Fragmentation of cyclobutoxychlorocarbene: The cyclopropylcarbinyl/cyclobutyl cations revisited

Article abstract of DOI:10.1002/poc.371

Fragmentations of cyclobutoxychlorocarbene (13, k_frag= 7.1 × 10⁵ s^-1) and cyclopropylmethoxychlorocarbene (14, k_frag = 7.6 × 10⁵ s^-1) in MeCN proceed to tight and distinct [R⁺ OC Cl^-] ion pairs, which collapse to different distributions of cyclopropylcarbinyl, cyclobutyl and allylcarbinyl chlorides. B3LYP/6-31G* calculations support these conclusions, affording computed fragmentation activation energies of 6.4 (13) and 3.0 (14) kcal mol^-1.

Full text of DOI:10.1002/poc.371

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2001; 14: 400–406
DOI:10.1002/poc.371
Fragmentation of cyclobutoxychlorocarbene: the cyclo-
†
propylcarbinyl/cyclobutyl cations revisited
Robert A. Moss,* Fengmei Zheng, Lauren A. Johnson and Ronald R. Sauers
Department of Chemistry, Rutgers, The State University, New Brunswick, New Jersey 08903, USA
Received 21 November 2000; revised 13 January 2001; accepted 19 January 2001
5
À1
ABSTRACT: Fragmentations of cyclobutoxychlorocarbene (13, k = 7.1 Â 10 s ) and cyclopropylmethoxy-
frag
5
À1
‡
À
chlorocarbene (14, kfrag = 7.6 Â 10 s ) in MeCN proceed to tight and distinct [R OC Cl ] ion pairs, which collapse
to different distributions of cyclopropylcarbinyl, cyclobutyl and allylcarbinyl chlorides. B3LYP/6–31G* calculations
support these conclusions, affording computed fragmentation activation energies of 6.4 (13) and 3.0 (14) kcal mol
Copyright  2001 John Wiley & Sons, Ltd.
À1
.
KEYWORDS: cyclobutoxychlorocarbene; fragmentation; cyclopropylcarbinyl cation; cyclobutyl cation
INTRODUCTION
example, reaction of cyclopropyldiazomethane with
ethereal benzoic acid affords the benzoate esters of 3, 4
and 5 in the distribution 79.2:13.6:7.2. The cyclopropyl-
to-cyclobutyl ratio, which is ꢀ1.2 in the nitrous acid
deamination reactions, increases to ꢀ5.8 in the ion pair
process.
2
The mechanistic interrelation of the cyclopropylcarbinyl
(
1) and cyclobutyl (2) cations is a classical triumph of
1
early physical organic chemistry. When either carbo-
cation is in a ‘free’ state and a polar solvent (e.g. as
generated from the alkyldiazonium cations produced by
We found that fragmentations of alkoxychlorocar-
3
benes lead to ion pairs that resemble those of
4
deaminative processes. Momentarily neglecting product
3
‘
(
memory effects’ due to cis or trans ROCCl precursors
memory effects due to precursor geometry are known for
5
nitrogen-separated ion pairs ), we can represent typical
carbene-derived and diazonium-derived ion pairs as 6
and 7, respectively. The similarity is apparent, even to the
isoelectronic character of the CO and N2 ‘leaving
groups.’ When cyclopropylmethoxychlorocarbene frag-
ments in MeCN at 23°C, the distribution of cyclopropyl-
the aqueous nitrous acid deaminations of the correspond-
ing amines), rapid 1,2-C shift rearrangements intercon-
vert 1 and 2, and ‘scramble’ labeled carbon atoms within
each carbocation. Nitrous acid deamination of either
cyclopropycarbinylamine or cyclobutylamine leads to
1
cyclopropylmethanol (3), cyclobutanol (4) and 3-butenol
1b
1c,2
(
5); a typical distribution is 52:44:4.
carbinyl chloride (8), cyclobutyl chloride (9) and 4-
chloro-1-butene (10) is 78:15:7 (cyclopropylcarbinyl:
When, however, the cyclopropylmethyldiazonium
cation is generated in a moderately polar solvent as part
‡
À
of an ion pair (e.g. RN2 O CR'), rapid anion capture of
2
the successor alkyl cation is facilitated, leading to a
product mixture rich in the cyclopropylmethyl ester. For
6
cyclobutyl = 5.2). This closely resembles the distribu-
tion of the esters formed from cyclopropyldiazomethane
2
and benzoic acid in diethyl ether (see above), suggesting
*
Correspondence to: R. A. Moss, Department of Chemistry, Rutgers,
The State University, New Brunswick, New Jersey 08903, USA.
E-mail: moss@rutchem.rutgers.edu
that ion pairs 6 and 7 (R = cyclopropylcarbinyl) are good
overall representations of the initial intermediates in
these processes.
†
Dedicated to Professor Hans-J o¨ rg Schneider in recognition of his 65th
birthday.
6
Contract/grant sponsor: National Science Foundation.
In our earlier work, however, we neither determined
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 400–406

FRAGMENTATION OF CYCLOBUTOXYCHLOROCARBENE
401
the rate constant for the fragmentation of cyclopropyl-
methoxychlorocarbene nor compared its kinetics and
product distributions with analogous data for cyclobu-
toxychlorocarbene. Here, we provide these results, in
addition to computational studies of these carbene
fragmentations, enabling us to refine our mechanistic
portrait of these reactions.
(with traces of 17) formed in addition to chlorides 8–10.
Product distributions for the fragmentations of 13 and 14
in 100% ethanol are also given in Table 1.
In Fig. 1, we depict the relative yields of chlorides 8–
10 and ethers 15 and 16 as a function of the mole fraction
of ethanol in MeCN for the fragmentation of 13. The
yield of cyclobutyl chloride (9) decreases smoothly,
whereas that of the corresponding ether (16) increases,
with increasing ethanol. Cyclopropylcarbinyl chloride (8)
and the analogous ether (15) present a similar profile.
Remarkably, the cyclopropylmethyl-to-cyclobutyl ratios,
RESULTS AND DISCUSSION
Precursor and products
Cyclobutanol (4) was converted to the cyclobutyl
isouronium triflate (11) by reaction with cyanamide and
7
8
trifluoromethanesulfonic acid. Graham oxidation of 11
with hypochlorite then afforded 3-cyclobutoxy-3-chloro-
diazirine (12), which was purified by chromatography
8
:9 and 15:16, remain reasonably constant at 1.4–1.6 for
RCl and 0.66–0.96 for ROEt over the range of ethanol
mole fractions, suggesting that the origins of the
chlorides and ethers lie in distinct ion pairs.
(
silica gel/pentane). The pentane solvent was removed
evaporatively and replaced with either MeCN or 1,2-
dichloroethane (DCE) for kinetic and product studies.
UV maxima were observed for 12 at 352 and 367 nm
(
pentane), 354 and 368 nm (MeCN) and 353 and 367 nm
(
DCE).
Photolysis (ꢀ > 320 nm) of diazirine 12 (Aꢀmax = 1.0)
in MeCN or DCE generated cyclobutoxychlorocarbene
(13), which fragmented to yield chlorides 8–10. These
products were identified by gas chromatography–mass
spectrometry (GC–MS) and GC comparisons with authen-
6
tic materials, while product distributions were ascertained
by capillary GC analysis. The same products were formed
from diazirine generated cyclopropylmethoxychlorocar-
6
bene (14). Table 1 gives the product distributions from
the fragmentations of carbenes 13 and 14.
Figure 1. Product distributions &%) from the fragmentation
of cyclobutoxychlorocarbene &13) in MeCN±EtOH as a
function of the mole fraction &X) of ethanol. See text for
product numbering
When 13 and 14 were generated by diazirine photolysis
6
6
in ethanol, or ethanol–MeCN mixtures, ethers 15 and 16
a
Table 1. Product distributions from fragmentations of carbenes 13 and 14
Product distribution (%)
b
b
Carbene
Solvent
8
9
10
8:9
15
16
15:16
1
3
MeCN
DCE
EtOH
MeCNc
MeCN
EtOH
55.3
64.5
37.4
73.3
78
34.3
29.8
23.9
17.1
15
10.3
5.6
8.5
9.6
7
1.6
2.2
1.6
4.3
5.2
7.4
12.9
29.1
17.4
22.3
0.74
1.3
1
4
41.0
5.5
2.1
a
b
c
Diazirine precursors were photolyzed at 25°C; product percentages refer to the total product.
Ratio of cyclopropylcarbinyl to cyclobutyl product.
Thermal generation of 14 at 25°C; from Ref. 6.
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 400–406

4
02
R. A. MOSS ET AL.
cyclopropylcarbinyl, ion pairs 6 give rise to slightly less
than half of the products. In both instances, ethers 15 and
1
6 comprise the balance of the product mixtures.
The ether product ratios (15:16) in Table 1 reveal only
small precursor memory effects and are much closer to
1.0 whether they stem from carbene 13 or 14. These
1
5:16 distributions approach the cyclopropylcarbinyl-to-
cyclobutyl ratios obtained from the ‘free’ cations formed
in nitrous acid deamination reactions (1.1–1.3).
1
,2
Clearly, the ultimate ionic precursors of chlorides 8 and
9
differ from those of ethers 15 and 16.
Interconversion of cations 1 and 2 within ion pairs 6 is
in competition with ion pair collapse, solvolysis by
À
ethanol and diversion to ion pairs in which OEt replaces
À
Cl . The dynamics of these processes are complex and
governed by only small enthalpy and entropy differences
between the different species. An additional complication
concerns the geometry of ion pair 6, which may initially
Figure 2. ROEt:RCl product ratio vs mole fraction of ethanol
in acetonitrile for the fragmentation of cyclobutoxychloro-
carbene &13)
3,6
arise as cis-6 or trans-6, depending on the geometry of
Moreover, the ion pair(s) involved in the fragmenta-
tions of carbene 13 must be, at least in part, distinct from
those arising from carbene 14. In the latter case (Table 1),
the 8:9 ratio of 4.3–7.4 is very much weighted toward
cyclopropylmethyl chloride, whereas from carbene 13,
the precursor ROCCl [where rotation around the central
À1
O—C bond is opposed by ꢀ15–18 kcal mol (1 kcal
1
1
= 4.184 kJ) due to partial double bond character].
3,6
We have speculated that cis-6, in which CO does not
‘insulate’ R from Cl , may be largely responsible for
‡
À
8
:9 is only 1.6–2.2, depending on solvent. Put another
the RCl product formed by ion pair return in ethanol,
‡
way, the 8:9 ratios ‘remember’ their origins: from the
cyclopropylmethyl precursor (14), chloride mixtures rich
in 8 result; from the cyclobutyl precursor (13), larger
whereas trans-6, in which CO is interposed between R
À
and Cl , may be more readily intercepted by ethanol,
mainly affording ROEt. No direct test of this idea is
available.
(
although not dominant) yields of 9 are obtained.
Just as the 8:9 ratio in MeCN (4.3–5.2) from 14, via 6,
resembles that obtained (5.6) from the ethereal cyclo-
2,6
propylmethyldiazonium benzoate ion pair (7), the 8:9
ratio found here for the fragmentation of carbene 13 (1.6
in MeCN) is similar to that observed (1.3) for the
analogous ester products formed from cyclobutyldiazo-
nium p-phenylazobenzoate ion pairs in 98% toluene –2%
ethanol. The analogy between the ion pairs arising from
carbene fragmentation and those from deaminative
reactions is thereby strengthened.
KINETICS
Although rate constants for the decompositions of most
alkyldiazonium cations are unknown because of their
12
exceedingly facile fragmentation, rate constants for the
fragmentation of ROCCl can be measured by laser flash
9
3,13
photolysis (LFP).
Absolute rate constants for the fragmentations of
14
We also note the persistence of RCl products in the
fragmentations of either 13 or 14, even in pure ethanol.
Figure 2 traces the ROEt-to-RCl product ratios as a
function of ethanol mole fraction in MeCN for fragmen-
tations of carbene 13. Analogous correlations have been
carbenes 13 and 14 were determined by LFP using the
15
pyridine ylide methodology. For example, LFP at
351 nm and 25°C of diazirine 12 in MeCN or DCE
(A ꢁ 1.0 at ꢀmax) in the presence of pyridine produced an
absorbance at 412 nm due to ylide 18. In MeCN,
6
published for the fragmentations of 14 and benzyloxy-
6,10
chlorocarbene.
At 100% ethanol, the RCl:ROEt ratios
are 2.3 from 13 and 0.94 from 14 (Table 1). Ion pairs 6,
with chloride counterions, therefore persist in ethanol
and, when R is initially cyclobutyl, account for about
two-thirds of the product mixture. When R is initially
correlation of the apparent rate constants for ylide
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 400–406

FRAGMENTATION OF CYCLOBUTOXYCHLOROCARBENE
403
a
Table 2. Rate constants for carbene fragmentation
À1 b
kfrag (s
)
Carbene
MeCN
DCE
5
5
5
4
13
14
7.1 Â 10
7.6 Â 10
3.2 Â 10
8.6 Â 10
a
At 25°C.
Errors, 10–15%.
b
Figure 3. B3LYP/6±31G* ground state and fragmentation
transition state for cis-cyclobutoxychlorocarbene &cis-13) in
SCI-PCM simulated MeCN. See text for computational
details
6
À1
formation, kobs (1.10–2.15 Â 10 s ), vs pyridine
concentration (1.65–7.42 M) was linear (seven points,
5
À1 À1
r = 0.994) with a slope of 1.59 Â 10 l mol
s
and a y-
5
À1
intercept of 7.14 Â 10 s . The former value is the
second-order rate constant for ylide formation from
carbene 13 and pyridine, whereas the latter value can be
equated with kfrag for 13 → 6 (R = cyclobutyl) because
the product studies show that only fragmentation
products arise from cyclobutoxychlorocarbene.
toxychlorocarbene fragmentation rate constant is about
twice that for benzyloxychlorocarbene (ꢀ4 Â
5
À1 13,16
10 s ),
and ꢀ10 times greater than k
for cyclo-
frag
4
À1
pentoxychlorocarbene (ꢀ8 Â 10 s ) (L. A. Johnson
and R. A. Moss, unpublished work). Small residual elec-
tronic stabilizations may therefore operate even in the
‘early’ fragmentation transition states of 13 and 14.
Kinetic data for the fragmentations of 13 and 14 in
both MeCN and DCE are given in Table 2. There is some
indication of a solvent effect between DCE (" = 10.7) and
the more polar MeCN (" = 36.6); kfrag is 2.2 (13) to 8.8
16
(
14) times smaller in DCE than in MeCN. However, the
COMPUTATIONALSTUDIES
important observation is the similarity of kfrag for 13 and
5
5
1
4, which fragment in, e.g., MeCN at 7 Â 10 –8 Â 10
In analogy with our previous computational studies of
À1
11a
s .
alkoxyhalocarbene fragmentation,
we computed
In contrast, in acetolysis reactions, cyclopropylcar-
binyl tosylate is considerably more reactive than
cyclobutyl tosylate: the respective enthalpies of activa-
ground-state and fragmentation transition states (TS)
for (cis) carbenes 13 and 14. All structures were fully
optimized by analytical gradient methods at the B3LYP/
6–31G* level using the Gaussian94 suite of programs
(the optimization used Gaussian94 Revision B.1 and
À1
tion are 16.7 (Ref. 17) and 30 (Ref. 18) kcal mol , with
À3 À1
corresponding rate constants at 50°C of ꢀ2.4 Â 10
s
17
19
(
extrapolated from the 25°C data ) and 3.4 Â
default convergence criteria and DFT calculations used
À5 À1 18
1
0
s , respectively. This rate differential arises
Becke’s three-parameter hybrid method using the LYP
20
mainly because the solvolysis of the cyclopropylcarbinyl
tosylate is strongly assisted by formation of the
correlation functional ). Computed (unscaled) gas-phase
energies were corrected for thermal effects at 298.15 K
and for zero-point energy differences. Normal coordinate
analyses confirmed the nature of the ground-and transi-
tion-state structures. To simulate the MeCN solvent
(" = 36.64) in which the fragmentations were studied, we
employed the SCI-PCM computational model, with full
geometry optimization.
‘
bisected’ cyclopropylmethyl cation, where electron
donation from the strained ring’s proximal p-rich s-
1a
bonds efficiently stabilizes the developing carbocation.
Fragmentations of carbenes 13 and 14, however,
proceed through early transition states and over low
activation barriers, where there is relatively little call for
the potential electronic stabilization available in the
The computed ground and transition states for
carbenes 13 and 14 (both in MeCN) appear in Figs 3
and 4, respectively. In each TS, the C—Cl and (alkyl)
C—O bonds are in the process of breaking: C—Cl
cyclopropyl ring of 14. Attempts to measure E for the
a
fragmentation of 13 in MeCN (À20 to 30°C) gave k
frag
values that showed little temperature dependence, and
˚
scattered about a nearly horizontal (E ꢁ 0) correlation
distances increase from 1.93 to 2.46 A (13) and from 1.94
a
˚
line. Scatter was also observed for k of 14 below 15°C.
to 2.33 A (14), while parallel increases occur for C—O
frag
˚
In our experience, these abortive Arrhenius correlations
distances: from 1.50 to 1.92 A (13) and from 1.52 to
À1
˚
imply low activation energies (E <3–4 kcal mol ).
1.85 A (14). Simultaneously, the (carbene) C—O bonds
a
˚
Indeed, computed E s for the fragmentations of 13 and 14
of 13 and 14 contract from 1.25–1.26 to 1.18–1.19 A, en
a
À1
˚
are 6.4 and 3.0 kcal mol , respectively, in MeCN (see
route to the CꢂO bond length of 1.128 A.
below).
Another point deserves mention: the close proximities
‡
Although the low activation energies and concomitant
early transition states for the fragmentations of 13 and 14
suppress differences in kfrag, we note that the cyclobu-
of the nascent chloride anions and C(ꢁ ) in the transition
˚
˚
states (3.37 A for 13 and 3.36 A for 14) point toward the
subsequent formation of (cis) ion pairs 6. Indeed, intrinsic
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 400–406

4
04
R. A. MOSS ET AL.
21a
by Koch et al. Small differences can be attributed to
2
1b
the presence of the counterion and carbon monoxide.
As in the fragmentation of 14, so too with 13, the
cyclobutyl C—O and C(carbene)—Cl separations con-
tinually increase (in vacuo) from ground state to
transition state to ion pair (for 13 the interatomic
distances are given for gas-phase ground state, transition
˚
state and ion pair), from 1.48 to 2.16 to 2.77 A for C—O
and 1.89 to 2.64 to 3.28 for C—Cl. The CO molecule is
clearly departing. The behavior of the cyclobutyl C—Cl
Figure 4. B3LYP/6±31G* ground state and fragmentation
transition state for cis-cyclopropylmethoxychlorocarbene
&cis-14) in SCI-PCM simulated MeCN. See text for computa-
separation, however, is different, contracting from
tional details
˚ ˚
.10 A in the transition state to 3.05 A in the ion pair
21b
3
(
Fig. 5). This contraction must, in part, reflect the absence
‡
of polar solvent in the calculation, making the C(ꢁ )—
Cl(ꢁ ) interaction even more important than it is in the
À
reactivity coordinate (irc) calculations were carried out
starting from the transition states of Figs 3 and 4. With
the cyclopropylcarbinyl system in MeCN (Fig. 4), the
energy and structure evolved in two directions: backward
toward ground-state carbene 14, and forward toward the
ion pair shown in Fig. 5(left), which displays the
geometry associated with the ‘bisected’ form of the
cyclopropylcarbinyl system (see above), where MeCN is
simulated, and stabilizes the ion pair {in the gas phase,
the ‘bicyclobutonium’ chloride ion pair is 5.7 kcal mol
À1
higher in energy than the carbene (13) from which it
derives. No doubt, in MeCN, the ion pair would be
significantly lower in energy. Note that the activation
energy for the fragmentation of 13 is computed to be
1a
cyclopropylcarbinyl cation.
Note that the C—Cl
À1
˚
separation in this ion pair is 3.48 A, not very different
much lower in MeCN (6.4 kcal mol ) than in vacuo
À1
˚
from the 3.36 A separation in the corresponding transi-
(20.3 kcal mol ), largely due to preferential stabilization
tion state (Fig. 4). In contrast, the C—O and C(car-
bene)—Cl separations of the ion pair (2.60 and 2.94 A,
of the polar transition state (Fig. 3) by MeCN. [The
˚
parallel computed reduction in E induced by solvation
a
respectively) greatly increase, relative to the analogous
in the fragmentation of 14 is 17.8 (vacuum) vs
À1
˚
separations in the fragmentation transition state (1.85 A
3.0 kcal mol
(MeCN)]}. In this latter case, the
‡
À
˚
and 2.33 A, respectively). Interestingly, the energy of the
C(ꢁ )—Cl(ꢁ ) distance increases slightly (from 3.36 to
˚
ion pair–CO assemblage is computed to be
3.48 A) on going from the transition state to the ion pair
À1
ꢀ
4.4 kcal mol lower than that of carbene 14. Thus, in
(see above).
MeCN, we calculate that cyclopropylmethoxychlorocar-
Clearly, the alkyl groups of carbenes 13 and 14 retain
distinctive characters in their fragmentation transition
states (Figs 3 and 4), and in the ion pairs that result from
the fragmentations (Fig. 5).
bene exothermically fragments to an ion pair, with the
latter separated from the carbene by a barrier of only
À1
3
.0 kcal mol .
An irc treatment of the transition state for the
Activation energies were obtained for the fragmenta-
tions of 13 and 14 in MeCN as differences between the
fragmentation of 13 (Fig. 3) proceeded only to the
ground-state geometry when the calculation was carried
out in simulated MeCN. In vacuum, the irc led backward
to the ground state and forward to the ion pair–CO
assemblage shown in Fig. 5(right), where the carbon
framework corresponds to that computed for the
bicyclobutonium ion. The geometries computed for the
ion pairs in Fig. 5 closely correspond to those computed
computed transition-state and ground-state energies;
À1
we obtained E = 6.4 (13) and 3.0 (14) kcal mol . A
a
small residue of the substantial acetolysis DDH‡
between cyclopropylcarbinyl and cyclobutyl tosylates
À1
(ꢀ13 kcal mol , see above) is apparent in the computed
À1
E s; DE ꢁ 3.4 kcal mol in favor of 14. However, the
a
a
computed E s are fairly low, consistent with the observed
a
rapid fragmentation rates of the carbenes (Table 2), and
with the ‘early’ transition states depicted in Figs 3 and 4.
The E s computed for the fragmentations of 13 and 14 in
a
MeCN are bracketed by those calculated for the other
1
1
(
(
cis) ROCCl in MeOH,
including Me CHOCCl
2
À1
À1
8.0 kcal mol ) and PhCH OCCl (1.4 kcal mol ).
2
CONCLUSIONS
Fragmentations of cyclobutoxychlorocarbene (13) and
cyclopropylmethoxychlorocarbene (14) in MeCN gen-
erate distinct ion pairs consisting of an alkyl cation,
Figure 5. Product endpoints of irc computations carried out
starting from the respective transition states: left, carbene 14
&
in MeCN); right, carbene 13 &gas phase)
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 400–406

FRAGMENTATION OF CYCLOBUTOXYCHLOROCARBENE
405
carbon monoxide and a chloride anion. Although
subsequent, reversible interconverting rearrangements
of the alkyl cations occur competitively with ion pair
collapse, the latter process dominates, so that distinct
product distributions of chlorides 8–10 are formed from
each carbene. Differences in product distributions persist,
although to a smaller extent, in the formation of ethers 15
and 16 from the fragmentations of carbenes 13 or 14 in
ethanol. Computational studies support the generation of
distinct ion pairs form each carbene, and further suggest
that the cationic components of the ion pairs derived from
harvested, washed with diethyl ether and dried in vacuo
to afford 54% of the title salt, m.p. 69–70°C.
1
H NMR (ꢁ, DMSO-d ): 8.30 (br, s, 4H NH ), 4.93 (m,
6
2
1H, CHO), 2.46, 2.07, 1.80, 1.54 (ms, 6H, cyclobutyl).
Anal. Calculated for C H F N O S: C, 27.26; H, 4.20;
6
11 3 2 4
N, 10.61. Found: C, 27.25; H, 4.26; N, 10.63%.
3-Cyclobutoxy-3-chlorodiazirine &12). The general
8
method of Graham was followed. To 3.5 g of LiCl in
100 ml of DMSO were added 1.0 g (3.8 mmol) of
isouronium salt 11 and 50 ml of pentane. The mixture
was cooled to 20°C and stirred magnetically. Then,
200 ml of 12% commercial aqueous sodium hypochlorite
solution (‘pool chlorine’), saturated with NaCl, were
slowly added. Stirring was continued for 15 min at 15°C
after the addition had been completed. The reaction
mixture was poured into 150 ml of ice–water in a large
separating funnel. The aqueous phase was removed and
the pentane layer was washed twice with 75 ml portions
1
3 or 14 resemble the bicyclobutonium cation or the
21a
bisected cyclopropylmethyl cation, respectively. The
rate constants for the fragmentations of carbenes 13 and
5
1
7
4 in MeCN were determined by LFP as 7.1 Â 10 and
5
À1
.6 Â 10 s , respectively. The similarity in rate
constants reflects the low activation energies of the
fragmentations, which are calculated (B3LYP/6–31G*)
À1
as 6.4 (13) and 3.0 (14) kcal mol in MeCN.
of ice–water and then dried for 2 h over CaCl at 0°C. The
2
EXPERIMENTAL
diazirine–pentane solution was purified by chromatogra-
phy over silica gel with pentane as eluent. Pentane was
removed by rotary evaporation and replaced by MeCN or
DCE to a volume of ꢀ30 ml. The UV maxima of 12 in
Solvents. Acetonitrile and pyridine (both Fisher, Certi-
fied, ACS) were dried by refluxing over CaH , followed
by distillation, and storage over 5A molecular sieves.
Dichloroethane (Aldrich, Certified, ACS) was used as
received. Pentane (Fisher, HPLC grade) was stored over
2
pentane, MeCN, and DCE are described in the text.
1
H NMR (ꢁ, CD CN): 4.2–4.4 (m, 1H, CHO), 1.9–2.0,
3
1.5–1.8, 1.2–1.5 (ms, 6H, cyclobutyl). Details of the
preparation of cyclopropylmethylisouronium tosylate
and of cyclopropylmethoxychlorodiazirine can be found
in Refs 6 and 24. Authentic samples of chloride products
8–10 and ether products 15 and 16, are also described in
these sources, and in references cited there in.
5
A molecular sieves.
22
Cyclobutanol . Cyclopropylcarbinol (10 g, 0.14 mol)
and 50 ml of 2.0 M aqueous HCl solution were heated at
8
5°C for 2 h until a clear solution was obtained. This
solution was extracted with 3 Â 20 ml of diethyl ether,
and the combined extract was dried over MgSO .
4
Filtration and rotary evaporation afforded 15 ml of liquid
that was distilled through a microscale spinning band
column. The fraction with bp 121–124°C was collected
Diazirine photolysis. Solutions of diazirine 12 in MeCN,
DCE or MeCN–EtOH (A = 1.0 at ꢀmax) were photolyzed
at 25°C for 1 h with a focused Oriel UV lamp, ꢀ
>320 nm (uranium glass filter). The products were
analyzed by capillary GC and GC–MS, using a 30 m
 0.25 mm i.d., 0.25 mm film thickness CP-Sil 5CB
(100% dimethylpolysiloxane) column at 25°C (4 min,
as cyclobutanol.
1
H NMR (200 MHz) (ꢁ, DMSO-d ): 4.02 (m, 1H,
6
CHOH), 2.10 (m, 2H), 1.80 (m, 2H), 1.20–1.60 (m, 2H)
[
the NMR spectrum (No. 1763) appears in the Sadtler
23
À1
collection; the b.p. is also given there as 123°C/
7
programmed to 80°C at 10°C min ). Products, which
33 mmHg].
were confirmed by GC and GC–MS comparisons to
6
authentic samples, are described above (cf. Table 1).
Cyclobutylisouronium tri¯uoromethanesulfonate &11).
This compound was prepared by the method in Ref. 7. In
a 50 ml round-bottomed flask, equipped with a stirring
Laser flash photolytic studies employed our LFP
system, which is described in detail elsewhere.
1
4
bar and protected with a CaCl tube, were placed 0.73 g
2
(
17.4 mmol) of cyanamide, 5.0 g (69.4 mmol) of cyclo-
butanol and 10 ml of dry THF. To this solution was added
.67 g (17.4 mmol) of trifluoromethanesulfonic acid. The
1
Acknowledgments
mixture was stirred magnetically at 25°C for 30 h, then
diluted with 200 ml of dry diethyl ether, sealed and
placed in the refrigerator. A light-brown oil formed,
which was separated and stored at 25°C for 1 week,
whereupon white crystals appeared. The crystals were
We are grateful to the National Science Foundation for
financial support. Access to the computational resources
of the Center for Computational Neuroscience at Rutgers
University (Newark) is much appreciated.
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 400–406

4
06
R. A. MOSS ET AL.
1
1
1
4. Moss RA, Johnson LA, Merrer DC, Lee GE Jr. J. Am. Chem. Soc.
999; 121: 5940.
5. Jackson JE, Soundararajan N, Platz MS, Liu MTH. J. Am. Chem.
Soc. 1988; 110: 5595.
6. Moss RA, Johnson LA, Yan S, Toscano JP, Showalter BM. J. Am.
Chem. Soc. 2000; 122: 11256.
REFERENCES
1
1
. (a) Lowry TH, Richardson KS. Mechanism and Theory in Organic
Chemistry (3rd edn). Harper and Row: New York, 1987; 454–463;
b) Mazur RH, White WN, Seminow DA, Lee CC, Silver MS,
(
Roberts JD. J. Am. Chem. Soc. 1959; 81: 4390; (c) Renk E, Roberts
JD. J. Am. Chem. Soc. 1961; 83: 878; (d) Richey HG Jr. In
Carbonium Ions, vol. III, Olah GA, Schleyer PvR (eds). Wiley–
Interscience: New York, 1972, 1201ff; (e) Wiberg KB, Hess BA Jr,
Ashe AJ III. In Carbonium Ions, vol. III, Olah GA, Schleyer PvR
1
1
7. Roberts DD. J. Org. Chem. 1964, 29: 294.
8. Roberts JD, Chambers VC. J. Am. Chem. Soc. 1951; 73: 5034.
19. Frisch MJ, Trucks GW, Schlegel HB, Gill PMW, Johnson BG,
Robb MA, Cheeseman JR, Keith T, Petersson GA, Montgomery
JA, Raghavachari K, Al-Laham MA, Zakrzewski VG, Ortiz JV,
Foresman JB, Cioslowski J, Stefanov BB, Nanayakkara A,
Challacombe M, Peng CY, Ayala PY, Chen W, Wong MW,
Andres JL, Replogle ES, Gomperts R, Martin RL, Fox DJ, Binkley
JS, Defrees DJ, Baker J, Stewart JP, Head-Gordon M, Gonzalez C,
Pople JA. Gaussian94 Revision B. 1. Gaussian: Pittsburgh, PA,
(
eds). Wiley–Interscience: New York, 1972, 1295ff.
2
3
4
5
6
7
. Moss RA, Shulman FC. Tetrahedron 1968; 24: 2881.
. Moss RA. Acc. Chem. Res. 1999; 32: 969.
. Moss RA. Acc. Chem. Res. 1974; 7: 421.
. Moss RA, Powell CE. J. Am. Chem. Soc. 1976; 98: 283.
. Moss RA, Ho GJ, Wilk BK. Tetrahedron Lett. 1989; 30: 2473.
. Moss RA, Kaczmarczyk GM, Johnson LA. Synth. Commun. 2000;
1
995.
3
0: 3233.
2
2
0. Becke AD. J. Chem. Phys. 1993; 98: 5648.
1. (a) Koch W, Liu B, DeFrees DJ. J. Am. Chem. Soc. 1988; 110:
8
9
. Graham WH. J. Am. Chem. Soc. 1965; 87: 4396.
. Applequist DE, McGreer DE. J. Am. Chem. Soc. 1960; 82: 1965.
0. Moss RA, Wilk BK, Hadel LM. Tetrahedron Lett. 1987; 28: 1969.
1. (a) Yan S, Sauers RR, Moss RA. Org. Lett. 1999; 1: 1603; (b)
Kesselmayer MA, Sheridan RS. J. Am. Chem. Soc. 1986; 108: 99–
7
325; (b) Saunders M, Laidig KE, Wiberg KB, Schleyer PvR. J.
1
1
Am. Chem. Soc. 1988; 110: 7652.
2
2
2. Lee CC, Cessna AJ. Can. J. Chem. 1980; 58: 1075.
3. Simons WW (ed). The Sadtler Handbook of Proton NMR Spectra,
vol. 2. Sadtler Research Laboratories: Philadelphia, PA, 1978; 760.
1
07, 844–845.
1
1
2. Zollinger H. Diazo Chemistry II. VCH: Weinheim, 1995; 241–304.
3. Moss RA, Ge C-S, Maksimovic L. J. Am. Chem. Soc. 1996; 118:
24. Ho G-J. PhD Dissertation, Rutgers University, New Brunswick,
NJ, 1992.
9
792.
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 400–406