Photochemistry of 1-isopropylcycloalkyl aryl ketones: Ring size effects, medium effects, and asymmetric induction

Article abstract of DOI:10.1021/ol050104k

(Chemical Equation Presented) The n = 0, 1, and 2 ketones shown above undergo Yang photocyclization in solution, but only the n = 1 analogues react this way in the solid state. Based on X-ray crystallography, these differences in reactivity are attributed to an unusually large distance for 1,4-hydroxybiradical cyclization in the solid state for the n = 0 and 2 ring systems, which leads to predominant reverse hydrogen transfer (rht). Enantiomeric excesses of up to 99% can be achieved in the case of the n = 1 system through the use of the solid-state ionic chiral auxiliary method of asymmetric synthesis.

Full text of DOI:10.1021/ol050104k

ORGANIC
LETTERS
2005
Vol. 7, No. 7
1315-1318
Photochemistry of 1-Isopropylcycloalkyl
Aryl Ketones: Ring Size Effects,
Medium Effects, and Asymmetric
Induction
Wujiong Xia,† John R. Scheffer,*,† Mark Botoshansky,‡ and Menahem Kaftory‡
Department of Chemistry, UniVersity of British Columbia, 2036 Main Mall,
VancouVer, B.C., Canada V6T 1Z1, and the Department of Chemistry, Technion,
Haifa 32000, Israel
scheffer@chem.ubc.ca
Received January 18, 2005
ABSTRACT
The n
)
0, 1, and 2 ketones shown above undergo Yang photocyclization in solution, but only the n
)
1 analogues react this way in the solid
state. Based on X-ray crystallography, these differences in reactivity are attributed to an unusually large distance for 1,4-hydroxybiradical
cyclization in the solid state for the n 0 and 2 ring systems, which leads to predominant reverse hydrogen transfer (rht). Enantiomeric
)
excesses of up to 99% can be achieved in the case of the n
)
1 system through the use of the solid-state ionic chiral auxiliary method of
asymmetric synthesis.
In an important 1974 publication, Lewis, Johnson, and
Johnson reported the solution-phase photochemistry of
1-methylcyclopentyl phenyl ketone (1, n ) 0) and 1-meth-
ylcyclohexyl phenyl ketone (1, n ) 1).1 In solution, both
compounds were shown to undergo Yang photocyclization2
via their axial and pseudoaxial conformers (1ax) to afford
the corresponding cyclobutanols (2, n ) 0, 1) (Scheme 1).
These reactions, which convert achiral reactants into chiral
products, appeared to be ideal for studying asymmetric
induction in organic photochemistry, and motivated by the
paucity of general methods of asymmetric photochemical
synthesis,3 we embarked on a program designed to achieve
this goal.
Scheme 1
† University of British Columbia.
‡ Technion.
The plan was to use the solid-state ionic chiral auxiliary
method of asymmetric synthesis developed in our laboratory
over the past several years.4 In this procedure, the reactant
is equipped with a carboxylic acid substituent to which an
optically pure ammonium ion (the ionic chiral auxiliary) is
attached by means of a salt bridge. Such salts are required
(1) Lewis, F. D.; Johnson, R. W.; Johnson, D. E. J. Am. Chem. Soc.
1974, 96, 6090.
(2) Yang cyclization refers to cyclobutanol formation in the Norrish type
II reaction and was first reported by: Yang, N. C.; Yang, D. H. J. Am.
Chem. Soc. 1958, 80, 1958. For a recent review, see: Wagner, P. J. In
CRC Handbook of Organic Photochemistry and Photobiology, 2nd ed.;
Horspool, W., Lenci, F., Eds.; CRC Press: Boca Raton, 2004; Chapter 58-
1.
10.1021/ol050104k CCC: $30.25
© 2005 American Chemical Society
Published on Web 03/11/2005

to crystallize in chiral space groups, which provide the
asymmetric environment responsible for chiral induction
when the compounds are photolyzed in the solid state.
A second requirement for the success of the solid-state
ionic chiral auxiliary method of asymmetric synthesis is that
the reactant must crystallize in a conformation favorable for
reaction. For photochemical reactions involving intramo-
lecular hydrogen atom abstraction, this requires a conforma-
tion in which the carbonyl oxygen is within approximately
2.7 ( 0.2 Å of a γ-hydrogen atom.5,6 This presented a
problem in the case of ketones 1 (n ) 0, 1) because the
photochemically reactive conformers 1ax are not the mini-
mum energy conformers1 and are therefore not likely to be
present in the crystalline state.7 Molecular mechanics cal-
culations, however, indicated that a simple change of the
R-methyl group in ketones 1 (n ) 0, 1) for the bulkier
isopropyl group would favor conformers having axial or
pseudoaxial ketone substituents, and for this reason, the
compounds chosen for study were the R-isopropyl-containing
ketones 3 (n ) 0, 1) as well as the corresponding seven-
membered ring homologue (n ) 2) (Scheme 2).8 In this
The required starting materials were synthesized by LDA-
induced alkylation of methyl cyclopentane-, cyclohexane-,
and cycloheptanecarboxylate with 2-bromopropane followed
by appropriate functional group modification of the ester
substituent. Prior to the solid-state studies, the photochemistry
of keto esters 3b (n ) 0, 1, 2) was investigated in solution.
In each case, irradiation through Pyrex in acetonitrile
produced the corresponding cyclobutanol (4b, n ) 0, 1, 2)
in excellent chemical yield (>95% by GC). These photo-
products were isolated by column chromatography and fully
1
characterized by 1D H, 1D NOE and 13C (BB and APT)
NMR, 2D 1H-1H COSY, 1H-13C HMQC and HMBC NMR,
FT-IR, HRMS, as well as by elemental analysis. The
cyclobutanols were found to have the same general structure
and stereochemistry as those reported by Lewis et al. for
cyclobutanols 2 (n ) 0, 1).1
Turning now to the solid-state studies, irradiation of
crystals of keto ester 3b (n ) 1) led exclusively to
cyclobutanol 4b (n ) 1), the same product formed in
acetonitrile. In the case of the n ) 0 and 2 keto esters,
however, there was a striking difference between the results
obtained in solution and those observed in the crystalline
state. No products were detectable by GC following pho-
tolysis of crystals of keto ester 3b (n ) 0), and solid-state
photolysis of keto ester 3b (n ) 2) led to the formation of
Norrish type I products 5b, 6, and 7 (Scheme 2).9 In neither
case were any cyclobutanol-type photoproducts formed.
Additional solid-state photochemical studies were carried
out on salts prepared from keto acids 3a (n ) 0, 1, 2) and a
variety of optically pure primary and secondary amines. The
procedure consisted of crushing ca. 5 mg of each salt between
two Pyrex microscope slides, taping the plates together,
sealing the resulting sandwiches under nitrogen in polyeth-
ylene bags, and irradiating the ensembles on both sides for
varying lengths of time by using a 450 W medium-pressure
mercury lamp. Following photolysis, the samples were
treated with ethereal diazomethane to form the corresponding
methyl esters and subjected to GC and chiral HPLC analysis.
The results were identical with those obtained when keto
esters 3b (n ) 0, 1, 2) were photolyzed in the crystalline
state, namely, no observable reaction for the 5-ring salts,
formation of cyclobutanol 4b in the case of the 6-ring salts,
and Norrish type I cleavage for the 7-ring salts.
Scheme 2
Table 1 shows the enantiomeric excess (ee) in which
cyclobutanol 4b (n ) 1) was formed in the solid-state
photolysis of salts of keto acid 3a (n ) 1). A total of 10
salts was investigated, of which four gave ee values greater
than 90%. Strikingly, two of these (the (R)-(-)-1-aminoindan
and the (S)-(-)-1-p-tolylethylamine salts), led to essentially
paper, we report the contrasting solution-phase and solid-
state photochemical behavior of these compounds as well
as highly successful asymmetric induction in the solid-state
photochemistry of salts of ketone 3a (n ) 1).
(3) For recent reviews dealing with asymmetric induction in organic
photochemistry, see: (a) Griesbeck, A. G.; Meierhenrich, U. J. Angew.
Chem., Int. Ed. 2002, 41, 3147. (b) Toda, F.; Tanaka, K.; Miyamoto, H. In
Molecular and Supramolecular Photochemistry; Ramamurthy, V., Schanze,
K. S., Eds.; Marcel Dekker: New York, 2001; Volume 8; Chapter 6. (c)
Everitt, S. R. L.; Inoue, Y. In Molecular and Supramolecular Photochem-
istry; Ramamurthy, V., Schanze, K. S., Eds.; Marcel Dekker: New York,
1999; Vol. 3, Chapter 2. (d) Feringa, B. L.; van Delden, R. A. Angew.
Chem., Int. Ed. 1999, 38, 3419.
(6) A third general requirement for successful solid-state reaction is that
the motions involved be compatible with the surrounding crystal lattice.
For a quantitative treatment of this aspect of solid-state photochemistry,
see: Zimmerman, H. E.; Nesterov, E. E. Acc. Chem. Res. 2002, 35, 77.
(7) Supporting this contention, ketone 1 (n ) 1) was found to be
photochemically unreactive in the solid state, presumably because it
crystallizes in conformation 1eq.
(4) Reviews: (a) Scheffer, J. R. Can. J. Chem. 2001, 79, 349. (b)
Scheffer, J. R. In Molecular and Supramolecular Photochemistry; Rama-
murthy, V., Schanze, K. S., Eds.; Marcel Dekker: New York, 2004; Vol.
11, pp 463-483.
(8) Competitive abstraction of a primary γ-hydrogen atom from the
isopropyl group was not expected to be a problem, and this proved to be
the case.
(9) Methyl terephthalate (5b) is presumably formed by air oxidation of
the corresponding aldehyde during workup.
(5) Ihmels, H.; Scheffer, J. R. Tetrahedron 1999, 55, 885.
1316
Org. Lett., Vol. 7, No. 7, 2005

Table 1. Enantiomeric Excess of Photoproduct 4b (n ) 1)
from Solid-State Photolysis of Chiral Salts of Keto Acid 3a (n
) 1)
amine
T (°C) conv (%) ee (%) [R]a
(R)-(-)-1-aminoindan
rt
rt
rt
rt
rt
rt
rt
-20
rt
-20
rt
-20
rt
-20
rt
-20
rt
-20
rt
32
71
100
43
100
81
99
100
95
63
63
56
95
100
75
61
98
100
66
100
100
99
99
99
99
99
98
91
98
94
98
41
49
13
53
11
28
5
+
(S)-(-)-1-p-tolylethylamine
(1S,2S)-(+)-pseudoephedrine
+
-
(1S,2S)-(+)-2-amino-3-methoxy-
1-phenyl-1-propanol
L-prolinamide
+
b
b
b
b
b
b
(1R,2R)-(-)-2-amino-1-phenyl-
1,3-propanediol
(S)-(+)-2-pyrrolidinemethanol
(R)-(-)-1-cyclohexylethylamine
7
11
6
(S)-(-)-N-methyl-1-phenyl-
ethylamine
(S)-(-)-1-phenylethylamine
rt
-20
1
a Sign of rotation at the sodium D line. b Not measured.
quantitative ee values at 100% conversion, a result that
illustrates the synthetic utility of the method.
With the photochemical results in hand, we next turned
to X-ray crystallography to help elucidate the observed solid
state/solution phase reactivity differences as well as to
provide a rationale for the asymmetric induction results.
Successful X-ray crystal structure determinations were car-
ried out for keto ester 3b (n ) 0), the 1-p-tolylethylammo-
nium salts of keto acids 3a (n ) 1) and 3a (n ) 2), and the
1-phenylethylammonium salt of keto acid 3a (n ) 1).10 In
accord with the molecular mechanics based conformational
analysis of these systems discussed above, in each case there
was a CdO‚‚‚Hγ contact within the 2.7 ( 0.2 Å range
established for successful γ-hydrogen atom abstraction.5 A
significant finding was that the carboxylate anion of the
1-phenylethylammonium salt of keto acid 3a (n ) 1)
contained equal amounts of two independent, mirror image-
related conformers in the crystal lattice, a phenomenon we
have termed conformational enantiomerism.11 For the other
three compounds, the crystal contained a single conforma-
tional enantiomer of the reactant. An ORTEP drawing of
each solid-state conformer is given in Figure 1.
Figure 1. Solid-state conformations of (a) keto ester 3b (n ) 0),
(b) 1-p-tolylethylammonium salt of keto acid 3a (n ) 1), (c)
1-phenylethylammonium salt of keto acid 3a (n ) 1), (d) confor-
mational enantiomer of salt (c), (e) 1-p-tolylethylammonium salt
of keto acid 3a (n ) 2). The letter d indicates the closest 01‚‚‚Hγ
abstraction distance, and D indicates the C1‚‚‚C4 cyclization
distance.
Conformational enantiomerism is undoubtedly the source
of the low ee resulting from irradiation of the (S)-(-)-1-
phenylethylammonium salt of keto acid 3a (n ) 1) in the
solid state (last entry, Table 1); half the molecules in the
crystal react to give one enantiomer of cyclobutanol 4b while
the other half lead to its optical antipode. The mechanism
involves selective abstraction of the nearer γ-hydrogen
atom,12 leading to a 1,4-hydroxybiradical that closes to a
single enantiomer of cyclobutanol 4b.13 A perfect racemate
(10) CCDC 260763-260766 contain the supplementary crystallographic
data for this paper. These data can be obtained free of charge from the
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif.
(11) (a) Cheung, E.; Kang, T.; Netherton, M. R.; Scheffer, J. R.; Trotter,
J. J. Am. Chem. Soc. 2000, 122, 11753. (b) Natarajan, A.; Wang, K.;
Ramamurthy, V.; Scheffer, J. R.; Patrick, B. Org. Lett. 2002, 4, 1443. We
note that such conformational enantiomers can also be considered as
conformational diastereomers when the presence of the optically pure
ammonium ion is taken into account.
(12) See the Supporting Information for a full listing of all hydrogen
abstraction distances and geometries.
Org. Lett., Vol. 7, No. 7, 2005
1317

is not formed because the two conformers are not exact
mirror images of one another. The abstraction distances are
slightly different (2.54 and 2.61 Å), and for this reason the
independent, mirror image-related conformers react at slightly
different rates. This interpretation is supported by the results
observed in the case of the (S)-(-)-p-tolylethylamine salt of
keto acid 3a (n ) 1) (second entry, Table 1). Here the crystal
contains a single conformational enantiomer of the carbox-
ylate anion, and all of the molecules in the crystal react to
form the same enantiomer of cyclobutanol 4b, thereby
leading to a high ee. These general reactivity patterns have
been found to apply to a number of other Yang photocy-
clization reactions studied by the solid-state ionic chiral
auxiliary method.4
Support for this interpretation comes from published and
unpublished work from our group on compounds that
unexpectedly failed to undergo Norrish/Yang type II pho-
tochemistry in the crystalline state despite having favorable
hydrogen atom abstraction geometries. The data are presented
graphically in Figure 2, and it is clear that the “unreactive”
Finally, we turn to a discussion of the possible reasons
for the unusual solid-state photochemical behavior of the
cyclopentyl and cycloheptyl systems. One possible explana-
tion for the fact that these compounds do not form cyclobu-
tanol photoproducts in the crystalline state is that hydrogen
atom abstraction fails for some reason. This is unlikely,
however, because the abstraction distances and geometries
in these cases are as good as or better than those of dozens
of compounds known to react in the solid state.14
We propose instead that cyclobutanol formation is not
observed in these cases because the cyclization distance, D
(the distance between the radical centers in the 1,4-hydroxy-
biradical), is too great. Making the reasonable assumption
that hydrogen abstraction occurs with minimum movement
of the associated heavy atoms,15 the value of D can be
approximated by the distance between the carbonyl carbon
and the γ-carbon in the ground state. From X-ray crystal-
lography, these distances are 3.24 and 3.26 Å for the
cyclopentyl and cycloheptyl compounds and 3.07 Å for the
cyclohexyl homologue. With hydrogen atom transfer known
to be reversible, both in solution16 and the solid state,15 these
numbers suggest that when D is greater than approximately
3.2 Å, the rate of cyclization is slowed to the point that it
cannot compete with reverse hydrogen transfer. As a result,
the cyclopentyl systems appear to be photochemically inert
and the cycloheptyl systems undergo competitive Norrish
type I photochemistry.17
Figure 2. Values of D for reactive and unreactive ketones.
systems have D values centered around 3.2 Å (3.2 ( 0.05
Å) while the reactive compounds have D values in the range
of 3.0 ( 0.09 Å.18 The fact that the cyclopentyl and
cycloheptyl systems undergo smooth cyclobutanol formation
in solution emphasizes the point that the 3.2 Å upper limit
to biradical cyclization applies only in the crystalline state.
In solution, where the biradicals have sufficient lifetime to
explore alternative conformations with reduced D values,
cyclization can once again become competitive with reverse
hydrogen transfer. Work is ongoing in our laboratory to
provide further experimental evidence on the validity of this
hypothesis.
Acknowledgment. We thank the Natural Sciences and
Engineering Research Council of Canada (NSERC) for
financial support.
Supporting Information Available: Synthesis of starting
ketones, photolysis procedures, characterization of photo-
products, table of hydrogen abstraction distances and ge-
ometries. This material is available free of charge via the
Internet at http://pubs.acs.org.
(13) The diastereoselectivity of this reaction is the result of least motion
closure with “retention” at the carbonyl carbon, a general feature of Yang
photocyclization reactions in the crystalline state.4
(14) Scheffer, J. R. In Molecular and Supramolecular Photochemistry:
Chiral Photochemistry; Ramamurthy, V., Inoue, Y., Eds.; Marcel Dekker:
New York, 2004; Vol. 11, pp 463-483.
OL050104K
(17) Still unresolved is the question of why the cycloheptyl systems
undergo type I photochemistry in the solid state whereas the cyclopentyl
systems do not.
(18) The compounds from which the data in Figure 2 are taken are given
in the Supporting Information.
(15) Chen, S.; Cheung, E.; Filson, H.; Netherton, M. R.; Patrick, B. O.;
Scheffer, J. R.; Scott, C.; Xia, W.; Braga, D.; Maini, L. J. Am. Chem. Soc.
2004, 126, 3511.
(16) Wagner, P. J. J. Am. Chem. Soc. 1967, 89, 5898.
1318
Org. Lett., Vol. 7, No. 7, 2005