Photoyellowing of peroxide-bleached lignin-rich pulps: A photochemical study on stilbene-hydroquinone chromophores issued from β-5 units of lignin during refining and (or) bleaching

Article abstract of DOI:10.1139/v99-230

The contribution of stilbene-parahydroquinone structures, formed from β-5 units during refining and bleaching, in photoreversion of peroxide- bleached high-yield pulps, was studied using 4'-benzyloxy-2,5-dihydroxy- 3,3'-dimethoxystilbene (I), as a model c

Full text of DOI:10.1139/v99-230

73
Photoyellowing of peroxide-bleached lignin-rich
pulps: a photochemical study on stilbene–
hydroquinone chromophores issued from β-5 units
of lignin during refining and (or) bleaching
Brigitte Ruffin and Alain Castellan
Abstract: The contribution of stilbene–parahydroquinone structures, formed from β-5 units during refining and bleach-
ing, in photoreversion of peroxide-bleached high-yield pulps, was studied using 4′-benzyloxy-2,5-dihydroxy-3,3′-
dimethoxystilbene (I), as a model compound. The various photoproducts formed during irradiation of I in solution or
adsorbed on cellulose paper were identified. Isomerization of the stilbene double bond and oxidation of the
hydroquinone into quinone were observed in solution, whereas in the solid state only the formation of the quinone was
found to operate. Our attempts to detect and quantify stilbene–hydroquinones forming from β-5 units created on perox-
ide bleached mechanical pulp were not met with success.
Key words: lignin, stilbene, hydroquinone, lignin model compound, pulp, photoyellowing.
Résumé : La contribution des structures stilbène–parahydroquinone, formées à partir des unités β-5 durant le raffinage
et le blanchiment de la pâte, au photojaunissement des pâtes mécaniques blanchies au peroxyde d’hydrogène a été
étudiée à l’aide du 4′-benzyloxy-2,5-dihydroxy-3,3′-diméthoxystilbène (I). Les photoproduits formés sous irradiation de
I en solution ou adsorbé sur papier de cellulose ont été identifiés. L’isomérisation de la double liaison stilbène ainsi
que l’oxydation de l’hydroquinone en quinone ont été observées en solution, tandis qu’en phase solide, seule la forma-
tion de la quinone est notée. La détection et le dosage des unités stilbène-hydroquinone dans la pâte mécanique
blanchie ne se sont pas révélés concluants.
Mots clés : lignine, stilbène, hydroquinone, composé modèle de la lignine, pâte, photojaunissement. Ruffin and
Castellan 83
Introduction
(11). All these chromophores, hydroquinones in particular,
were considered to have a key role in the photoyellowing of
bleached mechanical pulps because hydroquinones, non-
colored compounds, are easily oxidized into quinones (12),
(very colored species). Lee and Sumimoto (13) suggested
that parahydrobenzoquinone derivatives including a stilbene
or biphenyl or diphenyl ether conjugated structures were the
major chromophores responsible of the color reversion of
pulps.
In a recent study (14), we performed a semi-quantitative
evaluation, in peroxide-bleached high-yield pulps, of p-stilbene
phenols formed from β-1 units and demonstrated that they
are not the main contributors of the rapid photoyellowing of
these pulps. The present article describes a similar approach
on the photochemical behavior of chromophores including
both parahydroquinone and stilbene entities. In that respect,
the photochemical behavior of model I (Fig. 1) before and
after catalytic hydrogenation and (or) acetylation was estab-
lished and also a tentative evaluation of the content of this
type of structure in peroxide-bleached softwood mechanical
pulps was carried out.
High-yield pulps are obtained by mechanical refining of
wood. This treatment causes some chemical modifications in
the pulp and new chemical structures are formed. For exam-
ple, phenylcoumaran (1, 2) and β-1 units (3) are turned into
stilbenes by mechanochemistry, β-O-4 entities are converted
into parahydroxybenzaldehyde (4), and coniferyl alcohols
create α, β-unsaturated groups (5). Also acidic treatment
generates stilbenes and phenylcoumarones (6) very sensitive
to UV light (7, 8).
When lignin-rich pulps are bleached, new chromophores
very sensitive to UV light are created. For example by the
action of alkaline H2O2 bleaching solution, β-1 diol units are
turned into stilbenes (9), α-carbonyl phenols were converted
into hydroquinones (Dakin reaction) (10). In the same way,
coniferaldehydes are transformed into methoxyhydroquinones
Received August 6, 1999.
B. Ruffin and A. Castellan.1 Laboratoire de Chimie des
Substances Végétales, Institut du Pin, Université Bordeaux 1,
351, cours de la Libération, F-33405 TALENCE Cedex,
France.
Experimental
1Author to whom correspondence may be addressed.
Telephone: (33) 5 56 84 62 80. Fax: (33) 5 56 84 64 22.
e-mail: a.castellan@ipin.u-bordeaux.fr
1. General
Melting points were measured with a Mettler FP62 heat-
ing block. NMR spectra were recorded on a Bruker DMP
Can. J. Chem. 78: 73–83 (2000)
© 2000 NRC Canada

74
Can. J. Chem. Vol. 78, 2000
Fig. 1. Formulae of the studied compounds.
200 (1H, 13C) spectrometer (reference Me4Si, solvent CDCl3).
The IR and UV absorption spectra were respectively per-
formed on a Perkin–Elmer Paragon 1000 PC and an Hitachi
U-3300 spectrometers. Mass spectra were obtained using a
VG Micromass Autospec Q instrument.
The chemicals were obtained from Aldrich and were used
without further purification. The synthesized compounds
were purified by column chromatography on silica gel 60
(70–200 mesh) using the appropriate eluents. The solid
products were crystallized after chromatography.
16 mmol) was added to a suspension of the phosphonium
salt 2 (4.2 g, 8.0 mmol) in anhydrous THF (50 mL) under
N2 atmosphere and the mixture was stirred for 30 min at
0°C. The resulting red solution was treated with a solution
of aldehyde 1 (1.6 g, 8.0 mmol) in anhydrous THF (40 mL).
After being stirred for 3 h at room temperature, the THF was
removed under vacuum, the residue was dissolved in CH2Cl2
and treated with hydrochloric acid (10%). The aqueous layer
was extracted with CH2Cl2 (3 × 50 mL). The combined or-
ganic layers were washed with water, dried over MgSO4, fil-
tered, and concentrated. The residue was purified by
chromatography on silica gel (eluent: CH2Cl2–EP) affording
the expected E stilbene 3 as white crystals (1.15 g, yield
37%, mp 133°C). IR (KBr) ν : 950, 1010, 1080, 1140, 1260,
2. Syntheses
2.1 Stilbene derivatives
1
1520, 1600, 1690, 3400 cm–1. H NMR (CDCl3) δ: 3.96
4′-benzyloxy-2,5-dihydroxy-3,3′-dimethoxystilbene (I)
(s, 3H, OCH3), 3.99 (s, 3H, OCH3), 5.19 (s, 2H, OCH2), 6.50
(s, 1H, OH), 6.88 (d, 1H, J = 8.24, H5′), 7.04 (dd, 1H, J = 1.83
and 8.24, H6′), 7.13 (d,1H, J = 1.83, H2′), 7.2–7.5 (m, 8H,
ArH), 7.71 (d, 1H, J = 1.52 H4 or H6), 9.88 (s, 1H, CHO)
ppm. 13C NMR (CDCl3) δ: 56.1 (OCH3), 56.4 (OCH3), 71.0
4′-benzyloxy-5-formyl-2-hydroxy-3,3′-dimethoxystilbene (3):
Compounds 1 and 2 were prepared from vanillin by conven-
tional methods respectively described in ref. 15 and 16. A
n-butyl lithium solution in n-hexane (conc = 2.5 mol L–1,
© 2000 NRC Canada

Ruffin and Castellan
75
Scheme 1. Formation of stilbenes via mechanochemistry and then quinones by photooxidation (1, 2).
(OCH2), 106.9 (CH ar.), 109,6 (CH ar.), 113.9 (CH ar.),
114.6 (CH ar.), 119.7 (CH ar.), 120.1 (=CH), 124.1 (C ar.),
127.3 (CH Bn), 127.9 (CH Bn), 128.6 (CH Bn), 129.2
(C ar.), 130.6 (=CH), 130.8 (C ar.), 134.2 (C Bn), 147.3
(C ar.), 148.3 (C ar.), 148.6 (C ar.), 149.8 (C ar.), 191.2
(CHO) ppm.
1.2 mmol), and 4-dimethylaminopyridine (DMAP, 5 mg) for
4 h. The reacting mixture was treated with hydrochloric acid
(10%) and extracted with dichloromethane. The organic
layer was washed with water to neutrality, dried over mag-
nesium sulfate and evaporated under reduced pressure. The
reacting mixture, separated by chromatography (eluent:
CH2Cl2–Et2O, 95:5), followed by crystallization in ethanol
afforded the expected E stilbene III as white crystals
(204 mg, yield 93%, mp 158°C). IR (KBr) ν : 1030, 1080,
1140, 1180, 1210, 1270, 1510, 1590, 1760 cm–1. UV λmax
(MeOH)/nm : 320 (ε/L mol–1 cm–1 29 000). 1H NMR
(CDCl3) δ: 2.31 (s, 3H, CH3), 2.36 (s, 3H, CH3), 3.81 (s, 3H,
OCH3), 3.93 (s, 3H, OCH3), 5.18 (s, 2H, OCH2), 6.66
(d, 1H, H4, J = 2.75 Hz), 6.87 (d, 1H, H2′, J = 8.24 Hz), 6.91
(d, Hα, J = 15.26 Hz), 6.96–7.05 (m, 3H, H6, H6′, and H5′),
7.40 (d, Hα′, J = 15.26 Hz), 7.3–7.5 (m, 5H, ArH) ppm.
13C NMR (CDCl3) δ: 20.5 (CH3), 21.2 (CH3), 56.1 (OCH3),
56.2 (OCH3), 71.0 (OCH2), 104.9 (CH ar.), 110.1 (CH ar.),
110.2 (CH ar.), 113.9 (CH ar.), 119.4 (CH ar.), 119.9 (=CH),
127.2 (CH Bn), 127.9 (CH Bn), 128.6 (CH Bn), 130.5
(C ar.), 131.8 (=CH), 131.9 (C ar.), 135.1 (C ar.), 137.0
(C Bn), 148.5 (C ar.), 148.7 (C ar.), 149.7 (C ar.), 168.7
(C=O), 168.8 (C=O) ppm. MS (EI) m/z (%): 462 (M+., 65),
432 (12), 371 (30), 329 (76), 299 (28), 287 (31), 255 (54),
227 (12), 91 (100), 43 (66).
4′-benzyloxy-2,5-dihydroxy-3,3′-dimethoxystilbene (I): A
solution of compound 3 (600 mg, 1.4 mmol) in THF–water
(30mL and 8 mL) under nitrogen atmosphere was treated
with sodium carbonate (126 mg, 1.4 mmol) and oxygen per-
oxide (240 µL, 2.3 mmol). After being stirred at room tem-
perature for 3 h, the reaction mixture was reacted with 1 mL
of acetic acid. The aqueous layer was extracted with CH2Cl2
(3 × 30 mL). The combined organic layers were washed
with water, dried over MgSO4, filtered, and concentrated.
The residue was purified by crystallization (CH2Cl2–EP)
affording the expected E stilbene I as white crystals
(360 mg, yield 62%, mp 150°C). UV λ max (MeOH)/nm : 320
(ε/L mol–1 cm–1 19 600). IR (KBr) ν : 960, 1000, 1080,
1
1130, 1260, 1510, 1600, 3400 cm–1. H NMR (CDCl3) δ:
3.87 (s, 3H, OCH3), 3.94 (s, 3H, OCH3), 4.46 (s, 1H, OH),
5.14 (s, 2H, OCH2), 5.55 (s, 1H, OH) 6.36 (d, 1H, H4,
J = 2.75 Hz), 6.60 (d, 1H, H6, J = 2.75 Hz), 6.85 (d, 1H, H5′,
J = 8.22 Hz), 6.98 (dd, 1H, H6′, J = 2.15 and 8.22 Hz), 7.00
(d, 1H, Hα or Hα′, J = 16.45 Hz), 7.11 (d, 1H, H2′, J = 2.15 Hz),
7.24 (d, 1H, Hα or Hα′, J = 16.45 Hz), 7.25–7.6 (m, 5H,
ArH) ppm. 13C NMR (CDCl3) δ: 56.1 (OCH3), 56.2 (OCH3),
71.0 (OCH2), 98.7 (CH ar.), 103.4 (CH ar.), 109.5 (CH ar.),
114.0 (CH ar.), 120.0 (CH ar.), 120.8 (=CH), 123.8 (C ar.),
127.3 (CH Bn), 127.9 (CH Bn), 128.6 (CH Bn), 129.3
(=CH), 137.1 (C Bn), 137.6 (C ar.), 147.3 (C ar.), 148.0
(C ar.), 148.7 (C ar.), 149.8 (C ar.) ppm. MS (EI) m/z (%):
378 (M+., 65), 287 (100), 255 (14), 227 (14), 91 (48). HPLC–MS
(solvent acetonitrile, negative mode) m/z: 377 (M – H)–.
(Z)-4′-benzyloxy-2,5-dihydroxy-3,3′-dimethoxystilbene (IZ):
By irradiation with UV light (λ ≈ 350 nm), compound I
(50 mg, 0,13 mmol) in methanol (10 mL) solution degassed
by bubbling nitrogen was converted to compound IZ
([IZ]:[I] = 90:10. The NMR data on IZ were obtained on
the mixture after evaporation of the solvent and after sub-
tracting the contribution of I. The electronic absorption
spectrum of IQ was obtained with the diode array detector
of the HPLC system (eluent: methanol). The mass spectrum
of IQ was obtained using HPLC–MS technique with
acetonitrile as eluent for the HPLC and electrospray for the
ionization mode. UV λ max (MeOH): 300 nm. 1H NMR
(CDCl3) δ: 3.63 (s, 3H, OCH3), 3.81 (s, 3H, OCH3), 5.09
(s, 3H, OCH2), 6.27 (d, 1H, H4, J = 2.44 Hz), 6.35 (d, 1H,
2,5-diacetoxy-4′-benzyloxy-3,3′-dimethoxystilbene (III):
Compound I (200 mg, 0.43 mmol) in dichloromethane
(40 mL) was refluxed under nitrogen in presence of acetic
anhydride (1.2 mL, 1.2 mmol), triethylamine (0.17 mL,
© 2000 NRC Canada

76
Can. J. Chem. Vol. 78, 2000
1
H6, J = 2.44 Hz), 6.55 (s, 2H, CH=CH), 6.76 (s, 1H), 6.77
(d, 1H, J = 1.83 Hz), 6.88 (d, 1H, J = 1.53 Hz), 7.25–7.45
(m, 5H, ArH) ppm. 13C NMR (CDCl3) δ: 55.7 (OCH3), 56.1
(OCH3), 70.9 (OCH2), 98.8 (CH ar.), 106.9 (CH), 112.2
(CH), 113.4 (CH), 122.0 (CH), 122.9 (CH), 123.3 (CH),
127.3 (CH Bn), 127.4 (CH Bn), 127.9 (CH Bn), 128.6
(C ar.), 129.3 (C ar.), 130.2 (C ar.), 130.7 (C ar.), 137.1 (C ar.),
137.3 (C ar.), 147.4 (C ar.), 148.4 (C ar.) ppm. HPLC–MS
(solvent acetonitrile, negative mode); m/z: 377 (M – H)–.
2930 cm–1. H NMR (CDCl3) δ: 2.28 (s, 3H, CH3), 2.33
(s, 3H, CH3), 2.78 (s, 4H, CH2-CH2), 3.80 (s, 3H, OCH3),
3.86 (s, 3H, OCH3), 5.14 (s, 2H, OCH2), 6.54 (d, 1H, H4,
J = 2.75 Hz), 6.61 (d, 1H, H6, J = 2.75 Hz), 6.65 (dd, 1H, H6′,
J = 1.83 and 8.24 Hz), 6.69 (d, 1H, H2′, J = 1.83 Hz), 6.82
(d, 1H, H5′, J = 8.24 Hz), 7.25–7.50 (m, 5H, ArH) ppm.
13C NMR (CDCl3) δ: 20.6 (OAc), 21.2 (OAc), 32.4 (CH2),
35.6 (CH2), 55.9 (OCH3), 56.1 (OCH3), 71.2 (OCH2), 104.3
(CH ar.), 112.0 (CH ar.), 112.3 (CH ar.), 114.2 (CH ar.),
120.2 (CH ar.), 127.3 (CH Bn), 127.8 (CH Bn), 128.6 (CH
Bn), 134.6 (C ar.), 135.5 (C ar.), 135.7 (C ar.), 137.4 (C ar.),
146.5 (C ar.), 148.5 (C ar.), 149.5 (C ar.), 151.6 (C ar.),
168.9 (C=O), 169.4 (C=O) ppm. MS (EI): m/z (%): 464
(M+., 51), 422 (12), 380 (27), 331 (30), 289 (27), 257 (11),
227 (34), 195 (37), 183 (27), 137 (77), 91 (100), 43 (48).
Compound IV was also prepared in quantitative yield by
acetylation of compound II using the procedure described
for the synthesis of compound III.
2.2 Diphenyl ethane compounds
4′-benzyloxy-2,5-dihydroxy-3,3′-dimethoxydiphenylethane (II)
4′-benzyloxy-5-formyl-2-hydroxy-3,3′-dimethoxydiphenylethane
(4): Hydrogenation of compound 3 (500 mg, 1.28 mmol) in
ethyl acetate (100 mL) containing 50 mg of RhCl(PPh3)3
was accomplished by treating the mixture under hydrogen
pressure (180 bar, 1 bar = 105 Pa) for 5 h at 80°C. The mix-
ture, after filtration on a silica gel to removed the catalyst,
1
2,5,4′-trihydroxy-3,3′-dimethoxydiphenylethane (6)
afforded a white solid (500 mg, yield 100%). H NMR
(CDCl3) δ: 2.9 (s, 4H, CH2-CH2), 3.8 (s, 3H, OCH3), 3.9
(s, 3H, OCH3), 5.0 (s, 2H, OCH2), 6.5 (s, 1H, OH), 6.6–7.4
(m, 10H, ArH), 9.8 (s, 1H, CHO) ppm.
2,5-diacetoxy-4′-hydroxy-3,3′-dimethoxydiphenylethane (5):
Hydrogenation of III (300 mg, 0.65 mmol) in anhydrous
THF (50 mL) containing 30 mg of Pd–C was accomplished
by treating the mixture under hydrogen pressure (80 bar,
1 bar = 105 Pa) for 24 h at 60°C. The mixture, after filtration
on silica gel to removed the catalyst, afforded compound 5
4′-benzyloxy-2,5-dihydroxy-3,3′-dimethoxydiphenylethane (II):
A solution of compound 4 (300 mg, 0.8 mmol) in a mixture
of THF (20 mL)–water (5 mL) was treated with potassium
carbonate (72 mg, 0.8 mmol) and hydrogen peroxide
(150 µL, 1.4 mmol). After being stirred at room temperature
for 1 h, the reaction was stopped by addition of 1 mL of ace-
tic acid. The residue was reacted with a solution of hydro-
chloric acid (10%). The aqueous layer was extracted with
CH2Cl2. The combined organic layers were washed with wa-
ter, dried over MgSO4, filtered, and concentrated under vac-
uum. The residue was purified by crystallization in a mixture
of CH2Cl2–petroleum ether (1v:1v) affording the expected
stilbene II as a white solid (200 mg, yield 70%, mp 119°C).
UV λ max (MeOH)/nm : 270 (ε/L mol–1 cm–1 6800). IR (KBr)
ν : 730, 790, 1030, 1090, 1150, 1220, 1390, 1440, 1480,
1
(300 mg, yield 100%). H NMR (CDCl3) δ: 2.20 (s, 3H,
OAc), 2.25 (s, 3H, OAc), 2.38 (s, 4H, CH2-CH2), 3.72
(s, 3H, OCH3), 3.76 (s, 3H, OCH3), 6.1–7.0 (m, 5H, ArH)
ppm.
2,5,4′-trihydroxy-3,3′-dimethoxydiphenylethane (6): A solution
of 5 (300 mg, 0.8 mmol) in methanol (25 mL) was treated
with K2CO3 (350 mg, 4 mmol). After being stirred at 90°C
for 4h, the solution was reacted with a solution of hydro-
chloric acid (10%). The aqueous layer was extracted with
CH2Cl2. The combined organic layers were washed with wa-
ter, dried over MgSO4, filtered, and concentrated. The react-
1
ing mixture was analyzed by H NMR and by GC–MS after
1
1510, 2930, 3320 cm–1. H NMR (CDCl3) δ: 2.86 (s, 4H,
1
silylation. H NMR (CDCl3) δ: 2.20 (s, OAc), 2.41 (s, 4H,
CH2-CH2), 3.85 (s, 3H, OCH3), 3.88 (s, 3H, OCH3), 4.80 (s,
1H, OH), 5.16 (s, 2H, OCH2), 5.32 (s, 1H, OH), 6.17 (d,
1H, H4, J = 2.64 Hz), 6.34 (d, 1H, H6, J = 2.64 Hz), 6.72
(dd, 1H, H6′, J = 1.95 and 8.10 Hz), 6.77 (d, 1H, H2′,
J = 1.95 Hz), 6.84 (d, 1H, H5′, J = 8.10 Hz), 7.30–7.55
(m, 5H, ArH) ppm. 13C NMR (CDCl3) δ: 31.9 (CH2), 35.4
(CH2), 55.8 (OCH3), 55.9 (OCH3), 71.1 (OCH2), 97.3 (CH
ar.), 107.9 (CH ar.), 112.4 (CH ar.), 114.1 (CH ar.), 120.2
(CH ar.), 127.2 (CH Bn), 127.6 (CH Bn), 127.7 (C ar.),
128.4 (CH Bn), 135.4 (C ar.), 137.2 (C ar.), 137.3 (C ar.),
146.7 (C ar.), 148.3 (C ar.), 149.2 (C ar.) ppm. MS (EI) m/z
(%): 380 (M+., 30), 289 (35), 227 (40), 153 (33), 91 (100).
CH2-CH2), 3.75 (s, 3H, OCH3), 3.78 (s, 3H, OCH3), 6.0–7.0
(m, 5H, ArH) ppm. The observation of the peak at 2.20 ppm
reveals the presence of compound 6Ac (50%) in addition to
compound 6 (50%). This proportion was also revealed by
GC–MS when the mixture was silylated by treatment with
pyridine (10 µL), BSTFA (100 µL) in dichloromethane. MS
(EI) 6 silylated m/z (%): 433 ((M – 1)+, 33), 223 (14), 209
(100), 193 (11), 75 (12), 73 (34). MS (EI) 6Ac silylated m/z
(%): 476 (M+., 14), 434 (57), 209 (100), 179 (15), 75 (13),
73 (41). This spectra indicates the incorporation of two
trimethylsilyl groups in the molecule, the OH in the 2 posi-
tion being too hindered to be substituted by the bulky TMS
group.
2,5-diacetoxy-4′-benzyloxy-3,3′-dimethoxydiphenylethane (IV):
Hydrogenation of III (300 mg, 0.72 mmol) in ethyl acetate
(50 mL) containing 30 mg of RhCl(PPh3)3 was accomplished
by treating the mixture under hydrogen pressure (180 bar,
1 bar = 105 Pa) for 5 h at 80°C. After filtration on silica gel,
the solvent was removed under vacuum The compound IV
was obtained as a white solid (300 mg, yield 100%, mp
130°C). UV λmax (MeOH)/nm : 260 (ε/L mol–1 cm–1 6900).
IR (KBr) ν : 1030, 1090, 1140, 1180, 1220, 1510, 1760,
2.3 Formation of p-quinonoid compounds
Quinone diphenylethane (IIQ)
Oxidation with NaIO4 : A solution of compound II
(100 mg, 0.26 mmol) in a mixture of acetic acid (12 mL) –
water (8 mL) was treated with sodium periodate (85 mg,
0.40 mmol). After being stirred at room temperature for 4h,
© 2000 NRC Canada

Ruffin and Castellan
77
the organic layer was extracted with dichloromethane, dried
over MgSO4, and reduced under vacuum. The expected
compound IIQ was obtained as a pale yellow oil (100 mg,
yield 100%). IR (NaCl film) ν : 2930, 1680, 1650, 1600,
1510, 1450, 1260, 1230, 1025, 790, 760 cm–1. 1H NMR
(CDCl3) δ: 2.75 (s, 4H, CH2-CH2), 3.82 (s, 3H, OCH3), 3.88
(s, 3H, OCH3), 5.13 (s, 3H, OCH2), 5.88 (d, 1H, H4, J =
2.44 Hz), 6.44 (d, 1H, H6, J = 2.44 Hz), 6.65 (dd, 1H, H6′, J =
2.14 and 8.24 Hz), 6.73 (d, 1H, H2′, J = 2.14 Hz), 6.80 (d,
1H, H5′, J = 8.24 Hz), 7.2–7.5 (m, 5H, ArH) ppm.
stream. The paper sheets were irradiated with three black
light UV lamps (Mazdafluor TFWN 18) emitting mainly in
the 350 nm wavelength range. The distance between the
lamp and the sheets was 20 cm. The temperature of the irra-
diation was maintained at 25°C by a fan. The reflectance of
the sheets backed with a black and white backgrounds were
measured at appropriate irradiation time intervals. The ab-
sorption (K) given by the Kubelka–Munk theory (17) was
calculated and plotted as the difference between non irradi-
ated and irradiated samples versus irradiation time.
Oxidation with Fremy’s salt: A solution of compound II
(100 mg, 0.26 mmol) in methanol (20 mL) was treated with
Fremy’s salt (KSO3)2NO (110 mg, 0.42 mmol) in water
(5 mL) After being stirred at room temperature for 4h, the
organic layer was extracted by dichloromethane, dried over
MgSO4, and reduced under vacuum. The expected com-
pound IIQ was obtained as a pale yellow oil (100 mg, yield
100%).
4. Attempt to detect stilbene–hydroquinone
chromophore in bleached mechanical pulp
Peroxide bleaching of mechanical pulp (BMP) and
acetylation
Unbleached softwood mechanical pulp (20 g), kindly pro-
vided by Dr. Petit-Conil (Centre Technique du Papier,
Grenoble), was treated in a polyethylene bag with the
bleaching liquor (200 mL) constituted by hydrogen peroxide
(3% oven dry pulp, odp), NaOH (1.5% odp), MgSO4·7H2O
(0.5% odp), sodium silicate (3.5% odp), and DTPA (0.25%
odp) at 60°C for 4 h. Then the pulp was filtered and
bleached for a second time in the same conditions with fresh
liquor (200 mL). At the end of the bleaching, the pulp was
washed with distilled water and diluted hydrochloric acid to
pH 4.5. Before acetylation, the pulp (1 g) was made alkaline
by soaking it in a potassium carbonate solution (1 mol L–1)
for 15 min and filtrated under vacuum to remove the major
part of the water. The pulp was then exposed to acetic anhy-
dride vapor for 1 h and after was washed with water, ethanol
and water to pH 4.5. At the end, the pulp was dried at 30°C
for 24 h.
Quinone stilbene (IQ): A solution of compound I (10 mg,
0.026 mmol) in acetonitrile was oxidized into quinone IQ by
treatment under pressure of oxygen (20 bar, 1 bar = 105 Pa)
at 60°C for 5 h. After evaporation of the solvent, the mixture
1
was analyzed by HPLC and H NMR. Products I and IQ
were obtained in a 50:50 mixture. Various tentatives to iso-
late IQ as a pure compound by preparative chromatography
were not successful due to the high reactivity of the quinone.
1
So, the H NMR spectrum of IQ was obtained from the re-
acting mixture spectrum by subtracting the contribution of
compound I. The electronic absorption spectrum of IQ was
obtained with the diode array detector on the HPLC system
(eluent: methanol). The mass spectrum of IQ was obtained
using HPLC–MS technique with acetonitrile as the eluent
for the HPLC and electrospray for the ionization mode. This
analysis did not allow the observation of the (M – 1)– peak
as for I, but a peak was observed which corresponded to the
addition of acetonitrile followed by the elimination of CO.
Hydrogenation of the acetylated BMP
Hydrogenation of the pulp (300 mg) impregnated with a
solution of RhCl(PPh3)3 (15 mg) in ethyl acetate (3 mL) was
accomplished under 180 bar (1 bar = 105 Pa) of hydrogen
pressure for 4 h at 60°C in presence of 3 mL of ethyl acetate
in the bottom of the reactor as described previously (14). Af-
ter cooling the reactor, the pulp was washed with ethyl ace-
tate, methanol, and CH2Cl2 to remove the catalyst and then
washed with water and dried at 40°C for 24 h.
1
UV λmax (MeOH): 280 and 480 nm. H NMR (CDCl3) δ:
3.90 (s, 3H, OCH3), 3.96 (s, 3H, OCH3), 5.20 (s, 2H,
OCH2), 6.25 (d, 1H, H4, J = 2.70 Hz), 6.55 (d, 1H, H6,
J = 2.70 Hz), 6.8–7.1 (4H, ArH and =CH), 7.25–7.6
(m, 6H, ArH and = CH) ppm. HPLC–MS (solvent aceto-
nitrile, negative mode) m/z: 388 (M + CH3CN–CO–H)–.
Quinone stilbene (IZQ): The stilbene quinone IZQ formed
by irradiation of I in non-degassed solutions was character-
ized by HPLC equipped with a diode array detector (eluent:
methanol) and HPLC–MS technique with acetonitrile as
eluent for the HPLC and electrospray for the ionization
mode. As for IQ, this analysis did not allow the observation
of the (M – 1)– peak, but a peak was observed which corre-
sponds to the addition of acetonitrile followed by the elimi-
nation of CO. UV λmax (MeOH): 270 and 470 nm. HPLC–
MS (solvent acetonitrile, negative mode) m/z : 388 (M +
CH3CN–CO–H)–.
Thioacidolysis and Acidolysis with AlCl3
100 mg of pulp were treated with 20 mL of thioacidolysis
solution (10 mL EtSH + 2.4 mL BF3–Et2O + 80 mL dioxan)
at 100°C for 4 h under gentle magnetic stirring. After hydro-
lysis with diluted NaHCO3 solution until pH 3–4, the reac-
tion mixture was extracted with dichloromethane. The
organic phase was washed with water, dried over magnesium
sulfate, and evaporated under vacuum. The residue was dis-
solved in benzene (30 mL) and was treated with 10 mg of
AlCl3 at 90°C for 12 h, then the reacting mixture was hydro-
lyzed with water and extracted with CH2Cl2. The organic
phase was dried over magnesium sulfate concentrated under
vacuum and silylated with BTSFA (50 µL), pyridine (10 µL)
in THF (50 µL) at room temperature. The silylated mixture
was analyzed by GC–MS.
3. UV–Irradiation of models I, II, III and IV adsorbed
on cellulose matrix
Four paper sheets (3 cm × 3 cm) made from a kraft pulp
(30 g/m2) were separately impregnated with 2.0 mg of each
compound in CH2Cl2 (1 mL) and dried with a nitrogen
© 2000 NRC Canada

78
Can. J. Chem. Vol. 78, 2000
Fig. 2. Synthetic scheme of compounds I, II, III, and IV: (a) n-BuLi–DMF–THF; (b) 2, THF; (c) HCl 10%; (d) H2O2–Na2CO3–THF–
H2O; (e) Ac2O–NEt3–DMAP–CH2Cl2; (f) H2–Rh(PPh3)3Cl–ethyl acetate.
syntheses of other lignin stilbene models (14). The
hydroquinone I (E configuration) was obtained by Dakin re-
Results and discussion
action in mild conditions (room temperature and moderate
pH given by the use of sodium carbonate). Compound III
was synthesized in high yield (93%) by acetylation of com-
pound I, using the catalytic action of 4-N,N-dimethyl-
aminopyridine (DMAP) as for other lignin model compounds
(18). Hydrogenation of compound III, was realized using ex-
perimental conditions set up previously for a study on the role
of stilbenes issued from β-1 units on the photoyellowing of
lignin rich pulps (14). Compound II which was prepared by
catalytic hydrogenation of compound 3 followed by Dakin re-
action at moderate pH, was quantitatively transformed into
compound IV by acetylation.
1. Syntheses
The route chosen for the synthesis of compounds I, II, III,
and IV is outlined in Fig. 2. The preparation of compound I
which was similar to the Lee and Sumimoto (13) procedure,
consists in the Wittig coupling between the ylide, generated
by action of n-butyl lithium on the substituted benzyltriphenyl
phosphonium chloride 1 (14), and the substituted salicylalde-
hyde 2 (16). The protection of the aldehyde group, needed for
the coupling, was easily removed by treatment with hydro-
chloric acid at the end of the synthesis. Despite several at-
tempts to improve the Wittig coupling e.g., by changing the
reaction solvent, the yield was in our hands, lower (37%) than
the one given by Lee (65%) (13) for the synthesis of a similar
compound, but it was of the same order that we got for the
Compound IZ was produced as a major product by irradia-
tion of I in degassed methanol solution (Fig. 3). It was char-
© 2000 NRC Canada

Ruffin and Castellan
79
Fig. 3. Synthetic scheme of compounds IQ, IIQ, and IZ: (a) P(O2) = 20 bar; (b) NaIO4–acetic acid–water; (c) (KSO3)2NO–water.
Fig. 4. Synthetic scheme of compound 6: (a) H2–Pd–C–THF; (b) K2CO3–methanol.
1
acterized by H and 13C NMR and HPLC–MS (electrospray
ment were expected to be in the form of their acetates or
in the negative mode) spectroscopies (see experimental part).
trimethylsilyl derivatives. The mild deacetylation procedure,
very efficient for lignin models (18), and applied to com-
pound 5, produced a mixture of compounds 6 and 6Ac
(50%–50%). This mixture was silylated and used for the
GC–MS analysis of the pulps, because it is likely that these
two structures will be detected if the stilbene hydroquinone
entities would be present in the peroxide bleached pulp. It is
noticeable that silylation of compound 6 was incomplete, the
OH in the 2 position being too hindered to react with the
bulky trimethylsilyl group.
In contrast to compound II, which gave the p-quinone IIQ
by oxidation by sodium periodate or Fremy’s salt, compound
I did not produce the quinone IQ using the same reagents.
The latter compound, very reactive and sensitive to chro-
matographic phases and eluents was obtained by oxidation
of I in acetonitrile by oxygen under pressure (20 bar, 1 bar =
105 Pa) at 60°C (Fig. 3). The structure of IQ was determined
1
by H NMR and HPLC–MS (electrospray in the negative
mode) spectroscopies (see experimental part). The structure
of IZQ was assigned by comparison of its HPLC–UV spec-
trum and its HPLC–MS peaks with those of the quinone iso-
mer IQ (see experimental part).
In order to characterize the presence of stilbene–hydro-
quinone chromophores in bleached pulps, compounds 5 and
6 were synthesized. Following a procedure previously used
(14), and due to the high reactivity of the stilbene and
hydroquinone units, the pulp was acetylated and hydroge-
nated prior to thioacidolysis and acidolysis with aluminum
trichloride. The diphenylethane incorporating hydroquinone
entities obtained from the pulp after the degradative treat-
2. Electronic absorption and emission spectroscopies
Absorption spectra
The electronic absorption spectra of the stilbenes I, III
and of the diphenylethanes II, IV are reported on Fig. 5.
The two diphenylethanes present a significant absorption
above 300 nm, so they can absorb the solar light and so they
can give colored species and contribute to the yellowing of
peroxide-bleached high-yield pulps as already shown (19).
The presence of the stilbene double bond increases the ab-
© 2000 NRC Canada

80
Can. J. Chem. Vol. 78, 2000
Fig. 5. Electronic absorption spectra of compounds I, II, III, and IV in methanol solutions (left) and of compound I in acidic and
alkaline methanol solutions (right).
Fig. 6. Fluorescence emission spectra of compound I in ethanol solution. Left: neutral and acidic media; right: alkaline medium (λexc
320 nm).
=
region 330–430 nm. In contrast, the fluorophores at the ori-
gin of the emission of the pulp in the long wavelength range
(430–530 nm) are still unknown. It is remarkable that the
fluorescence emission, recorded for the stilbene I in neutral
or acidic methanol solutions (Fig. 6), displays emission in
this wavelength range (quantum yield ≈ 0.5 and lifetime ≈ 4
ns). This represents a bathochromic shift compared to other
non-cyclic stilbene lignin models (22) and cyclic stilbene
lignin model of the phenylcoumarone type (8b). In alkaline
solution, the emission of compound I is shifted to longer
wavelengths and its intensity is reduced, probably due to a
highest reactive deactivation channel from the phenolate anion.
sorption above 300 nm compared to diphenylethanes, and so
compound I which presents a very oxidizable phenol func-
tional group, should be very sensitive to UV–irradiation.
Moreover, phenols are in equilibrium with their phenolate
anions and this equilibrium depends on the pH of the me-
dium. In acidic or neutral media, the absorption of com-
pound I is the same (Fig. 5). In contrast, in alkaline medium,
compound I displays some absorption between 400 and
600 nm. The pulp industry usually produces papers with a
basic reserve such as calcium carbonate for long term stor-
age (20) but this alkaline treatment can induce a coloration
of the materials if structures like compound I, which is ion-
ized at pH 9, are present in lignin.
3. UV–irradiation of model I in solution
Fluorescence spectra
Compound I was irradiated in aerated methanol solutions
(concentration ≈ 10–3 mol L–1) with UV black lamps
(λ ≈ 350 nm). The irradiated solutions were analyzed by
HPLC at different irradiaton times and some chromatogram
traces are reported in Fig. 7.
The interpretation of fluorescence spectra from wood, pulp,
and paper samples is particularly cumbersome due to the
physical heterogeneity and chemical complexity of the
lignocellulosic matrix (21–23). Based on model studies (21,
22), it was shown that, in addition to cellulose emission, sev-
eral chromophores of the lignin polymer are emitting in the
The assignment of the HPLC peaks was carried out by
comparison with samples analyzed by NMR and mass spectro-
© 2000 NRC Canada

Ruffin and Castellan
81
Fig. 7. HPLC analyses of photoproducts formed during irradia-
tion of I in methanol.
Fig. 8. HPLC–UV analyses of photoproducts formed during irra-
diation of I in methanol.
tem are drawn in Fig. 8. One can observe a hypsochromic
shift (20 nm) for the absorption of the IZ isomer compared
to the E isomer (this shift was already found by Stomberg et
al. (24) on other lignin stilbene-like structures). Also a shift
to the short wavelength (10 nm) is observed for the quinone
IZQ compared to the E quinone IQ, both for nπ* visible and
for ππ* UV transitions.
It is known that Z stilbenes are able to photocyclize to
give phenanthrene derivatives (25). The dihydroxyphenanthrene
formed by oxidative photocyclization of compound IZ might
be transformed into p-phenanthrene quinone. The compound
formed after 15 min of irradiation and having a retention
time equal to 15 min might belong to this type of structure
(Fig. 9). The photochemistry of compound I in methanol or
acetonitrile solution is depicted in Fig. 9.
scopies. The structure of the phenanthrene quinone deriva-
tive was assumed according to its UV absorption spectrum.
The UV absorption spectra of the photoproducts which
were obtained with a diode array detector on the HPLC sys-
Fig. 9. Photoreactivity of compound I in solution.
© 2000 NRC Canada

82
Can. J. Chem. Vol. 78, 2000
Fig. 10. Irradiations of compounds adsorbed on cellulose paper. ∆K = absorption coefficient of irradiated paper – absorption coefficient
of non irradiated paper. (a), (b), (c), (d): ∆K vs. λ for compounds I, II, III, and IV respectively.; (e) ∆K (425 nm) vs. irradiation time;
(f) HPLC trace of extracted irradiated paper doped with compound I.
The kinetics confirm the great reactivity of compound I.
Its coloration is reduced by acetylation. Hydrogenation of
compound I followed by acetylation leads to a very stable
compound IV. In contrast, hydrogenation of I, which gives
hydroquinone structure II, is not sufficient enough to stop
the yellowing completely. The variations of ∆K observed for
compound II is reminiscent to what was observed for 2-
methoxyhydroquinone in 2-hydropropylcellulose films (19).
After irradiation, the cellulose paper incorporating model
I was extracted with solvent and analyzed by HPLC
(Fig. 10f). Only the formation of quinone IQ is observed
with no isomerization occurring. This is probably due the
hydrogen bonding between the product and the cellulose
4. UV–irradiation of model I, II, III, and IV adsorbed
on cellulose matrix.
The behavior of the lignin model compounds I, II, III,
and IV adsorbed on cellulose paper were comparatively irra-
diated with UV light (see experimental part). The
photocoloration was evaluated using the absorption coeffi-
cient given by the Kubelka–Munk theory (26). Fig. 10 (a–e)
shows the difference of absorption coefficient (∆K) of the
lignin models adsorbed on cellulose paper before and after
irradiation with UV light (λ ≈ 350 nm) ranging from 5 min
to 120 min. As observed for stilbenes formed from β-1 units
(14), hydrogenated and acetylated chromophores display less
variation of ∆K, which means less yellowing.
© 2000 NRC Canada

Ruffin and Castellan
83
matrix which locks the configuration of the stilbene double
bond. This confirms the key role of the environment in pho-
tochemical reactions.
of the stilbene–hydroquinone units formed from β-5 ele-
ments in lignin.
5. Search for stilbene–hydroquinone chromophores in
peroxide-bleached high-yield pulps
References
1. D.Y. Lee, M. Matsuoka, and M. Sumimoto. Holzforschung,
44, 415 (1990).
2. D.Y. Lee, M. Matsuoka, and M. Sumimoto. Fifth Intern.
Symp. Wood Pulp. Chem. Raleigh, Proceedings, 21 (1989).
3. Z. Wu, M. Matsuoka, D.Y. Lee, and M. Sumimoto. Mokuzai
Gakkaishi, 37, 164 (1991).
4. D.Y. Lee and M. Sumimoto. Holzforschung, 44, 347 (1990).
5. D.Y. Lee, S. Tachibana, and M. Sumimoto. Cellulose Chem.
Technol. 22, 201 (1988).
6. S. Li and K. Lundquist. Holzforschung, 53, 39 (1999).
7. A. Castellan, N. Colombo, A. Nourmamode, J. Zhu, D.
Lachenal, R.S. Davidson, and L. Dunn. J. Wood Chem.
Technol. 10, 461 (1990).
8. C. Noutary, P. Fornier de Violet, J. Vercauteren, and A.
Castellan. Res. Chem. Intermed. (a) 21, 223 (1995); (b) 21,
247 (1995).
9. (a) A.J. Nonni and C.W. Dence. Holzforschung, 42, 37 (1988);
(b) W.J. Leigh, T.J. Lewis, V. Lin, and J.A. Postigo. Can. J.
Chem. 74, 263 (1996).
10. S. Omori and C.W. Dence. Wood Sci. Technol. 15, 113 (1981).
11. G. Gellerstedt and R. Agnemo. Acta Chem. Scand. B34, 275
(1980).
In order to assess and quantify the contribution of
stilbene–hydroquinone chromophores in the photoreversion
of high-yield pulps, we have attempted to find and quantify
this structure in peroxide-bleached mechanical softwood
pulp. For this purpose, we have used the same experimental
procedure successfully set up for stilbenes emerging from
β-1 lignin structures (14). In addition to the previous proce-
dure, the pulp was first acetylated in the vapor phase in or-
der to block the reactive hydroquinone groups. Even so, it
was shown that catechol structures survive after the thio-
acidolysis (27). Therefore, after acetylation, hydrogenation,
thioacidolysis with EtSH – BF3–Et2O in dioxane, and
desulfurization with raney nickel, the organic residue was
treated with AlCl3 in benzene. The reacting mixture was an-
alyzed by GC–MS. Careful inspection of the mass spectra of
the chromatogram did not detect the silylated compounds 6
or 6Ac. So, it seems that a mechanical treatment is not
strong enough to produce structures resembling compound I
by mechanochemistry in high-yield pulps. However, we can
never neglect that the chemical treatment, used to liberate
the stilbene chromophore before analyses, is not efficient
enough if the chromophore is situated in the bulk of the
lignin polymer.
12. I. Forsskåhl, J. Gustafsson, and A. Nybergh. Acta Chem.
Scand. B35, 389 (1981).
13. D.Y. Lee and M. Sumimoto. Holzforschung, 45 (Supp.), 15
(1991).
Conclusions
14. B. Ruffin, A. Castellan, S. Grelier, A. Nourmamode, S. Riela,
and V. Trichet. J. Appl. Polym. Sci. 69, 2517 (1998).
15. J.G. Miller and C. Schuerch. Tappi J. 51, 273 (1968).
16. F. Nakatsubo and T. Higuchi. Mokuzai Gakkaishi, 25, 735
(1979).
17. J.A. Schmidt and C. Heitner. Tappi J. 76, 117 (1993).
18. C. Jaeger, A. Nourmamode, and A. Castellan. Holzforschung,
47, 375 (1993).
We studied stilbene–paraquinone models formed by mechano-
chemistry from β-5 units during grinding and disk-refining
of pulp. The contribution of these structures to the photo-
yellowing of bleached high-yield pulps was evaluated exam-
ining the photochemistry of model I which was hydrogenated
and (or) acetylated. Photoproducts formed during irradiation of
compound I were characterized. In particular, the isomeri-
zation of compound I into IZ was found to be the main pro-
cess in solution. In aerated solution, the stilbenes I and IZ
were oxidized in paraquinones IQ and IZQ respectively.
The isomerization of I was not observed when adsorbed on
cellulose paper, only the formation of quinone IQ was ob-
tained. Comparative irradiation of different models, ad-
sorbed on thin sheets of cellulose paper, showed the greatest
reactivity of model I. Hydrogenation coupled with acetyl-
ation produced an efficient stabilization of this compound.
The detection and quantification of stilbene–hydroquinone
units, present in peroxide-bleached mechanical pulp, could
not be determined. This might be due to inefficient
mechanochemistry in pulp, as suggested by Zhu’s studies
(28), or it might be due to inappropriate experimental condi-
tions necessary to liberate the acetylated hydroquinone
diphenylethane entities, these structures being characteristic
19. A. Castellan, A. Nourmamode, C. Jaeger, and I. Forsskåhl.
Nordic Pulp Pap. Res. J. 8, 239 (1993).
20. H.A. Carter. J. Chem. Educ. 73, 1068 (1996).
21. A. Castellan, H. Choudhury, R.S. Davidson, and S. Grelier. J.
Photochem. Photobiol. A: Chem. 81, 117 (1994).
22. H. Tylli, I. Forsskåhl, and C. Olkkonen. J. Photochem.
Photobiol. A: 87, 181 (1995).
23. J.A. Olmstead and D.J. Gray. J. Pulp Pap. Sci. 23, 571 (1997).
24. R. Stomberg, S. Li, K. Lundquist, and B. Albinsson. Acta
Crystallogr. Sect.C: Cryst. Struct. Commun. C54, 1929 (1998).
25. N.J. Turro. In Modern molecular photochemistry. Benjamin
Cumming Publ. Co., Menlo Park. (1978).
26. J.A. Schmidt and C. Heitner. Tappi J. 76, 117 (1993).
27. X. Pan, D. Lachenal, C. Lapierre, and B. Monties. J. Wood
Chem. Technol. 12, 135 (1992).
28. J.H. Zhu, C. Archer, F. McNab, M.P. Andrews, G. Kubes, and
D.G. Gray. J. Pulp Pap. Sci. 23, J305 (1997).
© 2000 NRC Canada