Stabilization of ortho-quinone methides by a bis(sulfonium ylide) derived from 2,5-dihydroxy-[1,4]benzoquinone

Article abstract of DOI:10.1016/j.tetlet.2008.02.045

The zwitterionic intermediates (2a) in the oxidation of ortho-alkylphenols (1) and bis(sulfonium ylide) 3 form reasonably stable 2:1-complexes (4), in which the ortho-quinone methide (oQM) moieties are not present in quinoid form with the exocyclic in-pla

Full text of DOI:10.1016/j.tetlet.2008.02.045

Available online at www.sciencedirect.com
Tetrahedron Letters 49 (2008) 2442–2445
Stabilization of ortho-quinone methides by a bis(sulfonium ylide)
derived from 2,5-dihydroxy-[1,4]benzoquinone
Anjan Patel a, Thomas Netscher b, Thomas Rosenau a,*
a BOKU University, Department of Chemistry, Muthgasse 18, A-1190 Vienna, Austria
b Research and Development, DSM Nutritional Products, PO Box 3255, CH-4002 Basel, Switzerland
Received 15 December 2007; revised 5 February 2008; accepted 8 February 2008
Available online 13 February 2008
Abstract
The zwitterionic intermediates (2a) in the oxidation of ortho-alkylphenols (1) and bis(sulfonium ylide) 3 form reasonably stable 2:1-
complexes (4), in which the ortho-quinone methide (oQM) moieties are not present in quinoid form with the exocyclic in-plane methylene
group, but as zwitterionic, aromatic conformer having an out-of-plane exocyclic methylene group. The complex 7 derived from the a-
tocopherol model compound PMC (5) was comprehensively characterized. As exemplarily demonstrated, the adducts can be advanta-
geously employed in organic synthesis as ‘stabilized oQMs’.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: ortho-Quinone methides; ortho-Alkylphenols; Aromaticity; Sulfonium ylides; Zwitterions
ortho-Quinone methides (oQMs, 2), obtained from their
parent ortho-methylphenols (1) by oxidation, are a class of
chemical intermediates that have not been isolated directly.
Their mere existence, however, is confirmed beyond any
doubt as there is abundant indirect evidence for their
in situ formation, which comes mainly from the identifica-
tion of their reaction products, as well as from trapping
reactions. In 1963, the first spectrophotometric observation
of a chromophore with supposed oQM structure was
reported.1 Also in some other cases fast UV spectroscopy2
at low temperatures was used to detect the occurrence of
intermediates, which had stronger UV-absorption than
starting materials and final products, and were conse-
quently attributed to oQM-type structures. In the previous
work, we had shown by the example of tocopherol-type
compounds that zwitterionic structures (2a) are quite tran-
sitory intermediates in the formation of the oQMs (2) from
the parent phenols.3 Direct NMR spectroscopic evidence
of an oQM was provided recently when the oQM derived
from a-tocopherol was temporarily stabilized by inter-
action with the zwitterionic N-methylmorpholine-N-oxide
at lower temperatures (ꢀ78 °C).4
Here, we would like to report the interaction of the zwit-
terionic precursors of ortho-quinone methides (2a) with
bis(sulfonium ylide) 3 and the resulting formation of rea-
sonably stable 2:1 complexes 4, which were unaltered at
ꢀ78 °C for 10 h and stable at room temperature under
inert conditions for as long as 15–30 min. The oQM was
produced by Ag2O oxidation of the respective ortho-
methylphenol in a solution containing 0.50–0.55 equiv of
bis(sulfonium ylide) 3. This compound is readily prepared
from 2,5-dihydroxy-[1,4]benzoquinone by reaction with
DMSO in acetic anhydride.5 Although the species inter-
acting with the ylide is actually the zwitterionic oxidation
intermediate 2a and not the oQM itself, the term ‘stabilized
oQM’ will be used in the following for complexes 4. Model
experiments were carried out with the a-tocopherol model
2,2,5,7,8-pentamethylchroman-6-ol (PMC, 5), which upon
oxidation formed the zwitterionic structure 6a that in turn
reacted with bis(sulfonium ylide) 3 to form the 2:1 complex
7.
*
Corresponding author. Tel.: +43 1 360066071; fax: +43 1 360066059.
E-mail address: thomas.rosenau@boku.ac.at (T. Rosenau).
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.02.045

A. Patel et al. / Tetrahedron Letters 49 (2008) 2442–2445
2443
In the complexes formed, both oQMs adopt a zwitter-
ionic, aromatic structure (2a) with the exocyclic methylene
group in perpendicular arrangement to the ring plane, sta-
bilized by the negatively charged oxygen in 3. Simulta-
neously, the negatively charged oxygens in the oQM
parts interact with the positively charged sulfur to provide
additional stabilization. Thus, the stabilized ortho-quinone
methides are evidently not present in their ‘traditional’
quinoid form with the exocyclic in-plane methylene group,
but as zwitterionic, aromatic structure having an out-of-
plane exocyclic methylene group. It should be noted that
the latter structure is not a resonance structure of the
oQM as such canonic structures differ only in the arrange-
ment of multiple bonds. However, the aromatics in 4 and
the oQMs (2) are distinguished by a different conformation
of the exocyclic methylene group in addition. Formation of
2 from 2a as present in complex 4 requires rotation of the
methylene group into the ring plane, which is coupled to
the immediate aromatic-to-quinone conversion.
The electrostatic interactions in complexes 4 were obvi-
ously sufficient to ‘favor’ the zwitterionic structure (2a) in a
manner that formation of the quinoid structure (2) was
‘suspended’, so that all reactions typical of oQMs in their
quinoid form (such as spiro-dimerizations or trapping in
[4+2]-cycloadditions, e.g., with ethyl vinyl ether)6 were
suppressed or at least slowed down. Decomposition of
the complex was immediate by fast heating to 40 °C or
above, best achieved by slowly adding a cold solution of
the complex into a heated solution of the coreactant. This
causes disintegration of the complex, immediate rotation of
the methylene group into the ring plane and thus formation
of the oQM which then shows the ‘classical’ chemistry of
such compounds (Scheme 1).
Complex 7, formed by Ag2O oxidation of PMC (5) in
the presence of 3, was isolated at ꢀ30 °C as an amorphous
addition product.7 It showed the exact ratio of 2:1 between
3 and 6a. Unfortunately, despite extensive trials, all
attempts to obtain (micro)crystalline 7 suitable for X-ray
or neutron scattering failed. A reliable image of the struc-
ture of the complex (Scheme 2) was obtained by refining
a quantum-chemical prediction (DFT, B3LYP 6-31G* level
of theory) of the crystal structure until optimum agreement
with powder diffraction data was reached.8 Proton NMR
spectroscopy of complex 7 in DMSO-d6 at 0 °C showed a
singlet (2H) at 5.85 ppm, corresponding to the exocyclic
methylene group. This peak showed a heteronuclear corre-
lation to a carbon at 191.8 ppm, and HMBC cross peaks at
129.9 ppm (2JH–C), 117.2 ppm (3JH–C), and 154.1 ppm
(3JH–C). The proton resonances of the 7a-CH3, 8b-CH3
methyl groups and the 4-CH2 methylene group at 11.8,
12.0, and 20.4 ppm indicated the presence of an aromatic
system.9 The high down-field shift of the carbon resonance
at 191 ppm for the exocyclic methylene group is especially
indicative of a cationic species,10 and the peak at 154 ppm
for C-5 agrees with a phenolate carbon, but not with a
quinoid carbonyl carbon.11 Also the bis(sulfonium ylide)
moiety was influenced, albeit rather weakly. The four mag-
netically equivalent methyl groups in bis(sulfonium ylide)
resonating at 3.02 ppm in DMSO-d6 appeared as two
_
CH2 in
plane
immediate
bond
rotation
H
O
O
H
H
C+
+
H
H
H
R
HO
O
R
ox. (Ag2O)
CH2 out of
plane
2 (oQM)
2a
O
O
5
6a
Ag2O
OH
O
S+
O
O
dimerization
or trapping
reaction
_
3 (0.52 eq.)
spiro-dimer (8)
_
T = -30 ºC
S+
CH3
T > 40 ºC
3
O
R
1
H
neat high-yield
O
C+
H
reaction with
S+
O
O
_
3 (0.5-0.55 eq.) in DMF
O
C-, O-, N-nucleophiles
O
_
Ag2O (10 eq.) in CHCl3,
_
_
O
T < -20 ºC
C+
H
O
O
S+
drop into heated
aprotic solution
T > 40 ºC,
S+
O
H
O
7
_
_
coreactant
O
S+
3
O
H
H
O
C+
_
S+
O
O
_
O
_
_
R
R
S+
O
C+
H
H
4
O
Scheme 1. Oxidation of ortho-alkylphenols (1) to the ortho-quinone
methide (2) via transient zwitterionic intermediates (2a) that can be
stabilized by forming complexes (4) with 2,5-dihydroxy[1,4]benzoquinone-
derived bis(sulfonium ylide) 3.
Scheme 2. Formula and molecular structure of the 2:1 complex 7, formed
between the zwitterionic oQM precursor derived from pentamethylchro-
manol 5 and bis(sulfonium ylide) 3.

2444
A. Patel et al. / Tetrahedron Letters 49 (2008) 2442–2445
singlets at 2.94 ppm and 2.98 ppm in complex 7. A fully
formed r-bond between the phenolic oxygen and the ben-
zylic methylene group can be excluded based on the NMR
data. However, through-space stabilization of the charges
is likely to be an important contribution to the stability
of the complex.
Compound 7 passes LC setups as single entity, but will
show signs of degradation at rt, so that it is best used at
lower temperatures. Purification of the oQM complexes is
unnecessary as their formation was near quantitative in
all cases if a slight excess of 3 and excess oxidant (5 equiv
of freshly prepared Ag2O) was used.12 Decomposition of
7 by heating in aprotic media to 40 °C afforded 3 and
spiro-dimer 8 quantitatively. In the presence of coreac-
tants, the zwitterionic forms of the oQMs can be conve-
niently derivatized as discussed in the following.
SiMe3
O
HO
O
Ag2O, 3
+
O
R
Me3Si
9
10 (1.1 eq.)
SiMe3
SiMe3
COOH
R = RR-C16H33
O
MeO
O
HO
O
R
O
R
12
11
Scheme 3. High-yield synthesis of 5-(c-tocopheryl)propionic acid (12)
employing the stabilized oQM derived from a-tocopherol in a hetero-
Diels–Alder reaction with inverse electron demand.
The stabilization of oQMs was preliminarily employed
in organic synthesis—an apparently very promising appli-
cation which, however, has yet to be fully exploited.8 The
general advantage is that oQMs can now be used and
handled like stable, stoichiometrically usable, dosable
reagents. They can be reacted in a controlled way without
the danger of immediate self-dimerization or other uncon-
trolled side reactions. Preparation of the complexes and
their further conversion can be separated, so that the deriv-
atization of the preformed, stabilized oQMs can be carried
out at later times, in different vessels or reaction media,
and under different reaction conditions. This way, high
yield conversions oQMs with coreactants in equimolar
amounts appear to become generally possible for the first
time, although it is advisable to optimize the conditions
for complex formation for each oQM used. By contrast,
oQMs were so far almost exclusively employed in trap-
ping-like scenarios, that is, the coreactant was present in
a large excess consuming the oQM immediately upon its
formation.
In four examples, the potential of the stabilization
approach in synthesis was demonstrated. 5-(c-Toco-
pheryl)propionic acid (12) was previously prepared in
72% yield by the reaction of the oQM derived from a-
tocopherol (9) with O-methyl-C,O-bis-(trimethylsilyl)-
ketene acetal (10), which was used as ‘trapping agent’ in
large excess.13 By introducing the stabilized oQM into a
solution of only 1.1 equiv of the ketene acetal at 40 °C a
near-quantitative reaction to ortho-ester derivative 11 was
observed after which two subsequent hydrolysis steps affor-
ded 11 in 93% overall yield.14 In this case, the oQM
released from the stabilized zwitterionic form reacted in
the ‘classical’ [4+2]-cycloaddition way as electron-deficient
hetero-analogous diene (Scheme 3).
favorably to the existing tedious isolation approaches
(Scheme 4). A similar alkylation approach was used to pro-
vide bis(5-c-tocopheryl)methane (17),17 a vitamin E oxida-
tion product and model compound which could be
synthesized so far in dissatisfying yields only. Applying
the ylide-stabilized form of the oQM derived from a-
tocopherol (9) as the alkylating agent toward equimolar
amounts of c-tocopherol (16), a non-optimized 78% yield
of 17 was obtained,18 which illustrated quite nicely the
potential of the stabilized oQMs as reagents in synthesis
(Scheme 4).
As the fourth example, o-hydroxybenzyl O-acetyl-sali-
cylate (saligenin O-acetyl-salicylate, 20), the aglycon in
salicoylsalicin, was synthesized which was so far not avail-
able from high-yield syntheses.19 This compound is closely
related to the b-glucosidase inhibitor salicortin, which is
OH
CH2
OH
C3H7
C3H7 OH
O
O
CH3
OMe
OH
HO
HO
Ag2O, 3
OMe
HO
OH
14
HO
O
O
C3H7 OH
13
15
C3H7
HO
O
R
O
R
16
HO
HO
Ag2O, 3
CH2
In a second example, the stabilized oQMs reacted as C-
alkylating agents. The anthelmintic margaspidin (15),15
was synthesized by introducing a solution of the stabilized
oQM from O-nor-methylaspidin (13) into a warm solution
of aspidin (14). The oQM liberated from the former acted
as alkylating agent toward the latter, and the target 1516
was obtained in a satisfying 84% yield which compared
HO
O
R
9
R = RR-C16H33
O
R
17
Scheme 4. High-yield synthesis of margaspidin (15) and bis(5-c-toco-
pheryl)methane (17) employing ylide-stabilized oQMs as alkylating agents
in Friedel–Crafts-type reactions.

A. Patel et al. / Tetrahedron Letters 49 (2008) 2442–2445
2445
NMR (DMSO-d6, 100 MHz, 18 mg/0.7 mL, 0 °C): 11.8 (7a-CH3),
12.0 (8b-CH3), 20.4 (4-CH2), 24.2 (S–Me), 24.4 (S–Me), 26.6
(2a-CH3), 27.3 (2a-CH3), 32.4 (3-CH2), 72.4 (2-C), 94.0 (C–S in
ylide), 116.2 (8-C), 117.2 (4a-C), 127.0 (7-C), 129.9 (5-C), 151.4
(8a-C), 154.1 (6-C), 172,2 (C–O in ylide), 191.8 (5a-CH2þ) 8.81,
123.01, 125.76, 145.04, 147.28. Anal. Calcd for C38H48O8S2 (696.9):
C, 65.49; H, 6.94; S, 9.20. Found: C, 65.42; H, 7.02; S, 9.02. It should
be noted that at rt, complex 7 is degraded into an equimolar mixture
of 3 and equimolar amounts of the spiro-dimer 8, which has the same
net composition. However, the correct microanalysis of that resulting
mixtures proves the purity of the parent complex 7 as well.
8. A comprehensive account of the structure of the complexes along with
computational treatments and synthetic applications will follow in
due course.
9. Due to the ring current effect, the resonances of the protons at C-7a,
C-8b, and C-4 experience a down-field shift in tocopherol (9) and
related derivatives, which evidently seems to be still operative in the
complex between 2 and 6a.
10. The 13C resonances of carbocations can range between 100 to above
300 ppm, see: Kalinowski, H. O.; Berger, S.; Braun, S. 13C NMR-
Spektroskopie; Georg Thieme: Stuttgart, 1984; p 370.
11. 13C resonances of quinoid carbons are usually found between 180 and
195 ppm, see Ref. 9, p 281.
12. General experimental procedure for the preparation of stabilized o-QMs
(4): Freshly prepared silver oxide (5 mmol) was suspended in ethyl
acetate or chloroform (150 mL) at temperatures between ꢀ10 °C and
ꢀ30 °C, maintaining this temperature throughout. A rt solution of
bis-ylide 3 (0.55 mmol) in DMSO (5 mL) was added at once. The
mixture was stirred vigorously and a solution of the ortho-alkylphenol
(1 mmol) in EtOAc or chloroform, respectively, was added dropwise
over about 10 min. After completion of the addition the mixture was
stirred for another 10 min and filtered through a cooled filter to
remove silver and excess silver oxide. In some cases the clear filtrate,
which is usually yellow, might be colored gray to black by colloidal
silver compounds, which does not influence reactivity or reaction
behavior. The cooled solution of the stabilized oQM was used for
further reactions. The oQMs were liberated by fast heating to 40 °C,
advantageously by dropping the solution into a solvent heated to
40 °C which contained the coreactant.
COOH
AcO
19
O
HO
O
OAc
HO
Ag2O, 3
18
20
Scheme 5. High-yield synthesis of (O-acetylsalicyl)saligenin (20) via the
stabilized oQMs from ortho-cresol undergoing nucleophilic substitution.
also an insecticide and a model for ‘biological’ oQMs. The
present stabilization allowed the reaction of the stabilized
oQM from ortho-cresol (18), in an O-alkylation process,
with 1.5 equiv of acetylsalicylic acid (19) to saligenin deriv-
ative 20 in 82% isolated yield20 (Scheme 5).
In all four examples, the formation of the dimerization
products of the oQMs was below 6%, which is remarkable
as the formation of (spiro-)dimers is usually the main reac-
tion path of oQMs in the absence of large excess of ‘exter-
nal’ coreactants. Further studies will have to exploit the
stabilization approach for organic synthesis. Similarly,
more general stabilization approaches, using the inter-
action of ionic species (as e.g., present in ionic liquids)
with the zwitterionic oxidation intermediate 2a, will have
to be tested.
Acknowledgment
The financial support of the Austrian Science Fund
(FWF), project P17426, is gratefully acknowledged.
References and notes
1. Merijan, A.; Gardner, P. D.; Shoulders, B. A. J. Org. Chem. 1963, 28,
2148–2149.
13. Rosenau, T.; Potthast, A.; Kosma, P.; Habicher, W. D. Synlett 1999,
3, 291–294.
2. See for example: (a) Chiang, Y.; Kresge, A. J.; Zhu, Y. J. Am. Chem.
Soc. 2001, 123, 8089–8094; (b) Foster, K. L.; Baker, S.; Brousmiche,
D. W.; Wan, P. J. Photochem. Photobiol. A: Chem. 1999, 129, 157–
163; (c) Wan, P.; Brousmiche, D. W. Pure Appl. Chem. 2001, 73, 529–
534; (d) Wan, P.; Barker, N. B.; Diao, L.; Fischer, M.; Shi, Y.; Yang,
C. Can. J. Chem. 1995, 74, 465–475.
14. NMR data were identical to those in Ref. 12. Purity: Anal. Calcd for
C
31H52O4 (488.75): C, 76.18; H, 10.72. Found: C, 76.06; H, 11.01.
15. (a) Penttila, A.; Kapadia, G. J. J. Pharm. Sci. 1965, 54, 1362–1364; (b)
Penttila, A.; Karpadia, G. J.; Fales, H. M. J. Am. Chem. Soc. 1965,
87, 4402–4403.
¨
16. NMR data were in line with previous reports: Ayra¨s, P.; Lo¨tjo¨nen, S.;
3. Rosenau, T.; Ebner, G.; Stanger, A.; Perl, S.; Nuri, L. Chem. Eur. J.
2005, 11, 280–287.
4. Rosenau, T.; Potthast, A.; Elder, T.; Kosma, P. Org. Lett. 2002, 4,
4285–4288.
5. Rosenau, T.; Mereiter, K.; Ja¨ger, C.; Schmid, P.; Kosma, P.
Tetrahedron 2004, 60, 5719–5723.
6. Van De Water, R. W.; Pettus, T. R. R. Tetrahedron 2002, 58, 5367–
5405. For the special case of the oQM derived from a-tocopherol see:
Ref. 4 and Rosenau, T.; Habicher, W. D. Tetrahedron 1995, 51, 7919–
7926.
7. Complex 7. 1H NMR (DMSO-d6, 400 MHz, 4 mg/0.6 mL, 0 °C): d
1.27 (s, 6H, 2 ꢁ 2a-CH3), 1.28 (s, 6H, 2 ꢁ 2a-CH3), 1.76 (t, 2H,
3J = 6.9 Hz, 2 ꢁ 3-CH2), 2.02 (s, 6H, 2 ꢁ 8b-CH3), 2.11 (s, 6H,
2 ꢁ 7a-CH3), 2.61 (t, 4H, 3J = 6.9 Hz, 2 ꢁ 4-CH2), 2.94 (s, 6H, 2 ꢁ
–S+–Me), 2.98 (s, 6H, 2 ꢁ –S+–Me), 5.85 (s, 4H, 2ꢁ, 5a-CH2þ), 13C
Widen, C. J. Org. Magn. Reson. 1981, 16, 209–213. Purity: Anal.
Calcd for C24H30O8 (446.50): C, 64.57; H, 6.72. Found: C, 64.76; H,
6.88.
17. Rosenau, T.; Adelwo¨hrer, C.; Kloser, E.; Mereiter, K.; Netscher, T.
Tetrahedron 2006, 62, 1772–1776.
18. NMR data were consistent with the literature: Schro¨der, H.;
Netscher, T. Magn. Reson. Chem. 2001, 39, 701–708. Purity: Anal.
Calcd for C57H96O4 (845.40): C, 80.98; H, 11.45. Found: C, 81.13; H,
11.32.
19. Clausen, T. P.; Keller, J. W.; Reichardt, P. B. Tetrahedron Lett. 1990,
31, 4537–4538.
20. NMR data agreed with those of the deacetylated derivative: Buss, T.,
Dissertation, Marburg University, Germany, 2005. Purity: Anal.
Calcd for C16H14O5 (286.28): C, 67.13; H, 4.93. Found: C, 67.02; H,
5.26.