065 (OCH3 )0,597 (OHAr )0,425 (OHAl )1,046 (OCO )0,183 (OOHCOOH )0,277 C9 H6,825 O0,560 N0,020 S0,065 (OCH3 )1,166 (OHAr )0,493 (OHAl )0,338 (OCO )0,141 (OOHCOOH )0,286 C9 H10,360 O2,880 N0,003 S0,070 (OCH3 )0,730 (OHAr )0,260 (OCO )0,354 (OOHCOOH )0,227 (HSO3 )0,330 C9 H6,705 O1,205 N0,002 (OCH3 )0,971 (OHAr )0,396 (OHAl )0,380 (OCO )0,260 (OOHCOOH )0,130 C9 H9,036 O2,270 N0,166 (OCH3 )0,646 (OHAr )0,289 (OHAl )0,515 (OCO )0,378 (OOHCOOH )0,083

OCH3 : Methoxyl groups; OHAr : aromatic phenolic hydroxyl; OHAl : aliphatic phenolic hydroxyl; OCO : carbonyl groups; OOHCOOH : carboxyl groups; HSO3 : sulfonate groups.

124

N.-E. El Mansouri, J. Salvad´o / Industrial Crops and Products 26 (2007) 116–124

these contradictory results, we selected aminolysis and non-aqueous titration with TnBAH as reliable methods. Non-aqueous titration with TnBAH can be used with all technical lignins and can determine not only phenolic OH but also carboxylic and sulfonate groups. Also, the methods used for the aliphatic hydroxyl groups are not comparable because the spectral technique is based on acetylation, which is incomplete for lignin, and because an overlapping signal affects reliability. The methods for quantifying carbonyl are also not comparable because one determines total carbonyl content and the other determines only some carbonyl structures. The oximating method is reliable for determining total carbonyl groups. The methods used to determine carboxylic groups are comparable and we selected non-aqueous titration and aqueous titration as reliable methods for the technical lignins in this study. By analyzing the various lignin functional groups, we determined their structural characteristics. Several analytical methods showed that the highest content of phenolic hydroxyl were in kraft and soda/anthraquinone lignins and that there was a high content in organosolv lignin but a relatively low content in ethanol process lignin and lignosulfonate. Kraft lignin had the highest content of aliphatic hydroxyl: the other lignin samples had low contents. Lignosulfonate and ethanol process lignin had the highest contents of carbonyl groups than the other lignins. Carboxyl groups analysis also showed that Kraft lignin and soda/anthraquinone were more highly degraded than the other lignins under study. In conclusion, the technical lignins analyzed in this study have different functional group contents. By combining elementary analysis and functional groups analysis, we can represent the expanded formulae C9 , which contains all the information about the structural environment of the lignins.

Acknowledgements The authors would like to thank Ligno-Tech Iberica, S.A., Santiago de Compostela University, the Centro de Investigaciones energ´eticas, medioambientales y tecnol´ogicas (CIEMAT) and Celulosa de Levante, S.A. (CELESA) for supplying the lignins. We would also like to express our sincere appreciation to the Rovira i Virgili University for their award of a scholarship, the Spanish Ministry of Science and Technology for providing finance under project number ENE200407624-C03-03, and the autonomous government of Catalonia also for providing finance under project number 2005SGR00580.

References Carmen, G.B., Dominique, B., Gosselink, R.J.A., Van Dam, J.E.G., 2004. Characterization of structure-dependent functional properties of lignin with infrared spectroscopy. Ind. Crops Prod. 20, 205–218. Chum, H.L., Johnson, D.K., Ratcliff, M., Black, S., Oiticica, B., Wallace, K., Schroeder, H.A., Robert, D., Sarkanen, K.V., 1985. Lignin characterization research: a process report. In: Biochemical conversion program Semi-annual Review Meeting. Solar Energy research Institute. Prepared for the U.S. department of Energy. Contract No. DE-AC02-83CH10093, 25–50. Dence, C.W., Lin, S.Y., 1992. General structural features of lignin. In: Lin, S.Y., Dence, C.W. (Eds.), Methods in Lignin Chemistry. Springer, Berlin, Heidelberg, pp. 3–6. Dence, C.W., 1992. Determination of carboxyl groups by non-aqueous potentiometric titration. In: Lin, S.Y., Dence, C.W. (Eds.), Methods in Lignin Chemistry. Springer, Berlin, Heidelberg, pp. 458– 463. El Mansouri, N.E., Salvad´o, J., 2006. Structural characterization of technical lignins for the production of adhesives: application to lignosulfonate, Kraft, Soda-anthraquinone, organosolv and ethanol process lignins. Ind. Crops Prod. 24, 8–16. Faix, O., 1992. Fourier transform infrared spectroscopy. In: Lin, S.Y., Dence, C.W. (Eds.), Methods in Lignin Chemistry. Springer, Berlin, Heidelberg, pp. 83–109. Faix, O., Andersons, B., Zakis, G., 1998. Determination of carbonyl groups of six round robin lignins by modified oximation and FTIR spectroscopy. Holzforschung 52, 268–274. Gosselink, R.J.A., Ab¨acherli, A., Semke, H., Malherbe, R., K¨auper, P., Nadif, A., Van Dam, J.E.G., 2004a. Analytical protocols of characterization of sulphur-free lignin. Ind. Crops Prod. 19, 271– 281. Gosselink, R.J.A., De Jong, E., Guran, B., Ab¨acherli, A., 2004b. Co-ordination network for lignin-standardization, production and applications adapted to market requirements (EUROLIGNIN). Ind. Crops Prod. 20, 121–129. Gosselink, R.J.A., Snijder, M.H.B., Kranenbarg, A., Keijsers, E.R.P., de Jong, E., Stigsson, L.L., 2004c. Characterization and application of NovaFiber lignin. Ind. Crops Prod. 20, 191–203. Lai, Y.Z., 1992. Determination of phenolic hydroxyl groups. In: Lin, S.Y., Dence, C.W. (Eds.), Methods in Lignin Chemistry. Springer, Berlin, Heidelberg, pp. 423–434. Lora, J.H., Glasser, W.G., 2002. Recent application of lignin: a sustainable alternative to nonrenewable materials. J. Polym. Environ. 10, 39–48. Lundquist, K., 1992. Proton (1 H) NMR spectroscopy. In: Lin, S.Y., Dence, C.W. (Eds.), Methods in Lignin Chemistry. Springer, Berlin, Heidelberg, pp. 242–247. Milne, T.A., Chum, H.L., Agblevor, F., Johonson, D.K., 1992. Standardized analytical methods. Part 2: Two-dimensional round robin on lignin analysis. Biomass Bioenergy 2 (1–6), 353– 366. Robert, D.R., Br¨unow, G., 1984. Quantitative estimation of hydroxyl groups in milled wood lignin from spruce and in dehydrogenation polymer from coniferyl alcohol using 13 C NMR spectroscopy. Holzforschung 38, 85–90. V´azquez, G., Antorrena, G., Gonz´alez, J., Freire, S., 1997. The influencing of pulping conditions on the structure of acetosolv eucalyptus lignins. J. Wood Chem. Technol. 17 (1 and 2), 147–162. Zakis, G.L., 1994. Functional Analysis of Lignins and Their Derivatives. Tappi Press, Atlanta, GA.