Aqueous hemin catalyzed sulfonium ylide formation and subsequent [2,3]-sigmatropic rearrangements

Article abstract of DOI:10.1039/c6gc02681h

A mild hemin catalytic system for sulfonium ylide generation via a metal carbenoid and a subsequent [2,3]-sigmatropic rearrangement reaction in aqueous solvent is well-established, with the assistance of cyclodextrin (CD) and Triton X-100. The protocol displays high catalytic activity with a broad substrate scope of aryl/alkyl allyl sulfides and diazo reagents, affording homoallyl sulfide products in up to 99% yield. Notably, this catalytic system is successful for water-insoluble allyl sulfides but ineffective for allyl amines or allyl ethers. Moreover, an unprecedented cascade reaction of sulfonium ylide formation, [2,3]-sigmatropic rearrangement and C-H insertion was reported.

Full text of DOI:10.1039/c6gc02681h

Green Chemistry
View Article Online
View Journal
COMMUNICATION
Aqueous hemin catalyzed sulfonium ylide
formation and subsequent [2,3]-sigmatropic
Cite this: DOI: 10.1039/c6gc02681h
rearrangements†
Received 26th September 2016,
Accepted 13th January 2017
Xiaofei Xu, Chang Li, Zhihao Tao and Yuanjiang Pan*
DOI: 10.1039/c6gc02681h
www.rsc.org/greenchem
A mild hemin catalytic system for sulfonium ylide generation via a ation of cost, safety and environmental sustainability, it is of
metal carbenoid and a subsequent [2,3]-sigmatropic rearrange- great significance to perform this reaction in water-based
ment reaction in aqueous solvent is well-established, with the green solvents. To the best of our knowledge, the first example
assistance of cyclodextrin (CD) and Triton X-100. The protocol dis- of the Doyle–Kirmse reaction carried out in aqueous solution
plays high catalytic activity with a broad substrate scope of aryl/ was reported by Wang and his co-workers in 2007 using Rh[II]
alkyl allyl sulfides and diazo reagents, affording homoallyl sulfide as the catalyst.5b
products in up to 99% yield. Notably, this catalytic system is suc-
Iron porphyrins (especially hemin) are abundantly distribu-
cessful for water-insoluble allyl sulfides but ineffective for allyl ted in nature, acting as a catalytic oxidation cofactor in
amines or allyl ethers. Moreover, an unprecedented cascade reac- different enzymes. In the last few years, Arnold12 and Fasan13
tion of sulfonium ylide formation, [2,3]-sigmatropic rearrangement have made big progress in hemoprotein-catalyzed carbenoid-
and C–H insertion was reported.
related reactions. In 2015, Fasan and co-workers reported
mutmyoglobin-catalyzed sigmatropic rearrangement of allyl
sulfide,13d which was subsequently further investigated system-
atically.13e Very recently, Arnold and co-workers demonstrated
that a sulfimidation/[2,3]-sigmatropic rearrangement sequence
can be conducted in whole cell containing cytochrome var-
iants.12b However, the utility of iron porphyrin in organic syn-
thesis is very limited, due to its low solubility in both organic
and aqueous solvents. Gross has previously reported the cata-
lytic activity of iron corroles and porphyrins for the reactions
of diazoacetates with nucleophilic substrates.9b Recently, we
developed hemin-catalyzed intermolecular N–H insertion of
diazo carbonyl compounds under aqueous conditions with the
assistance of cyclodextrin.14 With our continuing interest in
expanding the catalytic ability of hemin in carbenoid chem-
istry, herein, we established a highly efficient aqueous system
for [2,3]-sigmatropic rearrangement of sulfonium ylide derived
from iron-carbenoid and allyl sulfide, using hemin as a cata-
lyst, and cyclodextrin and Triton X-100 as promoters.
Firstly, phenyl allyl sulfide (1a) and ethyl diazo acetate
(EDA, 2a) were chosen as the model substrates to optimize the
reaction conditions (Table 1). When the hemoenzyme HRP
was employed as the catalyst in this reaction, no desired
product was detected (entry 1). However, when the reaction
was treated with 5 mol% of its cofactor, hemin, in water at
40 °C for 48 hours, the desired product 2-(phenylthio)-4-pente-
noic acid ethyl ester (3a) was obtained in 46% yield according
to 1H NMR analysis (entry 2). Encouraged by the results in our
previous work, we tried cyclodextrins as additives. As expected,
Diazo compounds have attracted great attention for decades,
because of their broad applications in organic synthesis via
thermal, photochemical, or catalytic processes with extrusion
of dinitrogen. Among all these reactions, metal carbenoids
generated through the reaction between transition metal-
catalysts and diazo compounds have been deeply explored to
achieve diverse transformations, such as C–H insertion, X–H
insertion (X = N, S, O, Si, B, etc.), cyclopropanation, and ylide
formation.1
The [2,3]-sigmatropic rearrangement of sulfonium ylide
generated from a transition metal carbenoid and allyl sulfide,
namely the Doyle–Kirmse reaction,2 has emerged as a powerful
synthetic strategy for C–C bond construction.3 Since its discov-
ery in the 1960s, numerous studies have reported that copper4-
and rhodium5-based catalysts are effective for this type of ylide
transformation. Subsequently, a variety of other transition-
metal complexes based on ruthenium,6 palladium,7 cobalt,8
iron,9 gold10 and silver11 have also been proven as capable
catalysts. However, most of the reported catalytic systems are
conducted in organic solvent under an inert atmosphere. As
water is a desirable solvent for chemical reactions in consider-
Department of Chemistry, Zhejiang University, Hangzhou 310027, People’s Republic
of China. E-mail: panyuanjiang@zju.edu.cn; Fax: +8657187951629;
Tel: +8657187951629
†Electronic supplementary information (ESI) available: Experimental pro-
cedures, characterization data. See DOI: 10.1039/c6gc02681h
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Communication
Green Chemistry
Table 1 Optimization of the reaction conditionsa
tions, our approach possesses a relatively lower E-factor of
about 1.7 (for details see the ESI†).
In most previously reported cases, aryl allyl sulfides mainly
result in the cyclopropanation products of a double bond or
the homo-coupling products of the diazo compound.6b
Additionally, iron porphyrin complexes have been proven to
be active catalysts for the cyclopropanation of alkenes with
EDA.16 However, no cyclopropanation or [1,2]-rearrangement
product was detected in our catalytic system. Only a small
amount of the byproduct EDA-dimer was observed.
Having identified the optimal conditions, we next examined
the substrate scope of this catalytic reaction system. Firstly,
various allyl sulfides were tested under these standard reaction
conditions using EDA as the carbene precursor as shown in
Table 2. All the allyl aryl sulfides underwent the Doyle–Kirmse
reaction smoothly, and the electronic properties of the substi-
tuent on the phenyl ring of the sulfide did not show an
obvious influence on the reactivity. Sulfides bearing electron-
donating or electron-withdrawing substituents at either the
meta or the para position on the phenyl ring were well toler-
ated, giving fairly good to excellent yields of the desired pro-
2a/ Catalysts
Entry 1a (mol %)
β-CD
Surfactant
(mol %) (mol %)
Yieldb
(%)
1
2
3
4
5
6
7
8
1
1
1
1.5
HRP (0.01)
Hemin (5)
Hemin (5)
—
—
—
20
20
20
20
20
—
5
10
20
20
20
20
20
20
—
—
—
—
—
—
—
—
—
—
—
—
0
46
60
0
1.5 Hemin (1)
1.5 Hemin (2.5)
1.5 Hemin (5)
1.5 Hemin (2.5)
1.5 Hemin (2.5)
1.5 Hemin (2.5)
1.5 Hemin (2.5)
65
77
71
65
65
68
77
83
57
52
70
93
9
10
11
12
13
14
15
16
2
2
2
2
2
Hemin (2.5)
Hemin (2.5)
Hemin (2.5)
Hemin (2.5)
Hemin (2.5)
TBAB (2.5)
PEG2000 (2.5)
NaDDBS (2.5)
Triton X-100
(2.5)
Triton X-100 (10) 58
Triton X-100
(2.5)
17
18
2
2
Hemin (2.5)
FeCl3·6H2O
—
20
0
19
2
Cu(II) protoporphyrin 20
IX
Triton X-100
(2.5)
0
Table 2 Hemin-catalyzed Doyle–Kirmse reaction of various sulfidesa,b
a Reaction was carried out with 3 mL of H2O at 40 °C for 48 h in a
thermo shaker. b Yield was determined by 1H NMR analysis of the
crude reaction mixture.
the yield of 3a improved to 60% when 20 mol% β-CD was
added to this reaction system (entry 3, the screening results of
different CDs are presented in Table S1, ESI†). Meanwhile, a
control reaction performed in the absence of hemin led to no
detectable product (entry 4), indicating that cyclodextrin itself
does not exhibit any catalytic ability. Further screening showed
that a 2.5 mol% and 20 mol% loading amount of hemin and
β-CD respectively gave a 77% yield of the product (entries
5–11). Subsequently, two equivalents of 2a promoted the reac-
tion to 83% yield (entry 12, see the ESI† for details).
Considering the aggregation of hemin in water, frequently
used surfactants, such as tetrabutylammonium bromide
(TBAB), PEG 2000, sodium dodecylbenzene sulfonate
(NaDDBS) and Triton X-100, were introduced in the reaction
system (entries 13–16). To our delight, when 2.5 mol% of the
non-ionic surfactant Triton X-100, which has been previously
employed as a dispersing agent for hemin,15 was added, the
yield of 3a further improved to 93% (entry 16), while other sur-
factants had no obvious effect on the transformation. A
control experiment in the absence of β-CD only gave rise to
58% product formation (entry 14), indicating that both cyclo-
dextrin and Triton X-100 contribute synergistically to the cata-
lytic system. Notably, the simple iron salt FeCl3·6H2O (entry
18) and Cu(II) protoporphyrin IX (entry 19) had no catalytic
effect in this reaction under the same reaction conditions,
revealing that hemin plays an irreplaceable role in this cata-
lytic system. Compared with the reported Doyle–Kirmse reac-
a Reaction conditions: EDA (0.6 mmol) was added to a water solution
of the sulfide (0.3 mmol), hemin (2.5 mol%), Triton X-100 (2.5 mol%)
and β-CD (20 mol%). The reaction was carried out at 40 °C for 48 h in
a thermo shaker. b Yield was determined by 1H NMR analysis of the
crude reaction mixture. Values in parentheses are isolated yields.
c Yield of the gram-scale reaction. d d.r. = diastereomeric ratio, deter-
mined by 1H NMR.
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Communication
ducts (3a–b and 3d–f), except for the para ethyl-substituted
and para nitro-substituted materials, giving moderate results
(3c and 3g). When the R1 group was 2-pyridyl or 2-naphthyl,
the corresponding rearranged products were obtained in high
yields (3h and 3i). Notably, alkyl allyl sulfides also exhibited
excellent reactivity toward the ylide-involved [2,3]-sigmatropic
rearrangement reaction (3n–p). Moreover, the effect of the sub-
stitutions on the allyl unit was also studied. Introduction of
methyl groups into the allyl unit had only marginal effect on
the reaction, resulting in 86% and 89% yields respectively for
3j and 3k, which is distinct from the silver-catalyzed proto-
col.11 Due to the steric effect, cinnamylic phenyl sulfide
afforded a 65% diastereomeric mixture of 3l in a 56 : 44 ratio.
In addition, a gram-scale synthesis of 3a was performed under
the optimized reaction conditions giving a 71% isolated yield
(1.678 g). Unfortunately, when we examined water-soluble
S-allyl-L-cysteine, neither the desired rearranged product (3q)
nor the N–H/O–H insertion product was detected. We rational-
ized that the aqueous hemin catalytic system might have some
“on water” effect,17 or the soluble substrate can’t enter the
hydrophobic cavity of β-CD. Further investigation showed that
the [2,3]-sigmatropic rearrangement reaction also worked well
for phenyl propargyl sulfide affording the corresponding
allene product 3r in 65% yield.
Scheme 1 Results from the reaction of EDA with diallyl sulfide.
Reaction conditions: EDA was added to a water solution of the sulfide
(0.3 mmol), hemin (2.5 mol%), Triton X-100 (2.5 mol%) and β-CD
(20 mol%). The reaction was carried out at 40 °C for 48 h in a thermo
shaker. a Yield was determined by 1H NMR analysis of the crude reaction
mixture.
When the diallyl sulfide was employed as a substrate, the
outcome of the reaction mainly relies on the amount of EDA
used (Scheme 1). The standard reaction conditions preferred
the single rearranged product (3s-1). In contrast, the major
product was shifted to the double-rearranged product (3s-2)
when 4 equivalents or 8 equivalents of EDA were added.
Notably, another unexpected minor product (3s-3) was also iso-
lated, which was formed through subsequent C–H insertion by
EDA at the tertiary carbon of a single rearranged product.
However, our attempts to further improve the efficiency of the
cascade process failed.
Table 3 Hemin-catalyzed Doyle–Kirmse reaction of different diazo
compoundsa
As an extension of this catalytic system, the diazo reagent
scope of the Doyle–Kirmse reactions was investigated under
the optimized conditions (Table 3). When (trimethylsilyl)di-
azomethane (TMSD) was subjected to the rearrangement
system, 3t was obtained in 69% yield in the presence of 5
equivalents of TMSD. Consistent with Plietker’s work,9a TMSD
showed lower reactivity than EDA, albeit with the opposite con-
clusion to that proposed by Van Vranken’s work on the
dppeFeCl2 system.9c,18 tert-Butyl diazoacetate also turned out
to be less reactive than EDA, furnishing the product in a rela-
tively modest yield (3u, 35%), presumably due to the steric
issue. However, fairly good yield (3v, 79%) was observed when
benzyl diazoacetate was used. Moreover, various aryl diazoace-
tates were also investigated, which turned out to be much
more sluggish than alkyl diazoacetates. Methyl phenyldiazo-
acetate (MPDA) and ethyl phenyldiazoacetate (EPDA), as accep-
tor–donor diazo species, led to the corresponding desired
product in good yield (83%, 78% for 3w and 3x respectively) at
a higher reaction temperature, 80 °C, in which the reaction
time was also elongated to 20 hours. In addition, the
rearrangement product of methyl naphthyldiazoacetate was
a Reaction conditions: the diazo compound (0.6 mmol) was added to a
water solution of 1a (0.3 mmol), hemin (2.5 mol%), Triton X-100
(2.5 mol%) and β-CD (20 mol%). The reaction was carried out at 40 °C
for 48 h in a thermo shaker. Yield was determined by 1H NMR analysis
of the crude reaction mixture. Values in parentheses are isolated
yields. b The molar ratio 1a/2t was 1/5. c The reaction was stirred at
80 °C for 20 h.
generated in moderate yield (3y, 59%) under standard reaction
conditions.
To further explore the potential of this catalytic system, allyl
phenyl amines and allyl ether were tested under the optimal
reaction conditions (Scheme 2). However, none of the products
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Communication
Green Chemistry
Acknowledgements
Financial support from the National Natural Science Foundation
of China (21327010, 21372199) is gratefully acknowledged.
Notes and references
Scheme 2 Results from the reactions of EDA with allyl amines and allyl
ether.
1 For selected reviews, see: (a) M. P. Doyle, Acc. Chem. Res.,
1986, 19, 348; (b) M. P. Doyle, Chem. Rev., 1986, 86, 919;
(c) A. Ford, H. Miel, A. Ring, C. N. Slattery, A. R. Maguire
and M. A. McKervey, Chem. Rev., 2015, 115, 9981;
(d) M. P. Doyle and D. C. Forbes, Chem. Rev., 1998, 98, 911;
(e) S.-F. Zhu and Q.-L. Zhou, Natl. Sci. Rev., 2014, 1, 580;
(f) Z. Zhang and J. Wang, Tetrahedron, 2008, 64, 6577;
(g) G. Maas, Organic Synthesis, Reactions and Mechanisms,
Springer, 1987.
2 (a) W. Kirmse and M. Kapps, Chem. Ber., 1968, 101, 994;
(b) M. P. Doyle, J. H. Griffin, M. S. Chinn and D. Van
Leusen, J. Org. Chem., 1984, 49, 1917.
3 For selected reviews, see: (a) Y. Zhang and J. Wang, Coord.
Chem. Rev., 2010, 254, 941–953; (b) T. H. West,
S. S. M. Spoehrle, K. Kasten, J. E. Taylor and A. D. Smith,
ACS Catal., 2015, 5, 7446–7479; (c) Z. Sheng, Z. Zhang,
C. Chu, Y. Zhang and J. Wang, Tetrahedron, 2016, DOI:
10.1016/j.tet.2016.11.045.
Scheme 3 Proposed mechanism.
4 (a) X. Zhang, Z. Qu, Z. Ma, W. Shi, X. Jin and J. Wang,
J. Org. Chem., 2002, 67, 5621; (b) M. Ma, L. Peng,
C. Li, X. Zhang and J. Wang, J. Am. Chem. Soc., 2005, 127,
15016.
expected from ammonium or oxonium ylide rearrangement
were observed. The N–H insertion product was mainly isolated
when the allylic secondary amine was employed (eqn (1)), in
line with previous research.6b Essentially no reaction occurred
when an allylic tertiary amine or an allyl methyl ether was sub-
jected to the standard conditions (eqn (2) and (3)).
5 (a) S. Kitagaki, Y. Yanamoto, H. Okubo, M. Nakajima and
S. Hashimoto, Heterocycles, 2001, 54, 623; (b) M. Liao and
J. Wang, Green Chem., 2007, 9, 184; (c) Z. Li, B. T. Parr and
H. M. L. Davies, J. Am. Chem. Soc., 2012, 134, 10942; (d) Z. Li
and H. M. L. Davies, J. Am. Chem. Soc., 2010, 132, 396.
6 (a) C. Poriel, Y. Ferrand, P. le Maux, C. Paul, J. Rault-
Berthelot and G. Simonneaux, Chem. Commun., 2003, 18,
2308; (b) G. Simonneaux, E. Galardon, C. Paul-Roth,
M. Gulea and S. Masson, J. Organomet. Chem., 2001, 617,
360.
7 K. L. Greenman, D. S. Carter and D. L. Vranken,
Tetrahedron, 2001, 57, 5219.
8 T. Fukuda, R. Irie and T. Katsuki, Tetrahedron, 1999, 55,
649.
9 (a) M. S. Holzwarth, I. Alt and B. Plietker, Angew. Chem.,
Int. Ed., 2012, 51, 5351; (b) I. Aviv and Z. Gross, Chem. –
Eur. J., 2008, 14, 3995; (c) D. S. Carter and D. L. Van
Vranken, Org. Lett., 2000, 2, 1303; (d) J. B. Perales,
N. F. Makino and D. L. Van Vranken, J. Org. Chem., 2002,
67, 6711.
It is well-established that the reaction of iron porphyrin
with diazo compounds would furnish iron-carbenoid species,
some of which have been characterized by X-ray crystallo-
graphy.19 More importantly, the process of the Doyle–Kirmse
reaction is strongly supposed to involve free sulfonium ylide
intermediates rather than the metal-associated ylide.4b,5a,8,20
Consequently, a proposed mechanism pathway is drawn in
Scheme 3 on the basis of previous reports. Firstly, the diazo
compound was transferred by hemin to form a carbenoid
intermediate. Then the carbenoid species reacts in situ with
allylic sulfides followed by dissociation of the catalytically
active hemin to generate a free sulfonium ylide, which could
then spontaneously undergo a five-membered, six-electron
transition state with an envelope conformation to deliver the
desired [2,3]-sigmatropic rearrangement product.
In summary, mild and efficient Doyle–Kirmse reactions
between allyl sulfides and diazo compounds in water were
developed with the synergistic introduction of hemin, Triton 10 P. W. Davies and S. J. C. Albrecht, Chem. Commun., 2008,
X-100 and β-CD. This catalytic system displays a broad func-
238.
tional-group tolerance and possesses a relatively lower E-factor. 11 P. W. Davies, S. J. C. Albrecht and G. Assanelli, Org. Biomol.
The mild reaction conditions and the use of eco-friendly water
Chem., 2009, 7, 1276.
as the solvent will greatly facilitate future sustainable chemical 12 (a) P. S. Coelho, E. M. Brustad, A. Kannan and F. H. Arnold,
catalysis and biocatalysis.
Science, 2013, 339, 307; (b) C. K. Prier, T. K. Hyster,
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Communication
C. C. Farwell, A. Huang and F. H. Arnold, Angew. Chem., 14 X. Xu, C. Li, Z. Tao and Y. Pan, Adv. Synth. Catal., 2015,
Int. Ed., 2016, 128, 4789; (c) H. Renata, R. D. Lewis, 357, 3341.
M. J. Sweredoski, A. Moradian, S. Hess, Z. J. Wang and 15 W. Liu and R. Guo, Photochem. Photobiol., 2010, 86, 835.
F. H. Arnold, J. Am. Chem. Soc., 2016, 138, 12527; 16 J. R. Wolf, C. G. Hamaker, J.-P. Djukic, T. Kodadek and
(d) Z. J. Wang, N. E. Peck, H. Renata and F. H. Arnold,
Chem. Sci., 2014, 5, 598.
13 (a) M. Bordeaux, V. Tyagi and R. Fasan, Angew. Chem., Int.
L. K. Woo, J. Am. Chem. Soc., 1995, 117, 9194.
17 S. Narayan, J. Muldoon, M. G. Finn, V. V. Fokin, H. C. Kolb
and K. B. Sharpless, Angew. Chem., Int. Ed., 2005, 44, 3275.
Ed., 2015, 127, 1764; (b) G. Sreenilayam and R. Fasan, 18 R. Prabharasuth and D. L. Van Vranken, J. Org. Chem.,
Chem. Commun., 2015, 51, 1532; (c) V. Tyagi and R. Fasan, 2001, 66, 5256.
Angew. Chem., Int. Ed., 2016, 128, 2558; (d) V. Tyagi, 19 Y. Li, J.-S. Huang, Z.-Y. Zhou, C.-M. Che and X.-Z. You,
R. B. Bonn and R. Fasan, Chem. Sci., 2015, 6, 2488; J. Am. Chem. Soc., 2002, 124, 13185.
(e) V. Tyagi, G. Screenilayam, P. Bajaj, A. Tinoco and 20 M. Reggelin, Sulfur-Mediated Rearrangements II, Springer,
R. Fasan, Angew. Chem., Int. Ed., 2016, 55, 13562.
2007.
This journal is © The Royal Society of Chemistry 2017
Green Chem.