Reaction of dicarbomethoxycarbene with thiophene derivatives

Article abstract of DOI:10.1021/jo802823s

Photolysis of derivatives of dimethylmalonate thiophene-S,C-ylide provides dicarbomethoxycarbene, which can react with thiophene to form dimethyl (2-thienyl)malonate. By generation of dicarbomethoxycarbene from the dibenzothiophene-based ylide in neat thi

Full text of DOI:10.1021/jo802823s

Reaction of Dicarbomethoxycarbene with Thiophene Derivatives
William S. Jenks,* Melanie J. Heying, Stacey A. Stoffregen, and Erin M. Rockafellow
Department of Chemistry, Iowa State UniVersity, Ames, Iowa 50011-3111
wsjenks@iastate.edu
ReceiVed December 30, 2008
Photolysis of derivatives of dimethylmalonate thiophene-S,C-ylide provides dicarbomethoxycarbene, which
can react with thiophene to form dimethyl (2-thienyl)malonate. By generation of dicarbomethoxycarbene
from the dibenzothiophene-based ylide in neat thiophene, it is shown that the thienylmalonate is not a
product of rearrangement of the thiophene ylide, in contrast to thermolysis results. Formation of the
thienylmalonate is suppressed by substitution of any sort in the 2- and 5-positions on the thiophene and
by substitution with an electron-withdrawing substituent.
Introduction
sulfoxides.14-19 Additionally, it has been shown that nitrenes
can be generated from analogous sulfilimine precursors.20-24
We recently communicated that S,C sulfonium ylides could
be used as photochemical precursors to carbenes.1 Photolysis
of compounds 1-4 in methanol or cyclohexene produces
reaction mixtures that contain products deriving from both
singlet and triplet dicarbomethoxycarbene, as illustrated in
Scheme 1. Of the four precursors, 1-3 cleanly produced the
corresponding thiophene derivative and apparently carbene-
derived products, while 4 produced a considerably more
complex mixture. This behavior is consistent with the analogous
photochemical deoxygenation of sulfoxides, in which direct
irradiation often produces very high chemical yields of sulfide
forderivativesofdibenzothiophene(andsometimesthiophene).2-13
In contrast, R-cleavage reactions tend to predominate for other
(11) Kumazoe, K.; Arima, K.; Mataka, S.; Walton, D. J.; Thiemann, T.
J. Chem. Res. (S) 2003, 60–61.
(12) Thiemann, T.; Ohira, D.; Arima, K.; Sawada, T.; Mataka, S.; Marken,
F.; Compton, R. G.; Bull, S.; Davies, S. G. J. Phys. Org. Chem. 2000, 13, 648–
653.
(13) Heying, M. J.; Nag, M.; Jenks, W. S. J. Phys. Org. Chem. 2008, 21,
915-924.
(14) Darmanyan, A. P.; Gregory, D. D.; Guo, Y.; Jenks, W. S. J. Phys. Chem.
A 1997, 101, 6855–6863.
(15) Guo, Y.; Darmayan, A.; Jenks, W. S. Tetrahedron Lett. 1997, 38, 8619–
8622.
(16) Guo, Y.; Jenks, W. S. J. Org. Chem. 1997, 62, 857–864.
(17) Still, I. W. J. In The Chemistry of Sulfones and Sulfoxides; Patai, S.,
Rappaport, Z., Stirling, C. J. M., Eds.; John Wiley & Sons Ltd.: New York,
1988; pp 873-887.
(18) Guo, Y.; Darmanyan, A. P.; Jenks, W. S. Tetrahedron Lett. 1997, 38,
8619–8622.
(19) Darmanyan, A. P.; Gregory, D. D.; Guo, Y.; Jenks, W. S.; Burel, L.;
Eloy, D.; Jardon, P. J. Am. Chem. Soc. 1998, 120, 396–403.
(20) Desikan, V.; Liu, Y.; Toscano, J. P.; Jenks, W. S. J. Org. Chem. 2007,
72, 6848–6859.
(1) Stoffregen, S. A.; Heying, M.; Jenks, W. S. J. Am. Chem. Soc. 2007,
129, 15746–15747.
(2) Wan, Z.; Jenks, W. S. J. Am. Chem. Soc. 1995, 117, 2667–2668.
(3) Gregory, D. D.; Wan, Z.; Jenks, W. S. J. Am. Chem. Soc. 1997, 119,
94–102.
(4) McCulla, R. D.; Jenks, W. S. J. Am. Chem. Soc. 2004, 126, 16058–
16065.
(5) Nag, M.; Jenks, W. S. J. Org. Chem. 2004, 69, 8177–8182.
(6) Nag, M.; Jenks, W. S. J. Org. Chem. 2005, 70, 3458–3463.
(7) Rockafellow, E. M.; McCulla, R. D.; Jenks, W. S. J. Photochem.
Photobiol., A 2008, 198, 45–51.
(8) Lucien, E.; Greer, A. J. Org. Chem. 2001, 66, 4576–4579.
(9) Thomas, K. B.; Greer, A. J. Org. Chem. 2003, 68, 1886–1891.
(10) Arima, K.; Ohira, D.; Watanabe, M.; Miura, A.; Mataka, S.; Thiemann,
T.; Valcarcel, J. I.; Walton, D. J. Photochem. Photobiol. Sci. 2005, 4, 808–816.
(21) Desikan, V.; Liu, Y.; Toscano, J. P.; Jenks, W. S. J. Org. Chem. 2008,
73, 4398–4414.
(22) Fujita, T.; Maeda, T.; Kim, B. J.; Tatami, A.; Miyamoto, D.; Kawaguchi,
H.; Tsuchiya, N.; Yoshida, M.; Kawashima, W.; Morita, H. J. Sulfur Chem.
2008, 29, 459–465.
(23) Morita, H.; Tatami, A.; Maeda, T.; Ju Kim, B.; Kawashima, W.;
Yoshimura, T.; Abe, H.; Akasaka, T. J. Org. Chem. 2008, 73, 7159–7163.
(24) Fujita, T.; Kamiyama, H.; Osawa, Y.; Kawaguchi, H.; Kim, B. J.; Tatami,
A.; Kawashima, W.; Maeda, T.; Nakanishi, A.; Morita, H. Tetrahedron 2007,
63, 7708–7716.
10.1021/jo802823s CCC: $40.75
Published on Web 03/02/2009
2009 American Chemical Society
J. Org. Chem. 2009, 74, 2765–2770 2765

Jenks et al.
SCHEME 1. Generation of Dicarbomethoxycarbene from
Photolysis of 1-4 and Trapping by Cyclohexene or
Methanola
SCHEME 2. Potential Mechanistic Pathways for Formation
of 8 on Photolysis of 1
a Adduct 8 is observed when 1 is the starting material.
compounds are illustrated in Scheme 2.) On the other hand,
treatment of ethyl diazoacetate with either Cu or Rh-based
catalysts, which leads to singlet metal carbenoids as intermedi-
ates, yielded a bicyclic product analogous to 14.29,30 Careful
reaction of dimethyl diazomalonate and thiophene with
Rh2(OAc)4 catalyst gave only 1, and the same reaction of methyl
diazoacetoacetate gave the appropriate analogues of 1, 8, and
14.30 Photolysis of dimethyl diazomalonate in thiophene was
reported to give a low yield of 1; an unreported yield of 3 was
obtained from dimethyl diazomalonate photolyzed in the pres-
ence of dibenzothiophene.31,32
In this paper, we report a series of experiments designed to
explore the chemistry of dicarbomethoxycarbene with thiophene
derivatives and to attempt to elucidate the mechanism of the
formation of 8 from among the several possibilities suggested
by these previous works. We show that the reaction does require
a diffusible intermediate, which we take to be dicarbomethoxy-
Consistent chemoselectivity in the photochemical reaction of
1-3 with cyclohexene (6 vs. 7) was interpreted as representative
of a common intermediate, presumably the singlet carbene 5.
However, when 1 was used as a precursor, the additional product
8 was observed. Photolysis of 1-3 in methanol gave differing
ratios of 9 to 10, which was interpreted by proposing that the
initial spin populations of the carbenes differed from one
precursor to another, and that spin equilibration was not faster
than trapping under these conditions.25 Again, when 1 was the
precursor, the adduct 8 was observed. No analogous adducts
were observed on photolysis of 2 or 3 in either solvent.
The mechanism for formation of 8 is of interest, in part
because closely related chemistry has led to varying products
discovered over the years.26-28 Thermolysis of 1 at reflux was
reported to result in formation of 8 without intervention of the
carbene (5), as demonstrated by the lack of trapping by olefins
or crossover from one thiophene to another.27,28 Thermolysis
of closely related derivatives (e.g., using the t-Bu, rather than
methyl, ester) led to the isolation of thiopyran derivatives
analogous to 13, presumably through intermediate 11.28 (These
(28) Bowles, T.; Gillespie, R. J.; Porter, A. E. A.; Rechka, J. A.; Rzepa,
H. S. J. Chem. Soc., Perkin Trans. 1 1988, 803–807.
(29) Gillespie, R. J.; Murray-Rust, J.; Murray-Rust, P.; Porter, A. E. A.
J. Chem. Soc., Chem. Commun. 1978, 83–84.
(30) Gillespie, R. J.; Porter, A. E. A. J. Chem. Soc., Perkin Trans. 1 1979,
2624–2628.
(31) Ando, W.; Yagihara, T.; Tozune, S.; Migita, T. J. Am. Chem. Soc. 1969,
91, 2786–2787.
(32) Ando, W.; Higuchi, H.; Migita, T. J. Org. Chem. 1977, 42, 3365–3372.
(25) This hypothesis will be addressed again in a separate publication.
(26) Meth-Cohn, O.; Van Vuuren, G. Tetrahedron Lett. 1986, 27, 1105–
1108.
(27) Gillespie, R. J.; Porter, A. E. A.; Wilmott, W. E. Chem. Commun. 1978,
85–86.
2766 J. Org. Chem. Vol. 74, No. 7, 2009

Dicarbomethoxycarbene
carbene, and that the intermediate is both electrophilic and
sensitive to steric interference.
Scheme 2 illustrates several potential pathways, largely drawn
from the literature. A unimolecular process, as suggested by the
reported thermal chemistry, might proceed via path a to give the
zwitterion 11. Final formation of 8 might be through spontaneous
or photochemical ring opening (path b) or by ring expansion to 13
(path c) followed by known chemistry to form 8.26-28
Our initial presumption for this photochemical reaction,
however, was that reaction of carbene 5 with thiophene would
produce the bicycle 14 (path d). This was based on the chemistry
of ethyl diazoacetate, Rh2(OAc)4, and thiophene,29,30 and on
the known photolysis of dimethyl diazomalonate in benzene,
which provided a 1:1.6 mixture of 16 and the equilibrating
mixture of 17 and 18.33
diffusible dicarbomethoxycarbene is shown and that our condi-
tions do not necessarily degrade all arene adducts of the carbene.
By NMR analysis of the reaction mixture, good yields of 17
and 1834-36 were observed to the exclusion of 16. Conversely,
only the benzyl insertion product 19 was observed in toluene,
to the exclusion of 20 (and/or its tautomer).37 Data are shown
in Table 1.
Once 14 is formed, the rearrangement to 8 must be addressed.
Heterolytic opening of 14 to give 12 might be spontaneous,
although this does not explain the isolability of the methyl
diazoacetoacetate derivative.28 Alternatively, a low barrier rear-
rangement via path f might provide the enol of the final product.
Finally, we posited that the specific regiochemistry of addition
might be attributable to a carbene addition mechanism that
resembles electrophilic addition. In this mechanism, 12 might
be generated directly, without intervention of 14, as illustrated
as path g in Scheme 2.
TABLE 1. Photolysis of Ylides 2 and 3 in Benzene and Toluene
product yieldsa
Cylopropanation
Net insertion products
ylide
solvent
products 17/18 or 20
16 or 19
2
3
2
3
benzene
benzene
toluene
toluene
73
79
85
76
a Relative to consumed starting material. Products identified by
comparison to literature spectra.
Results and Discussion
Further attempts were made to observe 14 or an analogue by
using 2-substituted or 2,5-disubstituted thiophenes. It was
reasoned that the absence of a labile hydrogen in the 2(or 5)-
position of the thiophene ring after cyclopropanation to give
21 would leave the bicyclic adduct unable to rearrange to the
formal C-X insertion product 23 through 22 (Scheme 3). It
further seemed that this tactic would block a direct rearrange-
ment, as shown for the parent system as path f in Scheme 2.
Of the mechanisms shown in Scheme 2, the most straight-
forward distinction to be made was to separate the unimolecular
and bimolecular pathways. This was accomplished by using the
ylide 3 as a precursor. Direct photolysis of 3 using broad
irradiation centered at 350 nm in neat thiophene resulted in a
near quantitative yield of 8. However, it became apparent that
the lamps being used would also photolyze 1, and thus a
sequential mechanism of 3 f 1 f 8 could not be ruled out. A
second set of experiments, using a Xe lamp filtered through a
monochromator centered at 365 nm ((12 nm linear dispersion),
was carried out. At these wavelengths, 1 does not absorb
significantly. Upon photolysis of 3 in thiophene, nearly equal
yields of 1 and 8 were obtained. Control experiments showed
that almost no secondary photolysis of 1 was possible on the
time scale of the formation of 8. Thus (1) formation of 8 does
not require formation of 1 as an intermediate step, (2) dicar-
bomethoxycarbene is almost certainly involved in formation of
8, and (3) the first “committed step” of the carbene toward 8,
whatever that may be, occurs at almost an identical rate as the
capture of the carbene as the ylide in thiophene.
SCHEME 3. Blocking of Rearrangement to the C-H
Insertion Product by Substitution on Thiophene
Two approaches to this experiment were taken. First the
analogues to 1 with 2,5-dimethyl (1a), 2,5-dichloro (1b), and
2-bromo (1g) substitutions on the thiophene (Scheme 4) were
prepared. Photolysis of any of these derivatives in CD3OD did
not result in observation of any bicyclic compound (analogue
Several experiments were performed to look for the presumed
initial adduct 14, even though it had not been observed in the
formation of 8 by other means. Photolysis of 1 by using 350
nm bulbs in degassed deuterated methanol at -30 °C was
followed by NMR, but there was no sign of any photolytic
product other than 8, 9, and 10.
(33) Jones, M., Jr.; Ando, W.; Hendrick, M. E.; Kulczycki, A., Jr.; Howley,
P. M.; Hummel, K. F.; Malament, D. S. J. Am. Chem. Soc. 1972, 94, 7469–
7479.
(34) Yang, M.; Webb, T. R.; Livant, P. J. Org. Chem. 2001, 66, 4945–4949.
(35) Berson, G. A.; Hartter, D. R.; Klinger, H.; Grubb, P. W. J. Org. Chem.
1969, 33, 1669–1671.
As a positive control, photolyses of 2 and 3 were carried out
in benzene and toluene. These experiments show again that
(36) Go¨rlitz, M.; Gu¨nter, H. Tetrahedron 1969, 25, 4467–4480.
J. Org. Chem. Vol. 74, No. 7, 2009 2767

Jenks et al.
SCHEME 4. Photolysis of Ylide 3 in Neat, Substituted thiophenesa
a The letter is used as an index to denote the substitution pattern.
TABLE 2. Photolysis of Ylide 3 in Substituted Thiophenes 24i a
of 14) or any analogue of 8. Instead, only the corresponding
thiophenes, dimethyl methoxymalonate (9), and dimethyl ma-
lonate (10) were observed. A similar result was also obtained
for the 3,4-dibromo analogue 1d.38
A more widely applicable approach to observing analogues
of 8 or 21 was envisioned, in which 3 would be photolyzed in
neat substituted thiophene. Several thiophenes (24a-i) were
selected, as shown in Scheme 4 and Table 2. However, in no
case was the bicyclic analogue 21 observed on direct photolysis
of 3 in any of the neat thiophenes.
The observed products are shown in Table 2, and discernible
patterns emerge. All three 2,5-disubstituted thiophenes resulted
in total suppression of the apparent insertion analogue of 8.
Indeed, even dibromo substitution at the 3- and 4-positions (3,4-
dibromothiophene, 24d) suppressed formation of 8d. Instead,
photolysis (using 365 nm irradiation) resulted only in formation
of the ylide analogues 1a-d, with the exception of a benzyl
insertion product 25a.
Thermolysis of 3 in thiophene at 80 °C yielded 8 quantita-
tively, whereas its thermolysis in 2,5-dichlorothiophene did not
yield any isolable products. Photolysis of 3 in 2,5-dichlo-
rothiophene provided 1b as the only isolable product (Table
2). The thermolysis of 1b in 2,5-dichlorothiophene was previ-
ously reported to provide a mixture of unidentifiable com-
pounds.27 All these results are reconciled if it is posited that
the thermolysis conditions make formation of 1 or 1b effectively
reversible. In the former case, eventually the carbene reacts with
thiophene irreversibly to form 8, and in the latter, the reaction
with 2,5-dichlorothiophene (or other compounds present in the
mixture) gives the unidentified products.
a The subscript i is used to indicate the appropriate modified structure,
e.g., 1a, 8a, and 25a for 2,5-dimethylthiophene (24a) as
the solvent. b Debromination of 1c was noted along with decomposition
of 3.
Several monosubstituted thiophenes were also used as
solvents for the photolysis of 3 in the current work. It was
reasoned that purely steric effects could allow for C-H insertion
on the “other” side of the thiophene, e.g., by cyclopropanation
across the C4-C5 (or C5-S) bond. In the cases where the
π-donation of the substituent is weak (Br, CN), only the ylide
analogue of 1 was observed and identified. On the other hand,
when the substituent was a better π-donor (CH3, Cl, OCH3),
the analogue of C-H insertion product 8 (with the substituent
on the other side of the thiophene ring) was formed.
It seems self-evident that there is a greater variability for the
relative rates of formation of 8 than for 1: without substitution
or with π-donation, formation of 8 competes effectively with
formation of 1. But if formation of 8 is slowed by electron-
withdrawing substituents (or inhibited by substitution at both
the 2- and 5-positions), the ylide is formed instead. These
substituent effects would appear to be a strong contraindication
of the pathways for formation of 8 that begin by addition to
give 11 and its analogues. Surely, 2,5-electron-donating sub-
stituents would be destabilizing to 11, and its formation would
be suppressed, relative to simple ylide formation.
Competitive kinetics were used to explore the substituent
effect further. Photolyses of 3 were done in solvent mixtures
containing thiophene and one of 2-methylthiophene, 2-chlo-
rothiophene, or 2-methoxythiophene. Product ratios were con-
verted into selectivities, and multiple solvent ratios were used
to ensure linearity. It was assumed that the product formation
was under kinetic control, so that the product selectivities could
be converted to relative rate constants. The results of these
experiments are given in Table 3. Plotting the log of the relative
(37) Small additional NMR peaks were observed, but they were not assignable
to compounds 13 or 14 in benzene or toluene, respectively.
(38) Unsuccessful attempts were made to prepare the corresponding ylide
from 2,5-dibromothiophene.
2768 J. Org. Chem. Vol. 74, No. 7, 2009

Dicarbomethoxycarbene
TABLE 3. Substituent Effects on the Relative Rate of Formation
of Analogues of 8
was optimized at HF, MP2, and B3LYP levels of theory with
various basis sets, and a search was undertaken for the
appropriate transition state. In spite of considerable effort, no
low-energy transition state that connected 14 and the enol could
be found. However, examination of computational and physical
models suggests that the orbital alignment for such a rearrange-
ment would be poor. Thus, while not finding a low-energy
transition state by computational means is by no means strong
evidence that one does not exist, we no longer consider
formation of 14 and its rapid decomposition by means of path
f to be a likely explanation for the experimental results.
Interestingly, previous workers report that photolysis of
dimethyl diazomalonate in the presence of vinyl sulfides and
thiophene resulted in only modest yields of sulfonium ylides,
not in any cyclopropanation.32,47 Though it is clear that the
photolysis of 3 in the presence of thiophene is cleaner than these
reported photolyses of dimethyl diazomalonate, we cannot
resolve the apparent conflict with our report, i.e., that no
cyclopropanation or other C-C bond formation was observed
in the earlier work.
a
b
substituent
σ+
kR/kH
log (kR/kH)
Cl
H
Me
OMe
0.11
0
-0.31
-0.78
0.6
1
11
30
-0.222
0.000
1.041
1.477
a The para values of σ+ were used because of the carbon-to-carbon
relationship. We are unaware of analogous thiophene-specific values.
b Relative rates were determined from average selectivities from multiple
ratios of unsubstituted to substituted thiophene.
rates vs σ+ (taken from benzene analogues;39 data are not
available for thiophene analogues) results in a slope of -2.0 (
0.4 (r2 ) 0.92). The σ+ values for 2-bromothiophene (σ+
)
0.15) and 2-cyanothiophene (σ+ ) 0.66) are consistent with
the pattern established in Table 3.40
These results clearly show that dicarbomethoxycarbene is an
electrophilic partner in the eventual formation of 8, but are not
definitive in other details. The sign of F, of course, shows that
partial positive charge is developed on the thiophene in the rate-
determining step. Magnitudes of F for electrophilic aromatic
substitution of toluene range from about -2.5 to about -12
with the most active electrophiles generally having the largest
magnitudes.41 However, the interpretation of the values is
complex, even for closely related reactions of this class. (See
for example refs 39 and 42.) The observed F value for
dicarbomethoxycarbene adding to thiophene of about -2.0 is
consistent with the carbene being a neutral electrophile in
reacting with the thiophene π-system. However, the experi-
mentally obtained F of -2.0 does not distinguish between
cyclopropanation to give 14 and direct formation of 12.
Regardless of the magnitude of F, though, classic electrophilic
aromatic substitution, in which the dicarbomethoxycarbene is
protonated before reaction with thiophene, can be ruled out
because no proton source is available in most of the reaction
conditions reported here. However, one could speculate that the
reaction of dicarbomethoxycarbene 5 with thiophene to form
12 might be a single step that begins with a transition state
similar to that of a classic cyclopropanation, but finds a low
energy path that diverts to 12 without formation of 14. If this
were the case, it might be argued that small steric differences
in monosubstituted thiophenes could direct the carbene away
from the substituted side of the thiophene, until it found either
the unsubstituted side or the sulfur atom.
Conclusions
Reaction between thiophene and dicarbomethoxycarbene
produces both the ylide 1 and the net C-H insertion product 8,
in almost equal part. Electronic donating substituents on the
2-position of the thiophene ring increase the relative yield of 8
at the unsubstituted 5-position. In contrast to certain thermal
reactions, formation of 8 is not a direct rearrangement of the
ylide 1. Furthermore, the substituent effects strongly imply that
an intermediate such as 11 is not involved in the formation of
8, either.
Beyond this, definitive conclusions are more difficult. Despite
observations of close analogues, 14 has, to the best of our
knowledge, never been detected in any experiment that generates
free dicarbomethoxycarbene (5) or a free carbenoid, or in which
1 is heated. Analogues deriving from the reaction of 5 and
benzene or between thiophene and slight structural variants of
5 have all been detected, though. This suggests that 14 may be
formed, but decomposes easily. We cannot disprove a low-
energy pathway that converts 14 to the monocyclic enol 15,
but did not find any positive computational evidence. Alterna-
tively, it is possible that 14 is never actually formed, but rather
that the addition of the carbene to the π bond of the thiophene
is diverted to a more “downstream” intermediate, e.g., 12,
through a very asynchronous transition state.
Another alternative that was considered was that 14 does
form, but undergoes a rapid intramolecular rearrangement to
an enol with ring-opening, i.e., path f in Scheme 2. Given that
no evidence for 14 was found, even at -30 °C, it was assumed
that the barrier to the rearrangement from 14 to 15 would have
to be low. The potential reaction would be pseudopericyclic
and thus not “forbidden” (see, for example refs 43-46). Thus,
a computational investigation was undertaken. Compound 14
Experimental Section
Materials. Unless otherwise noted, all solvents were the highest
purity commercially available, and reagents were used as received.
All NMR spectra were obtained with CDCl3 as solvent. The
formally carbanionic carbon typically does not show in the 13C
spectra.
Thiophene Ylides. In a small round-bottomed flask, 1 equiv
(approximately 5 mmol) of the thiophene derivative was mixed with
3 equiv of the dimethyl diazomalonate and 2 mg of rhodium acetate
dimer. In occasions when the thiophene derivative was a solid, a
minimal amount of 1,2-dichloroethane was added. The mixture was
allowed to stir until the color changed from dark green or a
(39) Smith, M. B.; March, J. B. March’s AdVanced Organic Chemistry, 6th
ed.; John Wiley and Sons: Hoboken, NJ, 2007.
(40) We did not find an appropriate value for trifluoromethyl.
(41) Stock, L. M.; Brown, H. C. J. Am. Chem. Soc. 1959, 81, 3323–3329.
(42) Knowles, J. R.; Norman, R. O. C.; Radda, G. K. J. Chem. Soc. 1960,
4885–4896.
(43) Birney, D. M.; Ham, S.; Unruh, G. R. J. Am. Chem. Soc. 1997, 119,
4509–4517.
(44) Birney, D. M.; Xu, S.; Ham, S. Angew. Chem., Int. Ed. Engl. 1999, 38,
189–193.
(45) Liu, R. C. Y.; Lusztyk, J.; McAllister, M. A.; Tidwell, T. T.; Wagner,
B. D. J. Am. Chem. Soc. 1998, 120, 6247–6251.
(46) Ross, J. A.; Seiders, R. P.; Lemal, D. M. J. Am. Chem. Soc. 1976, 98,
4325–4327.
(47) Cyclopropanes were obtained on photolysis of alkyl phenyl diaz-
omethanes in the presence of vinyl sulfides.
J. Org. Chem. Vol. 74, No. 7, 2009 2769

Jenks et al.
precipitate formed. Analysis by IR showed no diazo stretch. The
crude product was washed with hexane, decanted, washed again
with 50/50 ethyl acetate/hexane solution, and decanted again. The
remaining light green solid was then recrystallized from ethanol to
yield a white solid. Compounds 1-4 were previously reported by
us.1
broadly emitting low pressure fluorescent tubes centered at either
300 or 350 nm or a Xe arc lamp filtered through a monochromator
with (12 nm linear dispersion. Initial concentrations were 1-4
mM. The products 8e,28,29 8f, 8j,51 and 25 were identified by their
1H NMR spectra. Representative proton spectra of product mixtures
containing 8f and 25 are given in the Supporting Information. 8f:
1H NMR δ 7.23 (d, J ) 4.8 Hz, 1H), 7.01 (d, J ) 4.8 Hz, 1H),
Dimethylmalonate-2,5-dimethylthiophene-S,C-ylide (1a):48 Yield
1
1
85%. H NMR δ 6.95 (s, 2H), 3.76 (s, 6H), 2.16 (s, 6H).
4.90 (s, 1H), 3.86 (s, 6H). 25a: H NMR δ 6.64 (d, J ) 4.5 Hz,
Dimethylmalonate-2,5-dichlorothiophene-S,C-ylide (1b):32,49
1H), 6.51 (d, J ) 4.5 Hz, 1H), 4.58 (t, J ) 6.4 Hz, 1H), 3.73 (s,
6H), 2.36 (d, J ) 6.4 Hz, 2H), 2.31 (s, 3H).
1
Yield 72%. H NMR δ 7.02 (s, 2H), 3.78 (s, 3H), 3.62 (s, 3H).
Dimethylmalonate-2,5-bromothiophene-S,C-ylide (1c):32 Yield
Photolysis at Low Temperature. A methanol bath, chilled to
-30 °C by means of careful addition of dry ice, was used in
conjunction with a Rayonet minireactor and 300 nm bulbs. The
sample, held in an NMR tube, was taken directly from the bath
and examined in a precooled NMR probe at the same temperature.
Competition Experiments. A weighed mixture of thiophenes
or a neat thiophene was mixed with 4-6 mg of 3, degassed, and
photolyzed at 365 nm in the Rayonet until completion. Afterward,
the low-boiling thiophenes were gently evaporated, and the remain-
ing substances analyzed by 1H NMR. In the cases where the
thiophenes were too high boiling, the mixture was analyzed by
GCMS.
1
5.3%. H NMR δ 7.13 (s, 2H), 3.77 (s, 3H), 3.65 (s, 3H).
Dimethylmalonate-3,4-dibromothiophene-S,C-ylide (1d): Yield
1
23%. H NMR δ 6.99 (s, 2H), 3.70 (s, 6H). 13C δ 165.4, 129.0,
127.3, 51.9. HRMS calcd 371.8468, obsd 371.8449.
Dimethylmalonate-2-methylthiophene-S,C-ylide (1e):28,32 Yield
1
87%. H NMR δ 7.12 (d, J ) 3.6 Hz, 1H), 6.84 (d, J ) 5.4 Hz,
1H), 6.79 (dd J ) 5.4, 3.6 Hz 1H), 3.75 (s, 6H), 2.50 (s, 3H).
Dimethylmalonate-2-chlorothiophene-S,C-ylide (1f):48,50 Yield
1
77%. H NMR δ 7.18 (d, J ) 3.2 Hz, 1H), 7.05 (d, J ) 4.8 Hz,
1H), 6.91 (dd J ) 4.8, 3.2 Hz, 1H), 3.73 (s, 6H).
Dimethylmalonate-2-bromothiophene-S,C-ylide (1g):28,32 Yield
1
89%. H NMR δ 7.33 (d, J ) 6.0 Hz, 1H), 7.23 (d, J ) 3.2 Hz,
Thermolyses. Initial concentrations of ylide were 2-4 mM and
the solvent was purged with Ar. The sealed samples were placed
in an oven held at 80 °C and periodically analyzed over a time
scale of 12 -36 h.
1H), 7.17 (dd J ) 6.0, 3.2 Hz, 1H), 3.76 (s, 6H).
Dimethylmalonate-2-thiophenecarbonitrile-S,C-ylide (1h): Yield
1
88%. H NMR δ 7.21 (d, J ) 4.0 Hz, 1H), 7.02 (d, J ) 6.0 Hz,
1H), 6.95 (dd, J ) 6.0, 4.0 Hz, 1H), 3.72 (s, 6H). 13C NMR δ
171.4, 136.9, 133.3, 127.2, 116.1, 114.7, 60.6. Note: This reaction
was considerably less reliable than the others reported here.
Dimethylmalonate-2-methoxythiophene-S,C-ylide (1i): Yield
Acknowledgment. The authors acknowledge partial funding
of this work by National Science Foundation Grant CHE
0213375.
1
69%. H NMR δ 6.78 (dd, J ) 4.2, 3.0 Hz, 1H), 6.63 (d, J ) 4.2
Hz, 1H), 6.29 (d, J ) 3.0 Hz, 1H), 3.84 (s, 3H). 13C ΝΜR δ 171.4,
164.7, 123.6, 117.1, 60.6, 59.1. HRMS calcd 244.041, obsd 244.044.
Photolyses. All solutions were deoxygenated by bubbling with
Ar for a minimum of 15 min. Irradiations were done with either
Supporting Information Available: Spectral data. This
material is available free of charge via the Internet at
http://pubs.acs.org.
JO802823S
(48) Bien, S.; Kapon, M.; Gronowitz, S.; Hoernfeldt, A. B. Acta Chem.
Scand., Ser. B 1988, B42, 166–174.
(49) Bien, S.; Segal, Y. J. Org. Chem. 1977, 42, 1685–1688.
(50) Aggarwal, V.; Richardson, J. Sci. Synth. 2004, 27, 21–104.
(51) Weinstock, L. M.; Corley, E.; Abramson, N. L.; King, A. O.; Karady,
S. Heterocycles 1988, 27, 2627–2633.
2770 J. Org. Chem. Vol. 74, No. 7, 2009