Pyrolysis high-resolution gas chromatography-mass spectrometry ...

1 downloads 0 Views 458KB Size Report
(H. Tsai), the Cytogenctic Department (J. Hoppc) and the Max-Planck-. Institute for Experimental Medicine. Gottingen (B. Conolly and. F. Eckstein). Barber, M..
BIOCHEMICAL SOCIETY TRANSACTIONS

170 We are grateful for the help o f the G B F DNA Synthesis Group (H. Bliicker and R. Frank). the Enzymtechnology Department (H. Tsai), the Cytogenctic Department (J. Hoppc) and the Max-PlanckInstitute for Experimental Medicine. Gottingen (B. Conolly and F. Eckstein).

Barber, M.. Bordoli, R. S.. Elliott, G. J., Tyler, A. N.. Biel, J. & Green, B. N. (1984) Org. Muss Spec'/rom. I I . 182 189 Conolly. B. A.. Potter. B. V. L.. Eckstein, F.. Pingoud. A. & Grotjahn. L. (1984) Biochcwli,s/rj~23. 3443 3452 Fenselau, C., Yergey, J . & Heller, D . (1983) Inr. J . Muss Sptjctrom. /on Pl1y.Y. 53, 5 12 Grotjahn, L. & Steinert. H. (1985) in Mu.s.r Specrromerry in /hc H d r h uncl Li/c Sc,ic~ncw(Burlingame, A. L. & Castagnoli, N. Jr., eds.), pp. 597 61 5. Elsevier Scientific Publishers, Amsterdam Grotjahn. L. & Taylor, L. C. E. (1985) Org. Muss Spec/rom. 20, 146 159

Grotjahn, L.. Frank, R. & Bliicker. H . (1982) Nuc,/c,ic,Acids Ros. 10. 4671 4682 Le.//. 25, Grotjahn, L.. Frank. R. & Bliicker. H. (1984) Tc~/ruhi~c/ron 5373 5379 Grotjahn. L., Blocker. H. & Frank, R. (1985) Biomccl. Mu.vs Spc,c,/rom. 12. 514 526 Panico, M., Sindona, G . & Uccella. N . (1983) J . A m . ('licwi. SOC.105, 5607 5612 Richter, W. J.. Raschdorf. F. & Maerki. W. (1985) in MNSA Spc,c/romc'/rj, in /he Heulrh und Li/i, Scic~ncc~s(Burlingame. A . L. & Castagnoli, N . Jr.. eds.). pp. 193 208. Elsevier Scientific Publishers. Amsterdam Sindona, G., Uccella, N. & Weclawek; K . (19x2) J . Chc~m.Rc,s. Synop. 184 191 Soler. F. & Jankowski. K. (1985) Spe,i,/ro.vc, Inr. J . 4. 35 42 Velours. J.. Esparza, M.. Hoppe, J.. Sebald. W. & Guerin. B. (19x4) E M B O J . 3, 207 219 Received 9 July 1986

Pyrolysis high-resolution gas chromatography-mass spectrometry studies on beech wood: capillary high-resolution mass spectrometry of a beech lignin fraction JAAP J . BOON, ALOYS D. POUWELS and GERT B. EIJKEL FOM Institute ,for Atomic und Molecular Physics, Kruisluun 407, 1024 N A Amsterdum, The Netherlands Plant material and plant-derived organic matter fractions dominate the bio- and geo-cycles of the earth's ecosystem. Since ancient times they have played a central role in agriculture, energy, food technology and construction. Decomposed plant material contributes to sewage, soils and sediments and plays a role in the ecology of rivers, lakes, estuaries and oceans. Recently, interest has grown in the possibilities of manipulating the genetic material of plants and in trying to modify plant molecular architecture mainly for the purposes of the biotechnological industry. The understanding of the cell-wall structure in plant tissue is growing, although still mainly focused on the carbohydrate chemistry of the primary cell walls (Darvill et a/., 1980). The polyphenolic polymers such as lignin pose their own special analytical difficulties. There is a need for a general characterization method, which is applicable to all kinds of plant samples and which provides information on a molecular level. M.S. is an important general tool for the characterization of gases and liquids. Direct analysis of raw plant material or soil fractions is usually not possible without prior isolation and chemical workup. The analysis of solids such as plant cell-wall systems by m.s. is possible by prior thermal dissociation of the polymer system. Flash pyrolysis methods are generally applicable workup methods which can be combined with a g.c.-m.s. instrument for identification of the compounds or can be used without a chromatographic inlet for rapid fingerprinting. N o special requirements are necessary. except perhaps suitable sample size. In general, a large number of dissociation products characteristic of cellulose, hemicellulosic fractions (polyoses), polyphenolic polymer systems. e.g. lignin and tannin, and polyesters, such as cutin and suberin, are found in one analysis run of a sample. Apart from pyrolysis products, evaporation products, e.g. hydrocarbons, steroids, triterpenoids, can also be evaluated by regulation of the temperature of analysis. The complexity of these compound mixtures requires, on the one-hand, high-resolution capillary g.c. over a wide dynamic Abbreviations used: El. electron impact ionization; C1. chemical ionimtion: M WL. milled wood lignin.

range and, on the other hand, sophisticated m.s. Both electron impact (El) and chemical ionization ( I ) are used for the analysis of the thermal desorption and thermal dissociation products. High-resolution m.s., which allows the determination of the elementary composition of molecular ion and fragment ions of the compounds, is utilized for structural confirmation and for identification of unknowns. In this paper, we present data on beech wood and a soluble lignin fraction of beech wood. The milled wood vatica) was analysed by lignin (M W L) from beech (Fag pyrolysis capillary g.c.-m.s. un ynamic high-res?lution conditions.

Experimmtul Plunt materiuls. Beech wood and MWL were gifts from Dr. 0. Faix (Bundesforschungsanstalt fur Forst und Holzchemie, Hamburg, Federal Republic of Germany). The beech wood sample was solvent-extracted saw dust from Fugus syhuticu. The MWL was prepared from beech wood by milling for 160 h in water under a nitrogen atmosphere in a planetary ball mill. The purification procedure to isolate MWL has been described by Schweers & Faix (1973). Pyrolysis unif. Fig. 1 shows a schematic diagram of the pyrolysis unit designed by our group, which was used for the analysis of plant materials. The sample is applied from particulate suspension in water or from solution on to a ferromagnetic wire, which is inserted into a glass liner. This liner is placed into a pyrolysis unit (FOM 3-LX) and rests on a Kalrez interface directly above the entrace of the capillary column. The glass liner is surrounded by a ceramic tube kept at 160'C, which in turn is surrounded by a highfrequency coil. The unit is flushed with helium, which is also used as the carrier gas. The ferromagnetic wire is inductively heated to its Curie point within 0.1 s. Thermal energy is transferred to the sample, which evaporates and/or is pyrolysed. The volatile products are swept to the beginning of the capillary column. C'hromutogruphic conditions undni.s.. Beech wood and the MWL sample were both analysed at a Curie point temperature of 610 C (sample size about 30 x 10 'g). Pyrolysis g.c.-m.s. was performed with a Packard 438SGC and a JEOL DX-300/DA-5000 double focusing mass spectrometer with data sytem. Beech wood was analysed on a 50 m CP-SiI 5 CB fused silica capillary column (inner diam. 0.32mm, film thickness 1 pm). The MWL was analysed on a 60m DB-I fused silica capillary column (inner diam. I987

618th MEETING. LIVERPOOL

171 mined from the degree of separation between C, Hj,'C (benzene isotope) and C , H , N (pyridine) was about 8000. A continuous leak of P F K (Merck Chem. Co.) into the ion source was used for the calibration. The software in the J E O L data system uses P F K peaks for calibration of the sample mass peaks. The software package was used without modifications.

Carrier gas

Glass sample tube

,

Ferromagnetic wire with sample

Healed ceramic tube

High frequency coil

Interface ferrule

Fused silica capillary column

-1I Gas chromatograph

Fig. 1 . Schcrncric, diugrurn ofthi. FOM-3LX ('uric, point pj>rolp i s unit fOr splitkim injection of p y r o l j w t i , on c~upillrrj~ cdunins

0.32 mm, film thickness 0.2 pm). The column ended directly into the ion source. In LRMS-mode the g.c. oven was kept at 50°C during pyrolysis and subsequently programmed to 300°C a t a rate of 4"C/min. Compounds were ionized at 70eV. T h e acceleration voltage was 3 kV. The scan speed was 0.5 s/dec. Chemical ionization was performed in isobutane. In HRMS-mode the g.c. was programmed at a rate of 2"C/min. Ionization voltage was 70eV. the acceleration voltage was 3 kV. The scan speed was 4s/dec over a mass region from m/z 50 to 250. Dynamic resolution, deter-

_-10-

15

20

Rcsults und disc,ussion Figs. 2 and 3 show the partial total ion current traces of the pyrolysis -g.c.~-m.s.data of beech wood and its derived MWL. Peak numbers in these traces refer to the identified compounds listed in Table 1. The letter codes refer to pyrolysis products from polysaccharides, which are listed in the legend to Fig. 3. The beech wood sample was analysed under El and CI conditions. High-resolution data was obtained for the M W L pyrolysis products analysed under El conditions. Low-resolution spectra were compared with library spectra and spectra of standards. Elementary compositions of molecular ions and fragment ions of the pyrolysis products were derived from the high-resolution measurement. High-resolution data are reported in Table I as measured molecular ion o r fragment ion peaks with the accuracy b'riven in millimass units. The large dynamic range of the compounds due to their relative abundance and the large differences in intensity of fragmentation peaks in the spectra caused a higher inaccuracy in the more intense peaks. Practically all compounds detected with low-resolution m.s. could be detected with high-resolution m.s. despite the slower scanning rate of the magnet (g.c. conditions were adapted). Compounds with similar low-resolution spectra. e.g. syringylpropane (no. 50) and syringylethanol (no. 5 2 ) , which have to be identified in low-resolution m.s. by their relative retention time, were easily identified from their high-resolution m.s. data. Methoxydihydroxybenzene (no. 12), guaiacyl-prop-2-en- 1-one (no. 40) and 3-syringylprop-2-en-I-one (no. 62) have not been reported previously. Most of the identified compounds confirm earlier tentative identifications (Obst. 1983: Saiz-Jiminez & de Leeuw. 1986). A number ofcompounds. which remain as yet unidentified, are subject of further study. The spectra of nos. 29. 31, 50 and 5 I have many common characteristics. The elementary composition suggests C,H, side chain on guaiacyl o r syringyl unit. The furfural and the hydroxypentenolactone 1977). I t no. 3) are generated from xylanes (Ohnishi ('1 d., is possible that the acetic acid is released from xylanes as well because the beech xylanes are highly acetylated ( Fengel & Wegener, 1984).

7---

_ L

25

30

-

35

80

Fig. 2. Partitrl total ion cwrrrrit truce c?f'pj'r('lj'.~is-g.c,~ m.s. ckuto o f ' h i w l i M W L Peak numbers rclcr t o lignin pyrolysis products listed in Table I . VOl. 15

40

RT

I72

BIOCHEMICAL SOCIETY T R A N S A C T I O N S 30

100

35

40

_ - _ _ I

50

45

RT

55

Is

80 60

40 20 0

7

1000

1200

1400

1600

1800

2000

2200

2400

2600

2800

-

I -

3000

-

3200 3200.

3400 -3400

350 Scan

Fig. 3. Partial total ion current trace ofpyrolysis-g.c.-m.s. data of beech wood (Fagus sylvatica) analysed under low-resolution EI conditions Peak number refer to lignin pyrolysis products listed in Table 1. Polysaccharide pyrolysis products are indicated: A, unidentified xylane marker; B, 3-hydroxy-2-penteno-1.Slactone; C, 5-hydroxymethylfurfural; D, anhydroxylose; E, 1,4-didexoy-~-gluco-hexIenopyranos-3-ulose; F, anhydrohexose (Gal?); G , anhydroglucose (levoglucosan).

Table I . High-resolution data on milled wood beech lignin pyrolysis products P = phcnol unit; G = guaiacyl unit (C7 H7 02) = 2-methoxyphenol; S = syringyl unit (C8 H9 03) = 2.6-dimethoxyphenol. Commonly used trivial names: eugenol (no. 20), vanillic aldehyde (no. 22). isoeugenol (no. 27). acetovanillone (no. 30). acetosyringone (no. SS), conifcrylalcohol (no. 5 3 ) , sinapylalcohol (no. 66)*, base peak in m.s. Peak no.

I

-7 3 4

Identity Acetic acid Furfural 3-hydroxy-2-pentenoI .5-lactone Phenol

Peak area 329 9 6 3 4 6 I30

Measured mass

* 94.0404

-

1.5 -0.1

C6 H6 0 C5 H6

P H

2.9 I .8 - 2.4

C7 HX 0 2 C7 H8 0 2 C6 H5 0 2

P C G H

3.9 1.0 - 1.3 1.9

C8 C8 c7 C6 C7 C6

1 .0

C9 HI2 0 2

G C C

- 5.7

C9 H I 0 0 2 cx H7 0 2

G C=C

C8 H I 0 0 3 C7 H7 0 3 CI0 HI2 0 2 C9 H9 0 2

S H

1.2 - 3.3 1.6 7.5 4.4 2.9

C8 H8 0 3 C8 H7 0 3 CI0 HI2 0 2

1.8 2.2 0.4 - 2.6 - 3.4 - 0.2 0.9

C9 H I 2 0 3 C8 H9 0 3 C9 HI0 0 3 CX H9 0 2 CI0 HI2 0 2 C9 H9 0 2 C9 HI0 0 3

sc

C9 H I 0 0 3 C8 H7 0 3

G CO C

*

X

9 I0

Mcthylmethoxyphenol Mcthylmethoxyphenol Guaiacyl-methane

6 30

*

II I2

Dihydroxyhenzene Mcthoxydi hydroxyhenzene

14 42

* *

13 14

Methoxydihydroxyheniene Guaiacyl-ethane Methoxydihydroxybcnzcne Unknown ( F : nr z 153. 125) Guaiacyl-ethene

15

16 17 IX 19

Unknown ( F : tw: 109. 107) Syringol

*

9

5 312

20

3-Guaiacylprop- I -cnc

21 22

Guaiacylpropane Guaiacy lpropane

23 24 25

3-Guaiacylprop-2-ene ( c i s ) Unknown ( F : mi= 131) Syringyl-methane

18 6 69

26

Guaiacylcthanal

42

27

3-Guaiacylprop-2-ene ( rruns)

83

2X 29 30

Dimethoxybcnzaldehyde Spectrum same as no. 31 Guaiacyl-ethanone

3 3 72

33

*

* *

2

I I3

Structure

-

*

0-Methylphenol p-Methylphenol Guaiacol

14 3 2 101

Elementary composition

60 96.0198 114.0298

5 6 7

15

Accuracy

*

*

* * * *

*

66.0468 108 108.0604 124.0543 109.0266 I38 138.0720 138.0671 123.0433 110.0357 140.0448 125.0257 124 152.0848 I24 I68 150.0604 135.0420 I38 154.0604 139.0383 164.0804 149.0618 166 152.0549 151.0549 164.0866 I78 168.0768 153.0574 166.0634 137.0576 164.0803 149.060I 166.0638 162 166.0652 151.0380

-

1.4 1.9

-

1.1

- 2.5

- 2.6 - 2.6 -

-

2.2 - 2.0

C5 H5 0 2 C5 H6 0 3

HI0 0 2 HI0 0 2 H7 0 2 H6 0 2 HX 0 3 H5 0 3

G C P OH G OH

G C C=C

G C C C G CHO G C = C C (c)

G C CHO G C = C C(1)

1987

173

618th MEETING, LIVERPOOL Table I . Continued Peak no. 31

Identity

Structure unknown

32 33

Unknown ( F : nriz 151) Vanillic acid methyl ester

Peak area 16 5 8

34 35

Unknown ( F : mi:: 179. 136) 3-Guaiacylpropan-2-one

3 20

36

Syringyl-ethane

26

37 38 39

Unknown ( F : nil; 149. 117) Unknown ( F : ndz 163. 147. 131) Unknown ( F : mlz 165) Syringyl-ethene

218

Guaiacylprop-2-en- I -one

40

41 42 43

Guaiacylpropandione Unknown ( F : nil: 181, 125) 3-Syringylprop-I -ene

4 9 60

Syringylpropanc

45 46

Unknown (mi: 71, M??) Syringylaldeh yde

47

3-Syringylprop-2-ene (cis).

4 9 27 I 77

*

*

* * * * * * *

Spectrum similar to no. 66 48

3-Guaiacylprop-2-en01 (cis)

35

49 50

Unknown ( F : nl’; 195) Unknown structure

5 31

*

51

17

*

52

Unknown structure. Spectrum similar to no. 50 Syringylethanal

53

3-Syringylprop-2-ene (trans)

54

3-Guaiacylprop-2-enaI

55

Syringylethanone

101

56

3-Guaiacylprop-2-enol (trans)

322

57

Syringylpropan-2-one

40

58

Syringic acid methyl ester

28

59

Unknown structure

60 61

Unknown structure ( F : miz 182. 167) Syringylpropandione

62

3-Syringylprop- I -en-3-one

46

63

3-Syringylpropan- 1-01

14

64

3-Syringylprop-2-en- 1-01 (cis)

52

65

3-Syringylprop-2-enal

I33

66

3-Syringylprop-2-en-l-ol(/runs)

295

Vol. 15

55 I66

*

* *

70

7

* *

* * *

11

15

Accuracy

*

* * *

* *

147.0452 I80 182.0621 151.0380 I94 180.0862 137.0630 182.09I3 167.0699 I78 178 I80 180.0753 165.0481 178.0578 151.0402 194 196 194.0937 167.068I 196.1093 167.0722 ? 182.0563 18 1.0466 194.0857 179.0745 208.0826 I6 1.0574 180.0791 137.0603 2 10.0934 192.0802 177.0563 192.0770 177.0558 196.0714 167.0692 194.0904 179.0715 178.0600 163.0366 196.7680 181.0508 180.0791 137.0544 21 0.0874 167.0706 2 12.0605 181.0488 208.0715 16 I ,0642 210.0808 182.0520 224.0695 181.0448 208.0737 181.0506 2 12. I091 168.0737 210.0832 167.0716 208.0743 165.0504 2 10.0907 167.0723

Elementary composition

Structure

6.4 0.5

CIO HI0 0 2 C9 H7 0 2

G C=C=C’J

4. I 1.5

C9 HI0 0 4 C8 H7 0 3

G COO C

7.6 2.7 - 3.0 0.7

CI0 HI2 0 3 C8 H9 0 2 CIO HI4 0 3 C9 H I 1 0 3

G-C CO C

- 3.3

CIO HI2 0 3 C9 H9 0 3 CIO HI0 0 3 C8 H7 0 3

s c=c

* 162.0745

4 10

40

44

Measured mass

-

~

7.0 - 5.2 0.7

S C C

G CO C = C

G CO CO C ~

~

~

-

0.6 2.7 0.6 I .4

C11 H I 4 0 3 C9 HI1 0 3 C11 H I 6 0 3 C9 HI1 0 3

sC

1.6

C9 HI0 0 4 C9 H9 0 4 C11 HI4 0 3 C10 HI1 0 3 CII H I 2 0 4 CIO H9 0 2 CI0 HI2 0 3 C9 H9 0 2 CII H I 4 0 4 CI1 H I 2 0 3 CIO H9 0 3 CII H I 2 0 3 CIO H9 0 3 C10 HI2 0 4 C9 HI1 0 3 C10 HI4 0 3 CIO HI1 0 3 CIO H10 0 3 C9 H7 0 3 CIO HI2 0 4 C9 H9 0 4 CIO HI2 0 3 C8 H9 0 2 CII H I 4 0 4 C9 HI1 0 3 CIO HI2 0 5 C9 H9 0 4 C11 H12 0 4 CIO H9 0 2 C1 I H I 4 0 4 C9 H I 0 0 4 C11 H12 0 5 C9 H9 0 4 C11 H12 0 4 C9 H9 0 4 C11 HI6 0 4 C9 HI2 0 3 C11 HI4 0 4 C9 HI1 0 3 CII HI2 0 4 C9 H9 0 3 C11 H I 4 0 4 C9 HI1 0 3

S CHO

- 3.5 - 8.6

3.7 9 - 2.9 0.5 0.0 4.2 1.6 1.1 - 1.7 0.6 2.2 - 1.6 - 3.9 0.7 2.9 - 2.9 3.3 0.7 0.5 - 5.9 1.8 - 0.2 - 6.3 - 1.3 - 2.0 3.9 8.4 - 6.0 ~

~

~

~

1 .o

- 1.6

0.1 0.5 4.3 - 4.9 0.3 0.7 0.8 - 4.8 I .5 1.4

C=C

sccc

s C = C c (c) G C = C C OH

s C=C=C? S C CHO

s c = c c (t) G-C

= C-CHO

s CO c G C = C C OH s-C-co-c

s coo C

s CO CO C s co-C=C S C C C O H

S-C S-C

= C-C-OH

C-CHO

1

S-C=C-C OH

BIOCHEMICAL SOCIETY TRANSACTIONS

I74

Bwch M W L . The MWL is a dioxane-soluble lignin fraction from the wood, which does not necessarily reflect the whole lignin in the wood. The preparation is not entirely free from xylanes (approx. I YOimpurity). Its exact structure is not known, but models which explain most of the spectroscopic and chemical degradation data have been proposed (Nimz. 1974) Beech lignin is a dehydrogenative polymer of trunx-coniferyl and trans-sinapyl alcohol, which have undergone structural changes due to the coupling processes, displacement. addition and elimination reactions. At least ten intermonomeric linkage types exist. For lignin classification, chemical degradation methods are applied to determine the ratios between hydroxyphenyl, 2-methoxyhydroxyphenyl and 2,6-dimethoxyhydroxyphenyl monomeric units. Curie point pyrolysis is a dissociation method which activates the lignin polymer system by thermal energy. This energy is distributed over the various structural elements inducing dissociation of certain bonds and radical formation. Quenching of these phenolic radicals with hydrogen radicals results in a complex mixture of phenolic compounds which reflect the original linkages and the monomeric units involved. Presently, there is no real understanding about the mechanisms nor about the types of compounds which are generated from the various linkages. The major pyrolysis products in Table 1 are guaiacol, guaiacylethene, guaiacylaldehyde, coniferylaldehyde, coniferylalcohol, syringol, syringylethene, syringylaldehyde, rrLins-syringylprop-2-ene,syringylethanone, sinapylaldehyde and sinapylalcohol. Schemes for the formation of these compounds from 8-0-4- and p-I-linkages, and from phenylpropanoid units, have been proposed (Genuit & Boon, 1987)). The presence of the coniferyalcohol and sinapylalcohol, which are known to be highly thermosensitive (0.Faix, personal communication). points to relatively mild dissociation conditions in our pyrolysis unit. It is thought that these compounds are present in the original lignin as such, most probably bound by an ether linkage to the phenolic hydroxyl group. Very little disproportionation is observed to simple phenols, contrary to the data of Martin et (I/. (1979). The detection of methylesters of phenolic acids (nos. 33 and 58 in Table 1) may point to methylation reactions during pyrolysis. No dimeric units were detected, although they have been reported (Haider & Schulten, 1985) in pyrolysis experiments on lignin in the ion source of the mass spectrometer, but this is most likely caused by the use of a gas chromatograph as inlet. Beech M . O O ~ .Chemical analysis of beech wood (Fugus .sylvaticu) using classical methods of wood analysis yields 75% holocellulose, 43% cellulose, 30% polyoses, 20% pentosan and 22% lignin (Fengel & Wegener, 1984). The 0acetyl-4-0-methylglucuronoxylan, a common polyose in hardwood, is also an important constituent of beech wood. The xylose content of beech wood is reported to be 19%, with a 4.8% 4-0-methylglucuronic acid in the same sample (Fengel et ul., 1978). Polyoses are thought to be bound to lignin through the polyose side groups arabinose, galactose and 4-0-methylglucuronic acid by ether, ester and glycosidic linkages. The lignin in beech wood is a mixed guaiacyl/ syringyl lignin with a guaiacylisyringyl ratio of 4 / 5 6 (Nimz, 1974) (oxidation method). Pyrolysis-g.c.-m.s. of data on beech wood shows marker

compounds for xylanes, cellulose and lignin. The anhydroxylose (D) and the 3-hydroxy-2-penteno-1,5-lactone (B) are considered as xylane markers. Cellulose markers are the anhydroglucose (levoglucosan) (G), 1 ,Cdidexoy-u-glucohex- 1 -enopyranos-3-ulose (E) and several less specific compounds such as pyranones and furan derivatives (Shafizadeh, 1984; Van der Kaaden ('t a/., 1983). The presence of another anhydrohexose (F) with a shorter retention time than levoglucosan but with similar mass spectrometric characteristics points to another hexose marker (galactose ?). The highly branched galactan with rhamnose, arabinose, galacturonic acid and 4-0-methylglucuronic acid in beech wood in a possible origin of this compound. The lignin markers in the beech wood pyrolysate are a number of mono- and di-methoxyphenolic compounds. The coniferylalcohol and the synapylalcohol are perhaps the most characteristic dissociation products because they are the precursors from which lignin is formed. Sinapylalcohol is more abundant than coniferylalcohol. Dimethoxyphenolics are in general more predominant than the corresponding monomethoxy compounds in this pyrolysate. The major compounds are syringol, syringylethene, syringylaldehyde and the syringylpropenes. The abundance of the syringylethene and syringylaldehyde may be a reflection of the relatively high degree of [j-I-linkages in beech wood. The higher abundance of the various phenolic aldehydes could be caused by stronger dehydrogenating conditions when whole wood is pyrolysed. However, it is tempting to conclude that the lignin in the beech wood is richer in dimethoxy units than its derived MWL. Conclusion Combined Curie point pyrolysis-capillary g.c. ~-m.s.is an interesting microanalytical characterization method for plant material and its derived fractions. This work is part of the research programme of the Fundamenteel Onderzoek der Materie ( F O M ) with financial support from the Nederlandse Organisatie voor Zuiver Wetenschappelijk Onderzoek (ZWO). Darvill, A,. McNeil, M., Albersheim. A. & Delmer, D. P. (1980) in Thc Bioc~hmi.sri-yofPlunrs (Stumpf. P. K. & Conn, E. E., eds.). vol. I . pp. 92 163, Academic Press, New York Fengel, D. & Wegener, G. ( 1984) Wood: Chc,mis/ry. U//ru.s/ruc/urc,. Rcuctions. de Gruyter Vcrlag, Berlin Fengel, D., Wegener, G . , Heizmann, A . & Przylcnk, M. (1978) C d . C h m i . Tc.c,hnol. 12, 31 37 Genuit, W. J. L. & Boon, J. J . (1987) A n d . Chcm. in the press Haider. K. & Schulten. H-R. (1985) J . A n d . A p p / . f.vrol,vsi.s8. 317 326 Martin. F., Saiz-Jimenez. C . & Gonzalez-Vila, F. J. (1979) Ho/;fi)rscliung 33, 2 I&2 I 8 Nimz, H. H. (1974) Angm.. Chem. I n / . Ed Engl. 13, 313 323 Obst, J. R. (1983) Wood Chem. Tcchnol. 3. 377 381 Ohnishi, A,, Kato. K . & Takagi. E. (1977) Curhohydr. Res. 58. 387 394

Saiz-Jimenez. C. & de Leeuw, J. W. (1986) Org. Geochem. in the press. Schweers. W . & Faix, 0. (1973) Ho/;Jorschung 27. 208 212 Shafizadeh. F. (1984) in The Chemisrry / J / S ~ l iWood(Rowel1, d R., cd.), pp. 489 53 I , American Chemical Society, Washington Van der Kaaden, A,, Haverkamp, J., Boon. J . J. & d e Lecuw. J . W. ( 1983) J . A n d . A p p l . Pyro/ysi.s 5, 51 5 527 Received 9 July 1986

1987