SciELO - Scientific Electronic Library Online

vol.11 issue5Electrochemical studies of copper in N-N, dimethylformamide in the presence of water, ethanol and acetic acid as additivesSupercritical fluid extraction and chromatographic analysis (HRGC-FID and HRGC-MS) of Lupinus spp. alkaloids author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Journal of the Brazilian Chemical Society

Print version ISSN 0103-5053On-line version ISSN 1678-4790

J. Braz. Chem. Soc. vol.11 n.5 São Paulo Sept./Oct. 2000 



Piassava Fibers (Attalea funifera): NMR Spectroscopy of their Lignin


Adilson R. Gonçalvesa*, Ulf Schuchardtb, Maria L. Bianchib and Antonio A. S. Curveloc

a Departamento de Biotecnologia, Faculdade de Enghenharia Química de Lorena,
CP 116, 12600-000, Lorena - SP, Brazil
b Instituto de Química, Universidade Estadual de Campinas, CP 6154,
13083-970, Campinas - SP, Brazil
c Instituto de Química de São Carlos, Universidade de São Paulo, CP 780,
13560-970, São Carlos - SP, Brazil



Lignina de piaçava (Attalea funifera) foi analisada por espectroscopia RMN de hidrogênio e carbono-13. A natureza HGS dessa lignina foi confirmada, mas unidades pcumarato tambem estão presentes na estrutura da lignina. Os conteúdos de hidroxilas fenólicas e metoxilas por unidade fenilpropânica, determinados por RMN de 1H, foram 0,57 e 0,68, respectivamente e estão em concordância com os dados obtidos por métodos clássicos. A pequena extensão da condensação da lignina (35%) foi explicada pelo alto conteúdo de estruturas de álcool cinamílico, evidenciadas por RMN de 13C-DEPT. Uma fórmula C9 estendida foi determinada para a lignina da piaçava:


Lignin of piassava (Attalea funifera) was analyzed by 1H- and 13C-NMR spectroscopy. The HGS-nature of this lignin was confirmed but p-coumarate units are also present in the lignin. Methoxyl and phenolic hydroxyl contents per phenylpropanic unit were determined by 1HNMR spectroscopy to be 0.57 and 0.68, respectively and are in agreement with wet chemistry methods. A small extent of lignin condensation (35%) was explained by the high content of cinnamyl alcohol structures, evidenced by 13CNMR-DEPT. An extended C9formula for the piassava lignin was established to be

Keywords: piassava fibers, lignin, NMR-spectroscopy




In a previous paper lignocellulosic material of piassava fiber (Attalea funifera, a palm tree) was characterized by chemical analysis and submitted to alkaline solvolysis in dioxane/water1. The results concerning the piassava lignin were remarkable. The Klason lignin content was unusually high (ca. 45%) and the distillable oils, obtained by alkaline solvolysis in 40% yield (based on the raw material), consisted mainly of phenol, with some guaiacol, syringol and their alkyl derivatives. Methoxyl group analysis of the lignin showed a content of 0.54 OMe for each phenyl-propane unit, suggesting that the lignin of piassava is rich in hydroxyphenyl units.

In this work piassava lignin was submitted to further in-depth analysis using NMR spectroscopy. 1H- and 13C-NMR spectroscopy has been used as an alternative and non-degradative technique for the characterization of lignocellulosics. As shown in the literature2, hydrogen signals can be assigned in a 1H-NMR spectrum of acetylated lignin. The assignments of carbon signals in a 13C-NMR spectrum of lignin can also be improved by means of pulse sequence procedures such as DEPT. In the recent literature these techniques were used for monitoring production or reduction of lignin in pulps3 and for characterizing residual and isolated lignins from pulps4. Other modern techniques such as 1H-13C correlation spectra have been used for the characterization of grass lignins5. Since piassava lignin has not been well characterized until now, we present here the results of the NMR analysis of this lignin.


Materials and Methods

Extraction of lignin from piassava fibers

Piassava fibers obtained from palm tree (Attalea funifera), used in brooms, were exhautively washed with water and the extraction was performed with dioxan: HCl 2 mol L-1 (9:1) at reflux temperature, for 5 h. After evaporation of the solvent, extracted lignin was obtained and characterized for molecular weight distribution, FTIR and elementar analysis as described elsewhere1.

Nuclear Magnetic Resonance

The extracted piassava lignin was acetylated by a 1:1 pyridine/acetic anhydride solution at 50oC for 24 h in a sealed flask6.

The 1H-NMR spectrum of the acetylated lignin was recorded on a Bruker 300 MHz. The spectrum was measured at 24oC with 32K data points and 128 scans, using 30 mg of the sample dissolved in 1 mL of CDCl3 with TMS as internal standard. Pulse of 3.5 ms and relaxation delay 1 s. The 13C-NMR spectrum of the non-acetylated lignin was recorded on a Bruker 300 MHz at 24oC with 32K data points and 19096 scans, using 30 mg of the sample dissolved in 1 mL of DMSO (TMS as internal standard). Total spectrum measured by decoupling 1H at acquisition time (0.88 s), using the POWGATE software (Inverse Gated Decoupling); irradiation pulse of 6.4 ms, relaxation time of 5 s. After the quantitative analysis, the DEPT (Distortionless Enhancement by Polarization Transfer) experiment was performed to assign the regions according to the type of carbon atom. The spectra were measured in the same spectrometer at 24oC with 32K data points and 11000 scans and acquisition time 0.95 s. Three basic subspectra were recorded with pulses of 45o, 90o and 135o and the CH, CH2 and CH3 pure subspectra were calculated and analyzed according to Chen and Robert2,7.


Results and Discussion

1H-NMR spectrum of acetylated piassava lignin

The analyses of molecular weight distribution and FTIR spectra of the extracted lignin published in the previous work, showed that no contaminants (such as tannins or other extractives) were present in the sample1.

The results of the semiquantitative analysis of the 1H-NMR spectrum are shown in Table 1, where the hydrogen signals and areas are subdivided into the corresponding regions.



From the elemental analysis of extracted piassava1 (54.5% C; 5.84% H; 0.52% N), and subtracting the holocellulose content (54.73% measured as C6H10O5), we obtained the following atomic ratio for the piassava lignin: C2.51 H2.42 O0.79, where the nitrogen content was neglected and the oxygen content calculated by difference. In lignin chemistry the empirical formula of the macro-molecule is commonly given as a hypothetical hydroxy-phenyl structure unit. This is known as the C9-formula, with 6 carbon atoms at the benzene ring plus three carbon atoms of the propyl side-chain. Besides these 9 carbon atoms, only the carbon atom of the methoxyl units (OCH3) is present in the macromolecule. For the calculation of the C9formula the first step is to determine the methoxyl content and subtract it from the total carbon. This determination is easily made either by wet chemistry or by integration of the hydrogen signal of the 1H-NMR spectrum, since the methoxyl region is well defined.

The integration of the hydrogen signals (Table 1) results in a total area of 203.2 arbitrary units which corresponds to the 2.42 protons in the above-given ratio. The integration of the methoxyl region (d 3.55-3.95) results in 0.15 methoxyl units (0.44 hydrogen atoms divided by 3 hydrogen atoms per methoxyl group). Subtracting this value from the total carbon and oxygen atoms as well as subtracting the hydrogen atoms corresponding to the methoxyl units, we can establish a new relation: C2.36 H1.98 O0.64 (OCH3)0.15 . Making the carbon atoms to 9.00, the C9-formula was calculated to: C9.00 H7.53 O2.44 (OCH3)0.57 (Table 1). The calculated value for OCH3 per C9 unit is 0.57 which is in agreement with the value of 0.54 obtained in the previous paper1. These values are also in agreement with those obtained by Berns et al.8 from analytical pyrolysis of piassava fibers. These authors found a H/G/S ratio of 51/40/9, which correponds to an average methoxyl content of 0.58 OCH3 / C9 unit.

The condensation degree of the lignin can be calculated from the methoxyl content and the 1HNMR spectrum. For each C9 unit a maximum of 4 aromatic hydrogen atoms is expected, since the carbons at positions 1 and 4 are linked to the propyl chain and to the phenolic hydroxyl/ether linkage, respectively. Some of the aromatic positions are occupied by methoxyl groups and for the piassava lignin a maximum of 3.43 aromatic hydrogens (4.00 minus 0.57 OCH3) is expected. The aromatic region (d 6.25 to 7.90) corresponds to 2.25 hydrogens per C9 unit, thus allowing the determination of the condensation degree (34%) by the difference to the maximum aromatic hydrogens. This value is similar to that of lignins obtained by steam explosion, which ranges from 38% to 41%9. The condensations are mostly due to 5-5 and b1 linkages between phenylpropanic units.

Determination of phenolic hydroxyl groups by NMR (0.69/C9, Table 1) is also in agreement with the chromatographic method (0.72/C9)1. Additionally, the aliphatic OH groups amount to 0.86 per C9 unit. Thus the C9 formula can be expanded to

The amount of aldehyde and acidic protons is negligible (signal above d 9). This corresponds to a small content of terminal g-carbonyl and carboxyl groups, showing that most carbonyl groups are located in a and b positions of the propyl chains.

13C-NMR spectrum of the piassava lignin

The 13C-NMR spectrum of the extracted piassava lignin is shown in Figure 2. It has characteristic H signals found at d 161.5, 131.3, 116.0 and 121.32 and also typical pcoumarate signals in the region around d 150. Furthermore these signals have the highest intensities confirming the high H content of this lignin, evidenced by other techniques, or the high incorporation of p-coumarate units as attached moieties, as described for other lignins10,11.





The non-aromatic region (d 50-103) has a small intensity. Since each C9unit contains 6 aromatic and 3 aliphatic carbons, the integration of the non-aromatic region is expected to amount to about half of that of the aromatic region. In order to verify the presence of aliphatic structures such as cinnamic acids in the aromatic region, a DEPT spectrum was recorded to separate CH, CH2 and CH3 signals (Figure 2). The aromatic region (d 103-160) is only contaminated by C-a and C-b of cinnamyl-type structures, and the integration of these signals (d 122-130.5 and d 151.5 to 154.5, respectively) must be considered to be part of the non-aromatic region (carbons of propyl side chain). With respect to the C9-formula the aromatic region must correspond to 6 carbon atoms. The total integration of this region (d 103 to 155) is 72.0 arbitrary units which corresponds to 12.0 arbitrary units/carbon atom. Table 2 shows the results of the application of this factor to all regions of the 13CNMR spectrum.



Some information is first obtained by the integration of the region between d 103-107, which corresponds exclusively to C2 and C6 in s units (Table 2) and has an integration of 2.5 arbitrary units corresponding to 0.21 C2 and C6 per aromatic ring. This is equivalent to 0.105 S units per C9 unit or 10.5% S units in piassava lignin. Analytical pyrolysis furnished a similar value of 9%8.

Methoxyl content is calculated by integration of the d 55 to 58 region, giving a total area of 8.0 arbitrary units, which correspond to 0.67 OMe/C9. This value is 24% higher than the 0.54 OCH3 per C9 unit found by wet chemistry methods1.

A close analysis was performed with respect to the aromatic region. Table 2 shows that the integration of CH signals (tertiary carbons) amounts to 2.75 carbons. A corresponding analysis of 1H-NMR spectrum revealed only 2.25 hydrogens for this region. A higher intensity for the tertiary carbons could be attributed to the Nuclear Overhauser Effect, although the difference is high (22%). The values obtained from the 1H-NMR analysis are accepted as being correct.

The aliphatic region containing cinnamyl structures has an unusual integration area, corresponding to 1.5 carbons per C9 unit. This accounts for the small conden-sation of this lignin. It seems that the original structures were preserved in the piassava lignin biosynthesis, in spite of its high H content.



1H- and 13C-NMR spectroscopy was used in this work for piassava lignin characterization and good results were obtained. Information about methoxyl (0.57 OCH3 per C9 unit) and phenolic hydroxyl (0.69 OHph per C9 unit) contents agree with those obtained by other methods. An extended C9-formula was obtained . The piassava lignin has a HGS nature, with high content of p-coumarate, and the presence of cinnamyl alcohol structures was evidenced by 13C-NMR. The degree of piassava lignin condensation is low (34%), which explains its high reactivity under basic hydrolysis conditions.



Financial support from FAPESP and CNPq are acknowledged. The authors thank M. E. M. Coelho for technical assistance.



1. Schuchardt, U.; Bianchi, M. L.; Gonçalves, A. R.; Curvelo, A. A. S.; Biscolla, F. C.; Peres, L. O. Cellul. Chem. Technol. 1995, 29, 705.         [ Links ]

2. Chen, C. L.; Robert, D. In Methods in Enzimology, 161, Wood, W. A.; Kellog, S. T., Eds.; Academic Press; New York, 1988, p. 137.         [ Links ]

3. Labidi, A.; Robert, D.; Pla, F. Holzforschung 1993, 47, 213.         [ Links ]

4. Pan, X.; Lachenal, D.; Neirinck, V.; Robert, D. J. Wood Chem. Technol. 1994, 14, 483.         [ Links ]

5. Galkin, S.; Ärmmälahti, E.; Kilpeläinen, I.; Brunow, G.; Hatakka, A. Holzforschung 1997, 51, 130.         [ Links ]

6. Chen, C. L. In Methods in Lignin Chemistry, Lin, S.Y.; Dence, C.W., Eds.; Springer Verlag; Berlin, 1992, p. 409.         [ Links ]

7. Robert, D. In Methods in Lignin Chemistry, Lin, S.Y.; Dence, C.W., Eds.; Springer Verlag; Berlin, 1992, p. 250.         [ Links ]

8. Berns, J; Meier, D. and Faix, O. Analytical pyrolysis of piassava, unpublished results, personal communication.

9. Silva, F. T., Ph. D. Thesis, State University of Campinas, Campinas, Brazil, 1995.         [ Links ]

10. Ralph, J.; Hatfield, D. J. Agric. Food Chem. 1991, 39, 1426.         [ Links ]

11. Ralph, J.; Hatfield, R.D.; Quideau, S.; Helm, R.F.; Grabber, J.H.; Jung, H. -J. G. J. Am. Chem. Soc. 1994, 116, 9448.         [ Links ]


Received: July 14, 1999
Published on the web: August 31, 2000

FAPESP helped in meeting the publication costs of this article.


* e-mail:

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License