Fragmentation of alkoxychlorocarbenes in the gas phase

Article abstract of DOI:10.1016/S0040-4039(02)00438-0

In contrast to photolysis or thermal decomposition in solution, which is dominated by ionic reactions, flash vacuum pyrolysis of alkylchlorodiazirines in the gas phase generates radicals. The cyclopropylcarbinyl system is re-examined and the l-norbornylca

Full text of DOI:10.1016/S0040-4039(02)00438-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 3069–3071
Fragmentation of alkoxychlorocarbenes in the gas phase
Michael E. Blake,a Maitland Jones, Jr.,a,* Fengmei Zhengb and Robert A. Mossb,*
aDepartment of Chemistry, Princeton University, Princeton, NJ 08544, USA
bDepartment of Chemistry, Rutgers University, New Brunswick, NJ 08903, USA
Received 30 January 2002; accepted 15 February 2002
Abstract—In contrast to photolysis or thermal decomposition in solution, which is dominated by ionic reactions, flash vacuum
pyrolysis of alkylchlorodiazirines in the gas phase generates radicals. The cyclopropylcarbinyl system is re-examined and the
1-norbornylcarbinyl system is studied for the first time. © 2002 Elsevier Science Ltd. All rights reserved.
the products from ring expansion of 1-bicyclo-
[2.2.1]carbinyl compounds.5,6
Generation of alkoxyhalocarbenes in polar solvents
such as acetonitrile or ethanol,1a and even pentane,1b
leads to formation of ion pairs and the ultimate col-
lapse to products of retained and rearranged structure.
Two of the classic intermediates of physical organic
chemistry, the cyclopropylcarbinyl cation,2 and the 2-
norbornyl cation,3 have been recently revisited through
this reaction. Here we describe another classic ion, the
1-norbornylcarbinyl cation, and show that the mecha-
nism of carbene fragmentation strongly depends on the
medium; in the gas phase radicals, not ion pairs, are the
important intermediates.
The small amount of 2 presumably derives from a
bridged ion, or, more likely, from a minor amount of
displacement by chloride at the neopentyl carbon of 1.
In any event, the amount of retained product 2 is small.
When 1 is decomposed by flash vacuum pyrolysis
(FVP) at 500°C, the product mixture changes sharply.
The following figure gives normalized percentages;
there is about 15% of other products in the pyrolysis,
mostly formates and alcohols. Although we have not
been able to find a study of the 1-norbornylmethyl
radical, the pyrolysis product ratios are consistent with
a radical process in which the initially formed 5 is
captured by a chlorine atom before substantial rear-
rangement. The observation of substantial retention of
the bicyclo[2.2.1] skeleton is in keeping with the well-
known lack of an authentic 1,2-hydrogen or alkyl shift
in simple radicals.7
In accord with this scheme, when diazirine 1 is pho-
tolyzed in dichloroethane (DCE) at room temperature,
the products are the three chlorides, 2, 3, and 4. The
kinetic product of solvolysis of 1-norbornylcarbinyl
compounds is the bicyclo[2.2.2] system;4 thus, presence
of 3 in substantial amounts argues for some measure of
thermodynamic control in this reaction. Under thermo-
dynamic conditions the bicyclo[3.2.1] system dominates
* Corresponding authors.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00438-0

3070
M. E. Blake et al. / Tetrahedron Letters 43 (2002) 3069–3071
Where do rearranged compounds 3 and 4 come from?
We do not know for certain, but a likely possibility is a
measure of concerted migration during fragmentation.
Such a pathway is likely during the fragmentation of
propylcarbinyl radical generated by photolysis of the
diacyl peroxide, Sheldon and Kochi found about 60%
of rearranged materials.10 Our initial disappointment at
the near-absence of 9 was assuaged by the observation
that a fourth major product, 10 appeared in the
product mixture.
1-adamantylmethoxychlorocarbene,
which
affords
homoadamantyl chloride.8 Alternatively, 500°C may be
high enough to overcome the barrier to 1,2 alkyl
migration.
Compound 10, 1,4-dichlorobutane, is the product of
the radical addition of chlorine atoms to 9. Were
cations somehow involved in the formation of 10, the
product would have been 1,3-dichlorobutane, which is
not formed in any detectable amount. Thus, the forma-
tion of 10 becomes a lucky confirmation that the FVP
reaction is radical in nature. Even though the cyclo-
propyl-carbinyl radical and its partner chlorine atom
must be formed in proximity to each other, there is time
for ring opening before capture. The products are sta-
ble under the reaction conditions.
We have put these tentative mechanistic suggestions to
the test by re-examining the cyclopropylcarbinyl system
6, and comparing the FVP results to the known decom-
position in solution at room temperature.2 Moss, et al.
found the results shown above from either photolysis or
pyrolysis of 6 (25°C) in dichloroethane2 or pentane.1b
These results resemble closely the product spectrum
under conditions in which ion pairs are important
intermediates (7:8:9ꢀ79:14:7).2
It would seem that one has a choice in the decomposi-
tion of alkoxychlorodiazirines: reaction in solution
leads to ionic intermediates, whereas decomposition in
the gas phase generates radicals.
We anticipated that the product spectrum from FVP of
6 would resemble that from the cyclopropylcarbinyl
radical, not the cation, and be relatively rich in the
ring-opened minor product, 9.9 In a study of the cyclo-

M. E. Blake et al. / Tetrahedron Letters 43 (2002) 3069–3071
3071
Acknowledgements
4. Wilt, J. W.; Schneider, C. A.; Dabek, H. F., Jr.; Krae-
mer, J. J.; Wagner, W. J. J. Org. Chem. 1966, 31, 1543.
5. Sauers, R. R.; Ahlstrom, D. H. J. Org. Chem. 1967, 32,
2233.
6. For reviews see: (a) Walling, C. Molecular Rearrange-
ments; de Mayo, P., Ed.; Wiley: New York, 1963; Chap-
ter 7; (b) Wilt, J. W. Free Radicals; Kochi, J. K., Ed.
Wiley: New York, 1973; Chapter 8.
This work was supported at Princeton by the National
Science Foundation through Grant CHE-0073373 and
at Rutgers by the National Science Foundation through
Grant CHE-0091368.
7. Ref. 6b, p. 340ff.
References
8. (a) Moss, R. A.; Ge, C.-S.; Maksimovic, L. J. Am. Chem.
Soc. 1996, 118, 9792; (b) Moss, R. A.; Johnson, L. A.;
Yan, S.; Toscano, J. P.; Showalter, B. M. J. Am. Chem.
Soc. 2000, 122, 11256.
1. (a) Moss, R. A. Acc. Chem. Res. 1999, 32, 969; (b) Moss,
R. A.; Zheng, F., unpublished.
2. Moss, R. A.; Zheng, F.; Johnson, L. A.; Sauers, R. R. J.
9. Nonhebel, D. C. Chem. Soc. Rev. 1993, 347.
10. Sheldon, R. A.; Kochi, J. K. J. Am. Chem. Soc. 1970, 92,
4395.
Phys. Org. Chem. 2001, 14, 400.
3. Moss, R. A.; Zheng, F.; Sauers, R. R.; Toscano, J. P. J.
Am. Chem. Soc. 2001, 123, 8109.