Photochemistry of furyl- and thienyldiazomethanes: Spectroscopic characterization of triplet 3-thienylcarbene

Article abstract of DOI:10.1021/ja300927d

Photolysis (λ > 543 nm) of 3-thienyldiazomethane (1), matrix isolated in Ar or N₂ at 10 K, yields triplet 3-thienylcarbene (13) and α-thial-methylenecyclopropene (9). Carbene 13 was characterized by IR, UV/vis, and EPR spectroscopy. The conform

Full text of DOI:10.1021/ja300927d

Article
pubs.acs.org/JACS
Photochemistry of Furyl- and Thienyldiazomethanes: Spectroscopic
Characterization of Triplet 3-Thienylcarbene
Caroline R. Pharr,†,# Laura A. Kopff,† Brian Bennett,‡ Scott A. Reid,§ and Robert J. McMahon*,†
†Department of Chemistry, University of Wisconsin−Madison, 1101 University Avenue, Madison, Wisconsin 53706-1322, United
States
‡National Biomedical EPR Center, Department of Biophysics, Medical College of Wisconsin, 8701 Watertown Plank Road,
Milwaukee, Wisconsin 53226-0509, United States
§Department of Chemistry, Marquette University, 535 North 14th Street, Milwaukee, Wisconsin 53201-1881, United States
S
* Supporting Information
ABSTRACT: Photolysis (λ > 543 nm) of 3-thienyldiazomethane (1),
matrix isolated in Ar or N2 at 10 K, yields triplet 3-thienylcarbene (13)
and α-thial-methylenecyclopropene (9). Carbene 13 was characterized by
IR, UV/vis, and EPR spectroscopy. The conformational isomers of 3-
thienylcarbene (s-E and s-Z) exhibit an unusually large difference in zero-
field splitting parameters in the triplet EPR spectrum (|D/hc| = 0.508 cm−1, |E/hc| = 0.0554 cm−1; |D/hc| = 0.579 cm−1, |E/hc| =
0.0315 cm−1). Natural Bond Orbital (NBO) calculations reveal substantially differing spin densities in the 3-thienyl ring at the
positions adjacent to the carbene center, which is one factor contributing to the large difference in D values. NBO calculations
also reveal a stabilizing interaction between the sp orbital of the carbene carbon in the s-Z rotamer of 13 and the antibonding σ
orbital between sulfur and the neighboring carbonan interaction that is not observed in the s-E rotamer of 13. In contrast to
the EPR spectra, the electronic absorption spectra of the rotamers of triplet 3-thienylcarbene (13) are indistinguishable under our
experimental conditions. The carbene exhibits a weak electronic absorption in the visible spectrum (λmax = 467 nm) that is
characteristic of triplet arylcarbenes. Although studies of 2-thienyldiazomethane (2), 3-furyldiazomethane (3), or 2-
furyldiazomethane (4) provided further insight into the photochemical interconversions among C5H4S or C5H4O isomers,
these studies did not lead to the spectroscopic detection of the corresponding triplet carbenes (2-thienylcarbene (11), 3-
furylcarbene (23), or 2-furylcarbene (22), respectively).
understanding of the electronic structure of doped states of
thiophene derivativesmaterials that have achieved wide use in
conducting polymers and other electronic applications.18−20
Background. Shechter and co-workers reported detailed
product analyses associated with thermolysis of the isomeric
thienyldiazomethanes (1 and 2) and furyldiazomethanes (3 and
4) (Scheme 1).21,22 The intermediacy of the corresponding
INTRODUCTION
■
The chemistry of aryl and heteroaryl carbenes constitutes a
subject of longstanding interest in the field of organic reactive
intermediates.1−3 Basic relationships concerning the structure,
reactivity, and spectroscopy of these intermediates have been
deduced through product analyses, matrix-isolation spectro-
scopy, time-resolved spectroscopy, and computational studies.
For any given system, these relationships are strongly
influenced by the electronic ground state of the carbene
(singlet or triplet) and the magnitude of the singlet-triplet
energy gap. In the singlet series, the chemistry and spectros-
copy of arylchlorocarbenes (phenyl,4,5 pyridyl,6 furanyl,7,8
thienyl,9,10 and their benzo analogues11−14) have been
extensively investigated. Our own interest in arylcarbenes
focuses on the triplet series because these specieslacking the
halogen substituentare more directly relevant to the harsh
reaction environments encountered in combustion or in
astrochemistry.15 That the chemistry of atomic carbon is
important in astronomical environments invites attention to
aryl- and heteroaryl carbenes, as these intermediates may be
formed upon addition of a carbon atom to a stable, closed-shell
aryl substrate. In the current study, we focus on the isomeric
furyl- and thienylcarbenes, which have eluded detection and
characterization to the present time.16,17 These fundamental
studies of reactive thienyl intermediates are also germane to an
Scheme 1. Thienyl and Furyldiazomethanes (1−4)
thienyl- and furylcarbenes was inferred from the isolation of the
products of insertion into the C−H bond of cyclooctane and
the formal products of dimerization. Fragmentation of the ring,
to afford ring-opened products, was also observed. Shevlin and
workers later postulated the intermediacy of 2- and 3-
thienylcarbene in the reaction of atomic 13C with thio-
phene.23,24 Saito and co-workers devised intriguing systems in
Received: January 28, 2012
Published: March 30, 2012
© 2012 American Chemical Society
6443
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
a
Scheme 2. C5H4S Isomers and Their Computed Relative Energies
a
Energy (kcal/mol; ZPVE included). B3LYP/6-31G* level of theory.
a
Scheme 3. C5H4O Isomers and Their Computed Relative Energies
a
Energy (kcal/mol; ZPVE included). B3LYP/6-31G* level of theory.
which the thermal furylcarbene fragmentation reaction is driven
in the reverse direction under photochemical conditions. Thus,
an acyclic enynal or enynone will undergo photocyclization to a
2-furylcarbene derivative, which can be trapped in either an
inter- or intramolecular fashion.25 Thermal cyclization reactions
of azo-ene-ynes are analogous, proceeding via a heteroaryl
carbene intermediate.26−28 Albers and Sander elucidated key
aspects of the photochemistry and spectroscopy of C5H4S and
C5H4O isomers through their study of the photochemistry of 3-
thienyldiazomethane (1), 3-furyldiazomethane (3), and 2-
furyldiazomethane (4) under matrix isolation conditions.16,17
None of the carbene intermediates, however, were detected. In
terms of computational studies, McKee, Shevlin, and Zottola
described an insightful, comprehensive study of the C5H4S
potential energy surface.24 The mechanism of ring-opening of
2-furylcarbene, 2-thienylcarbene, and related compounds has
been the subject of several computational and theoretical
investigations.7,24,29−31 Herges29 described the reaction as a
coarctate transformation, while Birney31 interpreted it as a
pseudopericyclic process.
3, and additional details are provided as Supporting
Information. This methodology is adequate for providing
qualitatively reliable predictions of infrared spectra and relative
energies across a range of molecular structures. In terms of
relative energies, Density Functional Theory (DFT) methods
often overemphasize delocalization in conjugated π-electron
systems,32−34 and the B3LYP functional biases the calculation
of singlet−triplet energy gaps by ca. 1−3 kcal/mol by
underestimating the stability of the singlet, relative to the
triplet.35,36 Because the singlet−triplet energy gaps of the
thienylcarbenes are of particular interest to us, we sought to
corroborate the DFT predictions using ab initio methods. The
results of coupled-cluster calculations, performed at a moderate
level of theory (CCSD/cc-pVTZ), show good agreement with
those obtained by using density functional theory for 2- and 3-
thienylcarbene (Scheme 4). Both computational methods
predict a triplet electronic ground state for both carbenes.
Of the isomers that we investigated, the structurally
analogous thioaldehyde 5 (C5H4S) and aldehyde 16
(C5H4O) are the lowest energy structures on their respective
potential energy surfaces. Portions of these surfaces have been
investigated, previously, although the issue of conformational
isomerism (s-E, s-Z) in thioaldehydes (5, 6) or aldehydes (16,
17) has not been explored in detail.16,17,24 The (s-Z) rotamer is
commonly drawn in mechanistic schemes, but there is only one
case throughout the photochemistry of diazo compounds 1, 2,
3, and 4 where an (s-Z) rotamer of any pentenyne (s-Z-6) is
RESULTS
■
Computed Energies of C5H4S and C5H4O Isomers. To
provide context for interpreting our experimental observations,
we performed computational studies of the C5H4S and C5H4O
potential energy surfaces at the B3LYP/6-31G* level of theory.
Structures and relative energies are depicted in Schemes 2 and
6444
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
Scheme 4. Thienylcarbenes and Their Computed Relative
a
Energies
a
Energy (kcal/mol; ZPVE included).
actually observed in the matrix (see below). The computed IR
spectra of the (s-Z) and (s-E) rotamers are quite different and
allow clear analysis and differentiation of the peaks present in
the experimental spectrum.
Primary Photochemistry of 3-Thienyldiazomethane
(1). Irradiation of 3-thienyldiazomethane (1) (λ > 534 nm; N2,
10 K) gives a mixture of (s-Z)-α-thial-methylenecyclopropene
(9), (s-E)-3-thienylcarbene (s-E-13), (s-Z)-3-thienylcarbene (s-
Z-13), and a minor amount of a species tentatively assigned as
1H-2-thiabicyclo[3.1.0]hexa-3,5-diene (12) (Scheme 5; Figure
S1).37 IR assignments are based upon comparison of
experimental spectra with computed IR spectra (B3LYP/6-
31G*), as well as comparison with previously reported spectral
data of 9 and 12.16,38 EPR and UV/vis experiments, performed
under analogous conditions, provide strong support for the
assignment of triplet 3-thienylcarbene (s-E- and s-Z-13). The
EPR spectrum of 13 affords direct evidence for the triplet
species, and the spectroscopic features are generally consistent
Figure 1. EPR spectra. Top: Spectrum of triplet (s-E) and (s-Z) 3-
thienylcarbene (13), obtained upon irradiation (λ > 571 nm, 22 h) of
3-thienyldiazomethane (1) (Ar, 15 K). Bottom: Spectrum of triplet 13
obtained upon continued irradiation (λ > 571 nm, 22 h; λ > 534 nm,
4.5 h; λ > 497 nm, 17.5 h; λ > 472 nm, 22.5 h) of diazo compound 1.
(Note difference in y-axis scales.)
Comparison of the electronic absorption spectra of 3-
thienyldiazomethane (1) and triplet 3-thienylcarbene (13)
shows that diazo compound 1 absorbs at slightly longer
wavelength (broad absorption with λmax = 494 nm) than
carbene 13 (λmax = 467 nm). (The 494-nm absorption of diazo
compound 1, which was measured in acetonitrile solution, is
not observable in the matrix spectrum depicted in Figure 2
because of the low concentration.) It is possible, therefore, to
irradiate diazo compound 1 under long-wavelength, broadband
conditions without inducing secondary photochemistry in the
incipient 3-thienylcarbene (13). Long-wavelength irradiation of
3-thienyldiazomethane (1) exhibits a subtle wavelength
dependence, which we tentatively ascribe to differential
photoreactivity of s-E and s-Z conformers of the diazo
with those expected for an arylcarbene (Figure 1).39−41
A
detailed analysis of the EPR spectrum will be provided as part
of the Discussion section. The electronic absorption spectrum
exhibits the weak visible absorption features that are character-
istic of triplet aryl carbenes (and the benzyl radical) (Figure
2).39 The spectroscopic assignment of triplet 3-thienylcarbene
(13) is further supported by the wavelength-dependent
photochemistry, in which the IR, UV/vis, and EPR signals
attributed to 13 disappear upon photoexcitation into the visible
absorption feature at λmax ca. 467 nm (Figures 2 and 3).
Scheme 5. Photochemistry of 3-Thienyldiazomethane (1) (N2, 10 K)
6445
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
Figure 3. IR subtraction spectrum showing spectral changes observed
upon selective irradiation (λ = 467 10 nm, 4 h) of a matrix that
contains a mixture of (3-thienyl)diazomethane (1), (s-Z)-α-thial-
methylenecyclopropene (9), 1H-2-thiabicyclo[3.1.0]hexa-3,5-diene
(12), and triplet 3-thienylcarbene (13) (N2, 10 K). The spectrum
shows the disappearance of 3-thienylcarbene (13) and the growth of 9.
Under these irradiation conditions, diazo compound 1 exhibits
virtually no change in concentration, while bicyclic compound 12
displays only a very small increase (X = the major peak associated with
12). Computed IR spectra for both conformers of triplet 3-
thienylcarbene (13) are superimposed at the bottom of the figure,
while the computed IR spectrum for (s-Z)-9 is shown at the top of the
figure.
Figure 2. Electronic absorption spectra. Dotted line: Matrix containing
(3-thienyl)diazomethane (1) (Ar, 10 K) before irradiation. (The
visible absorption of 1 is not observable because of the low
concentration.) Solid line: Spectrum of (s-E) and (s-Z) triplet 3-
thienylcarbene (13) observed upon photolysis of diazo compound 1
(λ > 472 nm, 19 h). Dashed line: Spectrum showing the disappearance
of carbenes (s-E)-13 and (s-Z)-13 upon irradiation (λ > 444 nm, 22
h).
unsuccessful, they help us understand the photochemical
reactivity of this system. An important observation is that the
methylenecyclopropene derivative (9) is always formed, even
under conditions where triplet 3-thienylcarbene (13) is stable
to the irradiation conditions. Plausible mechanisms for the
formation of 9, under these conditions, involve a hot ground-
state reaction of carbene 13 or a reaction in the excited state of
diazo compound 1. Both explanations have ample precedent in
the field of carbene chemistry. The first mechanistic scenario
posits that N2 loss from the excited state of 3-thienyldiazo-
methane (1) yields carbene 13 with excess vibrational energy,
and that 9 arises from a hot ground-state reaction of carbene
13. To probe for the involvement of a hot ground-state
carbene, we performed the photolysis of diazo compound 1 in
Ar and N2 matrices, as well as a 2-methyltetrahydrofuran (2-
MeTHF) glass. These media are increasingly effective in
vibrational cooling of guest molecules, owing to an increasing
density of vibrational states.43,44 We found, however, that the
matrix medium does not influence the photochemistry, as
judged by the fact that the intensity of the EPR signal of triplet
3-thienylcarbene (13) was not enhanced (relative to Ar) in
either an N2 matrix or a 2-MeTHF glass. Thus, we do not favor
the mechanism involving hot ground-state 3-thienylcarbene
(13). Such an explanation would also seem to be at odds with
the relatively high thermal barrier (>30 kcal/mol) separating
singlet 3-thienylcarbene (13) and α-thial-methylenecyclopro-
pene (9).24 The more likely explanation, therefore, involves the
excited state of the diazo compound. As observed in several
other systems,45,46 the excited state of diazo compound 1 may
partition between two pathways: N2 loss, to give 3-
thienylcarbene (13), and N2 loss with concomitant rearrange-
ment, to give α-thial-methylenecyclopropene (9) without the
intervention of the carbene.
compound. Irradiation of diazo compound 1 at λ > 571 nm
results in a mixture of s-E and s-Z conformers of triplet 3-
thienylcarbene (13), with the s-Z conformer growing more
quickly, as measured by the greater intensity of the X2 and Y2
signals in the EPR spectrum (Figure 1 and Figure S8). Both
conformers continue to grow upon irradiation at slightly
shorter wavelength (λ > 534, > 497, and > 472 nm), but the X2
and Y2 signals of the s-E conformer grow more quickly, such
that the s-E- conformer of triplet 3-thienylcarbene (13)
becomes the major conformer in the EPR spectrum (Figure
1). Since both conformers appear to be stable to the irradiation
conditions, and each disappears rapidly upon irradiation at λ >
444 nm or λ = 467 10 nm, we ascribe the differential rates of
formation to the differential absorption and/or quantum yield
for carbene formation from the conformational isomers of 3-
thienyldiazomethane (1). (TD-DFT calculations predict the
electronic absorption spectra for the conformational isomers of
3-thienyldiazomethane (1) to be similar (see the Supporting
Information).)
With the EPR and visible spectra of triplet 3-thienylcarbene
(13) in hand, we expended considerable effort in an attempt to
optimize the experimental conditions to enhance the
production of carbene 13 and/or minimize the formation of
methylenecyclopropene derivative 9. (These efforts included
extensive studies of the wavelength dependence of the
photolysis of 3-thienyldiazomethane (1), as well as broad-
band “flash” irradiation.42) Although our attempts were largely
6446
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
a
Table 1. Computed Electronic Absorption Spectra of Carbenes
b
(s-E)-3-thienyl carbene (13)
(s-Z)-3-thienyl carbene (13)
phenyl carbene
phenyl carbene (expt)
λmax
f
λmax
f
λmax
f
λmax
triplet (T0)
388
355
333
328
322
321
295
278
995
316
307
303
285
278
256
252
0.0021
0.0022
0.0006
0.0006
0.0196
0.0001
0.0002
0.1056
0.0013
0.0039
0.0037
0.0132
0.0130
0.0044
0.1976
0.0048
389
356
337
328
322
314
302
278
1045
317
314
307
291
268
263
256
0.0020
0.0016
0.0002
0.0005
0.0196
0.0002
0.0010
0.1045
0.0018
0.0001
0.0111
0.0098
0.0007
0.0279
0.0549
0.1549
373
354
337
313
310
306
274
243
1140
320
311
308
285
272
271
260
0.0000
0.0012
0.0015
0.0001
0.0324
0.0002
0.0973
0.1607
0.0012
0.0007
0.0214
0.0086
0.0036
0.0082
0.0315
0.2462
430
300
246
singlet (S1)
a
b
(TD)M06/aug-cc-pVTZ//B3LYP/6-31G*; λmax (nm), f = oscillator strength. Eight lowest energy excitations reported. Reference 47.
Electronic Absorption Spectrum of Triplet 3-Thienyl-
> 363 nm) results in photoisomerization of the s-Z conformer
to the s-E conformer (Figure S4). At slightly shorter wavelength
(λ > 330 nm), the IR bands of the s-E conformer continue to
grow, and the IR bands of the acyclic isomers, pent-2-en-4-
ynethial (5 and 6), slowly appear (Figure S5). At λ > 280 nm,
another acyclic isomer, propargyl thioketene (8), appears in the
IR spectrum, accompanied by the decrease in the IR
absorptions of 5, 6, and 9 (Figure S6). Our findings
corroborate those described earlier by Sander and co-workers,16
while adding some additional detail in terms of the conforma-
tional isomerism of pent-2-en-4-ynethial (5 and 6) and
methylenecyclopropene derivatives (9). Our discovery of the
shorter-wavelength chemistry that affords thioketene 8 is also
new.
Possible Thermal Chemistry of Triplet 3-Thienylcar-
bene (13). Allowing a matrix containing triplet 3-thienylcar-
bene (13) to stand in the dark at 15 K for an extended period
of time (30−80 h) affords a small, but reproducible, decrease in
the EPR signal of (s-E)-13 (Figures 4 and S10) and the visible
absorption of 13 (Figure S12). These data suggest that the (s-
E) conformer of triplet 3-thienylcarbene (13) undergoes a very
slow thermal reaction at 15 K. IR experiments were performed
under identical conditions to try to determine the product of
this reaction, but the low concentration of triplet 3-
thienylcarbene (13) in the matrix, and the presence of several
other photoproducts, did not permit us to quantify carbene
disappearance or product growth. In general, measurements of
reaction kineticswhether by EPR, UV/vis, or IRwere
problematic because of a combination of low signal intensity
and very slow reaction rates. Thermal reactions of reactive
intermediates trapped in matrices at cryogenic temperatures are
not uncommon, and these processes typically occur via a
quantum mechanical tunneling mechanism48−50even in cases
where the reaction is formally forbidden by virtue of a change
in spin multiplicity.51,52 We consider the cyclization of (s-E) 3-
thienylcarbene (13) to 2-thiabicyclo[3.1.0]hexa-3,5-diene (12)
to be a likely candidate for such a process, although we cannot
exclude other possibilities (Scheme 6). The rearrangement of
triplet 13 to singlet 12, which does not involve extensive
motion of the heavy atoms, is computed to be slightly
exothermic (ca. 2 kcal/mol).
carbene (13). 3-Thienylcarbene (13) exhibits a weak
electronic absorption in the visible region of the spectrum
(λmax = 467 nm) that is characteristic of triplet arylcarbenes
(and the benzyl radical) (Figure 2).39 The vibronic structure
associated with the corresponding electronic transition in triplet
phenylcarbene (λmax = 430 nm) has been analyzed in detail.47
Time-dependent density functional theory calculations (TD-
DFT) do a reasonably good job of reproducing the general
features of the absorption spectrum of triplet phenylcarbene
(Table 1), although the energy of the visible excitation is
overestimated. Analogous calculations for triplet 3-thienylcar-
bene (13) predict the visible absorption to be red-shifted
relative to phenylcarbene (Table 1), in accord with
experimental observation. The absorption spectra of the
rotamers of triplet 3-thienylcarbene (13) are indistinguishable
under our experimental conditionsa result that is also borne
out by the TD-DFT calculations. In this respect, triplet 3-
thienylcarbene (13) is unlike the related singlet arylchlor-
ocarbenes (furanyl,7,8 thienyl,9,10 and their benzo ana-
logues11−13), in which the absorption spectra of the conforma-
tional isomers exhibit significant differences. In the lowest
singlet state, however, the absorption spectra of the conformers
of 3-thienylcarbene are predicted to be readily distinguishable
(λmax 995 nm vs 1045 nm; Table 1).
Photochemistry of 3-Thienylcarbene (13) and Other
C5H4S Isomers. Photoexcitation into the visible absorption
feature (λmax ca. 467 nm) of triplet 3-thienylcarbene (13), using
either broadband (λ > 444 nm) or narrow-band (λ = 467 10
nm) irradiation conditions, leads to the rapid disappearance of
the IR, UV/vis, and EPR signals of carbene 13 and the growth
of (s-Z)-α-thial-methylenecyclopropene (9), as well as the
growth of a small amount of 1H-2-thiabicyclo[3.1.0]hexa-3,5-
diene (12) (Figures 3 and S3). This behavior explains the
inability of Albers and Sander to observe carbene 13 in their
study of the photochemistry of 3-thienyldiazomethane (1).16
Our findings establish that carbene 13 is not stable to the
photolysis conditions used in the earlier experiment (λ > 435
nm).
The UV/vis spectrum of 9 exhibits an absorption at ca. 300−
400 nm (Figure S11). Irradiation into this absorption feature (λ
6447
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
triplet gap of 4.1 kcal/mol,24 but we are not overly confident in
that prediction, either. Energies reported in the earlier study
were estimated by using an additivity scheme that utilized both
QCISD(T) and MP2 energies. The application of MP2
methodology to carbenes, however, is problematic because of
spin contamination. It is evident from the data in Table 1 of ref
24 that the QCISD(T) single-point calculation predicts a triplet
ground state for 11, while two different MP2 single-point
calculations predict a singlet ground state.53 We are inclined to
discount the additivity scheme because of its reliance on MP2
data, and focus on the results afforded by CCSD (Scheme 4),
QCISD(T),24 or B3LYP (Scheme 4), which each predict a
triplet ground state with small singlet−triplet gap.
Spectroscopic Data for C5H4S Isomers. 3-Thienylcar-
bene (s-E-13): IR (N2, 10 K) 744 s cm−1; UV/vis (Ar, 10 K)
λmax 441, 449, 457, 467 nm; EPR (Ar, 15 K), |D/hc| = 0.508
cm−1, |E/hc| = 0.0554 cm−1, Z1 = 1954.7 G, X2 = 4250.0 G, Y2 =
6550.4 G, Z2 = 8823.2 G, microwave frequency = 9.491 GHz.
3-Thienylcarbene (s-Z-13): IR (N2, 10 K) 744 s cm−1; UV/vis
(Ar, 10 K) λmax 441, 449, 457, 467 nm; EPR (Ar, 15 K), |D/hc|
= 0.579 cm−1, |E/hc| = 0.0315 cm−1, Z1 = 2770.0 G, X2 = 5000.0
G, Y2 = 6337.0 G, Z2 = 9600.0 G, microwave frequency = 9.491
GHz. (s-E)-E-Pent-2-en-4-ynethial (5): IR (Ar, 10 K) 3319 s,
1575 s, 1362 w, 1255 m, 1163 m, 1159 m, 976 w, 971 w, 962 m,
776 w, 694 w, 643 w, 640 w, 631 w cm−1. (s-E)-Z-Pent-2-en-4-
ynethial (6): IR (Ar, 10 K) 3317 s, 2109 w, 1560 s, 1404 s,
1258 w, 1218 m, 1130 s, 1031 s, 862 w, 854 w, 777 s, 700 m,
643 s, 629 m, 465 w cm−1. (s-Z)-Z-Pent-2-en-4-ynethial (6): IR
(Ar, 10 K) 3327 s, 2109 w, 1554 m, 1410 m, 1364 w, 1257 m,
1128 m, 1134 m, 944 w, 858 w, 638 s, 629 m, 465 w cm−1.
Propargyl thioketene (8): IR (N2, 10 K) 3316 m, 1779 s, 1318
w, 1253 w, 645 w, 637 w, 613 w cm−1; IR (Ar, 10 K) 3324 w,
1781 s, 1315 w, 1245 w cm−1. (s-Z)-α-Thial-methylenecyclo-
propene (9): IR (N2, 10 K) 3141 w, 1718 s, 1498 w, 1474 m,
1367 m, 1358 m, 1189 w, 1169 m, 1009 w, 775 m, 696 w, 646
w, 629 w cm−1. (s-E)-α-Thial-methylenecyclopropene (9): IR
(N2, 10 K) 3143 w, 1719 s, 1714 m, 1496 s, 1480 m, 1412 w,
1362 w, 1353 w, 1190 m, 1164 m, 1104 w, 1007 w, 912 w, 908
w, 843 w, 834 w, 718 w cm−1.
Photochemistry of 3-Furyldiazomethane (3). Irradi-
ation of 3-furyldiazomethane (3), under the same conditions
that led to the successful generation of triplet 3-thienylcarbene
(13) from 3-thienyldiazomethane (1) (λ > 571 nm or > 534
nm; Ar, 10 K), failed to afford IR or EPR features attributable
to triplet 3-furylcarbene (22) (Scheme 8). Long-wavelength
irradiation (λ > 571 nm) of 3 yields a mixture of (α-
formyl)methylenecyclopropene (s-Z- and s-E-19) and an
unidentified species (U) exhibiting an IR band in the acetylenic
C−H stretching region (3429 m, 1636 m, 1476 m cm−1, Figure
S16). Irradiation at λ > 399 nm drives (s-Z)-19 away, while (s-
E)-19 and U continue to grow (Figure S17). Subsequent
irradiation (λ > 330 nm) causes 19 to decrease in intensity and
a mixture of E-pent-2-en-4-ynal (16) and Z-pent-2-en-4-ynal
(17) to appear. As in the case of the sulfur analogues, the pent-
2-en-4-ynals (16 and 17) are present in the s-E conformation
Figure 4. EPR transitions of triplet (s-E) 3-thienylcarbene (13) (Ar, 15
K). Pink: Carbene 13 formed upon irradiation of (3-thienyl)-
diazomethane (1) (λ > 472 nm, 26 h). Blue: Spectrum showing
decrease in signal intensity after standing in the dark at 15 K (46 h).
Scheme 6. Possible Tunneling Reaction of 3-Thienylcarbene
(13)
Photochemistry of 2-Thienyldiazomethane (2). Irradi-
ation of 2-thienyldiazomethane (2), under the same conditions
that led to the successful generation of triplet 3-thienylcarbene
(13) (λ > 534 nm; Ar, 10 K), failed to afford IR or EPR
features attributable to triplet 2-thienylcarbene (11) (Scheme
7). In accord with the results of earlier studies,16 the ring-
opened product, Z-pent-2-en-4-ynethial (6), is the only species
observed by IR spectroscopy (Figure S13). Careful examination
of the spectrum establishes that both s-E and s-Z rotamers of 6
are present. Further irradiation (λ > 363 nm) results in cis−
trans photoisomerization of Z-pent-2-en-4-ynethial (6) and E-
pent-2-en-4-ynethial (5) (present only as the s-E conformer;
Figure S14). Irradiation at λ > 237 nm affords propargyl
thioketene (8) (Figure S15).
The nature of the electronic ground state of 2-thienylcarbene
(11)singlet or tripletremains unclear at this time. Our
DFT calculations predict a triplet ground state with a small
singlet−triplet gap (1−2 kcal/mol) for both conformers of 2-
thienylcarbene (11) (Scheme 2). We are not overly confident
in the prediction of the ground state multiplicity, since the two
states lie so close in energy and this computational method-
ology typically underestimates the energy of the singlet, relative
to the triplet, by 1−3 kcal/mol.35,36 McKee’s calculations
predict a singlet ground state for carbene 11 with a singlet−
Scheme 7. Photochemistry of 2-Thienyldiazomethane (2) (Ar, 10 K)
6448
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
10 K) 1743 w, 1690 s, 1665 w, 1511 m, 1146 w, 1016 w, 799 w
cm−1. Unknown (U): IR (Ar, 10 K) 3429 m, 1636 m, 1476 m
cm−1.
Scheme 8. Photochemistry of 3-Furyldiazomethane (3) (Ar,
10 K)
DISCUSSION
■
EPR Spectrum and Electronic Structure of Triplet 3-
Thienylcarbene (13). The triplet EPR spectrum obtained
upon irradiation of 3-thienyldiazomethane (1) (Figure 1)
exhibits several unusual features, with respect to the spectra of
triplet arylcarbenes.39,55−57 Conformational isomerism, a heavy
atom effect of sulfur, and site effects in the matrix may all be
manifest in the spectrum, and a proper interpretation is not
readily apparent upon cursory inspection. The assignment of
the individual EPR transitions to major and minor conforma-
tional isomers derives from the differential rate of appearance as
a function of photolysis wavelength (λ > 571 vs λ > 472 nm;
Figure 1 and Figure S8), as well as comparison with EPR
spectra of other 2- and 3-thienylcarbene derivatives obtained in
subsequent investigations.58,59 The zero-field splitting param-
eters of the major and minor isomers were calculated in the
usual way, using the best fit of the EPR transitions with the spin
Hamiltonian under the assumption that gx = gy = gz = ge.60 The
striking result is the large difference in magnitude of the D
values between the major (D = 0.508 cm−1) and minor (D =
0.579 cm−1) isomers observed in the triplet EPR spectrum.
Simulation. Simulation of the triplet EPR spectrum, using
the program XSophe (Bruker), was accomplished by manually
varying values of D, E, and g to give an optimal fit. The
experimental and simulated spectra exhibit excellent agreement
(Figure 5). The best fit was obtained with g = 2.0 and the zero-
field parameters D and E given in Table 2. In the experimental
spectrum, the major isomer exhibits subtle features in the X2
and Y2 transitions, which we attribute to inequivalent sites in
the argon matrix. These features are well-reproduced, in the
simulated spectrum, by two species (A and A′) that differ only
slightly in terms of E value (Table 2). Although the g = 2 signal
in the experimental spectrum appears to be intense, its
contribution in terms of percent of spins is negligible (0.3%),
as established by the simulation. The quality of the simulated
spectrum, and the excellent agreement of the zero-field
parameters determined from the experimental spectrum and
from the simulated spectra give confidence in the correctness of
the analysis.
(Figure S18). The IR assignments for 16, 17, and 19 are in
good agreement with those reported previously.17
The photochemical appearance and disappearance of
compound U is reproducible across multiple experiments, yet
the identity of this species eludes us. We initially suspected
propargyl ketene (18)the oxygen analogue of propargyl
thioketene (8) that was observed in the photochemistry of the
related sulfur-containing system. The observed IR absorptions
of U, however, are inconsistent with the IR spectrum computed
for ketene 18. (The apparent absence of ketene or carbonyl
stretching vibrations in the spectrum of U is telling.) We
considered several other C5H4O isomers that contain a terminal
alkyne moiety (Scheme S2), but none of the computed IR
spectra shows a good correlation with the experimentally
observed bands of U. We also considered the possibility of a
photofragmentation reaction, but the observed IR absorption at
3429 cm−1 does not correspond to that of either acetylene54 or
diacetylene.
Photochemistry of 2-Furyldiazomethane (4). Under
gentler conditions than those that led to the successful
generation of triplet 3-thienylcarbene (13) (λ > 571 nm; Ar,
10 K), irradiation of (2-furyl)diazomethane (4) (λ > 613 nm;
Ar, 10 K) failed to afford IR or EPR features attributable to
triplet 2-furylcarbene (22) (Scheme 9). In accord with the
Scheme 9. Photochemistry of 2-Furyldiazomethane (4) (Ar,
10 K)
That the EPR spectrum of triplet 3-thienylcarbene (13)
could be interpreted and simulated by using g = 2.0 was not
obvious to us, at the outset, because of the possible influence of
a heavy-atom effect by sulfur. Sulfur-containing radicals have
been widely studied in biological systems.61 In cases where an
unpaired electron is localized on sulfur, g values range from
1.96 to 2.073.62 In the case of carbene 13, the fact that little
spin density resides at sulfur (see below), and that the nature of
the EPR spectrum is much more sensitive to the values of D
and E than to the value of g, suggest that any manifestation of
spin−orbit coupling on g, through the heavy-atom effect of
sulfur, is minimal.
results of earlier studies,17 irradiation of diazo compound 4 (λ >
613 nm; Ar, 10 K) causes fragmentation of the furan ring,
yielding Z-pent-2-en-4-ynal (17) (Figure S19). Subsequent
irradiation at λ > 444 nm gives rise to E-pent-2-en-4-ynal (16)
(Scheme 9, Figure S20).
Spectroscopic Data for C5H4O Isomers. (s-E)-E-Pent-2-
en-4-ynal (16): IR (Ar, 10 K) 3313 m, 1710 s, 1594 m, 1264 w,
1124 s, 963 m, 716 w, 641 m, 583 w cm−1. (s-E)-Z-Pent-2-en-4-
ynal (17): IR (Ar, 10 K) 3324 m, 2845 w, 2826 w, 2743 w,
1699 s, 1582 m, 1266 w, 1096 w, 909 m, 643 w cm−1. (s-Z)-(α-
Formyl)methylenecyclopropene (19): IR (Ar, 10 K) 1749 s,
1647 s, 1494 m, 1344 w, 1114 w, 1080 w, 729 w, 705 m, 602 m
cm−1. (s-E)-(α-Formyl)methylenecyclopropene (19): IR (Ar,
Spectroscopic Assignments. The preceding analysis
provides the zero-field splitting parameters for the major and
minor species observed in the matrix, but it does not establish
the spectroscopic assignments. In making the assignments, we
consider several interpretations of the EPR spectrum (Figure
4). In the first, which is the assignment that we ultimately favor,
we hypothesize that the transitions for the major and minor
isomers arise from s-E/s-Z conformational isomerism of triplet
6449
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
Scheme 10. Plausible Carbene Rearrangements
dene (10). We therefore reject the assignment of 10 as a carrier
of a triplet EPR signal.
In the third interpretation of the EPR spectrum, we
hypothesize that the transitions for the major and minor
isomers arise from the presence of a mixture of triplet 2-
thienylcarbene (11) and 3-thienylcarbene (13). (In this
scenario, the doubling of spectral transitions might arise
through either s-E/s-Z conformational isomerism or multiple
matrix sites.) That triplet 2-thienylcarbene (11) cannot be
generated by direct irradiation of (2-thienyl)diazomethane (2)
does not exclude the possibility that triplet 11 might be formed
and observable under different reaction conditions (e.g.,
photochemical or hot molecule rearrangement of 3-thienylcar-
bene (13) or 4-thiacyclohexa-1,2,5-triene (7); Scheme 10).
Although this scenario seems improbablegiven the very low
barrier for ring-opening of singlet 2-thienylcarbene (11)we
cannot rigorously exclude it.
Figure 5. Bottom: EPR spectrum of triplet (s-E) and (s-Z)-3-
thienylcarbene (13), obtained upon irradiation (λ > 472 nm, 22 h) of
3-thienyldiazomethane (1) (Ar, 15 K). Top: Simulated spectrum of
triplet 3-thienylcarbene (13) (see text).
Conformational Isomerism. Here, we follow the analysis
of Roth and co-workers assigning geometric isomers for triplet
arylcarbenes.55−57 If the conformational isomers of a triplet
carbene experience different dipolar coupling between the
unpaired electrons, the conformers will exhibit different D
values. Differences in dipolar coupling may arise as a
consequence of differences in spin density at the β-positions
of a triplet carbene in which one unpaired electron is
delocalized across a π-electron system. Thus, we computed
natural spin densities for both triplet 3-thienylcarbene (13) and
3-furylcarbene (23) (Table 3). The finding of a large difference
in spin density at the β-positions of the carbene (C2 = +0.42,
C4 = +0.13) is consistent with a large difference in the D values
of the conformational isomers of 3-thienylcarbene (13) (0.508
cm−1 vs 0.579 cm−1). Empirical correlations establish that the
larger D value arises when the sp orbital of a triplet carbene is
oriented anti with respect to the β-position bearing the larger
spin density.55−57 Since C2 is the β-position bearing higher spin
density in triplet 3-thienylcarbene (13), the s-Z conformer of
13 is assigned as the isomer with the larger D value (minor
isomer in Figure 5), and the s-E conformer of 13 is assigned as
the isomer with the smaller D value (major isomer in Figure 5).
It is interesting to note that the difference in D values between
the two conformers of 3-thienylcarbene (ΔD = 0.071 cm−1) is
substantially larger than the difference observed for conformers
of most arylcarbenes (ΔD ≈ 0.02−0.03 cm−1), vinylcarbenes
(ΔD ≈ 0.05 cm−1), and α-carbonylcarbenes (ΔD ≈ 0.05
cm−1).55−57 The value rivals that reported for the conformers of
benzoyl phenyl carbene (ΔD = 0.08 cm−1).
Table 2. Zero-Field Splitting Parameters (in cm−1) of Triplet
3-Thienylcarbene (13)
experimental
XSophe simulation
conformer
|D/hc|
|E/hc|
|D/hc|
|E/hc|
% spins
A
0.508
0.0554
0.505
0.505
0.575
0.05454
0.05707
0.03128
56
28
16
0.3
A′
B
0.579
0.0315
g = 2
3-thienylcarbene (13). (In other words, the EPR spectra of the
conformational isomers are quite different.) In this scenario, the
spectral doubling of the X2 and Y2 transitions is attributed to
matrix site splitting.
In the second interpretation of the EPR spectrum, we
hypothesize that the transitions for the major isomerfor
which the X2 and Y2 transitions exhibit features of spectral
doublingarise from conformational isomers of triplet 3-
thienylcarbene (13). (In other words, the EPR spectra of the
conformational isomers are quite similar.) We hypothesize that
the transitions for the minor isomerwhich are quite
differentarise from a different triplet species. In our mind, a
plausible candidate is 4-thia-2,5-cyclohexadienylidene (10)
(Scheme 10), which could be readily envisioned to form
upon photochemical or hot molecule rearrangement of 3-
thienylcarbene (13) and bicyclic compound 12. DFT
calculations by McKee et al.24 and by us (Scheme 2), however,
predict a singlet ground state for 4-thia-2,5-cyclohexadienyli-
In an effort to obtain independent evidence bearing on the
assignment of the conformational isomers of triplet 3-
6450
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
most important conclusion from these studies is the absence of
a readily identifiable, dominant interaction between these
centers. The analysis does reveal a weak interaction between
the filled sp orbital at the carbene carbon and the empty
antibonding σ orbital of the C2−S1 bond in the s-Z isomer
(Figure 6). This interaction was absent in the s-E isomer. The
Table 3. Natural Spin Densities and Zero-Field Splitting
Parameters for Triplet 3-Thienylcarbene (13) and Triplet 3-
Furylcarbene (23)
a
Figure 6. Natural Bond Orbital analysis of triplet 3-thienylcarbene
(13).
energy of this interaction is not large (0.37 kcal/mol), but it
does reveal a subtle distinction in the electronic structure of the
conformational isomers.
In summary, our analysis of conformational isomerism in
triplet 3-thienylcarbene (13) is commensurate with a conven-
tional interpretation, in which differences in spin density at the
nonequivalent ortho positions give rise to differences in dipolar
spin−spin coupling. NBO analyses provide a tenuous hint
concerning differential electronic interactions between sulfur
and the carbene moiety in the s-E and s-Z conformers of triplet
13. It should be noted, however, that nothing in the analysis
rigorously requires the assignment of conformers to be correct.
The structural assignment is based upon model systems that do
not contain a heavy atom.57 Literature precedent, although
sparse, suggests that the spin−orbit contribution in a system
that contains a heavy atom will be dominant.64 It is conceivable
that a significant spin−orbit contribution could reverse the
predictions of which conformer has the larger or smaller D
value, thereby reversing the assignments of s-E and s-Z
conformers of triplet 3-thienylcarbene (13).
Absence of the Other Carbenes. The inability to
generate and characterize triplet 3-furylcarbene (23) was the
biggest disappointment for us in the current investigation. The
electronic absorption spectrum of 3-furyldiazomethane (3) is
quite similar to the corresponding sulfur analogue (1), and we
did not anticipate a significant difference in the photochemistry
of these diazo compounds. For both oxygen- and sulfur-
containing carbenes (23 and 13), DFT calculations predict
triplet ground states for both s-E and s-Z conformers, with small
singlet−triplet energy gaps (ca. 2−5 kcal/mol) (Schemes 2−4).
The computed singlet−triplet gaps of the 3-furylcarbenes (23)
are slightly smaller than those of the 3-thienylcarbenes (13)
the species that are spectroscopically observed. One conclusion
to be gleaned from the computational data shown in Schemes 2
and 3 is that the methylenecyclopropene derivatives, which are
formed as the major products during the initial photolysis, lie
significantly lower in energy than the singlet carbenes for the
oxygen series (ca. 24 kcal/mol lower), compared to the sulfur
series (ca. 14 kcal/mol lower). This situation is consistent with
a scenario in which a vibrationally hot 3-furylcarbene (23)
would be more prone to suffer rearrangement to methyl-
enecyclopropene (19) than the corresponding 3-thienylcarbene
(13)relative to undergoing vibrational cooling by the matrix.
The inability to generate and characterize 2-thienylcarbene
(11) or 2-furylcarbene (22) is less surprising. DFT calculations,
by us and by others,30,31 predict a singlet ground state for both
a
NBO spin densities at B3LYP/6-31G*; zero-field splitting parameters
at B3LYP/EPRIII.
thienylcarbene (13), we sought to utilize newly developing
methods for the first-principles calculation of zero-field splitting
parameters. Neese’s ORCA program shows considerable
promise for the analysis of a variety of high-spin organic
reactive intermediates.63−66 Unfortunately, we find that the
ORCA calculations for the zero-field splitting parameters of
triplet 3-thienylcarbene (13) are not robust. The computed
values of D and E are much too sensitive to the basis set and
level of theory. These problems are not manifest in calculations
of the corresponding 3-furylcarbene (23) and therefore appear
to be related to the presence of the second-row atom (sulfur).
For species composed only of first-row elements, the spin−spin
interaction term makes the dominant (sole) contribution to D.
These systems appear to be well-handled by ORCA. For
species that include heavier elements, Neese has shown that the
spin−orbit interaction term makes the dominant contribution
to D.64
Comparison of 3-thienylcarbene (13) and 3-furylcarbene
(23) is instructive. The computed zero-field splitting
parameters of 3-furylcarbene (23) exhibit the expected effects
of conformational isomerism (Table 3), and the prediction that
the (s-Z) conformer displays the larger D value is consistent
with the qualitative expectations described in the preceding
section. The difference in D values between the two conformers
of 3-furylcarbene (23) (ΔD = 0.025 cm−1), however, is
significantly smaller than that of 3-thienylcarbene (13) (ΔD =
0.071 cm−1). While the very good agreement between the D
values for the (s-E) conformers of these carbenes is
undoubtedly fortuitous (0.508 cm−1 observed for 13; 0.506
cm−1 computed for 23), it serves to focus attention on the fact
the experimental and computed D values for the (s-Z)
conformers differ substantially (0.579 cm−1 observed for 13;
0.531 cm−1 computed for 23). Although the computed spin
densities in both systems are quite similar (Table 2), there
appears to be an interaction in the (s-Z) conformer of triplet 3-
thienylcarbene (13) that renders it uniquely different, in terms
of its zero-field splitting parameter D.
We employed NBO calculations to investigate the possible
electronic interactions between the sulfur heteroatom and the
triplet carbene center in 3-thienylcarbene (13). Perhaps the
6451
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
diagonalization method was utilized (with 1600 partitions and 8
segments).
conformers of 2-furylcarbene (22). The recent analysis by
Herges and Haley reveals a substantial thermodynamic
stabilization of heteroaryl carbenes.28 Ring-opening of 2-
thienylcarbene (11) or 2-furylcarbene (22) to the correspond-
ing pentenynal is exothermic, relative to the singlet carbenes, by
20 kcal/mol in the sulfur case and 30 kcal/mol in the oxygen
case. The computed barriers for ring-opening of syn- and anti-2-
furylcarbene (22) are very low (2−4 kcal/mol).30,31 Herges29
described the reaction as a coarctate transformation, while
Birney31 interpreted it as a pseudopericyclic process.
Independent of the classification, in terms of orbital symmetry,
the reaction is one that may be envisioned to occur via a
tunneling mechanism (at least at cryogenic temperature). Our
experiments, of course, do not establish that either 2-
thienylcarbene (11) or 2-furylcarbene (22) is directly involved
in the photochemistry of the corresponding diazo compounds.
The literature contains many examples in which reactions occur
in an excited state of a diazo compound, circumventing a
carbene intermediate altogether.45,46
Preparation of Tosylhydrazones. The tosylhydrazone precur-
sors to compounds 1, 2, 3, and 4 were synthesized by using a modified
method given by Katritzky.78 p-Toluenesulfonhydrazide (4.65 g, 25
mmol) was added to a solution of the corresponding aldehyde (25
mmol) in 30 mL of methanol and refluxed for 12 h. The solution was
diluted with water (50 mL) and the precipitate was collected by
suction filtration. The crude product was recrystallized from methanol.
Thiophene-3-carboxaldehyde Tosylhydrazone. White crys-
tals, yield 70%; mp 155−157 °C (lit.22 mp 157.5−159 °C); H NMR
1
(Me2SO-d6) δ 2.35 (s, 3H), 7.27 (d, J = 4.6 Hz, 1H), 7.39 (d, J = 8.5
Hz, 2H), 7.56 (dd, J = 4.9, 2.8 Hz, 1H), 7.75 (d, J = 8.2 Hz, 2H), 7.80
(d, J = 1.9 Hz, 1H), 7.93 (s, 1H), 11.28 (s, 1H); MS (ESI) (MH+)
281.1; MS (EMM) (MH+) calcd 281.0413, measured 281.0413.
Thiophene-2-carboxaldehyde Tosylhydrazone. White crys-
tals, yield 83%; mp 145−146 °C (lit.22 mp 142−143.5 °C); H NMR
1
(Me2SO-d6) δ 2.29 (s, 3H), 7.06 (dd, J = 4.9, 3.6 Hz, 1H), 7.35 (dd, J
= 3.5, 1.1 Hz, 1H), 7.40 (d, J = 8.3 Hz, 2H), 7.59 (d, J = 4.1 Hz, 1H),
7.71 (d, J = 8.2 Hz, 2H), 8.08 (s, 1H); MS (ESI) (MH+) 281.1; MS
(EMM) (MH+) calcd 281.0413, measured 281.0410.
Furan-3-carboxaldehyde Tosylhydrazone. Tan crystals, yield
36%; mp 118−120 °C (lit.22 mp 116−119 °C); 1H NMR (Me2SO-d6)
δ 2.36 (s, 3H), 6.62 (s, 1H), 7.39 (d, J = 8.1 Hz, 2H), 7.68 (s, 1H),
7.73 (d, J = 8.1 Hz, 2H), 7.85 (s, 1H), 8.02 (s, 1H), 11.26 (s, 1H); MS
(ESI) (MH+) 265.1 MS (EMM) (MH+) calcd 265.0642, measured
265.0638.
Furan-2-carboxaldehyde Tosylhydrazone. Light tan crystals,
yield 44%; mp 113−115 °C (lit.22 mp 125−126 °C); 1H NMR
(Me2SO-d6) δ 2.31 (s, 3H), 6.55 (dd, J = 3.3, 1.9 Hz, 1H), 6.79 (d, J =
3.2 Hz, 1H), 7.40 (d, J = 8.1 Hz, 2H), 7.71 (d, J = 8.3 Hz, 2H), 7.76
(d, J = 1.6 Hz, 1H), 7.77 (s, 1H), 11.43 (s, 1H); MS (ESI) (MH+)
265.1; MS (EMM) (MH+) calcd 265.0642, measured 265.0654.
Preparation of Tosylhydrazide Sodium Salts. Sodium hydride
(1.1 equiv, 60% suspension in mineral oil) was added to the
corresponding tosylhydrazone (1 equiv) in THF and the mixture was
allowed to stir at room temperature for 1 h. The resulting precipitate
was filtered and dried on a vacuum line. This reaction was run on a
500 mg scale with respect to the tosylhydrazone.
Preparation of Diazo Compounds 1−4. The diazo compounds
were generated by heating the corresponding tosylhydrazone sodium
salts to 110 °C under vacuum. The highly colored diazo compound
was collected on a coldfinger at −78 °C and rinsed into a deposition
tube with CH2Cl2. After solvent removal under vacuum at −41 °C, the
diazo compound was deposited onto a cold window (IR, UV/vis) or
copper rod (EPR) with a constant flow of argon or nitrogen. This
procedure yielded the matrix isolated samples of diazo compounds 1,
2, 3, or 4.
CONCLUSIONS
■
Triplet 3-thienylcarbene (13) has been generated upon
irradiation of 3-thienyldiazomethane (1) (λ > 534 nm) and
characterized by IR, UV/vis, and EPR spectroscopy. The s-E
and s-Z conformers of the triplet carbene exhibit substantially
different zero-field splitting parameters (D = 0.508 and 0.579
cm−1, respectively), which arise as a consequence of a large
difference in spin density at the two ortho positions in the
thiophene ring. Despite finding experimental conditions for the
successful generation of triplet 3-thienylcarbene, the conditions
were not transferable for generating triplet 2-thienylcarbene
(11) or the isomeric 2- or 3-furylcarbenes (22, 23).
METHODS SECTION
■
1
General. H NMR spectra (300 MHz) were obtained in Me2SO-
d6; chemical shifts (δ) are reported as ppm downfield from internal
SiMe4. Mass spectra and exact mass measurement (EMM) were
obtained by using electrospray ionization (ESI). Thiophene-3-
carboxaldehyde, thiophene-2-carboxaldehyde, furan-3-carboxaldehyde,
and furan-2-carboxaldehyde were purchased from commercial sources
and purified by vacuum distillation. The matrix isolation apparatus and
technique have been described previously,67,68 and additional details
are provided as Supporting Information.
Computational Methods. Optimized geometries, energies, and
infrared intensities were obtained at the B3LYP/6-31G* level of
theory, using the Gaussian software package.69 The nature of
stationary points was confirmed by calculation of the harmonic
vibrational frequencies, which also provided zero-point vibrational
energy (ZPVE) corrections. Vibrational frequencies were not scaled.
Geometries, harmonic vibrational frequencies, and singlet−triplet
energy gaps for 2- and 3-thienylcarbene (11 and 13) were also
computed by using coupled-cluster methodology, CCSD,70 which
includes single and double excitations. The correlation-consistent cc-
pVTZ basis set was used.71−73 All CCSD calculations were done in the
frozen-core approximation with the CFOUR program system.74
Natural bond orbital (NBO) calculations were performed at the
B3LYP/6-31G* level of theory using the NBO program.75 Electronic
absorption spectra were computed at the B3LYP/6-31G* geometries,
using time-dependent density functional theory methods (M06 and
CAM-B3LYP) and the aug-cc-pVTZ basis set, as implemented in
Gaussian09.69 First-principles calculations of zero-field splitting
parameters D and E were performed at the B3LYP/6-31G*
geometries, using the B3LYP functional and the EPRIII basis set, as
implemented in the ORCA program.76 Simulations of EPR spectra of
randomly oriented triplets were performed with use of the XSophe
program,77 as supplied by Bruker Instruments. The matrix
Procedure for Determining Extinction Coefficients for Diazo
Compounds 1−4. Diazo compounds 1−4 were formed via the
procedure given above, but were rinsed from the coldfinger with
CD3CN instead of CH2Cl2 to permit the use of 1H NMR
spectroscopy. An NMR sample was made with 1 mL of the original
sample and the remaining sample was used to make stock solutions at
varying concentrations. The NMR solution was spiked with a known
amount of benzene to serve as an internal standard. This allowed the
molar quantity of diazo compound to be determined, based upon the
integration of the peaks in the spectrum. Finally, dividing the moles of
diazo compound in the original sample by the 1 mL sample size used
for NMR spectroscopy, the molarity of the original solution could be
determined. The different stock solutions were used to determine λmax
in the visible and ultraviolet regions of the UV/vis spectrum and the
Beer−Lambert law was used to determine extinction coefficients.
Samples and solutions were kept in dry ice when they were not being
used, in order to minimize decomposition. Solutions slowly warmed to
room temperature during the course of NMR or UV/vis spectroscopic
measurements.
(3-Thienyl)diazomethane (1). Red liquid. UV/vis (CH3CN) λmax
(nm) (ε) (L mol−1 cm−1) 494 (5.98), 260 (2820), 229 (2950), 192
(2010); IR (N2, 10 K) 2065 s, 1597 w, 1544 w, 1534 w, 1435 w, 1404
6452
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
w, 859 w, 841 w, 763 m, 709 w, 622 w, 485 w, 445 w cm−1; 1H NMR
(CD3CN) δ 5.31 (s, 1H), 6.78 (dd, J = 1.3, 5.0 Hz, 1H), 6.83 (dd, J =
1.3, 3.0 Hz, 1H), 7.44 (dd, J = 3.0, 5.0 Hz, 1H).
(4) Moss, R. A.; Turro, N. J. In Kinetics and Spectroscopy of Carbenes
and Biradicals; Platz, M. S., Ed.; Plenum Press: New York, 1990, pp
213−238.
(2-Thienyl)diazomethane (2). Red liquid. UV/vis (CH3CN) λmax
(nm) (ε) (L mol−1 cm−1) 502 (8.13), 297 (8620), 204 (6120); IR (Ar,
10 K) 2071 s, 1597 w, 1523 w, 1449 w, 1381 w, 1306 w, 1077 w, 803
w, 680 w cm−1; 1H NMR (CD3CN) δ 5.54 (s, 1H), 6.86 (dd, J = 1.0,
3.5 Hz, 1H), 6.99 (dd, J = 3.5, 5.0 Hz, 1H), 7.15 (dd, J = 1.0, 5.0 Hz,
1H).
(5) Jones, M., Jr.; Moss, R. A. In Reactive Intermediate Chemistry;
Moss, R. A., Platz, M. S., Jones, M., Jr., Eds.; Wiley: Hoboken, NJ,
2004, pp 273−328.
(6) Moss, R. A.; Jang, E. G.; Kim, H.-R.; Ho, G.-J.; Baird, M. S.
Tetrahedron Lett. 1992, 33, 1427−1430.
(7) Khasanova, T.; Sheridan, R. S. J. Am. Chem. Soc. 1998, 120, 233−
234.
(8) Khasanova, T.; Sheridan, R. S. Org. Lett. 1999, 1, 1091−1093.
(9) Baird, M. S.; Bruce, I. J. Chem. Res., Synop. 1990, 134−135.
(10) Moss, R. A.; Fu, X.; Ma, Y.; Sauers, R. R. Tetrahedron Lett. 2003,
44, 773−776.
(11) Khasanova, T.; Sheridan, R. S. J. Am. Chem. Soc. 2000, 122,
8585−8586.
(12) Nikitina, A.; Sheridan, R. S. J. Am. Chem. Soc. 2002, 124, 7670−
7671.
(13) Nikitina, A. F.; Sheridan, R. S. Org. Lett. 2005, 7, 4467−4470.
(14) Wang, J.; Sheridan, R. S. Org. Lett. 2007, 9, 3177−3180.
(15) Kaiser, R. I. Chem. Rev. 2002, 102, 1309−1358.
(16) Albers, R.; Sander, W. J. Org. Chem. 1997, 62, 761−764.
(17) Albers, R.; Sander, W. Liebigs Ann./Recl. 1997, 897−900.
(18) Osaka, I.; McCullough, R. D. Acc. Chem. Res. 2008, 41, 1202−
1214.
(3-Furyl)diazomethane (3). Orange-red liquid. UV/vis (CH3CN)
λmax (nm) (ε) (L mol−1 cm−1) 491 (8.51), 251 (8050), 211 (5260); IR
(Ar, 10 K) 2066 s, 1594 w, 1509 w, 1416 w, 1368 w, 1170 w, 1065 w,
1
1028 w, 874 w, 763 w, 589 w cm−1; H NMR (CD3CN) δ 5.03 (s,
1H), 6.29 (s, 1H), 7.34 (m, 1H), 7.49 (m, 1Hvery small coupling).
(The NMR sample was spiked with a known quantity of benzene to
enable the determination of the solution concentration of diazo
1
compound. The H NMR resonance of benzene interferes with the
7.34-ppm resonance of diazo compound 3, precluding an accurate
determination of chemical shift/integration/coupling constant.)
(2-Furyl)diazomethane (4). Orange-red liquid. UV/vis (CH3CN)
λmax (nm) (ε) (L mol−1 cm−1) 496 (16.3), 278 (10800); IR (Ar, 10 K)
2076 s, 1593 w, 1516 w, 1428 w, 1007 w, 927 w, 884 w, 767 w, 713 w,
1
657 w, 593 w cm−1; H NMR (CD3CN) δ 5.36 (s, 1H), 6.03 (d, J =
3.3 Hz, 1H), 6.45 (dd, J = 2.0, 3.3 Hz, 1H), 7.42 (dd, J = 0.7, 2.0 Hz,
1H).
(19) Handbook of Thiophene-Based Materials: Applications in Organic
Electronics and Photonics; Perepichka, I. F., Perepichka, D. F., Eds.;
Wiley: Chichester, UK, 2009.
(20) Sun, J.; Zhang, B.; Katz, H. E. Adv. Funct. Mater. 2011, 21, 29−
45.
(21) Hoffman, R. V.; Shechter, H. J. Am. Chem. Soc. 1971, 93, 5940−
5941.
(22) Hoffman, R. V.; Orphanides, G. G.; Shechter, H. J. Am. Chem.
ASSOCIATED CONTENT
■
S
* Supporting Information
Matrix-isolation spectra (IR, UV/vis, EPR) associated with the
photolyses of diazo compounds 1−4; details concerning
XSophe simulation of triplet EPR spectra; computed energies,
harmonic vibrational frequencies, infrared intensities, and
Cartesian coordinates of all computed structures. This material
is available free of charge via the Internet at http://pubs.acs.org.
Soc. 1978, 100, 7927−7933.
(23) Pan, W.; Balci, M.; Shevlin, P. B. J. Am. Chem. Soc. 1997, 119,
5035−5036.
(24) McKee, M. L.; Shevlin, P. B.; Zottola, M. J. Am. Chem. Soc.
2001, 123, 9418−9425.
AUTHOR INFORMATION
■
(25) Nakatani, K.; Adachi, K.; Tanabe, K.; Saito, I. J. Am. Chem. Soc.
1999, 121, 8221−8228.
Corresponding Author
*mcmahon@chem.wisc.edu
(26) McClintock, S. P.; Shirtcliff, L. D.; Herges, R.; Haley, M. M. J.
Org. Chem. 2008, 73, 8755−8762.
Present Address
#Department of Chemistry, Carroll College, 1601 North
Benton Ave, Helena, MT 59625
(27) Shirtcliff, L. D.; McClintock, S. P.; Haley, M. M. Chem. Soc. Rev.
2008, 37, 343−364.
(28) Young, B. S.; Kohler, F.; Herges, R.; Haley, M. M. J. Org. Chem.
̈
Notes
2011, 76, 8483−8487.
The authors declare no competing financial interest.
(29) Herges, R. Angew. Chem., Int. Ed. 1994, 33, 255−273.
(30) Sun, Y.; Wong, M. W. J. Org. Chem. 1999, 64, 9170−9174.
(31) Birney, D. M. J. Am. Chem. Soc. 2000, 122, 10917−10925.
(32) Seburg, R. A.; McMahon, R. J.; Stanton, J. F.; Gauss, J. J. Am.
Chem. Soc. 1997, 119, 10838−10845.
ACKNOWLEDGMENTS
■
We gratefully acknowledge financial support from the National
Science Foundation (NSF-1011959). We also acknowledge
NSF support for Departmental facilities used in this research:
EPR spectrometer (NSF-9013030), computing facilities (NSF-
0091916 and NSF-0840494), and NMR instrumentation (NSF-
0342998). S.A.R. acknowledges a Way-Klingler sabbatical
fellowship from Marquette University. We thank Prof. Thomas
Brunold (UW-Madison) for assistance with the ORCA
calculations.
(33) Plattner, D. A.; Houk, K. N. J. Am. Chem. Soc. 1995, 117, 4405−
4406.
(34) Woodcock, H. L.; Schaefer, H. F.; Schreiner, P. R. J. Phys. Chem.
A 2002, 106, 11923−11931.
(35) Woodcock, H. L.; Moran, D.; Brooks, B. R.; Schleyer, P. v. R.;
Schaefer, H. F. I. J. Am. Chem. Soc. 2007, 129, 3763−3770.
(36) Perrotta, R. R.; Winter, A. H.; Coldren, W. H.; Falvey, D. E. J.
Am. Chem. Soc. 2011, 133, 15553−15558.
(37) Our ability to observe the IR absorption of triplet 3-
thienylcarbene (13) relied on the use of a nitrogen matrix instead of
an argon matrix. IR experiments carried out in argon failed to reveal
any bands attributable to the carbene, due to the unfortunate overlap
between a strong absorption of 3-thienyldiazomethane (1) (disappear-
ing) and the strongest absorption of triplet 3-thienylcarbene (13)
(appearing). The nitrogen matrix shifts the peaks just enough that the
absorption of carbene 13 can be observed.
REFERENCES
■
(1) Wentrup, C. In Methoden der Organischen Chemie (Houben-Weyl);
Regitz, M., Ed.; G. Thieme: Stuttgart, Germany, 1989; Vol. E19b, pp
824−1021.
(2) Kinetics and Spectroscopy of Carbenes and Biradicals; Platz, M. S.,
Ed.; Plenum: New York, 1990.
(38) The IR bands of the minor species in the matrix are weak
(3) Reactive Intermediate Chemistry; Moss, R. A., Platz, M. S., Jones,
M., Jr., Eds.; Wiley: Hoboken, NJ, 2004.
(Figure S1), and hence the comparison with the computed spectrum
6453
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454

Journal of the American Chemical Society
Article
of 1H-2-thiabicyclo[3.1.0]hexa-3,5-diene (12) does not establish the
structural assignment with a high degree of certainty. We therefore
considered alternative structural assignments. 4-Thiacyclohexa-1,2,5-
triene (7) is another low-energy isomer on the C5H4S potential energy
surface, but the computed IR spectrum of 7 exhibits poor agreement
with the experimental IR bands (Figure S2). Each of the other C5H4S
isomers (Scheme 2) was also considered, but, in the end, the
computed spectrum of 12 remained the most plausible assignment.
(39) Sander, W.; Bucher, G.; Wierlacher, S. Chem. Rev. 1993, 93,
1583−1621.
(68) Seburg, R. A.; McMahon, R. J. J. Am. Chem. Soc. 1992, 114,
7183−7189.
(69) Frisch, M. J. et al. Gaussian 09; Revision B.1, 2009.
(70) Purvis, G. D.; Bartlett, R. J. J. Chem. Phys. 1982, 76, 1910−1918.
(71) Dunning, T. H. Jr. J. Chem. Phys. 1989, 90, 1007−1023.
(72) Woon, D. E.; Dunning, T. H. Jr. J. Chem. Phys. 1993, 98, 1358−
1371.
(73) Dunning, T. H. Jr.; Peterson, K. A.; Wilson, A. K. J. Chem. Phys.
2001, 114, 9244−9253.
(74) Stanton, J. F.; Gauss, J.; Harding, M. E.; Szalay, P. G. CFOUR,
Coupled-Cluster Techniques for Computational Chemistry; with con-
tributions from A. A. Auer, R. J. Bartlett, U. Benedikt, C. Berger, D. E.
Bernholdt, Y. J. Bomble, L. Cheng, O. Christiansen, M. Heckert, O.
(40) Trozzolo, A. M.; Wasserman, E. In Carbenes; Moss, R. A., Jones,
M., Jr., Eds.; Robert E. Krieger: Malabar, FL, 1983; Vol. II, pp 185−
206.
́
Heun, C. Huber, T.-C. Jagau, D. Jonsson, J. Juselius, K. Klein, W. J.
Lauderdale, D. A. Matthews, T. Metzroth, D. P. O’Neill, D. R. Price, E.
Prochnow, K. Ruud, F. Schiffmann, W. Schwalbach, S. Stopkowicz, A.
(41) Kirmse, W. Carbene Chemistry, 2nd ed.; Academic Press: New
York, 1971.
(42) In this case, flash photolysis refers to shining high intensity (200
nm) light on the sample for 10 s, 30 s, 1 min, or 5 min and taking a
scan to see if carbene signal is produced. In the case of 3-
thienyldiazomethane, there was no conversion to carbene under
these conditions.
Tajti, J. Vaz
́
quez, F. Wang, J. D. Watts and the integral packages:
MOLECULE (J. Almlof and P. R. Taylor), PROPS (P. R. Taylor),
ABACUS (T. Helgaker, H. J. Aa. Jensen, P. Jørgensen, and J. Olsen),
̈
and ECP routines by A. V. Mitin, C. van Wullen; www.cfour.de.
̈
(75) Glendening, E. D.; Badenhoop, J. K.; Reed, A. E.; Carpenter, J.
E.; Bohmann, J. A.; Morales, C. M.; Weinhold, F. NBO 5.0;
Theoretical Chemistry Institute, University of Wisconsin: Madison,
Wisconsin, 2001.
(43) LeBlanc, B. F.; Sheridan, R. S. J. Am. Chem. Soc. 1985, 107,
4554−4556.
(44) LeBlanc, B. F.; Sheridan, R. S. J. Am. Chem. Soc. 1988, 110,
7250−7252.
(76) Neese, F.; Wennmohs, F.; Becker, U.; Ganyushin, D.; Hansen,
A.; Liakos, D. G.; Kollmar, C.; Kossmann, S.; Petrenko, T.; Reimann,
C.; Riplinger, C.; Sivalingam, K.; Valeev, E.; Wezisla, B. ORCA, 2.8 ed.
(45) Tomioka, H.; Suzuki, S.; Izawa, Y. J. Am. Chem. Soc. 1982, 104,
1047−1050.
(46) Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Am.
Chem. Soc. 2007, 129, 2597−2606.
Lehrstuhl fur Theoretische Chemie: Bonn, 2010; http://www.thch.
̈
uni-bonn.de/tc/orca/.
(47) Matzinger, S.; Bally, T. J. Phys. Chem. A 2000, 104, 3544−3552.
(48) Zuev, P. S.; Sheridan, R. S.; Albu, T. V.; Truhlar, D. G.; Hrovat,
D. A.; Borden, W. T. Science 2003, 299, 867−870.
(49) Schreiner, P. R.; Reisenauer, H. P.; Ley, D.; Gerbig, D.; Wu, C.-
H.; Allen, W. D. Science 2011, 332, 1300−1303.
(50) McMahon, R. J. Science 2003, 299, 833−834.
(51) McMahon, R. J.; Chapman, O. L. J. Am. Chem. Soc. 1987, 109,
683−692.
(77) Hanson, G. R.; Noble, C. J.; Benson, S. In High Resolution EPR:
Application to Metalloenzymes and Metals in Medicine; Hanson, G.,
Berliner, L., Eds.; Springer: Dordrecht, The Netherlands, 2009; Vol.
28, pp 105−174.
(78) Katritzky, A. R.; Tymoshenko, D.; Nikonov, G. N. J. Org. Chem.
2001, 66, 4045−4046.
(52) Bucher, G.; Sander, W. J. Org. Chem. 1992, 57, 1346−1351.
(53) In the article by McKee et al., the energy for singlet 2-
thienylcarbene is computed for the (s-Z) isomer, while the energy for
triplet 2-thienylcarbene is computed for the (s-E) isomer. We noted
this discrepancy when comparing our absolute energies to those
reported in Table 1 of McKee’s article, and we confirmed this
interpretation by inspecting the Cartesian coordinates provided in the
Supporting Information supplied with the article.
(54) Aycard, J.-P.; Allouche, A.; Cossu, M.; Hillebrand, M. J. Phys.
Chem. A 1999, 103, 9013−9021.
(55) Hutton, R. S.; Roth, H. D.; Schilling, M. L. M.; Suggs, J. W. J.
Am. Chem. Soc. 1981, 103, 5147−5151.
(56) Hutton, R. S.; Roth, H. D. J. Am. Chem. Soc. 1982, 104, 7395−
7399.
(57) Roth, H. D.; Hutton, R. S. Tetrahedron 1985, 41, 1567−1578.
(58) Pharr, C. R. Ph.D. Dissertation, University of Wisconsin, 2008.
(59) Pharr, C. R.; McMahon, R. J., manuscript in preparation.
(60) Wasserman, E.; Snyder, L. C.; Yager, W. A. J. Chem. Phys. 1964,
41, 1763−1772.
(61) Fontecave, M.; Ollagnier-de-Choudens, S.; Mulliez, E. Chem.
Rev. 2003, 103, 2149−2166.
(62) Bolman, P. S. H.; Safarik, I.; Stiles, D. A.; Tyerman, W. J. R.;
Strausz, O. P. Can. J. Chem. 1970, 48, 3872−3876.
(63) Sinnecker, S.; Neese, F. J. Phys. Chem. A 2006, 116, 12267−
12275.
(64) Neese, F. J. Chem. Phys. 2007, 127, 164112.
(65) Sander, W.; Grote, D.; Kossmann, S.; Neese, F. J. Am. Chem.
Soc. 2008, 130, 4396−4403.
(66) Grote, D.; Finke, C.; Kossmann, S.; Neese, F.; Sander, W.
Chem.Eur. J. 2010, 16, 4496−4506.
(67) McMahon, R. J.; Chapman, O. L.; Hayes, R. A.; Hess, T. C.;
Krimmer, H.-P. J. Am. Chem. Soc. 1985, 107, 7597−7606.
6454
dx.doi.org/10.1021/ja300927d | J. Am. Chem. Soc. 2012, 134, 6443−6454