Activation energies for the 1,2-carbon migration of ring-fused cyclopropylchlorocarbenes

Article abstract of DOI:10.1016/j.tetlet.2005.03.209

Fused-ring cyclopropylchlorocarbenes 2 and 3 ring expand to fused-ring chlorocyclobutenes 10 and 11, respectively, with activation energies of 3-4 kcal/mol and activation entropies of ～-20 e.u. Carbon tunneling appears to be unimportant at low temperatures.

Full text of DOI:10.1016/j.tetlet.2005.03.209

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 4137–4141
Activation energies for the 1,2-carbon migration of ring-fused
cyclopropylchlorocarbenes
Gaosheng Chu,a Robert A. Moss,a,* Ronald R. Sauers,a,*
Robert S. Sheridanb,* and Peter S. Zuevb
aDepartment of Chemistry and Chemical Biology, Rutgers, The State University of New Jersey, New Brunswick, NJ 08903, USA
bDepartment of Chemistry, University of Nevada, Reno, NV 89557, USA
Accepted 22 February 2005
Abstract—Fused-ring cyclopropylchlorocarbenes 2 and 3 ring expand to fused-ring chlorocyclobutenes 10 and 11, respectively, with
activation energies of 3–4 kcal/mol and activation entropies of ꢀꢁ20e.u. Carbon tunneling appears to be unimportant at low
temperatures.
ꢀ 2005 Elsevier Ltd. All rights reserved.
The rearrangements of cyclopropylcarbenes are compli-
cated and have been extensively investigated.1 Some
simplification is possible upon substitution of a chlorine
atom at the divalent carbon center, which stabilizes the
cyclopropylcarbene by electron donation into the car-
beneÕs vacant p orbital.2 In particular, the 1,2-carbon
migration of cyclopropylchlorocarbene (1) to chloro-
cyclobutene (Eq. 1) has attracted significant attention.
The rate constant for the rearrangement was determined
by laser flash photolysis.
for concurrent carbene dimerization.3b However, a sig-
nificantly lower value of 3.0 0.4 kcal/mol was obtained
by four different methods: direct observation of 1; deter-
mination of k1,2 by ylide methodology; by analysis of the
kinetics of competitive carbene addition to an alkene;
and from product analysis of temperature dependent,
competitive rearrangement and alkene additions of 1.6
Dimerization was found not to be a complicating factor
under the reaction conditions.6
Early computational studies gave DHà ꢀ 8 kcal/mol (in
vacuo) for the ring expansion of 1,4,6 similar to the high-
er experimental value of Ea (7.4 kcal/mol).3b More re-
cently, an exhaustive state-of-the-art theoretical study
of reaction (1) reinforced the divergence between the
computed and ÔlowerÕ experimental Ea; the computed
EaÕs were 10.7 kcal/mol in vacuo or 8.5 kcal/mol in iso-
octane.7 An even larger discrepancy was found for the
experimentally determined8 and computed Ea of the
analogous ring expansion of cyclopropylfluorocarbene
to fluorocyclobutene (4.2 vs 11.9 kcal/mol, respec-
tively).9 Finally, both experiment10 and theory7,10 agree
that carbon atom tunneling is not important in the ring
expansion of carbene 1 at ambient temperature.
Cl
..
C
Cl
ð1Þ
1
(LFP) employing a diazirine precursor for carbene 1,
and following the kinetics either by direct observation
of the carbene at 250nm 3,4 or by ÔindirectÕ monitoring
of the competitive growth of its N-ylide formed by reac-
tion with added pyridine.3–5 Values of k1,2 ranging from
3.8 · 105 to 9 · 105 sꢁ1 (at 20–23 ꢁC in isooctane) were
observed, depending on method and source.3,4 The acti-
vation energy of the rearrangement, however, remains
somewhat uncertain. A value of 7.4 0.4 kcal/mol was
determined by direct observation of 1 after correction
In view of the continuing divergence between our experi-
mental values of Ea for the ring expansion of 1 and the
corresponding computational values, we examined the
analogous ring expansion of two novel ring-fused cyclo-
propylchlorocarbenes, 2 and 3. The results are commu-
nicated here.
Keywords: Carbenes; Kinetics.
*
Corresponding authors. Tel.: +1 732 445 2206; fax: +1 732 445
5312; e-mail: moss@rutchem.rutgers.edu
0040-4039/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.03.209

4138
G. Chu et al. / Tetrahedron Letters 46 (2005) 4137–4141
The kinetics of the 2 ! 10 and 3 ! 11 rearrangements
H
H
were examined by LFP using the pyridine ylide method.5
LFP20 at 351 nm and 25 ꢁC of diazirine 8 (A350 ꢀ 1 in
DCE) in the presence of pyridine produced an ylide
absorbance at (kmax) 385 nm. A correlation of the
apparent rate constants for ylide formation, kobs
(2.8 · 106–6.4 · 107 sꢁ1) versus pyridine concentration
(0.02–2.47 M), was linear (10 points, r = 0.999) with a
slope of 2.6 · 107 Mꢁ1 sꢁ1, equivalent to the rate con-
stant for ylide formation, ky, and a Y intercept of
5.71 · 105 sꢁ1; cf. Figure 1. The intercept represents
the sum of rate constants for processes that destroy car-
bene 2 when [pyridine] = 0. Given that alkene 10 is the
principal product (>80%) formed from carbene 2, we
can take the extrapolated Y intercept of Figure 1 as a
reasonable estimate of k1,2 for the 2 ! 10 rearrange-
ment. Repetition of the determination gave a second
Cl
Cl
C
..
C
..
3
2
The diazirine precursors for carbenes 2 and 3 were both
prepared by the same sequence of reactions, illustrated
for carbene 2 in Scheme 1. Thus, cyclopentene was con-
verted to exo- and endo-bicyclo[3.1.0]hexane-6-carboxyl-
ate ethyl esters 4 with ethyl diazoacetate (Rh2(OAc)4,
CH2Cl2, 6 h, rt, N2 atm, 96%).11 Without purification,
the esters were hydrolyzed with epimerization, and then
acidified to yield exo-acid 512 (KOH, aq EtOH, refl., 6 h,
then HCl, mp 58–59 ꢁC, 72%), which was reesterified to
exo-methyl ester 6 (MeOH, H2SO4, refl., 6 h, 94%). The
ester was converted to amidine 7 by the Gielen–Garigi-
pati reaction13 (MeAl(Cl)NH2 toluene, 80 ꢁC, 40h,
then MeOH, rt, 1 h, mp 206–208 ꢁC from i-PrOH/ace-
tone, 56%; Anal. C, H, N). Compounds 4–7 were each
characterized by 1H and 13C NMR spectroscopy. Final-
ly, 7 was oxidized to diazirine 8 by aqueous NaOCl
(12.5% Ôpool chlorineÕ, LiCl, DMSO, pentane, 0–15 ꢁC,
15 min, ꢀ45%).14 Diazirine 8 was purified by chroma-
tography over silica gel, and characterized by NMR,15
IR (1566 cmꢁ1, film, N@N), and UV (kmax 346 nm,
pentane).
value of k1,2 = 5.87 · 105 sꢁ1, leading to an average
value of (5.79 0.08) · 105 sꢁ1
.
21–23
In a similar way, we determined k1,2 for carbene 3. Here,
ylide formation was observed at kmax 375 nm, and the
quality of the kinetic data is illustrated in Figure 2,
where ky = 3.40 · 107 Mꢁ1sꢁ1, and the Y intercept
(k1,2) = 1.90 · 106 sꢁ1. Again, correlation of kobs and
[pyridine] is linear, with r = 0.998 for 8 points. Repeti-
tion of the experiment led to an average value of
k1,2 = (1.89 0.02) · 105 sꢁ1 for the 3 ! 11 rearrange-
ment. The values of k1,2 that we observe for carbenes
In a precisely parallel series of reactions, norbornene
was transformed to diazirine 9,16 which was also purified
by chromatography, and characterized by NMR,17 IR
(1565 cmꢁ1, film, N@N), and UV (kmax 346 nm, pen-
tane). Diazirines 8 and 9 were decomposed either photo-
chemically (350nm, 25 ꢁC) or thermally (16 h, 70 ꢁC) in
CDCl3 or 1,2-dichloroethane (DCE) to give the 1,2-C
migration products 10 or 11, anticipated from the inter-
mediate carbenes 2 or 3, respectively. The products were
7x107
y = 5.71 e5 + 2.60 e7 x, R = 0.999
6x107
5x107
4x107
3x107
2x107
1x107
0
characterized by capillary GC, GC–MS, and 1H and 13
C
NMR spectroscopy. For 6-chlorobicyclo[3.2.0]hept-6-
ene (10), we observed the vinyl proton at d 5.69, as well
as the allylic ring-junction protons at d 3.2 and 3.09.18
Similarly, for 3-chloro-exo-tricyclo[4.2.1.02,5]nona-3-
ene (11), we found the vinyl proton at d 5.74 and the
allylic ring-junction protons at d 2.72 and 2.41.19 In
the photolysis of 8, about 17% of cyclopentene was
formed by the fragmentation of carbene 2; fragment-
ation was much reduced in the photolysis of 9, but other
impurities lowered the yield of 11 to about 81%.
0.0
0.5
1.0
1.5
[Pyr] / M
2.0
2.5
Figure 1. kobs (sꢁ1) for ylide formation between carbene 2 and pyridine
versus [pyr] (M); see text.
H
COOEt
MeOH
H2SO4
KOH, refl,
6 h, then H+
N2=CHCOOEt
Rh2(OAc)4
COOH
4
5
4
5
H
H
H
6
NaOCl, LiCl
DMSO, H2O
MeAl(Cl)NH2
then MeOH
N
N
3
COOMe
CNH2
Cl
NH2+, Cl-
2
1
6
7
8
Scheme 1.

G. Chu et al. / Tetrahedron Letters 46 (2005) 4137–4141
4139
14.2
3.0x107
2.5x107
2.0x107
1.5x107
1.0x107
5.0x106
0.0
y = 1.90 e6 +3.40 e7 x, R = 0.998
y = 20.33 - 1.91x, R = 0.983
14.0
13.8
13.6
13.4
13.2
13.0
12.8
12.6
12.4
12.2
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
[Pyr] / M
3.2
3.4
3.6
3.8
4.0
4.2
103 / T (K)
Figure 2. kobs (sꢁ1) for ylide formation between carbene 3 and pyridine
versus [pyr] (M); see text.
Figure 4. Arrhenius correlation for the 3 to 11 rearrangement in DCE;
ln k1,2 (sꢁ1) versus 1/T (Kꢁ1).
2 (5.8 · 105 sꢁ1) and 3 (1.9 · 105 sꢁ1) are comparable to
the range of k1,2 values (3.8–9 · 105 sꢁ1) reported
for the parent cyclopropylchlorocarbene, 1; see above.
Although the present results do not explain the discrep-
ancy between the experimental6 and computed7 values
of Ea for the rearrangements of 1, they parallel the
ÔlowÕ experimental Ea of ꢀ3 kcal/mol, and suggest that
the separation between the measured and computed
actizvation energies is Ôpersistent.Õ Indeed, we computed
DH298 for the rearrangement of carbene 2 to product
10 at the B3LYP/6-31G(d) level. Our value of
11.5 kcal/mol (in vacuo) compares well with DH2z98
10.1 kcal/mol (HDFT, in vacuo) reported for rearrange-
ment of the parent carbene (Eq. 1),7 reinforcing the per-
sistence of a ÔgapÕ between the ÔlowÕ experimental and
ÔhighÕ computed EaÕs or DHàÕs for these rearrangements.
Arrhenius LFP studies of the carbene rearrangements
were conducted over the temperature range of 243 (3)
or 253 (2) to 303 K. Correlations of ln k1,2 versus 1/T
are reproduced in Figure 3 for carbene 2 and Figure 4
for carbene 3. The quality of the data is good,
with r = 0.994 (7 points) for 2, and r = 0.983 (8 points)
for 3. The corresponding activation parameters are
2 ! 10: Ea = 3.1 kcal/mol, A = 4.3 · 108 sꢁ1
,
DSz298
ꢁ21 e.u.; and 3 ! 11: Ea = 3.8 kcal/mol, A = 6.8 ·
108 sꢁ1, and DS2z98 ꢁ 20e.u. Both sets of activation
parameters are similar, and they are also similar to those
we measured for the ring expansion of the parent
carbene 1 (Ea = 3 kcal/mol, DSà ꢀ ꢁ20to ꢁ24 e.u.).6
As a control experiment, the photolysis of diazirine 8
was examined at ꢁ30to ꢁ40 ꢁC. The product mixture,
dominated by 10, was unchanged from the room
temperature photolysis.
8
5
H
3
Cl
Cl
4
N
N
6
1
Cl
7
2
10
11
9
Me
..
..
14.8
C
F
C
Cl
y = 19.87 -1.58 x, R = 0.994
14.6
13
12
14.4
14.2
14.0
13.8
13.6
Heavy atom quantum mechanical tunneling (QMT) is
important in the ring expansions of 1-methylcyclobutyl-
fluorocarbene (12) to 1-fluoro-2-methylcyclopentene,24
and of noradamantyl chlorocarbene (13) to 2-chloro-
adamantene.25 In contrast, calculations suggest that car-
bon QMT is unimportant in the ring expansion of 1.7
Our preliminary experiments with matrix-isolated 2
and 3 similarly failed to demonstrate QMT at low tem-
peratures. Irradiation (334 nm, 75 min) of diazirine 8
isolated in a N2 matrix (ca. 1:400) at 9 K cleanly pro-
duced carbene 2. DFT calculated vibrational spectra
(B3LYP/6-31+G**) of ca. 50:50 syn/anti-mixture of 2
13.4
3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4.0
103 / T (K)
Figure 3. Arrhenius correlation for the 2 to 10 rearrangement in DCE;
ln k1,2 (sꢁ1) versus 1/T (Kꢁ1).

4140
G. Chu et al. / Tetrahedron Letters 46 (2005) 4137–4141
3. (a) Bonneau, R.; Liu, M. T. H.; Rayez, M. T. J. Am.
Chem. Soc. 1989, 111, 5973; (b) Liu, M. T. H.; Bonneau,
R. J. Phys. Chem. 1989, 93, 7298.
4. Ho, G.-J.; Krogh-Jespersen, K.; Moss, R. A.; Shen, S.;
Sheridan, R. S.; Subramanian, R. J. Am. Chem. Soc. 1989,
111, 6875.
gave a good fit to the experimental IR. Concurrently, a
broad absorption in the UV/vis at 400–600 nm (max
490nm) was observed to grow in. No changes were
observed in the IR or UV/vis spectra of 2 after standing
in the dark for 7 days at 8 K.
5. For the ylide methodology, see: Jackson, J. E.; Soundar-
arajan, N.; Platz, M. S.; Liu, M. T. H. J. Am. Chem. Soc.
1988, 110, 5595.
6. Moss, R. A.; Ho, G.-J.; Shen, S.; Krogh-Jespersen, K.
J. Am. Chem. Soc. 1990, 112, 1638.
7. Albu, T. V.; Lynch, B. L.; Truhlar, D. G.; Goren, A. C.;
Hrovat, D. A.; Borden, W. T.; Moss, R. A. J. Phys. Chem.
A 2002, 106, 5323.
8. Moss, R. A.; Ho, G.-J.; Liu, W. J. Am. Chem. Soc. 1992,
114, 959.
9. A repetition of the experimental study gave Ea = 5.1
1.1 kcal/mol for the ring expansion of cyclopropyl-
fluorocarbene.7
10. Moss, R. A.; Liu, W.; Krogh-Jespersen, K. Tetrahedron
Lett. 1993, 34, 6025.
However, irradiation of 2 at 546 nm for 30min resulted
in its disappearance with the simultaneous growth of a
series of new bands, some of which belonged to chloro-
acetylene (2110cm ꢁ1) and cyclopentene (assigned by
comparison to the IR spectrum of an authentic sample).
Other new bands, although of low intensity, were consis-
tent with the IR spectra calculated for 10.26 Interest-
ingly, irradiation of 2 at 435 nm caused more rapid
disappearance of the IR bands that fit predictions for
the carbene conformer with H and Cl syn (strongest
bands at 1323 and 734 cmꢁ1), compared to the bands
assigned to anti-2 (strongest at 1386 and 646 cmꢁ1).
Similarly, photolysis of matrix-isolated 9 (ca. 1:400, N2,
8 K) at 334 nm gave a mixture of syn and anti-carbene 3,
matching the calculated IR spectra. Carbene 3 showed
UV/vis absorption very similar to 2 with a broad band
in the 400–600 nm region and a maximum at 500 nm.
Although no changes in the spectra could be detected
after the matrix was maintained in the dark for 5 days
at 8 K, 3 could be easily destroyed by irradiation at
546 nm for 3 h, producing norbornene (assigned by
comparison to the IR spectrum of an authentic sample),
chloroacetylene, and 11.26 Again, 435 nm irradiation
caused selective ꢁr1eaction of syn-3 (strongest bands at
1126 and 824 cm ) compared to anti-3 (strongest bands
at 1136 and 651 cmꢁ1).27
11. Cf. Doyle, M. P.; Bagheri, V.; Wandless, T. J.; Harn, N.
K.; Brinker, D. A.; Eagle, C. T.; Loh, K.-L. J. Am. Chem.
Soc. 1990, 112, 1906.
12. Jung, M. E.; Fahr, B. T. J. Org. Chem. 2000, 65, 2239.
13. (a) Gielen, H.; Alonso-Alija, C.; Hendrix, M.; Niewo¨hner,
U.; Schauss, D. Tetrahedron Lett. 2002, 43, 419; (b)
Garigipati, R. S. Tetrahedron Lett. 1990, 31, 1969.
14. Graham, W. H. J. Am. Chem. Soc. 1965, 87, 4396.
1
15. Diazirine 8. H NMR (400 MHz, CDCl3, d): 0.96–1.1 (m,
1H, H3), 1.29 (d, J = 1 Hz, 2H, H1, H5), 1.39 (t, J = 1 Hz,
1H, H6), 1.54–1.62 (m, 1H, H30), 1.64–1.78 (m, 4H, H2,
H4). 13C NMR (300 MHz, CDCl3, d): 20.52, 25.08, 25.12,
27.40, 49.04.
16. The initial addition reaction between norbornene and
ethyl diazoacetate was catalyzed by CuCN: Sauers, R. R.;
Sonnet, P. E. Tetrahedron 1964, 20, 1029.
17. Diazirine 9. 1H NMR (400 MHz, CDCl3, d): 0.68 (d,
J = 11 Hz, 1H, H8), 0.77 (d, J = 2.4 Hz, 2H, H2, H4), 0.87
(dt, J = 11 Hz, 1.8 Hz, H80), 1.22–1.48 (m, 4H, H6, H7),
1.57 (t, J = 2.4 Hz, 1H, H3), 2.29 (s, 2H, H1, H5). 13C
NMR (300 MHz, CDCl3, d): 20.40, 22.98, 28.43, 29.09,
35.87, 48.91.
In conclusion, fused-ring cyclopropylchlorocarbenes 2
and 3 ring-expand to fused-ring chlorocyclobutenes 10
and 11 with activation energies of ꢀ3–4 kcal/mol and
activation entropies ꢀꢁ20e.u. These parameters are
similar to those observed for the parent cyclopropyl-
chlorocarbene 1,6 suggesting that the difference between
the experimental values and the computed Ea of
ꢀ8.5 kcal/mol7 for the ring expansion of 1 is real and
may be significant.
18. Alkene 10. 1H NMR (400 MHz, CDCl3, d): 1.14–1.28 (m,
2H, H3), 1.50–1.90 (m, 4H, H2, H4), 3.09 (m, 1H, H1),
3.32 (dd, J = 3.2, 7.6 Hz, 1H, H5), 5.69 (d, J = 0.4 Hz, 1H,
H6). 13C NMR (300 MHz, CDCl3, d): 22.90, 24.54, 26.64,
44.37, 54.15, 117.80, 131.42. MS (m/e): 128, 130[ M+,
(M+2)+].
1
19. Alkene 11. H NMR (400 MHz, CDCl3, d): 1.03 (m, 2H,
Acknowledgements
H9), 1.47 (m, 2H, H7), 1.59 (m, 2H, H8), 2.02 (t,
J = 1.6 Hz, 1H, H6), 2.10(t, J = 1.6 Hz, 1H, H1), 2.41
(d, J = 3.2 Hz, 1H, H5), 2.72 (d, J = 3.2 Hz, 1H, H2), 5.74
(s, 2H, H4). 13C NMR (300 MHz, CDCl3, d): 22.99, 28.29,
30.49, 32.84, 36.37, 46.34, 56.38, 117.42, 131.56. MS (m/e):
154, 156 [M+, (M+2)+].
We are grateful to the National Science Foundation
(Rutgers and Nevada) and the donors of the Petroleum
Research Fund (Nevada) for financial support. We
thank the National Center for Computer Applications
for time on the IBM P Series 690(to R.R.S.).
20. See: Moss, R. A.; Johnson, L. A.; Merrer, D. C.; Lee, G.
E. J. Am. Chem. Soc. 1999, 121, 5940, for a description of
the LFP system. The 1000 W Xe monitoring lamp has
since been replaced by a 150W pulsed Xe lamp.
References and notes
21. Some alkene 10 could arise directly by rearrangement
concerted with nitrogen loss of diazirine 8.22,23 To estimate
this contribution, 8 was photolyzed in 2-methyl-2-butene
and the reaction products were examined by capillary GC.
Formation of ꢀ61% of the cyclopropane adducts of
carbene 2 and 2-methyl-2-butene was verified by GC–MS;
the proportion of cyclobutene 10 was ꢀ39%. The extensive
quenching of 10 by the added alkene proves that most of
the 10 formed in the absence of alkene (>60%) stems from
1. (a) For leading references see: Huang, H.; Platz, M. S.
J. Am. Chem. Soc. 1998, 120, 5990; Fernamberg, K.;
Snoonian, J. R.; Platz, M. S. Tetrahedron Lett. 2001, 42,
8761; (b) Platz, M. S. Adv. Carbene Chem., Brinker, U. H.,
Ed. 1998, 2, 133f.
2. Moss, R. A.; Fantina, M. E. J. Am. Chem. Soc. 1978, 100,
6788.

G. Chu et al. / Tetrahedron Letters 46 (2005) 4137–4141
4141
carbene 2, and not from an excited state of diazirine 8. The
pyridine ylide LFP kinetics are therefore valid as a
measure of the rearrangement of carbene 2. The method
is independent of excited diazirine because it follows
reactions of the carbene; the excited diazirine does not live
long enough to react with pyridine.23
26. Only minor amounts of the fragmentation or rearrange-
ment products of the carbenes were observed in initial
irradiations of matrix isolated diazirines 8 or 9, suggesting
at most only minor excited state diazirine participation at
low temperature. See also, note 21.
27. TD B3LYP/6-31+G** calculations predict shorter wave-
length absorptions for the syn conformations of carbenes 2
and 3 (488 and 489 nm, respectively), compared to the anti
conformers (553 and 555 nm, respectively). Preliminary
experiments suggest that irradiations at 590nm induce
selective photochemistry of the anti carbenes. Mixtures of
both rearrangement and fragmentation products are
observed from 2 and 3 irrespective of the irradiation
wavelength, although further experiments probing
whether the syn and anti conformers of the carbenes
may favor different products are in progress.
22. Reviews: (a) Bonneau, R.; Liu, M. T. H. Adv. Carbene
Chem., Brinker, U. H., Ed. 1998, 2, 1f; (b) Ref. 1b; (c)
Merrer, D. C.; Moss, R. A. Adv. Carbene Chem., Brinker,
U. H., Ed. 2001, 3, 53f.
23. Moss, R. A.; Liu, W.; Krogh-Jespersen, K. J. Phys. Chem.
1993, 97, 13413.
24. Zuev, P. S.; Sheridan, R. S.; Albu, T. V.; Truhlar, D. G.;
Hrovat, D. A.; Borden, W. T. Science 2003, 299, 867.
25. Moss, R. A.; Sauers, R. R.; Sheridan, R. S.; Tian, J.; Zuev,
P. S. J. Am. Chem. Soc. 2004, 126, 10196.