3-Trimethylsilylcyclobutylidene. the γ-effect of silicon on carbenes

Article abstract of DOI:10.1021/ja400747u

3-Trimethylsilylcyclobutylidene was generated by pyrolysis of the sodium salt of the tosylhydrazone derivative of 3-trimethylsilylcyclobutanone. This carbene converts to 1-trimethylsilylbicyclobutane as the major product. A labeling study shows that this intramolecular rearrangement product comes from 1,3-hydrogen migration to the carbenic center and not 1,3-silyl migration. Computational studies show two carbene minimum energy conformations, with the lower energy conformation displaying a large stabilizing interaction of the carbene center with the rear lobe of the C3-Si bond. In this conformation, the trimethylsilyl group cannot migrate to the carbene center, and the most favorable process is 1,3-hydrogen migration. When the carbene is generated photochemically in methanol, it reacts by a protonation mechanism giving the highly stabilized 3-trimethylsilylcyclobutyl carbocation as an intermediate. When generated in dimethylamine as solvent, the carbene undergoes preferred attack of this nucleophilic solvent from the back of this C-Si rear lobe stabilized carbene.

Full text of DOI:10.1021/ja400747u

Article
pubs.acs.org/JACS
3‑Trimethylsilylcyclobutylidene. The γ‑Effect of Silicon on Carbenes
Xavier Creary*
Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, Indiana 46556, United States
S
* Supporting Information
ABSTRACT: 3-Trimethylsilylcyclobutylidene was generated by pyrolysis of the sodium salt of the tosylhydrazone derivative of
3-trimethylsilylcyclobutanone. This carbene converts to 1-trimethylsilylbicyclobutane as the major product. A labeling study
shows that this intramolecular rearrangement product comes from 1,3-hydrogen migration to the carbenic center and not 1,3-
silyl migration. Computational studies show two carbene minimum energy conformations, with the lower energy conformation
displaying a large stabilizing interaction of the carbene center with the rear lobe of the C3−Si bond. In this conformation, the
trimethylsilyl group cannot migrate to the carbene center, and the most favorable process is 1,3-hydrogen migration. When the
carbene is generated photochemically in methanol, it reacts by a protonation mechanism giving the highly stabilized 3-
trimethylsilylcyclobutyl carbocation as an intermediate. When generated in dimethylamine as solvent, the carbene undergoes
preferred attack of this nucleophilic solvent from the back of this C−Si rear lobe stabilized carbene.
Scheme 1. Carbocation Stabilization by the Trimethylsilyl
Group
INTRODUCTION
■
The effect of a β-silyl group on carbocation 1 has been extensively
studied.1 These cations are formed much more readily (up to
1011 times faster) than unsilylated analogues.2 β-Silyl cations 1
are stabilized by a very favorable interaction between the cationic
center and the adjacent carbon−silicon bond. Computational
studies,3 as well as an X-ray crystallographic study,4 are consistent
with this mode of stabilization, as opposed to a potential
stabilization mode involving a silicon bridged intermediate such
as 2. We have been interested in subtle aspects of the β-silyl
effect.5 Recently,6 we have also studied the effect of a silyl group
one carbon further removed from a developing carbocationic
center, that is, “the γ-silyl effect”. Shiner7 and Grob8 had earlier
examined this phenomenon and found modest rate enhance-
ments in solvolytic reactions when the silyl group was in the γ-
position relative to a developing cationic center. We have found a
remarkable rate enhancing effect in the substrate 5, where the
solvolysis rate of 5 is even greater than that of the β-silyl substrate
3.5 This effect was attributed to a γ-silicon stabilized carbocation
6 formed by assistance involving the rear lobe of the carbon−
silicon orbital in a “W” fashion (Scheme 1). Computational
studies5 also provided evidence for this type of γ-trimethylsilyl
stabilization.
hydrogen migration occurred to the carbene center.9a The
bicyclo [2.2.2] systems 14 and 16 were also used as a probe for γ-
trimethylsilyl interactions.9b The rearranged product 15
indicates that the γ-trimethylsilyl group interacts quite effectively
Our laboratory has also extensively studied the effect of both β-
and γ-silyl groups on a variety of carbenes.9 Early studies showed
that β-trimethylsilyl groups are prone to migrate to carbene
centers, as in carbenes 79a and 9.9c By way of contrast, a labeling
study of the β-silylcarbene 11 showed that both trimethylsilyl and
Received: January 28, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja400747u | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
with a carbenic center and that this migration is more effective
than that of an unactivated hydrogen. The product 17 derived
from the carbene 16 suggests that the trimethylsilyl group also
activates hydrogen toward migration to a carbenic center
(Scheme 2).
Scheme 3. Generation of Carbene 18 via the Bamford−
Stevens Reaction
Scheme 2. Rearrangements of Trimethylsilyl-Substituted
Carbenes
With these carbocation and carbene studies in mind, we
therefore wanted to examine the carbene 18. What role will the
TMS group play in stabilization of this carbene? What
rearrangement products will be formed from this carbene?
How does the chemistry of carbene 18 compare to that of the
desilylated analogue, cyclobutylidene, 19? Reported here are the
results of these studies.
Figure 1. Partial 13C NMR spectrum of the products of pyrolysis of the
sodium salt 22.
26. These products presumably arise by way of the intermediate
carbene 18.11
The formation of the minor products 25 and 26 is reminiscent
of the behavior of the parent cyclobutylidene, 19, where
methylenecyclopropane and cyclobutene are formed in an
85:15 ratio.11,12 Of more interest is the major bicyclobutane
product 24.13 To determine the origin of this product, that is,
which group (H or TMS) had migrated to the carbene center, a
labeling study was necessary. Generation of a labeled carbene
required a 3-trimethylsilylcyclobutanone with the label either in
the C1 or in the C3 position. Because 3-trimethylsilylcyclobu-
tanone can be prepared from trimethylsilylketene,14 labeled 3-
trimethylsilylcyclobutanone was needed. We chose to carry out
such a synthesis using 13C depleted CO as the starting material
(Scheme 4).15 Carbon monoxide (99.95% 12C) is commercially
available and is less expensive than 13C enriched CO. The
synthesis began with the knowledge the carbonyl groups of
Co2(CO)8 rapidly exchange with 14CO.16 Hence 13C depleted
12CO was bubbled through a solution of Co2(CO)8, and then
trimethylsilyldiazomethane was added. Subsequent purging of
RESULTS AND DISCUSSION
■
3-Trimethylsilylcyclobutanone, 20, served as the starting ma-
terial for generation of the carbene 18 (Scheme 3). The standard
Bamford−Stevens reaction10 was used to generate the
diazocompound 23 from the tosylhydrazone 21. Thus, reaction
of 20 with tosylhydrazine followed by deprotonation of the
tosylhydrazone 21 with sodium methoxide and pyrolysis of the
salt 22 leads to the in situ generated diazocompound 23. Loss of
molecular nitrogen from 23 under the thermal conditions leads
to the bicyclobutane 24 as the major product, as shown in Figure
1. Also produced is a smaller amount of the methylenecyclopro-
pane 25, as well as a trace amount of 3-trimethylsilylcyclobutene,
́
the mixture with 12CO using a reaction developed by Ungvary17
B
dx.doi.org/10.1021/ja400747u | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 4. Synthesis of Labeled 3-
Trimethylsilylcyclobutanone
Scheme 5. Generation and Rearrangement of Labeled
Carbene 18*
led to the formation of 13C depleted product, that is, labeled
trimethylsilylketene 27. Addition of diazomethane to 27 gave a
mixture of silylated cyclobutanones 28 and 29, from which pure
28 could be isolated. The 13C NMR spectrum of 28 was devoid of
the carbonyl signal at δ 208.3, which appears in the spectrum of
20.
Conversion of labeled 28 to the tosylhydrazone salt 30 was
straightforward. The vacuum pyrolysis products from 30 were
analyzed by 13C NMR spectroscopy (decoupled spectrum with
no NOE) to determine the position of the 12C label, and the
result is shown in Figure 2. The signal at δ = 2.8 ppm has
conformation 18a (Figure 3) is highly puckered with a very short
distance (1.709 Å) between the carbene center and C3. The C3−
Si bond is also somewhat elongated (1.901 Å) relative to the C−
Si bond in bicyclobutane 24 (1.853 Å), but not as long as in
cation 6 (2.000 Å). Also of interest are the ring bonds of 18a,
where the C1−C2 bond (1.460 Å) is significantly shorter than
the C2−C3 bond (1.592 Å). It is of interest to compare the
structure of 18a to that of the unsubstituted cyclobutenylidene,
19. A previous computational study on 19 has been carried out
using various levels of theory.19 These calculations show that 19
is a puckered species with a significant interaction between the
carbene center and C3. NBO analysis confirms a bonding
transannular interaction. At the M062X/6-311+G** computa-
tional level in this Article (Figure 3), 19 is more puckered and the
C1−C3 bond (1.704 Å) is even shorter than previously
calculated. The similar C1−C2 and C2−C3 bond lengths in
19 and 18a, as well as the similar C1−C3 distance, suggest similar
carbene stabilizing interactions occur in 19 and in 18a.
It is proposed that stabilization of carbene 18a involves an
interaction of the carbene vacant 2p orbital with the C2−C3 σ-
orbitals. This leads to a shortening of the C1−C2 bond (1.460 Å)
as well as a lengthening of the C2−C3 bonds (1.592 Å). An
additional interaction with the rear lobe of the C3−Si bond is
also suggested. This leads to a lengthening of the Si−C3 bond. A
representation of these interactions is given in Figure 4. The
shortening of C1−C2 and lengthening of C2−C3 in 19, relative
to 18a, reflect the loss of the additional stabilization from the rear
lobe of the C3−Si bond in 18a.
A more quantitative measure of the stabilization of 18a by the
trimethylsilyl group comes from the isodesmic calculations
shown in Scheme 6. A comparison with the simple carbene
cyclohexylidene, 32, suggests that net stabilization of 18a is quite
large. Relative to the parent cyclobutylidene 19, 18a is stabilized
by an additional 7.5 kcal/mol. While this additional trimethylsilyl
group stabilization is substantial, it is not as large as that
calculated for the carbocation 6, where additional TMS
stabilization of 6 relative to the cyclobutyl cation amounts to
22.6 kcal/mol.
A second energy minimum, 18b, has been located at various
levels of theory. At the M062X/6-311+G** level, this carbene is
also highly puckered, with a short C1−C3 bond distance (1.678
Å). As in 18a, the C1−C2 bond in 18b is quite short, while the
C2−C3 bond is quite long. It therefore appears that some of the
same stabilizing interactions occur in 18b as in 18a and 19, but
conformation 18b is somewhat higher energy than 18a. It is
suggested that the higher energy of 18b is due to the lack of the
Figure 2. Partial 13C NMR spectrum of the products of pyrolysis of the
sodium salt 30.
essentially disappeared, while the signal at δ = −8.4 ppm remains
strong. The signal at δ = 2.8 ppm is due to the carbon attached to
the bridgehead hydrogen, and hence this is the labeled (12C)
carbon atom. The product is therefore bicyclobutane 31; that is,
hydrogen has migrated to the carbenic center of 18* (Scheme 5).
Computational Studies. Calculations have been used to
gain insight into modes of stabilization of carbene 18, as well as
the reasons for the observed hydrogen migration to the carbene
center. Two conformational energy minima for 18 have been
located at the M062X/6-311+G** level.18 The lower energy
C
dx.doi.org/10.1021/ja400747u | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 3. M062X/6-311+G** calculated structures of carbenes 18a, 18b, and 19.
possible transition states, 37 and 38, for formation of the trace
amount of trimethylsilylcyclobutene 26. These latter three
transition states are all of significantly higher energy than that of
34, which leads to the major product. Hence, these computa-
tional studies are qualitatively consistent with the experimental
observation of bicyclobutane 24 as the major product from 18 via
1,3-hydrogen migration to the carbene center. The computa-
tional studies also suggest that 18a is a highly stabilized carbene
with relatively large barriers to intramolecular rearrangement. As
such, it should have a relatively long lifetime prior to
intramolecular rearrangement. Finally, it should be noted that
these M062X/6-311+G** transition state energies of 36−38
(Figure 7) are quite high relative to carbene 18a and may be
overestimated by this computational method. B3LYP/6-
311+G** relative energies are lower, but trends are analogous.21
Photolysis of Tosylhydrazone Salts in Methanol and
Dimethylamine. To gain further insights into the behavior of
carbene 18, the tosylhydrazone salt 22 was irradiated in methanol
solvent (Scheme 7). The major product of this photolysis was the
cis-methyl ether 39.22 Careful examination of the product
revealed that a trace (2%) of the bicyclobutane 24 was also
formed. These products undoubtedly arise from photoextrusion
of p-toluenesulfinate anion to generate diazocompound 23 as a
transient intermediate. There are a number of mechanisms by
which 23 could be converted to the product 39. These
mechanistic suggestions should address this stereospecificity of
the reaction.
Figure 4. Stabilizing interactions in carbenes 18a and 18b.
additional rear lobe stabilization that is present in conformation
18a.
The energy diagram shown in Figure 5 implies that
conformations 18a and 18b do not readily interconvert. The
transition state 33 (Figure 6) for interconversion of these
confomations lies 16.5 kcal/mol above 18a. In this flattened
structure, there is little or no interaction of the carbene vacant
orbital with the σ-bonds of the cyclobutane or with the rear lobe
of the Si−C σ-bond. This transition state gives an approximation
of the net stabilization of 18a by interaction of the carbene center
with the cyclobutane σ bonds and the rear lobe of the Si−C
orbital.
The transition state for hydrogen migration in 18a to give the
bicyclobutane 24 has also been located. This transition state 34
lies 10.4 kcal/mol above 18a. This is a rather large barrier for
hydrogen migration,20 but it is still below the transition state for
inversion of the carbene 18a to 18b. There is also a rather
significant barrier for trimethylsilyl migration in 18b, and the
transition state 35 for such a migration lies 7.0 kcal/mol above
18b. Finally, the transition state 36 for formation of the minor
methylenecyclopropane 25 has been located, along with the two
Photoinitiated loss of nitrogen from 23 could generate the
carbene 18, and the ether 39 would be the product of formal
insertion into the OH bond of methanol. The reaction of
Scheme 6. Isodesmic Reactions of Carbene 18a and Cation 6
D
dx.doi.org/10.1021/ja400747u | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 5. M062X/6-311+G** calculated energy diagram for conversion of carbenes 18a and 18b to trimethylsilylbicyclobutane 24.
Figure 6. M062X/6-311+G** calculated structures of transition states 33, 34, and 35.
Figure 7. M062X/6-311+G** calculated structures of transition states 36, 37, and 38.
carbenes with alcohols has been previously reviewed.23 This
formal insertion can occur by two stepwise processes or,
theoretically, by a concerted process. Alternatively, studies have
shown that diazocompounds can be protonated by alcohol
solvent to give diazonium ions (path A), which can serve as
precursors to ether products by way of carbocation inter-
mediates.23 This mechanism bypasses carbene intermediates.
These potential mechanistic pathways are summarized for
diazocompound 23 in Scheme 8.
To shed light on these mechanistic possibilities, the photolysis
of 22 was carried out in an equimolar CH3OH/CH3OD mixture,
and the extent of deuterium incorporation in the product was
measured. The ratio of 39 to the mono deuterated analogue was
3.1 0.1. This substantial isotope effect argues in favor of either
protonation of diazocompound 23 (path A) or protonation of
carbene 18 (path B). A precedent is the directly measured
isotope effect for reaction of dimethoxycarbene, MeO−C−OMe,
with CH3OH and CH3OD, where the isotope effect is 3.3.24 Our
isotope effect of 3.1 also argues against path D as the source of 39
E
dx.doi.org/10.1021/ja400747u | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 7. Photochemical Reaction of Tosylhydrazone Salt 22
in Methanol
Scheme 9. Solvolytic Reaction of trans-Mesylate 44 in
CD3CO2D
Scheme 8. Potential Mechanisms for Formation of Ether 39
from Diazocompound 23
orbital with the C−Si bond as well as the C2−C3 sigma bonds.
Hence, an interaction of 18 with methanol results in initial
proton transfer to the filled carbene orbital. Subsequent ion-pair
capture of methoxide from the rear of delocalized cation 6 gives
the observed stereochemistry as shown in Scheme 8.
To carry out the reaction under conditions where protonation
of diazocompound 23 was improbable, dimethylamine was used
as solvent. Because of the low solubility of the sodium salt 22 in
this solvent, the more soluble lithium salt 49 was irradiated in
neat dimethylamine. The major product of this reaction was the
amine 50, which was also formed in a stereospecific fashion. Also
formed was about 20% of a formal carbene dimer 51. It is highly
unlikely that the insertion product 50 arises from protonation of
the diazocompound 23 by the dimethylamine. Stepwise
protonation of carbene 18 by dimethylamine to give an ion
pair is also unlikely.
The product 50 speaks to the stereochemistry of reaction of
carbenes with nucleophiles.27 The stereospecificity of the
reaction suggests that dimethylamine approaches carbene 18
from the back of a delocalized carbene, as shown in Scheme 10,
and further supports the proposed delocalized bonding in the
carbene 18. The overall reaction could be a concerted insertion
process proceeding through transition state 52. A stepwise
process analogous to path D of Scheme 8, involving an
intermediate 53 analogous to Platz’s pyridinium ylids,28 cannot
be eliminated. In other words, the nucleophilic carbene 18 in
methanol has become an electrophilic carbene in dimethylamine;
that is, 18 is ambiphilic. To support the suggestion that carbene
18 can become electrophilic, the salt 49 was irradiated in a
pyrrolidine/diethylamine mixture. This led to a mixture of amine
products 54 and 55 where carbene 18 reacted preferentially with
the more nucleophilic pyrrolidine29 over diethylamine by a factor
of 3.0.
because this mechanism, which involves no OH/OD bond
breaking in the product determining step, predicts essentially no
isotope effect. Along these lines, the OH insertion reaction of
phenylcarbene, Ph−C−H, with CH3OH/CH3OD, which
probably proceeds by path D, gives a very small isotope effect.25
The nonlinear transition state involved in path C should also
result in a relatively small isotope effect.26
The stereochemistry of ether 39 is consistent with the
intermediacy of carbocation 6. Our previous studies5 under
solvolytic conditions have shown that the 3-trimethylsilylcyclo-
butyl cation 6 captures nucleophile from the rear of this
delocalized cation to give exclusive cis-product. The ion-pair 6 is
therefore an attractive possibility in the photolysis of 22. A choice
between path A (diazonium ion) and path B (carbene
protonation) is more problematic. Here, we rely on some of
our previous data to argue in favor of mechanism B. If the
diazocompound 23 were to undergo protonation, there is no
apparent reason why only diazonium ion 40 would be formed.
Some of the isomeric diazonium ion 43 should also be formed.
We have previously studied solvolytic ionization of the analogous
mesylate 44 (Scheme 9), and this leads to exclusively rearranged
products 45−47 in the cyclopropylcarbinyl-homoallylic carbo-
cation manifold.6 If diazonium ion 43 were formed, then loss of
nitrogen should result in analogous rearranged products. Because
no rearranged methyl ethers are formed when 22 is irradiated in
methanol, diazonium ion 43 is not involved. Hence, it is unlikely
that the isomeric diazonium ion 40 is involved.
CONCLUSIONS
■
The carbene 18 rearranges to give the bicyclobutane 24 as well as
smaller amounts of the methylenecyclopropane 25 and cyclo-
butene 26. A labeling study shows that the bicyclobutane product
is exclusively derived from trans-annular 1,3-hydrogen migration
to the carbene center. Computationally, there are two isomeric
carbenes, 18a and 18b, with the more stable form 18a lying in an
It is now suggested that carbene 18 has some nucleophilic
character due to the rear lobe interaction of the carbene vacant
F
dx.doi.org/10.1021/ja400747u | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
flask containing the salt was placed in an oil bath, and the temperature of
the oil bath was gradually raised to 80 °C. The receiver flask was then
cooled in a dry ice−acetone bath as the temperature of the oil was then
raised gradually to 160 °C. At about 140 °C, decomposition occurred as
evidenced by a pressure increase. The products of the pyrolysis collected
in the cold receiver flask. The yield of distillate was 116 mg (51% yield).
The products 24,30 25,31 and 2632 were identified by NMR spectral
comparisons with authentic samples.
Scheme 10. Photochemical Reaction of Tosylhydrazone Salt
49 in Dimethylamine
Preparation of Labeled 3-Trimethylsilylcyclobutanone 28. A
solution of 220 mg of Co2(CO)8 in 75 mL of pentane in a three-neck
flask was cooled in a water bath to about 10 °C under an argon
atmosphere. 13C depleted CO (99.95% 12C; Cambridge Isotope
Laboratories) from a lecture bottle was slowly bubbled into the stirred
mixture over a 2 h period. The volume of 12CO used was about 2 L. At
the end of this period, 1.0 g of distilled TMSCHN2 (Caution: safety
hazard)33 was added to the solution. 12CO was again periodically
bubbled through the mixture over a 3 h period at 10 °C. A total of about
3.5 g of 12CO was used for the entire procedure.
A Vigreux column was attached to the reaction flask, along with a
receiver flask that was cooled in a dry ice/acetone bath. The pressure was
gradually lowered to 320 mm, and the volatile material was condensed in
the receiver flask. The pot was heated using an oil bath to 40−50 °C as
the pentane and trimethylsilylketene distilled. The pressure was
gradually reduced to 15 mm, and the reaction flask was distilled to
dryness as the receiver flask was thoroughly cooled in the dry ice bath.
A solution of diazomethane in ether was prepared from Diazald (6.0
g) in 60 mL of ether by dropwise addition to 3.0 g of KOH in 4.8 mL of
water and 17 mL of carbitol and 10 mL of ether. The diazomethane/
ether solution was cooled in a dry ice/acetone bath, and the solution of
trimethylsilylketene prepared above in pentane was added in a single
portion to the cold diazomethane solution. The solution was allowed to
gradually warm to room temperature. After about 30 min at room
temperature, most of the solvents were removed using a rotary
evaporator. The residue was transferred to a 10 mL flask fitted with a
short path distillation head. The products were distilled at 15 mm
pressure. The yield of the distilled mixture of 28 and 29 was 756 mg
(42% yield).
The mixture of ketones 28 and 29 was placed in a flask with about 0.5
mL of ether. Water (20 mL) was then added followed by 41 mg of
K2CO3. The mixture was then stirred vigorously for 2.5 h and then
transferred to a separatory funnel using 25 mL of pentane. The pentane
phase was separated, dried over Na2SO4, decanted from the drying
agent, and the pentane was removed using a rotary evaporator. The
residue was distilled to give 335 mg of pure 28, bp 70 °C (15 mm). The
1H NMR spectrum of 28 was identical to that of 20. The 13C NMR
energy well such that there is a relatively large barrier to
interconversion of these isomeric carbenes. The carbene 18a is
highly stabilized by the σ-bent bonds of the cyclobutane as well as
a significant rear lobe interaction with the trimethylsilyl group.
There are also relatively large barriers to intramolecular
rearrangement. Carbene 18 is effectively trapped by methanol
to give cis-3-trimethylsilyl-1-methoxycyclobutane, via the highly
stabilized 3-trimethylsilylcyclobutyl cation. Dimethylamine also
effectively traps carbene 18 to give cis-trimethylsilyl-1-
dimethylaminocyclobutane. The stereochemistry of this product
suggests preferential approach of dimethylamine from the back
of the rear lobe stabilized carbene 18.
EXPERIMENTAL SECTION
■
General. NMR spectra were recorded on a Varian DirectDrive 600
MHz spectrometer. HRMS measurements were carried out using a
Bruker MicroTOF-II spectrometer (electrospray ionization source with
time-of-flight mass analyzer).
spectrum (Supporting Information) showed no carbonyl signal at 208.3
ppm.
Preparation of Tosylhydrazone 21. Tosylhydrazine (340 mg;
1.826 mmol) was placed in a flask and stirred as 255 mg (1.792 mmol) of
3-trimethylsilylcyclobutanone in about 2 mL of methanol was rapidly
added. The mixture was stirred, and the tosylhydrazine dissolved in a few
minutes. The tosylhydrazone 21 crystallized after about 40 min. After 9
h at room temperature, most of the methanol was removed using a
rotary evaporator, and the solid residue was slurried with about 3 mL of
cold 50% ether in pentane. The solvent was then decanted, and the
residue was dried under aspirator vacuum to give 540 mg (97% yield) of
21, mp 121−122 °C (dec). 1H NMR of 21 (CDCl3): δ 7.82 (d, J = 8.3
Hz, 2H), 7.44 (br s, 1H), 7.31 (d, J = 8.3 Hz, 2H), 2.99 (m, 1H), 2.86 (m,
1H), 2.65 (m, 1H), 2.50 (m, 1H), 2.42 (s, 3H), 1.56 (t of t, J = 10.8, 7.4
Hz, 1H), −0.08 (s, 9H). 13C NMR of 21 (CDCl3): δ 1.60.4, 144.1, 135.4,
129.7, 127.9, 35.0, 32.6, 21.6, 13.2, −4.0. HRMS (ESI) (MH+)
calculated for C14H23N2O2SSi 311.1244, found 311.1271.
Preparation and Pyrolysis of Tosylhydrazone Salt 22.
Tosylhydrazone 21 (557 mg, 1.797 mmol) was placed in a 25 mL
flask, and 3.5 mL of 0.579 M NaOCH3 in methanol (2.026 mmol) was
added. After 20 min, the methanol solvent was removed using a rotary
evaporator, and the pressure was maintained at 15 mm for 4 h. The last
traces of methanol were removed using a vacuum pump at 0.2 mm for 3
h. The solid tosylhydrazone salt 22 was broken with a spatula, and the
flask containing the dry salt was then fitted with a short path distillation
head and a receiver flask. The pressure was maintained at 0.2 mm as the
Preparation and Pyrolysis of Labeled Tosylhydrazone Salt
30. The procedure for preparation of the tosylhydrazone salt 30 from
ketone 28 was identical to that described above to prepare the unlabeled
salt 22. The pyrolysis procedure was also identical. The 13C NMR
spectrum of the pyrolysis product is shown in Figure 2. The spectrum
was recorded without a NOE by turning the decoupler off except during
signal acquisition (dm = “nny” using the Varian simple 2-pulse
sequence). A relaxation delay (d1) of 90 s between pulses was also
included to ensure complete relaxation of all 13C signals.
Photolysis of Tosylhydrazone Salt 22 in Methanol. Tosylhy-
drazone 21 (42.3 mg; 0.136 mmol) was placed in a vial, and 250 μL of
0.579 M NaOCH3 (0.145 mmol) in CH3OH was added via syringe. The
mixture was stirred with a micro stir bar to dissolve the tosylhydrazone
21, and an additional 1.75 mL of methanol was added. The methanol
solution was placed in a 5 mm NMR tube, and the air-cooled tube was
irradiated with a Hanovia 450 W source for 15 min. The solution was
then transferred to a vial, and 5 mL of water was added followed by 1.6
mL of C6D6. The mixture was stirred to extract the product into the
C6D6. The aqueous phase was separated, and the C6D6 phase was then
washed with two additional portions of water and then dried over
Na2SO4. NMR spectra of the products were determined in the C6D6
extract. The 1-trimethylsilylbicyclobutane 24 (2%) was identified by
spectral comparison with an authentic sample. 1H NMR of 39 (C6D6): δ
G
dx.doi.org/10.1021/ja400747u | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
3.73 (t of t, J = 8.0, 6.9 Hz, 1H), 3.04 (s, 3H), 2.90 (m, 2H), 1.77 (m,
2H), 0.90 (t of t, J = 11.7, 8.1 Hz, 1H), −0.06 (s, 9 H). 13C NMR of 39
(C6D6): δ 74.8, 54.5, 31.7, 12.4, −3.3. HRMS (EI) (M+) calculated for
C8H18OSi 158.1121, found 158.1149.
phase was discarded. The C6D6 phase was then extracted with three
additional portions of water. The C6D6 phase was then extracted with
1.2 mL of 0.5 M HCl, and the aqueous extract was then separated. Solid
K2CO3 was then carefully added to the aqueous extract until the mixture
became basic. The mixture was then re-extracted with C6D6, and the
C6D6 phase was dried over Na2SO4 and analyzed by 1H NMR. The ratio
of the two products 54 and 55 was determined by 1H NMR from the
relative areas of the multiplets at 3.07 ppm (55) and 3.91 ppm (54). The
ratio of 54:55 was 3.0. Details are given as Supporting Information.
Computational Studies. Ab initio molecular orbital calculations
were performed using the Gaussian 09 series of programs.18 Structures
were characterized as energy minima via frequency calculations that
showed no negative frequencies or as transition states that showed one
negative frequency.
Photolysis of Tosylhydrazone Salt 22 in CH3OH/CH3OD. A
mixture of 4.003 g of CH3OH and 4.132 g of 99.5% CH3OD was
prepared. The dry tosylhydrazone salt 22 was prepared as described
above from 42.5 mg of 21 and 250 μL of 0.574 M NaOCH3 in methanol.
The salt 22 (20 mg) was placed in a vial, and 1.1 mL of the CH3OH/
CH3OD mixture was added. The mixture was transferred to an NMR
tube under argon, and the air-cooled tube was irradiated with a Hanovia
450 W source for 22 min. The contents of the tube were poured into a
vial containing 3 mL of water, and 1 mL of C6D6 was then added. The
mixture was stirred, and the aqueous phase was separated. The C6D6
extract was washed with an additional three portions of water and then
dried over Na2SO4. The ratio of the two ether products 39 and 39-d1 was
determined by 1H NMR from the relative areas of Ha (3.73 ppm) and Hb
(1.77 ppm). The ratio of 39:39-d1 was 3.1 0.1 in duplicate runs.
Details are given as Supporting Information.
ASSOCIATED CONTENT
■
S
* Supporting Information
Complete ref 18, the M062X/6-311+G* calculated structures,
energies, and Cartesian coordinates of 6, 18a, 18b, 19, 32, 32, 33,
34, 35, 36, 37, and 38, 1H and 13C NMR spectra of 21, 28, 39, 50,
54, and 55 as well as the 1H and 13C NMR spectra of the products
of vacuum pyrolysis of 22, the products of photolysis of 22 in
CH3OH/CH3OD, and the products of photolysis of 48 in
pyrrolidine/diethylamine. This material is available free of charge
via the Internet at http://pubs.acs.org.
Photolysis of Tosylhydrazone Salt 49 in Dimethylamine.
Tosylhydrazone 21 (49.4 mg; 0.159 mmol) was placed in a small vial,
and 324 μL of 0.516 M LiOCH3 in methanol (0.167 mmol) was added.
After the mixture was stirred for a few minutes at room temperature, the
methanol solvent was removed using a rotary evaporator. Evacuation at
15 mm pressure was continued for 10 h, and, during this time, the salt 49
solidified. The dry salt 49 was crushed with a spatula, and 7.0 mg was
placed in an NMR tube under argon. The tube was cooled in an ice/
acetone bath, and 1.25 mL of gaseous dimethylamine was condensed
into the cold tube. The tube was sealed under argon and shaken to
dissolve the salt 49. The air-cooled tube was then irradiated with a
Hanovia 450 W source for 5 min. The tube was cooled in ice, opened,
and the contents were poured into 4 mL of ice water. The mixture was
extracted with 1.2 mL of C6D6, and the aqueous phase was discarded.
The C6D6 phase was then washed with two additional portions of water
and then extracted with 2 mL of 1% aqueous HCl. The acidic extract was
separated, neutralized with solid Na2CO3, and re-extracted with C6D6.
AUTHOR INFORMATION
■
Corresponding Author
creary.1@nd.edu
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) For reviews and early studies, see: (a) Colvin, E. W. Silicon in
Organic Synthesis; Butterworths: London, 1981. (b) White, J. M. Aust. J.
Chem. 1995, 48, 1227. (c) Ushakav, S. N.; Itenberg, A. M. Zh. Obshch.
Khim. 1937, 7, 2495. (d) Sommer, L. H.; Dorfman, E.; Goldberg, G. M.;
Whitmore, F. C. J. Am. Chem. Soc. 1946, 68, 488. (e) Sommer, L. H.;
Bailey, D. L.; Whitmore, F. C. J. Am. Chem. Soc. 1948, 70, 2869.
(f) Sommer, L. H.; Baughman, G. A. J. Am. Chem. Soc. 1961, 83, 3346.
(g) Davis, D. D.; Jacocks, H. M., III. J. Organomet. Chem. 1981, 206, 33.
(2) (a) Lambert, J. B.; Wang, G.-t.; Finzel, R. B.; Teramura, D. H. J. Am.
Chem. Soc. 1987, 109, 7838. (b) Lambert, J. B.; Chelius, E. C. J. Am.
Chem. Soc. 1990, 112, 8120. (c) Lambert, J. B. Tetrahedron 1990, 46,
2677. (d) Lambert, J. B.; Liu, X. J. Organomet. Chem. 1996, 521, 203.
(e) Lambert, J. B.; Zhao, Y.; Emblidge, R. W.; Salvador, L. A.; Liu, X.; So,
J. H.; Chelius, E. C. Acc. Chem. Res. 1999, 32, 183.
1
The C6D6 phase was dried over Na2SO4 and analyzed by NMR. H
NMR of 50 (C6D6): δ 2.60 (t of t, J = 8.6, 7.0 Hz, 1H), 2.00 (s, 6H), 1.93
(m, 2H), 1.72 (m, 2H), 1.15 (t of t, J = 11.5, 8.2 Hz, 1H), −0.02 (s, 9H).
13C NMR of 50 (C6D6): δ 62.8, 41.6, 29.3, 14.2, −3.3. HRMS (ESI)
(MH+) calculated for C9H22NSi 172.1516, found 172.1505.
Preparation of Amines 54 and 55. Authentic samples of amines
54 and 55 were prepared in the same fashion as amine 50 by photolysis
of tosylhydrazone salt 49 in pyrrolidine and diethylamine as solvent,
respectively, as previously described.
(3) (a) Wierschke, S. G.; Chandrasekhar, J.; Jorgensen, W. L. J. Am.
Chem. Soc. 1985, 107, 1496. (b) Ibrahim, M. R.; Jorgensen, W. L. J. Am.
Chem. Soc. 1989, 111, 819.
(4) Klaer, A.; Saak, W.; Haase, D.; Muller, T. J. Am. Chem. Soc. 2008,
130, 14956.
1H NMR of 54 (C6D6): δ (2.91, t of t, J = 8.6, 7.0 Hz, 1H), 2.34 (m,
4H), 1.98 (m, 2H), 1.85 (m, 2H), 1.61 (m, 4H), 1.27 (t of t, J = 11.5, 8.4
Hz, 1H), 0.00 (s, 9H). 13C NMR of 54 (C6D6): δ 60.4, 50.7, 29.4, 24.0,
15.9, −3.2. HRMS (ESI) (MH+) calculated for C11H24NSi 198.1673,
found 198.1695.
̈
(5) (a) Creary, X.; O’Donnell, B. D.; Vervaeke, M. J. Org. Chem. 2007,
72, 3360. (b) Creary, X.; Kochly, E. D. J. Org. Chem. 2009, 74, 2134.
(6) Creary, X.; Kochly, E. D. J. Org. Chem. 2009, 74, 9044.
(7) (a) Shiner, V. J., Jr.; Ensinger, M. W. J. Am. Chem. Soc. 1986, 108,
842. (b) Davidson, E. R.; Shiner, V. J., Jr. J. Am. Chem. Soc. 1986, 108,
3135. (c) Shiner, V. J., Jr.; Ensinger, M. W.; Rutkowske, R. D. J. Am.
Chem. Soc. 1987, 109, 804. (d) Shiner, V. J., Jr.; Ensinger, M. W.;
Huffman, J. C. J. Am. Chem. Soc. 1989, 111, 7199.
(8) Grob, C. A.; Sawlewicz, P. Tetrahedron Lett. 1987, 28, 951.
(9) (a) Creary, X.; Wang, Y.-X. Tetrahedron Lett. 1989, 19, 2493.
(b) Creary, X.; Wang, Y.-X. J. Org. Chem. 1994, 59, 1604. (c) Creary, X.;
Butchko, M. A. J. Org. Chem. 2001, 66, 1115. (d) Creary, X.; Butchko, M.
A. J. Am. Chem. Soc. 2001, 123, 1569. (e) Creary, X.; Butchko, M. A. J.
Org. Chem. 2002, 67, 112.
1H NMR of 55 (C6D6): δ 3.07 (t of t, J = 9.0, 6.9 Hz, 1H), 2.43 (q, J =
7.1 Hz, 4H), 1.96 (m, 2H), 1.74 (m, 2H), 1.14 (t of t, J = 11.6, 8.1 Hz,
1H), 0.93 (t, J = 7.1 Hz, 6H), −0.01 (s, 9H). 13C NMR of 55 (C6D6): δ
58.8, 42.7, 30.3, 15.3, 11.7, −3.1. HRMS (ESI) (MH+) calculated for
C11H26NSi 200.1829, found 200.1855.
Photolysis of Tosylhydrazone Salt 48 in Pyrrolidine/Diethyl-
amine. A mixture of pyrrolidine (930 mg) and Et2NH (953 mg) was
prepared, and 6.7 mg of the dry salt 49 was dissolved in 954 mg of this
mixture. The solution was placed in a 5 mm NMR tube under argon, and
the air-cooled tube was irradiated with a Hanovia 450 W source for 10
min. The tube was opened, and the contents were poured into 4 mL of
water. The mixture was extracted with 1.2 mL of C6D6, and the aqueous
H
dx.doi.org/10.1021/ja400747u | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(10) For a discussion of the tosylhydrazone salt pyrolysis method of
diazocompound and carbene generation, see: (a) Bamford, W. R.;
Stevens, T. S. J. Chem. Soc. 1952, 4735. (b) Kaufman, G. M.; Smith, J. A.;
Vander Stouw, G. G.; Shechter, H. J. Am. Chem. Soc. 1965, 87, 935.
(c) Creary, X. Org. Synth. 1986, 64, 207.
(11) Photolyses of diazocompounds can lead to excited-state
diazocompounds, followed by concerted loss of nitrogen and rearrange-
ment to give “carbene-like” products; see: Pezacki, J. P.; Pole, D. L.;
Warkentin, J.; Chen, T.; Ford, F.; Toscano, J. P.; Fell, J.; Platz, M. S. J.
Am. Chem. Soc. 1997, 119, 3191 and references therein. While such a
concerted loss N2 from 23 and rearrangement (bypassing carbene 18)
cannot be ruled out, it is unlikely in our thermal rearrangement.
(12) (a) Friedman, L.; Shechter, H. J. Am. Chem. Soc. 1960, 82, 1002.
(b) Smith, J. A.; Shechter, H.; Bayless, J.; Friedman, L. J. Am. Chem. Soc.
1965, 87, 659. (c) Sulzbach, H. M.; Platz, M. S.; Schaefer, H. S., III;
Hadad, C. M. J. Am. Chem. Soc. 1997, 119, 5682.
(26) A non-linear proton transfer isotope effect has lower KIEs, see:
Anslyn, E. V.; Dougherty, D. A. Modern Physical Organic Chemistry;
University Science Books: Sausalito, CA, 2006; p 428.
(27) A reviewer has suggested that steric effects might account for the
preferential additional of dimethylamine and methanol to carbene 18a.
This is highly unlikely because ketone 20 adds nucleophiles from both
directions. Additionally, transition state steric effects for these
completely stereospecific additions to 18a would have to be much
greater that steric effects in the potential products.
(28) Jackson, J. E.; Platz, M. S. In Advances in Carbene Chemistry;
Brinker, U. H., Ed.; JAI Press Inc.: Greenwich, CT, 1994; Vol. 1, pp 89−
160.
(29) Kanzian, T.; Nigst, T. A.; Maier, A.; Pichl, S.; Mayr, H. Eur. J. Org.
Chem. 2009, 6379.
(30) (a) Baird, M. S.; Buxton, S. R.; Mitra, M. M. Acta Chem. Scand.
1992, 46, 669. (b) Cohen, H. M. J. Organomet. Chem. 1967, 9, 375.
(31) Sternberg, E.; Binger, P. Tetrahedron Lett. 1985, 26, 301.
(32) Tian, Z.; Fattahi, A.; Lis, L.; Kass, S. R. J. Am. Chem. Soc. 2006, 128,
17087.
(33) Although trimethylsilyldiazomethane is claimed to be non-
explosive and non-mutagenic, it is quite toxic. It should be handled as if it
were a highly toxic and potentially explosive substance. See: Org. Synth.
Coll. 1993, 8, 612. See note added in September 2009.
(13) For a recent example of solvolytic generation of a bicyclobutane
from the 1-trifluoromethyl-3-trimethylsilyl cation, see: Kelly, C. B.;
Colthart, A. M.; Constant, B. D.; Corning, S. R.; Dubois, L. N. E.;
Genovese, J. T.; Radziewicz, J. L.; Sletten, E. M.; Whitaker, K. R.; Tilley,
L. J. Org. Lett. 2011, 13, 1646.
(14) Zaitseva, G. S.; Bogdanova, G. S.; Baukov, Yu. I.; Lutsenko, D. F. J.
Organomet. Chem. 1976, 121, C21.
(15) This 13C depleted labeling technique is less commonly used in
mechanistic studies. For previous examples, see: (a) Brinker, U. H.;
Fleischhauer, I. Tetrahedron 1981, 37, 4495. (b) Dehmlow, E. V.;
Balschukat, D.; Schmidt, P. P.; Krause, R. J. Chem. Soc., Chem. Commun.
1986, 1435.
(16) Basolo, F.; Wojcicki, A. J. Am. Chem. Soc. 1961, 83, 520.
(17) Neszta Ungvar
7.
́ ́
i, T.; Kegl, F. U. J. Mol. Catal. A: Chem. 2004, 219,
(18) Frisch, M. J.; et al. Gaussian 09, revision A.02; Gaussian, Inc.:
Wallingford, CT, 2009.
(19) Stracener, L. L.; Halter, R. J.; McMahon, R. J.; Castro, C.; Karney,
W. L. J. Org. Chem. 2000, 65, 199.
(20) By way of contrast, the M062X/6-311+G** calculated barrier for
1,3-hydrogen migration in the 2-norbornyl carbene is only 1.6 kcal/mol.
The calculated barrier for 1,2-hydrogen migration in cyclohexylidene is
2.7 kcal/mol.
(21) A reviewer has suggested that steric effects could account for the
energy difference between 18a and 18b as well as the observed hydrogen
migration because a sterically locked 18a cannot migrate the TMS
group. We have now carried out computational studies on the t-butyl
analogues of 18a and 18b (ii and iii). These computational studies
indicate that steric effects are not very important because iii is actually of
lower energy than ii. The lower energy of 18a is therefore an electronic
effect and not a steric effect.
(22) The stereochemistry of 39 is based on coupling information. Jab
0 Hz in the cis-isomer 39, while trans-isomer i shows Jab = 1.3 Hz.
=
(23) (a) Kirmse, W. In Advances in Advances in Carbene Chemistry;
Brinker, U. H., Ed.; JAI Press Inc.: Greenwich, CT, 1994; Vol. 1, pp 1−
57. See also: (b) Kirmse, W.; Meinert, T.; Modarelli, D. A.; Platz, M. S. J.
Am. Chem. Soc. 1993, 115, 8918. (c) Kirmse, W.; Meinert, T. J. Chem.
Soc., Chem. Commun. 1994, 1065.
(24) Moss, R. A.; Shen, S.; Wlostowski, M. Tetrahedron Lett. 1988, 29,
6417.
(25) Tomioka, H.; Ozaki, Y.; Izawa, Y. Bull. Chem. Soc. Jpn. 1983, 56,
1239.
I
dx.doi.org/10.1021/ja400747u | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX