Asymmetric biomimetic oxidations of phenols using oxazolidines as chiral auxiliaries: The enantioselective synthesis of (+)- and (-)-dehydrodiconiferyl alcohol

Article abstract of DOI:10.1002/poc.1089

Stereoselective bimolecular radical coupling reactions of phenylpropenoid phenols are described. Evans's 2-oxazolidinone 11a-d derivatives of ferulic acid were prepared and oxidized to give dimeric benzofuran neolignan structures 12-13a-d in 40-50% overal

Full text of DOI:10.1002/poc.1089

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2006; 19: 592–596
Published online 2 August 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/poc.1089
Asymmetric biomimetic oxidations of phenols using
oxazolidines as chiral auxiliaries: the enantioselective
synthesis of (þ)- and (ꢀ)-dehydrodiconiferyl alcoholy
1
2
Maurizio Bruschi,2 Marco Orlandi,2 Bruno Rindone,2 Petteri Rummakko and Luca Zoia
*
1Department of Chemistry, FIN-00014, University of Helsinki, Finland
2Department of Environmental Sciences, University of Milan, Piazza della Scienza, 1, 20126 Milan, Italy
Received 10 October 2005; revised 29 March 2006; accepted 13 April 2006
ABSTRACT: Stereoselective bimolecular radical coupling reactions of phenylpropenoid phenols are described.
Evans’s 2-oxazolidinone 11a–d derivatives of ferulic acid were prepared and oxidized to give dimeric benzofuran
neolignan structures 12–13a–d in 40–50% overall yields. The chiral phenols were dimerized either enzymatically with
hydrogen peroxide and horseradish peroxidase (HRP) or with silver oxide. The enantioselectivity after reductive
cleavage of the chiral auxiliaries to give dehydrodiconiferyl alcohol ranged from 18% to 62% enantiomeric excess.
The conformational analysis and the activation energy using semiempirical PM3 calculations on the intermediate
quinomethides is used to explain the observed stereoselectivity. Copyright # 2006 John Wiley & Sons, Ltd.
KEYWORDS: enantioselection; lignans; horseradish peroxidase; oxidative phenol coupling; conformational analysis;
semiempiric calculations
INTRODUCTION
Moderate stereoselectivity had been observed by
Charlton et al.9 in the reaction of chiral sinapate esters
with FeCl3 to give aryltetralin lignan structures. Moreover,
the biosynthetic pathway to enantiopure (þ)-pinoresinol
lignan has been elucidated quite recently. In fact, a protein
has been isolated from Forsythia suspensa and found to be
responsible for the formation of enantiomerically pure
(þ)-pinoresinol from E-coniferyl alcohol.10
Ferulic acid derivatives 1 are dimerized to dehydrodimers
7 and 8 (Scheme 1) to give almost exclusively the racemic
trans isomer if R is achiral.1 This has been suggested
recently to be the result of diastereocontrol in the
cyclization of the intermediate quinomethide2,3 deriving
from the dimerization of a persistent phenoxy radical.4,5
An initial single electron transfer from the starting ferulic
acid derivative 1 gives the intermediate p-complex 2,
which undergoes carbon–carbon bond formation in a
reversible way6,7 to give the isomeric quinomethides 3–6,
possessing, respectively, the (E) and (Z) configuration at
the exocyclic quinonoid double bond and the (R) and (S)
configuration at the stereogenic C-3 carbon.
We demonstrated recently that stereocontrol in
enzymatic oxidative coupling reactions of phenylprope-
noidic phenols is possible. In fact, ferulic acid amides
with chiral amino acids were oxidized with hydrogen
peroxide in the presence of horseradish peroxidase (HRP)
as the catalyst to give the dehydrodimers 7 and 8(R ¼ —
C(O)—NH—CH(COOH)—C6H5) in different amounts.
Chromatographic separation and hydrolysis allowed to
show that 65% enantioselectivity had been obtained.8
In a more recent approach,2 stereoselective bimole-
cular radical coupling of enantiopure phenylpropenoidic
phenols starting from enantiopure ferulic acid amides
with Oppolzer camphor sultam gave the enantioselective
phenol coupling with enantiomeric excess 80–84%. The
conformational analysis of the quinomethide intermedi-
ates showed the reasons of the diastereoselectivity in the
formation of reaction products.
EXPERIMENTAL
Melting points were determined with a Bu¨chi apparatus
and are uncorrected. IR spectra were recorded with a FT-
IR Jasco spectrophotometer. Mass spectra were measured
by the direct injections system mode with positive
electron impact with a VG 7070 EQ instrument. 1H NMR
spectra were taken with
a
Bruker AMX 300
*Correspondence to: B. Rindone, Dipartimento di Scienze dell’Am-
biente e del Territorio, Universita’ di Milano-Bicocca, Piazza della
Scienza, 1, I-20126 Milano, Italy.
instrument (in CDCl3 solutions). Chemical shifts are
given as ppm from tetramethylsilane and J values are
given in Hz. Optical rotations were recorded on a Perkin
Elmer 241 polarimeter at the sodium D line at 25 8C.
E-mail: Bruno.Rindone@unimib.it
ySelected paper presented at the 10th European Symposium on Organic
Reactivity, 25–30 July 2005, Rome, Italy.
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 592–596

BIOMIMETIC OXIDATIONS OF PHENOLS
593
Compounds 10a–b showed: 1H NMR: 7.58 (d,
J ¼ 15.0, 1H), 7.18–7.10 (m, 16H), 6.57–6.60 (m, 3H),
6.31 (d, J ¼ 15.0, 1H), 4.67 (m, 1H), 4.35, (dd, J ¼ 9,9 Hz,
1H), 4.20 (dd, J ¼ 9,9 Hz, 1H), 3.95 (s, 3H), 2.77 (m-2H);
IR (nujol): 3200, 1462 cmꢀ1; compounds 10c–d showed:
1H NMR: 7.55 (d, J ¼ 15.0, 1H), 7.18–7.10 (m, 16H),
6.57–6.60 (m, 3H), 6.31 (d, J ¼ 15.0, 1H), 4.97 (m,
J ¼ 9,9 Hz ,1H), 4.73, (dd, J ¼ 9,9 Hz, 1H), 4.30 (dd,
J ¼ 9,9 Hz, 1H), 3.95 (s, 3H), IR (nujol): 3200,
1462 cmꢀ1
.
Oxidative phenol coupling of compound
10(a–d) catalyzed by horseradish
peroxidase (HRP)
A solution of 10(a–d) (1.0 mmol) in the appropriate
solvent (14 mL) and 0.02 M phosphate/citric acid buffer
pH 3.5 (4.0 mL) was added of a 0.86 M aqueous hydrogen
peroxide solution (0.60 mL, 0.5 mmol) and aqueous HRP
(0.93 mL, 837 U) at 0 8C in small portions over 15 min.
The mixture was then stirred at 0 8C. After 4 h, a saturated
aqueous NaCl solution (20 mL) was added. The organic
solvent was then removed under reduced pressure, and the
resulting solution was extracted with AcOEt (4 ꢁ 20 mL).
The combined organic extracts were washed with 10%
aqueous NaHCO3 (25 mL), then with water (25 mL), and
dried over Na2SO4. The solvent was evaporated under
reduced pressure, and the residue was purified by silica
gel flash cromatography (eluent toluene-AcOEt, gradient
mode, from 4:1 to 1:1) yielding a mixture of phenylcou-
marans 11(a–d) and 12(a–d).
Scheme 1
HPLC analysis were performed on a WATERS 600 E
instrument by using an HP 1040 Diode Array Detector.
Ag2O-promoted oxidative phenol
coupling of compound 10(a–d)
Silver(I)oxide (0.18 g, 0.8 mmol) was added to a solution
of 11(a–d) (0.5 mmol) in dry CH2Cl2 (5.0 mL) under
argon atmosphere at room temperature. After stirring for
24 h, the mixture was filtered through celite and
evaporated under reduced pressure. The residue was
purified by silica gel flash chromatography (eluent
toluene-AcOEt, gradient mode, from 4:1 to 1:1) yielding
a mixture of phenylcoumarans 11(a–d) and 12(a–d). The
two diastereomers were then separated by preparative
HPLC (isocratic mode, CH3CN—H2O 1:1). A 40% yield
was obtained.
Preparation of compound 10(a–d)
A mixture of ferulic acid 9 (427 mg, 2.2 mmol) , 2-chloro-
1-methylpyridine iodide (670 mg, 2.6 mmol), and the
appropriate chiral oxazolidinone (2.4 mmol) were dis-
solved in 6 mL of dry CH2Cl2 under nitrogen atmosphere.
A solution of triethylamine (0.3 mL 5.6 mmol), dissolved
in 2 mL of dry CH2Cl2, was added dropwise for 15 min.
The mixture was stirred at 80 8C for 120 h, then 10 mL of
a saturated aqueous NaCl solution were added. The
organic phase was separated, then washed with a pH 4
solution of aqueous HCl (10 mL), with a 10% aqueous
NaHCO3 solution (10 mL), with water (10 mL), and
finally dried over Na2SO4. The solvent was evaporated
under reduced pressure, and the residue was purified by
silica gel flash chromatography, (eluent CH2Cl2-acetone
7:3). Yields >40%.
1
Compound 12c showed H NMR: 7.58 (d, J ¼ 15.0,
1H), 7,28–7.20 (m, 10H), 6.55–6.60 (m, 5H), 6.31 (d,
J ¼ 15.0, 1H), 6.12 (d, J ¼ 8.0, 1H), 5.90 (dd, J ¼ 9,9, 1H),
5.70 (dd, J 9,9, 1H), 4.97 (dd, J ¼ 9,9 Hz, 1H), 4.80 (dd,
J ¼ 9,9 Hz, 1H), 4.73 (dd, J ¼ 9,9 Hz, 1H), 4.60 (dd,
J ¼ 9,9 Hz, 1H), 4.30 12 (d, J ¼ 8.0, 1H), 3.95 (s, 3H),
3.90 (s, 3H), IR (nujol): 3200, 1462 cmꢀ1; HREIMS of
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 592–596

594
M. BRUSCHI ET AL.
12c calculated for C22H44O8N2: 433.1559; found
433.1573.
Reductions of phenylcoumarans 12(a–d) or
13(a–d) to dehydrodiconiferyl alcohol 7–8
(R = CH2OH)
Phenylcoumarans 12(a–d) or 13(a–d) (0.024 mmol) were
dissolved in dry THF (5 mL) under argon at ꢀ78 8C.
LiBH4 (1 mg, 0.054 mmol) was suspended in dry THF
(1.0 mL) and added to the reaction mixture, which was
further stirred for 2 h at ꢀ78 8C. After diluting with
aqueous THF (10 mL), aqueous 0.1 M ammonium
chloride (5 mL) was added. The mixture was extracted
with AcOEt (2 ꢁ 10 mL), and the combined organic
extracts were washed with water (10 mL), then dried over
Na2SO4. The solvent was evaporated under reduced
pressure, and the residue was analysed by HPLC with a
chiral column (Chiralcell OF; isocratic mode, eluent
hexane-isopropanol 1:1). The absolute configurations of
dehydrodiconiferyl alcohols 7–8 (R ¼ CH2OH) were
determined by comparison with literature data.11
RESULTS AND DISCUSSION
Two chiral 2-oxazolidinones,12 namely (R-) and (S-)
2-phenyl- and (R-) and (S-) 2-benzyloxazolidinine
were used to induce asymmetry in these oxidative
dimerizations.
Scheme 2
Chiral compounds 10a–d were prepared by using a
‘‘one pot’’ reaction from ferulic acid 9 and chiral 2-
oxazolidinones in the presence of 2-chloro-1-methylpyr-
idine iodide with a yield of 46% (Scheme 2). Compounds
10a–d were thus oxidized using HRP/H2O2 or with silver
oxide6,13 to give the diastereoisomeric dehydrodimers
11–12a–d. Enzymatic oxidations were performed in
aqueous buffer systems and acetone or dioxane was used
as a co-solvent. Silver oxide oxidations were performed in
dry dichloromethane. After dimerization, oxazolidinone
fragments were removed by reduction with LiBH4/THF14
to give the mixture of enantiomeric dehydrodiconiferyl
alcohols 7–8 (R ¼ CH2OH).
prevented the reaction completely. It seems that the use of
Lewis acids to control conformational rotamers via
chelative interactions is not possible in phenol oxidations.
The temperature and solvent system had only small effect
to the observed stereoselectivity.
In our experiments, chiral auxiliaries with opposite
chirality gave opposite chirality of the resulting trans-
dehydrodimer 11–12a–d, thus indicating that the
observed diastereoselection was due to the influence of
the stereogenic centers of the chiral auxiliaries. Hence,
the nature of the product-determining step was investi-
gated in order to understand the reasons of the observed
diastereoselectivity in the formation of the diastereoiso-
meric dehydrodimers 11–12a–d.
These quinomethides undergo nucleophilic attack of
the phenolic oxygen to the quinomethide double bond to
give cyclization to the final phenylcoumarans 11–12c–d
Two diastereofaces are involved in this reaction. Hence,
chiral induction from the intermediate quinomethide may
result in diastereoselectivity in the cyclization reaction
and the stereochemistry of the product phenylcoumarans
11–12c–d may be predicted if the energy of activation of
Table 1 shows the results thus obtained. In these
experiments, the sterically smaller chiral auxiliaries
2-phenyloxazolidinone gave considerably higher selec-
tivities (e.e. 62%) compared to the corresponding
2-benzyl oxazolidinone chiral auxiliary (e.e. 21%). This
observation is opposite to Evans results in Diels–Alder
and alkylation reactions15 and the results of Sibi et al. in
radical allylation reactions with similar auxiliaries.16 The
attempts to use of Lewis acid additives in phenol
oxidations to restrict the rotational freedom of N-
acyloxazolidinones failed. In fact, addition of MgBr2
or Zn-triflate to Ag2O oxidation reaction mixture of 10a
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 592–596

BIOMIMETIC OXIDATIONS OF PHENOLS
595
Table 1. Oxidations of chiral phenols
Absolute configuration
of the major enantiomer 7–8
Chiral auxiliary
Oxidant
Solvent
T/ 8C
e.e %
(R)-benzyl (10a)
(R)-benzyl (10a)
(S)-benzyl (10b)
(S)-benzyl (10b)
(R)-phenyl (10c)
(R)-phenyl (10c)
(R)-phenyl (10c)
(S)-phenyl (10d)
HRP/H2O2
Ag2O
HRP/H2O2
Ag2O
HRP/H2O2
HRP/H2O2
Ag2O
Dioxane/buffer pH 3.5 8/2
CH2Cl2
Dioxane/buffer pH 3.5 8/2
CH2Cl2
Acetone/Buffer pH 3.5 8/2
Acetone/buffer pH 3.5 8/2
CH2Cl2
25
25
25
ꢀ20
ꢀ20
0
2R,3S-(ꢀ)
2R,3S-(ꢀ)
2S,3R-(þ)
2S,3R-(þ)
2R,3S-(ꢀ)
2R,3S-(ꢀ)
2R,3S-(ꢀ)
2S,3R-(þ)
21
18
21
20
62
62
53
59
ꢀ20
HRP/H2O2
Acetone/buffer pH 3.5 8/2
0
the cyclization reaction is nearly the same for all the
cyclization reactions leading to trans-phenylcoumarans.
The conformational analysis on rotating the C2–C3
and Ca’–C3 bond in the intermediate quinomethides
3–6 was performed using the semiempirical PM3
Hamiltonian. This analysis allowed to predict the
amount of each isomeric quinomethide at the
equilibrium. The potential energy surfaces for the
most stable intermediates are shown in Figs. 1 and 2.
Here, using (R)-2-phenyloxazolidinone as the chiral
auxiliary, the isomer having chirality (R) at C3, Z-
arrangement of the double bond and two (R)-2-
phenyloxazolidinone fragments (the 3-SERR isomer)
resulted about 2 kcal molꢀ1 more stable than the 3-
RERR isomer).
(2R,3S) dehydrodiconiferyl alcohol 8 (R ¼ CH2OH). This
prediction is in line with the observation that (2R,3S)
reaction product was the major isomer experimentally
found with e.e. 52–63%.
In the same way, the phenolic oxygen of few lower
energy conformers of the less stable 3-RERR quino-
methide isomer will attack the (si) face of the
quinomethide double bond and this predicts for the
formation of (2S,3R) dehydrodiconiferyl alcohol 7
(R ¼ CH2OH), which was in fact formed in minor
amounts.
As expected, opposite results were obtained using (S)-
2-phenyloxazolidinone as the chiral auxiliary. Here, the
3-SESS isomer of the intermediate quinomethide 3–6 was
more than 1 kcal molꢀ1 less stable than the 3-RZSS
isomer.
The 3-RZSS isomer may be predicted to attack the (si)
face of the quinomethide double bond to give the (2S–3S)
phenylcoumaran 11d. The subsequent reduction with
LiBH4 will then afford the (2S,3R) dehydrodiconiferyl
The phenolic oxygen of several minimum energy
conformers of the most stable 3-SERR isomer is suitably
located to attack the (re) face of the double bond of the
quinomethide to give the (2R–3R) phenylcoumaran 12c.
The subsequent reduction with LiBH4 will then afford the
Figure 1. Potential energy surface for quinomethide
Figure 2. Potential energy surface for quinomethide
3-SERR for R ¼ (R)-2-phenyloxazolidinone. Angle
w
is
3-RERR for R ¼ (R)-2-phenyloxazolidinone. Angle
w is
Ca0-C3; angle u is C2–C3
Ca0-C3; angle u is C2–C3
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 592–596

596
M. BRUSCHI ET AL.
alcohol 7 (R ¼ CH2OH). This is in line with the
observation that (2S,3R)-(ꢀ) reaction product was the
major isomer experimentally found with e.e. 59%.
The 3-SESS isomer may be predicted to attack the (re)
face of the quinomethide double bond to form, after
reduction, (2R,3S)-(þ) dehydrodiconiferyl alcohol 8
(R ¼ CH2OH), which was in fact formed in minor
amounts.
structures thus providing a new approach to the synthesis
of valuable lignans.
Acknowledgements
We thank our student Anna Bianchi for technical help.
These predictions require that trans-cyclization reac-
tions have similar activation energy. The activation
energy for the transformation of both 3-SERR and
3-RERR phenolate anions to the corresponding dimers
REFERENCES
¨
1. Chioccara F, Poli S, Rindone B, Pilati T, Brunow G, Pietikainen P,
¨ ¨
was calculated by PM3 to be 8.32 and 6.67 kcal molꢀ1
respectively.
,
Setala H. Acta Chem. Scand. 1993; 47: 610–616.
2. Orlandi M, Rindone B, Molteni G, Rummakko P, Brunow G.
Tetrahedron 2001; 57: 371–378.
Moreover, the fact that cis-cyclization was never
observed in spite of the fact that the quinomethide
conformations undergoing this reaction are sometimes
relatively stable implies that cis-cyclization involves
higher activation energy than the trans-cyclization.
This was clear observing the results from (R)-2-
benzyloxazolidinone as the chiral auxiliary. Many
conformations of the most stable isomers predict for
the cis-cyclization, which was never observed. Hence, in
the case of the more flexibile benzyloxazolidinone
moiety, quinomethide stability is unable to predict the
enantioselectivity of the reaction.
In conclusion, these results show that chiral auxiliaries
provide significant levels of diastereoselection in bimo-
lecular coupling reactions of phenoxyl radicals, and this
results in enantioselection in the final product. Evaluation
of the stability of the intermediate quinomethides allows
to predict the degree of enantioselection. It is expected
that this methodology could be extended to various lignan
3. Chance M, Powers L, Poulos T. Biochemistry 1986; 25: 1266–
1270.
4. Ward RS. Nat. Prod. Rep. 1997; 14: 43–74.
5. Hiray N, Okamoto M, Udagawa H, Yamamuro M, Kato M,
Koshimizu K. Biosci. Biotech. Biochem. 1994; 58: 1678–1684.
6. Amstrong DA, Cameron C, Nonhebel DC, Perkins PG. J. Chem.
Soc., Perkin Trans. 21983; 563–568.
7. Bolzacchini E, Meinardi S, Orlandi M, Rindone B, Brunow G,
Pietikainen P, Rummakko P. In Green Chemistry, Tundo P. Anastas
P (eds). Oxford University Press: London, 2000.
8. Bolzacchini E, Brunow G, Meinardi S, Orlandi M, Rindone B,
¨ ¨
Rummakko P, Setala H. Tetrahedron Lett. 1998; 39: 3291–3294.
9. Bogucki DE, Charlton JL. Can. J. Chem. 1997; 75: 1783–1794.
10. Davin LB, Wang HB, Crowell AL, Bedgar DL, Martin DM,
Sarkanen S, Lewis NG. Science 1997; 275: 362–366.
11. Hiray N, Yamamuro M, Koshimizu K, Shinozaki M, Takimoto A.
Plant Cell. Physiol. 1994; 35: 691–695.
12. Rummalko P, Brunow G, Orlandi M, Rindone B. Synlett 1999;
333–335.
13. Quideau S, Ralph J. Holzforschung 1994; 48: 12–22.
14. Penning TD, Djuric SW, Haack RA, Kalish VJ, Miyashiro JM,
Rowell BW, Yu SS. Synth. Commun. 1990; 20: 307–312.
15. Evans DA, Chapman KT, Bisaha J. J. Am. Chem. Soc. 1988; 110:
1238–1256.
16. Sibi MP, Ji J. Angew. Chem. Int. Ed. Engl. 1996; 35: 190–192.
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 592–596