Stereoselective synthesis of 2-aminocyclobutanols via photocyclization of α-amido alkylaryl ketones: Mechanistic implications for the Norrish/Yang reaction

Article abstract of DOI:10.1021/ja0111941

A series of chiral N-acylated α-amino p-methylbutyrophenone derivatives 1a-1h was synthesized from α-amino acids via a three-step procedure. These substrates were photolyzed in benzene and gave Norrish II and Norrish I cleavage products as well as the N-acylated 2-aminocyclobutanols that derive from γ-hydrogen abstraction and 1,4-triplet biradical combination (Yang cyclization). The products were formed with characteristic Yang/cleavage ratios. The quantum yields for the photodecomposition of the N-acetyl-protected substrates 1b,e,f were moderate (12-26%); the diastereoselectivities of the cyclobutanol formation were remarkably high for all substrates. High diastereospecificity was observed for the isoleucine derivatives (2S,3S)-1g and allo-(2S,3R)-1g; the Yang reaction dominated the photochemistry of allo-1g, whereas 1g gave preferentially Norrish II cleavage. The role of hydrogen bonding as one of the stereo-directing effects was demonstrated by comparison of the cyclization efficiency of the valine derivative 1e with 1h,i,j. Also, aromatic β-keto esters gave the Yang cyclization products in low yields. The diastereoselectivity of the cyclobutanol formation was rationalized by a three-step mechanism where every step is connected with one distinct stereochemical induction mechanism: (a) diastereoselective hydrogen abstraction, (b) conformational equilibration of the 1,4-tetramethylene biradicals (as calculated by semiempiric methods) controlled by hydrogen bonding, and (c) diastereoselective biradical combination (versus cleavage) influenced by spin-orbit coupling controlled intersystem crossing geometries.

Full text of DOI:10.1021/ja0111941

Published on Web 12/20/2001
Stereoselective Synthesis of 2-Aminocyclobutanols via
Photocyclization of r-Amido Alkylaryl Ketones: Mechanistic
Implications for the Norrish/Yang Reaction
Axel G. Griesbeck* and Heike Heckroth
Contribution from the Institute of Organic Chemistry, Greinstrasse 4, D-50939 Ko¨ln, Germany
Received May 14, 2001
Abstract: A series of chiral N-acylated R-amino p-methylbutyrophenone derivatives 1a-1h was synthesized
from R-amino acids via a three-step procedure. These substrates were photolyzed in benzene and gave
Norrish II and Norrish I cleavage products as well as the N-acylated 2-aminocyclobutanols that derive from
γ-hydrogen abstraction and 1,4-triplet biradical combination (Yang cyclization). The products were formed
with characteristic Yang/cleavage ratios. The quantum yields for the photodecomposition of the N-acetyl-
protected substrates 1b,e,f were moderate (12-26%); the diastereoselectivities of the cyclobutanol formation
were remarkably high for all substrates. High diastereospecificity was observed for the isoleucine derivatives
(2S,3S)-1g and allo-(2S,3R)-1g; the Yang reaction dominated the photochemistry of allo-1g, whereas 1g
gave preferentially Norrish II cleavage. The role of hydrogen bonding as one of the stereo-
directing effects was demonstrated by comparison of the cyclization efficiency of the valine derivative 1e
with 1h,i,j. Also, aromatic â-keto esters gave the Yang cyclization products in low yields. The diastereo-
selectivity of the cyclobutanol formation was rationalized by a three-step mechanism where every step is
connected with one distinct stereochemical induction mechanism: (a) diastereoselective hydrogen
abstraction, (b) conformational equilibration of the 1,4-tetramethylene biradicals (as calculated by semiempiric
methods) controlled by hydrogen bonding, and (c) diastereoselective biradical combination (versus cleavage)
influenced by spin-orbit coupling controlled intersystem crossing geometries.
Introduction
â-CH positions can be triggered either by exchange of the γ-CH
carbon by a heteroatom or another functional group6 or simply
Photochemical C-H-abstraction reactions initiated by elec-
tronically excited triplet carbonyl species are one of the
archetype reactions in organic photochemistry. The spin-
isomeric singlet excited states also undergo Norrish II reactions
with hydrogen-transfer reactions in less than 100 fs and
secondary processes on the subpicosecond time scale.1 The time
domain for triplet reactions is shifted into the nanosecond
region,2 and therefore spectral and kinetic information concern-
ing the excited state reactivity and the structures of the
intermediate 1,4-triplet biradical (tetramethylenes) have been
available for more than two decades.3,4
The following rules of thumb are appropriate for a multitude
of intramolecular C-H-abstraction (Norrish type II) reactions
initiated by electronically excited triplet carbonyls:5 (a) Optimal
hydrogen abstraction geometry in flexible substrates is achieved
in a six-membered H-transfer geometry, thus implying a strong
preference for γ-hydrogen-activation. Less accessible δ- or
by peralkylation of the γ-carbon.7 The Scheffer group has
intensively investigated the geometrical parameters which guide
efficient versus nonefficient hydrogen transfer and reported the
bandwidth of the critical O‚‚‚‚‚HCγ distance and the bond angles
involved.8,9 These data originated from profound solid-state
investigations which were recently also modified to give highly
enantioselective photochemistry using Norrish II reactions
followed by Yang10 cyclization.11-13 A similar study was
conducted by us for the liquid-phase hydrogen-transfer reactions
of N-alkylated phthalimides.14 (b) There are three competing
pathways following the primary hydrogen transfer from the
γ-CH position: cyclization of the biradical to give the cyclo-
butanol (Yang reaction), cleavage of the central C2-C3 single
(6) For aspartic acid derived γ-ketoamides: Griesbeck, A. G.; Heckroth, H.;
Schmickler, H. Tetrahedron Lett. 1999, 40, 3137-3140.
(7) For N-phthaloyl tert-leucine derivatives: Griesbeck, A. G.; Mauder, H.;
Mu¨ller, I. Chem. Ber. 1992, 125, 2467-2475.
(8) Scheffer, J. R.; Ihmels, H. Tetrahedron 1999, 55, 884-907.
(9) Leibovitch, M.; Olovsson, G.; Scheffer, J. R.; Trotter, J. J. Am. Chem.
Soc. 1998, 120, 12755-12769.
* To whom correspondence should be addressed. Fax: 0049-221-470
5057. E-mail: griesbeck@uni-koeln.de.
(1) De Feyter, S.; Diau, E. W.-G.; Zewail, A. H. Angew. Chem. 2000, 112,
266-269; Angew. Chem., Int. Ed. Engl. 2000, 39, 260-263.
(2) Scaiano, J. C. Acc. Chem. Res. 1982, 15, 252-258.
(3) Johnston, L. J.; Scaiano, J. C. Chem. ReV. 1989, 89, 521-547.
(4) Caldwell, R. A. In Kinetics and Spectroscopy of Carbenes and Biradicals;
Platz, M. S., Ed.; Plenum Press: New York, 1989; pp 77-116.
(5) For the CH-transfer chemistry of electronically excited carbonyl compounds
in general, see: Wagner, P. J.; Park, B.-S. Org. Photochem. 1991, 11, 227-
366.
(10) Yang, N. C.; Yang, D.-H. J. Am. Chem. Soc. 1958, 80, 2913.
(11) Cheung, E.; Kang, T.; Netherton, M. R.; Scheffer, J. R.; Trotter, T. J. Am.
Chem. Soc. 2000, 122, 11753-11754.
(12) Cheung, E.; Netherton, M. R.; Scheffer, J. R.; Trotter, J.; Zenova, A.
Tetrahedron Lett. 2000, 41, 9673-9677.
(13) Cheung, E.; Kang, T.; Raymond, J. R.; Scheffer, J. R.; Trotter, J.
Tetrahedron Lett. 1999, 40, 8729-8732.
(14) Griesbeck, A. G.; Henz, A.; Kramer, W.; Wamser, P.; Peters, K.; Peters,
E.-M. Tetrahedron Lett. 1998, 39, 1549-1552.
9
396 VOL. 124, NO. 3, 2002 J. AM. CHEM. SOC.
10.1021/ja0111941 CCC: $22.00 © 2002 American Chemical Society

Stereoselective Synthesis of 2-Aminocyclobutanols
A R T I C L E S
Scheme 1. Three-Step Process of the Norrish II/Yang Reaction
Scheme 2. Synthesis of Starting Materials
bond to give an enol and the corresponding alkene (Norrish II
cleavage), and hydrogen back-transfer to reconstitute the starting
material in its electronic ground state.15 In all of these cases,
the spin conservation rules require triplet/singlet intersystem
crossing (ISC) because the formation of triplet excited products
is energetically improbable. (c) In singlet photochemical reac-
tions, stereoselectivity of C-C bond-forming steps is often
controlled by the optimal geometries for radical-radical
combinations, whereas in triplet photoreactions, the geometries
most favorable for ISC are considered to be of similar
relevance.16 These geometries can be quite different from the
former ones due to differences in spin-orbit coupling (SOC)
values. We have described this phenomenon for [2 + 2]-pho-
tocycloadditions of carbonyl compounds with cycloalkenes17,18
and allylic alcohols.19 The “90° rule” can be used as a useful
approximation to estimate the reactive conformers.20 After
transition from the triplet to the singlet potential energy surface,
immediate product formation is expected. Thus, the ISC is
expected to proceed concerted with the formation of a new bond
or the cleavage of the primarily formed single bond (Scheme
1). Concerning the chemoselectivity (cyclization versus frag-
mentation) as well as the stereochemistry of the cyclobutanol
formation, the processes a-c were expected to disclose different
influences.
Scheme 3. Photochemistry of 1a-1l in Benzene
Results and Discussion
To “separate” these effects and to investigate similar com-
pounds with small but defined changes in substrate and
intermediate structures, we conceptualized a series of R-branched
butyrophenone derivatives. In line with our recent investigations
on the photochemistry of N-activated R-amino acids,21-23 we
transformed N-acylated R-amino acids into C-activated R-amino
p-methylbutyrophenone derivatives 1 and studied their photo-
chemistry.24 Herein we report full details of this study and a
comparison with modified substrates and â-keto esters 4.
The substrates 1a-1f and 1h-1l were synthesized by a two-
step one-pot procedure from N-acyl R-amino acids via the
corresponding acyl chlorides. Because of the harsh reaction
conditions (PCl5 chlorination), racemic material was produced
in all cases. This was easily recognized from the reaction of
isoleucine where a 1:1 mixture of 1g and its diastereoisomer
allo-1g was produced. To obtain the pure diastereoisomers, a
milder method using oxalyl chloride for the primary chlorination
step was applied for the synthesis of these two derivatives
(Scheme 2).
(15) In the presence of leaving groups at the R-position, 1-acylated 1,3-biradicals
can be generated; for a recent application for the synthesis of cyclopropanes,
see: Wessig, P.; Mu¨hling, O. Angew. Chem. 2001, 113, 1099-1101; Angew.
Chem., Int. Ed. 2001, 40, 1064-1066.
Photochemical reactions were studied by irradiating 4 mM
solutions in benzene with phosphor-coated mercury low-pressure
lamps emitting at λ ) 350 ( 20 nm under a nitrogen atmosphere
at 15 °C. The crude reaction mixtures were analyzed by NMR
as well as gas chromatography (GC) and subsequently separated
by flash chromatography. The cyclobutanols 2 were isolated
(when formed) from the slow eluting fractions. In all cases
except for the proline derivative 1l, the acetylamino
acetophenone 3 was detected as the Norrish II fragmentation
product (Scheme 3). The corresponding alkenes were detected
by GC. The degree of conversion was >95% for all substrates
1, with a mass balance after chromatography of 75-80%
(16) Griesbeck, A. G.; Fiege, M.; Bondock, S.; Gudipati, M. S. Org. Lett. 2000,
2, 3623-3625.
(17) Griesbeck, A. G.; Mauder, H.; Stadtmu¨ller, S. Acc. Chem. Res. 1994, 27,
70-76.
(18) Griesbeck, A. G.; Buhr, S.; Fiege, M.; Schmickler, H.; Lex, J. J. Org. Chem.
1998, 63, 3847-3854.
(19) Griesbeck, A. G.; Bondock, S. J. Am. Chem. Soc. 2001, 123, 6191-6192.
(20) Carlacci, L.; Doubleday, C., Jr.; Furlani, T. R.; King, H. F.; McIver, J. W.
J. Am. Chem. Soc. 1987, 109, 5323-5329.
(21) Griesbeck, A. G. Chimia 1998, 52, 272-283.
(22) Griesbeck, A. G. Liebigs Ann. Chem. 1996, 1951-1958.
(23) Griesbeck, A. G.; Kramer, W.; Oelgemo¨ller, M. Synlett 1999, 1169-1178.
(24) Preliminary communication: Griesbeck, A. G.; Heckroth, H.; Lex, J. J.
Chem. Soc., Chem. Commun. 1999, 1109-1110.
9
J. AM. CHEM. SOC. VOL. 124, NO. 3, 2002 397

A R T I C L E S
Griesbeck and Heckroth
Scheme 4. Three-Step Reaction Protocol of the tert-Leucine
Table 1. Product Composition from Substrates 1a-1ka
Derivative 1f
Norrish II
cleavage (3)
cyclization/
fragmentation-ratio
Norrish I
cleavage
substrate
Yang (2)
1a
1b
1c
1d
1e
1f
allo-1g
1g
1h
1i
1j
1k
33
40
35
35
8
0
0.8
1.4
1.4
9.3
1.7
>95
0
67
22
12
15
18
20
<5
<5
22
5
33
48
49
74
50
30
>95
>95
13
30
65
65
5.0
2.2
0
20
40b
group. This accounts for the assumption that after primary
hydrogen transfer from the γ-CH position, the skew triplet
biradical rearranges to a hydrogen-bonded structure that is
conformationally flexible with respect to rotation about the
central C2-C3 bond but not about the C1-C2 single bond. In
Scheme 4, the reaction pathway for the tert-leucine derivative
1f is shown.
65b
0
1
a From H NMR analysis of the crude reaction mixture, (2%, ca. 5%
detection limit. b Further products could not be characterized.
In this case, the cyclization/fragmentation ratio is 1.7, similar
to that for the leucine derivative 1d (1.4). Modification of the
hydrogen-bond interaction was investigated by application of
the four N-acyl derivatives of valine: N-acetyl (1e), N-benzoyl
(1h), N,N-succinimidyl (1i), and N-trifluoroacetyl (1j). The first
three substrates were expected to exhibit similar hydrogen-
bonding efficiency between the acyl substituent and the newly
formed hydroxy group, whereas the trifluoroacetyl group should
be less active in hydrogen-bonding interactions. This presump-
tion correlates with hydrogen-bond basicity values â2H reported
by Abraham and Platts for the groups CH3(CO)OR ) 0.45, CF3-
(CO)OR ) 0.25, and CH3(CO)NHR ) 0.73.27 Thus, strong
hydrogen-bond activity was expected for substrates 1a-1i, a
weaker hydrogen-bond activity was expected for the ester
derivatives 4a,b (vide infra), and the weakest hydrogen-bond
activity was expected for 1j. The idea that hydrogen bonding
can function as a highly relevant tool in directing the stereo-
chemistry of carbon-carbon-bond-forming steps has already
been widely discussed for photochemical reactions involving
triplet biradicals28-30 and also established on theoretical grounds.31
As expected, the cyclization reaction dominated the photochem-
istry of the substrates 1e,h,i, whereas no Yang product was
observed from the trifluoroacetyl derivative 1j. Thus, not only
is the stereoselectivity of the product formation influenced by
hydrogen bonding but also the chemoselectivity of the reaction,
that is, the cyclization/fragmentation ratio. This concept is also
reflected in the quantum yields for H-transfer-initiated processes
of 1b,e,f (vide infra).
Figure 1. Yang cyclization products from amino acid derivatives: ste-
reoselectivity.
(besides the methionine derivative 1k which gave mainly 3 and
ca. 30% decomposition products).
The relative product composition is summarized in Table 1.
Norrish type I cleavage was observed for all substrates except
for the isoleucine derivatives 1g and allo-1g. The Norrish I
cleavage products (N-alkyl acetamides) were detected by NMR
and GC and compared with authentic samples. The product from
the Norrish type I cleavage of the methionine derivative 1k was
not detected probably due to secondary electron-transfer reac-
tions which have also been detected for triplet-excited phthal-
imide derivatives of methionine25 and cysteine derivatives.26 The
proline derivative 1l resulted in Norrish II cleavage and gave
minor amounts of the bicyclic Yang product 2l. All cyclobutanes
2b-2j were formed with diastereoselectivities >95%. The
relative configurations of these products were established by
NMR methods (NOE effects and coupling constants) and X-ray
structure analyses of the cyclo-butanols 2e,f as well as 2g.24
By comparison between the NMR spectral data for these
compounds and for the norvaline- and norleucine-derived
products 2b and 2c, respectively, as well as the leucine-derived
product 2d, the relative configuration of these cyclobutanols
was unambiguously proven (Figure 1). The acylamino substitu-
ent (C-1) is always located cis, a methyl group at C-4 (in 2e
and 2g) is located trans, and a substituent at C-3 (in 2b,c,g) is
located cis with respect to the hydroxy substituent at C-2.
To interpret the stereochemical results mechanistically, the
different selectivity-directing effects are discussed separately.
(A) Selectivity Connected with Hydrogen-Bonding Inter-
actions. The Norrish II/Yang products 2 were always formed
cis selective with respect to the acylamino and the hydroxy
(B) Selectivity Connected with the Hydrogen-Transfer
Step. After electronic excitation and intersystem crossing to the
triplet nπ* state of the carbonyl compound, a chairlike
conformation initiates the hydrogen-transfer process.32,33 The
geometrical prerequisites for step A (Scheme 1) have been
extensively investigated by Scheffer and co-workers for liquid-
(27) Abraham, M. H.; Platts, J. A. J. Org. Chem. 2001, 66, 3484-3491.
(28) Hasegawa, T.; Arata, Y.; Endoh, M.; Yoshioka, M. Tetrahedron 1985, 41,
1667-1673.
(29) Steiner, A.; Wessig, P.; Polborn, K. HelV. Chim. Acta 1996, 79, 1843-
1862.
(30) Griesbeck, A. G.; Heckroth, H.; Schmickler, H. Tetrahedron Lett. 1999,
40, 3137-3140.
(31) Lindemann, U.; Wulff-Molder, D.; Wessig, P. J. Photochem. Photobiol.,
A 1998, 119, 73-83.
(25) Griesbeck, A. G.; Mauder, H.; Mu¨ller, I.; Peters, K.; Peters, E.-M.; von
Schnering, H. G. Tetrahedron Lett. 1993, 34, 453-456.
(26) Griesbeck, A. G.; Hirt, J.; Kramer, W.; Dallakian, P. Tetrahedron 1998,
54, 3169-3180.
(32) Dorigo, A. E.; McCarrick, M. A.; Loncharich, R. J.; Houk, K. N. J. Am.
Chem. Soc. 1990, 112, 7508-7514.
(33) Sauers, R. R.; Edberg, L. A. J. Org. Chem. 1994, 59, 7061-7066.
9
398 J. AM. CHEM. SOC. VOL. 124, NO. 3, 2002

Stereoselective Synthesis of 2-Aminocyclobutanols
A R T I C L E S
Table 2. Relative Energiesa and C1‚‚‚C4 Distancesb of Triplet
Scheme 5. Three-Step Reaction Protocol of the Valine Derivative
1e
1,4-Biradical from 1a-1g
substrate
syn
anti
syn(2)
1a
1b
1e
1f
allo-1g
1g
1.7 (3.17)
0.1 (3.10)
0.0 (2.92)
0.8 (2.94)
0.0 (2.91)
1.7 (3.28)
0.0 (3.81)
0.0 (3.81)
1.5 (3.80)
0.2 (3.83)
1.8 (3.77)
0.0 (3.85)
5.7 (3.25)
3.0 (3.18)
4.9 (3.19)
0.0 (3.10)
2.4 (3.17)
1.0 (3.20)
a In kcal/mol, PM3-UHF, AMBER52 preoptimized, para-methyl was
omitted. b In brackets, in Å.
mol (Scheme 4). Hydrogen bond distances CO‚‚‚HO were
between 1.80 and 1.85 Å (Figure 2).
and solid-phase photolyses.8 Following the classical assumption
of an interaction between the reactive CH bond and the py
orbital at the carbonyl oxygen,34 the terminal carbonyl substitu-
ent (the tolyl group in substrates 1) can adopt a pseudoequatorial
position. The optimal conformation for efficient hydrogen
transfer overrides substantial differences in γ-CH dissociation
energies; for example, primary, secondary, and tertiary γ-CH
(1f/1b/1d) were activated in a similar fashion and with similar
quantum yields (vide infra). Additionally, as exemplified for
the valine derivative 1e, diastereotopic methyl groups are
perfectly differentiated, and the chairlike transition state with
one methyl group in an equatorial position is preferred (Scheme
5). This selectivity in hydrogen transfer is reflected in the relative
configuration at C-4 of the cyclobutanol 2e.
(C) Selectivity Connected with Triplet 1,4-Biradical Con-
formational Equilibration. The lifetimes of 1-hydroxytetra-
methylenes reported in the literature3 are in the 50 ns region
and thus long enough to enable conformational flexibility at
room temperature. The syn biradical, which is formed after
rearrangement of the primarily formed skew biradical, is in
equilibrium with the anti conformer by rotation about the C2-
C3 bond. The equilibrium constant, the hydrogen-back-transfer
efficiency (from the syn structure), and the orbital orientation
(p,p relative to the central C-C single bond) control the
cyclization/cleavage ratio. The assumption is widely accepted
that, after intersystem crossing, the singlet biradicals maintain
conformational memory of their triplet precursors;35,36 that is,
the anti 1,4-biradical conformer gives exclusively cleavage
products, whereas the syn 1,4-biradicals can cleave as well as
cyclize. The assumption of triplet biradical equilibration and
subsequent ring closure versus bond cleavage reflects nicely
the experimental as well as the theoretical results. We observed
three types of reaction modes concerning the Yang cyclization/
cleavage ratios: (a) preferred cleavage (1a and 1g), (b) nearly
equal amounts of cleavage and cyclization (1f), and (c) preferred
cyclization (1e and allo-1g). Semiempirical calculations (PM3)37,38
of the three relevant 1,4-biradical conformers syn, syn(2), and
anti (Table 2) resulted in absolute differences of less than 3
kcal/mol. For 1a and 1g, the anti biradicals are more stable with
approximately 1.6 kcal; for 1e and allo-1g, the syn biradicals
are more stable with approximately 1-1.5 kcal. In the case of
the biradicals from the tert-leucine derivative 1f, the energies
of the possible three conformers are identical within (0.3 kcal/
The stereospecificity detected for the Yang cyclization of the
isoleucine derivatives (2S,3S)-1g and (2S,3R)-allo-1g is also in
agreement with the photoenol generation and spectroscopy from
1g and 1b, respectively.39 Whereas from 1g the photoenol
(deriving from the Norrish II cleavage) could be easily detected,
the amount of enol transient formed from photolysis of 1b was
significantly smaller. The diastereoisomeric starting materials
1g and allo-1g already exhibit two stereogenic centers and two
chemically different γ-CH positions [in contrast to 1e,f where
only homotopic (for 1f) or diastereotopic (for 1e) methyl groups
delivered the γ-hydrogens]. The biradical situation is depicted
for allo-1g in Scheme 6; in this case, the two different γ-CH
positions available for H transfer are clearly differentiated in
bond dissociation energies, and the H transfer is favored from
the methylene group independently of the position of the less
reactive methyl group (equatorial/axial).
The biradical conformers relevant for the chemoselectivity
of the Norrish II reaction of allo-1g are the syn and the anti
isomer [syn(2) has one gauche interaction more than syn], with
the syn isomer approximately 1 kcal/mol more stable than the
anti.40 The situation for the isoleucine system 1g was like a
mirror image; the anti isomer is approximately 1-1.7 kcal/mol
more stable than both possible syn conformers (Scheme 7).
These numbers nicely reflect chemical intuition with respect to
the gauche interaction heat of formation increments.41
Further evidence for the role of geometrical factors came from
an analysis of the C1-C4 distance in the triplet biradical which
resulted from the PM3 calculations (Table 2). In most cases
where cyclobutane formation was observed, this distance in one
of the syn biradicals was between 3.1 and 2.9 Å. The
intermediates from amino butyric acid (1a) and isoleucine (1g),
however, that reacted solely in photofragmentation, showed
reduced C1-C4 interactions with shortest C-C distances of
3.17 and 3.28 Å, respectively.42
(D) Selectivity Connected with the Terminal Carbon-
Carbon Bond-Forming Step. After primary hydrogen transfer
and conformational rearrangement, the terminal C-C-bond-
forming step leads to the cyclobutanols 2. Prior to this step as
well as the C2-C3 fragmentation step, the triplet biradicals have
to undergo spin inversion to cross into the singlet energy surface.
In singlet photochemical reactions, stereoselectivity is often
(34) Zimmerman, H. E. Science 1966, 153, 837-844.
(35) Scaiano, J. C. Tetrahedron 1982, 38, 819-824.
(36) Wagner, P. J.; Zand, A.; Park, B.-S. J. Am. Chem. Soc. 1996, 118, 12856-
12857.
(37) Stewart, J. P. P. J. Comput. Chem. 1989, 10, 209-220.
(38) The PM3 method has been used for calculation of triplet biradicals, for
example: Suishu, T.; Shimo, T.; Somekawa, K. Tetrahedron 1997, 53,
3545-3556.
(39) Chiang, Y.; Griesbeck, A. G.; Heckroth, H.; Hellrung, B.; Kresge, A. J.;
Meng, Q.; O’Donoghue, A. M.; Richard, J. P.; Wirz, J. J. Am. Chem. Soc.
2001, 123, 8979-8984.
(40) Density-functional theory (DFT) calculations gave the same trends but
resulted in higher energy differences: unpublished results.
(41) Cohen, N.; Benson, S. W. Chem. ReV. 1993, 93, 2419-2438.
(42) Optimal distances of 3.02-3.08 Å have been reported: Cheung, E.;
Netherton, M. R.; Scheffer, J. R.; Trotter, J. Org. Lett. 2000, 2, 77-80.
9
J. AM. CHEM. SOC. VOL. 124, NO. 3, 2002 399

A R T I C L E S
Griesbeck and Heckroth
Figure 2. PM3 calculated triplet biradicals (3BR) from substrates 1e, 1f, 1g, and allo-1g. Low energy conformers: 1e-syn, allo-1g-syn, and 1g-anti. High
energy conformers: 1e-anti and 1g-syn.
Scheme 6. Three-Step Reaction Protocol of the Isoleucine
Derivative allo-1g
Figure 3. 90° conformational situation for the triplet 1,4-biradical:
stereoselectivity of the C2-C4 bond formation.
Scheme 8. Three-Step Reaction Protocol of the Norvaline
Derivative 1b
Scheme 7. Three-Step Reaction Protocol of the Isoleucine
Derivative 1g
important contribution for SOC modulation. Already the un-
branched substrates 1b and 1c gave the corresponding cyclobu-
tanols 2b and 2c with high simple diasteroselectivity but low
controlled by the optimal geometries for radical-radical
cyclization/fragmentation ratios (shown for the isovaline deriva-
combinations, whereas in triplet photoreactions, the geometries
tive 1b in Scheme 8).
most favorable for intersystem crossing (ISC) are considered
Optimal φ values for strong SOC are in the range of 90° (
to be of similar relevance. These geometries can be quite
10°.20,44 This model presupposes conformational mobility at this
different from the former ones due to differences in spin-orbit
intermediate stage in contrast to the reaction of singlet excited
coupling (SOC) values. We have described this phenomenon
carbonyl states where conical intersections guide the substrates
to the cycloaddition products nearly barrier-free. A view along
the central C-C bond of the intermediate triplet 1,4-biradical
(Figure 3) shows the important contribution to this high degree
for [2 + 2]-photocycloadditions.16-18 The situation is very much
the same for the terminal step of the Yang reaction. Calculations
resulted in large SOC values for syn geometries and lower
values for anti geometries.43 Assuming that in the syn conformer
of stereocontrol: increasing gauche interactions are expected
the hydroxy group localized at the benzylic radical is strongly
for a biradical approach from the same half space as the
hydrogen bonded and thus the radical geometry at C-1 is largely
fixed, bond rotation about the C3-C4 single bond makes the
(44) A more sophisticated and theoretically more advanced description of the
conformational dependence of SOC in 2-oxatetramethylenes: Kutateladze,
(43) Klessinger, M. Pure Appl. Chem. 1997, 69, 773-778.
A. G. J. Am. Chem. Soc. 2001, 123, 9279-9282.
9
400 J. AM. CHEM. SOC. VOL. 124, NO. 3, 2002

Stereoselective Synthesis of 2-Aminocyclobutanols
A R T I C L E S
Table 3. Quantum Yields for Photoprocesses of 1b,e,fa
independent of hydrogen-bonding effects) still prevails, whereas
the efficiency of the cyclobutanol formation decreases and also
the stereoselectivity of the terminal C-C bond formation step
with 4b (which are both depending on hydrogen-bonding
effects). These results are corroborated by the results reported
by Hasegawa et al. for the Yang cyclization reactions of
R-alkylated benzoyl ethyl acetates.48
substrate
ΦT
ΦC
ΦF
1b
1e
1f
0.12
0.26
0.23
0.04
0.19
0.11
0.05
0.02
0.08
a In benzene, (0.02, by valerophenone actinometry.
Scheme 9. Photochemistry of the 2-Benzoylbutyric Acid Ethyl
Esters 4a,b
Conclusion
The set of alkyl-substituted R-amino p-methylbutyrophenone
derivatives 1a-1g was highly useful to describe the different
chemo- and stereoselection stages in the Norrish/Yang reaction
sequence. The three decisive steps (a) hydrogen transfer, (b)
biradical equilibration and hydrogen-bond stabilization, and (c)
ISC-coupled bond formation versus bond cleavage became
visible in (a) selective hydrogen transfer from diastereotopic
γ-CH, (b) asymmetric 1,2-induction at the stage of the biradical
connected with intramolecular hydrogen bonding, and (c)
asymmetric induction coupled with selective biradical cycliza-
tion due to SOC-determined ISC geometries.49 The comparison
of the diastereoisomeric isoleucine derivatives50 1g and allo-
1g is a useful method for evaluation of the reactivity modulation
at the stage of the triplet 1,4-biradicals.
hydroxymethyl or acetoxymethyl group, respectively (vide
supra). The aryl substituent at C-1 repulses a large substituent
at C-4 when oriented orthogonal with respect to each other,
and thus the preferred geometry for ISC is expected to be such
as depicted in Figure 2 which leads to the correct product
configuration.
Quantum Yield Studies. The quantum yields for the pho-
todecomposition of substrates 1b, 1e, and 1f were determined
using a merry-go-round apparatus45 with valerophenone as the
actinometer.46 The product composition was measured as a
function of time by GC, and the total quantum yields ΦT for
substrate photodecomposition are separated into quantum yields
for cyclization ΦC and quantum yields for Norrish II fragmenta-
tion ΦF (Table 3).
The trend in quantum yields is in agreement with literature
values for R-branched alkylaryl ketones. In our cases, the
quantum yields for H-transfer-initiated reactions, that is, the
Norrish II cleavage and the Yang cyclization, are much higher
than for simple butyrophenones. For substrates without a further
R-substituent, Φc of 4-7% was reported, whereas R-methylated
compounds show increased cyclization efficiences with ΦC of
9-12%.47 The increased Φc and ΦF values for 1b,e,f account
for the hydrogen-bonding interaction at the stage of the triplet
1,4-biradical which slows down the hydrogen back-transfer. The
valine derivative 1e has one of the largest quantum yields known
for the Norrish/Yang reaction with Φc ) 19% and also a
synthetically useful chemical yield of 52% for the 2-aminocy-
clobutanol 2e which is formed in diastereomerically pure form.
Photochemistry of â-Keto Esters 4a,b. Also, aromatic
â-keto esters 4a,b gave the Yang cyclization products 5a,b in
low yields (Scheme 9). The isopropyl derivative (by alkylation
of benzoyl ethyl acetate) 4a gave upon photolysis in benzene
mainly Norrish II cleavage (6), in contrast to the analogous
substrate 1e. Similar to 1e, the cyclobutanol 5a was formed with
high diastereoselectivity. The selectivity also vanished in the
case of the isoleucine derivative 4b where a 2:1 mixture of
diastereoisomers (trans/cis with respect to the C3, C4-dimethyl
substitution pattern) was isolated. Both trends account for a
reduced influence of the intermediary hydrogen-bond stabiliza-
tion;27 the diastereoselectivity connected with hydrogen transfer
from one of the diastereotopic methyl groups in 4a (which is
Experimental Section
1
General Remarks. H NMR: Bruker AC 250 (250 MHz), Bruker
AC 300 (300 MHz). 13C NMR: Bruker AC 250 (63.4 MHz), Bruker
AC 300 (75.5 MHz), carbon multiplicities were determined by
distortionless enhancement by polarization transfer (DEPT). UV-vis:
Hitachi U-3200. Mass spectroscopy (EI or CI): Finnigan Incos 500;
m/z (rel. %). HR-mass spectroscopy (FAB): Finnigan MAT H-SQ 30.
Column chromatography: silica gel (Merck) 60-230 mesh; petroleum
ether (PE, 40-60 °C), ethyl acetate (EA). All melting points were
determined with a Bu¨chi melting point apparatus (type Nr. 535) and
are uncorrected. Combustion analyses: Elementar Vario EL. Rayonet
chamber photoreactors RPR-208 (8 × 3000 Å lamps, ca. 800 W, λ )
300 ( 10 nm) and RPR-100 (16 × 3500 Å lamps, ca. 400 W, λ ) 350
( 20 nm) were used for irradiations.
Substrates 1a-1l. General Procedure. N-[1-(4-Methyl-benzoyl)-
propyl]-acetamide (1a). To a suspension of 6.25 g (15.0 mmol) of
phosphorus pentachloride in 40 mL of toluene was added under external
cooling a solution of 2.0 g (13.8 mmol) of N-acetyl 2-aminobuytric
acid in 20 mL of toluene. After heating to 60 °C for 2 h and cooling
to room temperature, 4.0 g (30 mmol) of aluminum chloride was added,
and, after stirring overnight at room temperature, the yellow-orange
mixture was poured into 100 mL of ice and extracted with 3 × 50 mL
of methylene chloride. After washing the organic phase with 5%
aqueous sodium bicarbonate and drying over magnesium sulfate, the
solvent was evaporated, and the residue was purified by column
chromatography (silica gel, chloroform/PE 2:1). 1b-1l were synthesized
1
likewise and characterized by H NMR and 13C NMR (Table 4), IR,
and HR-MS.
(2S,3S)-N-[2-Methyl-1-(4-methyl-benzoyl)-butyl]-acetamide (1g)
and (2S,3R)-N-[2-Methyl-1-(4-methyl-benzoyl)-butyl]-acetamide (allo-
1g). To a solution of 1.0 g (5.8 mmol) of N-acetyl (L) isoleucine in 25
mL of dichloromethane under nitrogen cooled to 0 °C was added over
1 h a solution of 1 mL (11.6 mmol) of oxalyl chloride and 1 drop of
pyridine in 5 mL of dichloromethane. After stirring at room temperature
(48) Hasegawa, T.; Arata, Y.; Endoh, M.; Yoshioka, M. Tetrahedron 1985, 41,
1667-1674.
(49) Griesbeck, A. G.; Heckroth, H. Res. Chem. Intermed. 1999, 25, 599-608.
(50) We have also used this stereochemical probe in phthalimide photochem-
istry: Griesbeck, A. G.; Mauder, H. Angew. Chem. 1992, 103, 97-99;
Angew. Chem., Int. Ed. Engl. 1992, 31, 73-75.
(45) Moses, F. G.; Liu, R. S. H.; Monroe, B. H. Mol. Photochem. 1969, 1, 245.
(46) Wagner, P. J.; Kemppainen, A. E. J. Am. Chem. Soc. 1972, 94, 7495-
7499.
(47) Lewis, F. D.; Hillard, T. A. J. Am. Chem. Soc. 1972, 94, 3852-3856.
9
J. AM. CHEM. SOC. VOL. 124, NO. 3, 2002 401

A R T I C L E S
Griesbeck and Heckroth
Table 4. NMR Spectral Data for Substrates 1a-1l and 4a,ba,b
1H NMR (CDCl3, 300 MHz)
13C NMR (CDCl3, 75.5 MHz)
CH3: 8.8, 21.6, 23.2
1a
1b
1c
1d
0.81 (t, 3H, J 7.5), 1.66 (m, 1H), 2.01 (m, 1H), 2.07 (s, 3H),
2.40 (s, 3H), 5.56 (m, 1H), 6.57 (d, 1H, J 8.3), 7.28 (d, 2H, J 8.4),
7.47 (d, 2H, J 8.4)
CH2: 26.3
CH: 54.5, 128.7, 129.4
Cq: 131.8, 144.8, 169.7, 198.5
CH3: 13.8, 21.7, 23.4
0.84 (t, 3H, J 7.5), 1.30 (m, 2H), 1.54 (m, 1H), 1.90 (m, 1H),
2.04 (s, 3H), 2.40 (s, 3H), 5.59 (m, 1H), 6.49 (d, 1H, J 8.7),
7.28 (d, 2H, J 8.4), 7.47 (d, 2H, J 8.4)
CH2: 18.2, 35.7
CH: 53.5, 128.7, 129.6
Cq: 131.9, 144.9, 169.7, 198.7
CH3: 13.8, 21.7, 23.3
0.80 (t, 3H, J 6.9), 1.24 (m, 4H), 1.54 (m, 1H), 1.90 (m, 1H),
2.03 (s, 3H), 2.40 (s, 3H), 5.57 (m, 1H), 6.49 (d, 1H, J 8.1),
7.28 (d, 2H, J 8.4), 7.47 (d, 2H, J 8.4)
CH2: 22.4, 26.9, 33.3
CH: 53.6, 128.7, 129.6
Cq: 131.9, 144.9, 169.6, 198.7
CH3: 21.7, 21.8, 23.3, 23.4
CH2: 25.0
CH: 43.1, 51.9, 128.7, 129.5
Cq: 131.9, 144.8, 169.7, 199.3
CH3: 16.7, 19.9, 21.6, 23.4
CH: 31.9, 57.7, 128.7, 129.5
Cq: 132.6, 144.7, 170.1, 199.1
CH3: 21.6, 23.4, 26.9
CH: 58.5, 128.7, 129.4
Cq: 35.6, 135.0, 144.6, 169.6, 200.6
CH3: 11.9, 16.1, 21.6, 23.8
CH2: 38.6
CH: 27.1, 57.5, 128.6, 129.4
Cq: 132.2, 144.7, 170.1, 199.0
CH3: 11.5, 13.6, 21.6, 23.2
CH2: 38.4
0.88 (t, 3H, J 6.9), 1.24 (m, 4H), 1.54 (m, 1H), 1.90 (m, 1H),
2.03 (s, 3H), 2.40 (s, 3H), 5.57 (m, 1H), 6.49 (d, 1H, J 8.1),
7.28 (d, 2H, J 8.4), 7.47 (d, 2H, J 8.4)
1e
0.72 (d, 3H, J 6.1), 0.98 (d, 3H, J 6.0), 2.02 (s, 3H), 2.11 (m, 1H),
2.34 (s, 3H), 5.50 (dd, 1H, J 4.4, 9.0), 6.40 (d, 1H, J 9.0),
7.23 (d, 2H, J 8.2), 7.82 (d, 2H, 8.2)
0.91 (s, 9H), 2.01 (s, 3H), 2.41 (s, 3H), 5.52 (d, 1H, J 9.4),
6.35 (d, 1H, J 9.4), 7.26 (d, 2H, J 8.2), 7.89 (d, 2H, J 8.2)
1f
allo-1g
0.94 (d, 3H, J 6.8), 0.99 (d, 3H, J 7.3), 1.17 (m, 1H), 1.46 (m, 1H),
1.81 (m, 1H), 1.99 (s, 3H), 2.43 (s, 3H), 5.68 (dd, 1H, J 3.4, 8.9),
6.40 (d, 1H, J 8.9), 7.38 (d, 2H, J 8.2), 7.88 (d, 2H, J 8.2)
1g
1h
1i
0.78 (t, 3H, J 7.5), 0.92 (d, 3H, J 6.8), 1.0 (m, 1H), 1.30 (m, 1H),
1.89 (m, 1H), 2.01 (s, 3H), 2.39 (s, 3H), 5.56 (dd, 1H, J 5.1, 8.9),
6.53 (d, 1H, J 8.9), 7.24 (d, 2H, J 7.9), 7.89 (d, 2H, J 7.9)
CH: 27.0, 56.3, 128.7, 129.5
Cq: 132.2, 144.6, 169.9, 198.9
CH3: 16.7, 19.9, 21.5
CH: 32.1, 58.1, 126.9, 128.0,
128.4, 129.4, 131.4
0.80 (d, 3H, J 7.0), 1.04 (d, 3H, J 6.8), 2.29 (m, 1H), 2.36 (s, 3H),
5.74 (dd, 1H, J 4.4, 8.7), 7.01 (d, 1H, J 8.7), 7.22-7.93 (m, 9 H)
Cq: 132.5, 134.2, 144.7, 167.3, 198.8
CH3: 18.7, 20.8, 21.5
CH2: 27.6
0.82 (d, 3H, J 6.8), 1.06 (d, 3H, J 6.6), 2.37 (s, 3H), 2.60 (br s, 4H),
2.88 (m, 1H), 5.01 (d, 1H, J 9.0), 7.19 (d, 2H, J 8.4), 7.67 (d, 2H, J 8.4)
CH: 26.1, 59.9, 127.7, 129.2
Cq: 133.4, 143.7, 176.3, 194.1
CH3: 17.1, 19.4, 21.4, 115.1 (CF3)
CH: 32.3, 59.4, 128.4, 129.7
Cq: 133.1, 144.9, 157.6, 200.1
CH3: 15.4, 21.6, 23.3
1j
0.78 (d, 3H, J 6.9), 1.04 (d, 3H, J 6.8), 2.30 (m, 1H), 2.36 (s, 3H),
5.57 (dd, 1H, J 4.5, 8.8), 7.23 (d, 2H, J 8.2), 7.85 (d, 2H, J 8.2)
1k
1.81 (m, 1H), 2.01 (s, 3H), 2.08 (s, 3H), 2.20 (m, 1H), 2.41 (s, 3H),
2.50 (m, 2H), 5.71 (m, 1H), 7.28 (d, 1H, J 8.2), 7.91 (d, 2H, J 8.2)
CH2: 29.9, 33.3
CH: 52.8, 128.8, 129.5
Cq: 131.5, 145.1, 169.7, 197.8
CH3: 21.6, 24.5
1l
2.10 (m, 3H), 2.13 (s, 3H), 2.31 (m, 1H), 2.41 (s, 3H), 3.58 (m, 1H),
3.76 (m, 1H), 5.50 (dd, 1H, J 3.5, 8.9), 7.26 (d, 3H, J 8.2), 7.89 (d, 2H, J 8.2)
CH2: 21.9, 29.4, 48.2
CH: 60.9, 128.5, 129.6
Cq: 131.9, 144.1, 169.3, 196.8
CH3: 14.0, 20.4, 21.0
4a
4b
0.92 (d, 3H, J 6.6), 1.03 (d, 3H, J 6.8), 1.15 (t, 3H, J 7.2), 2.64 (m, 1H),
4.07 (d, 1H, J 9.4), 4.11 (q, 2H, J 7.2), 7.46 (m, 2H), 7.56 (m, 1H), 8.01 (m, 2H)
CH2: 61.2
CH: 61.6, 128.6, 128.8, 133.4
Cq: 136.9, 169.1, 194.7
CH3: 11.0, 14.0, 16.4
CH2: 26.9, 59.9
CH: 34.9, 61.2, 128.5, 128.7, 133.4
Cq: 137.1, 169.1, 194.8
0.91 (t, 3H, J 6.8), 1.03 (d, 3H, J 6.8), 1.16 (t, 3H, J 7.2), 1.31 (m, 2H),
2.46 (m, 1H), 4.11 (d, 1H, J 7.0), 4.16 (q, 2H, J 7.2), 7.46 (m, 2H),
7.56 (m, 1H), 8.04 (m, 2H)
a 13C NMR, multiplicities determined by DEPT. b All compounds were characterized by correct combustion analysis.
for 2 h, the solvent was evaporated, and the residue was dissolved in
20 mL of toluene and treated with 1.59 g (12 mmol) of aluminum
chloride. After stirring for 24 h at room temperature, the yellow-orange
mixture was poured into 100 mL of ice and extracted with 3 × 50 mL
of methylene chloride. After washing the organic phase with 5%
aqueous sodium bicarbonate and drying over magnesium sulfate, the
solvent was evaporated, and the residue was purified by column
chromatography (silica gel, chloroform/PE 2:1). Allo-1g was synthesized
likewise (Table 4).
2-Benzoyl-3-methyl-butyric Acid Ethyl Ester (4a). A solution of
5.55 g (28.9 mmol) of ethyl benzoyl acetate and 2.36 g (34.7 mmol)
of sodium ethoxide in 50 mL of ethanol was heated to reflux for 3 h.
After cooling to room temperature, 13.6 mL (144.5 mmol) of
2-bromopropane was added, and the solution was heated to reflux for
an additional 2 h. After filtration and evaporation of 80% of the solvent,
100 mL of methylene chloride was added and extracted with 3 × 50
mL of water and 2 × 50 mL of 4 N HCl. After drying over magnesium
sulfate, the solvent was evaporated, and the residue was purified by
9
402 J. AM. CHEM. SOC. VOL. 124, NO. 3, 2002

Stereoselective Synthesis of 2-Aminocyclobutanols
A R T I C L E S
column chromatography (silica gel, PE/EE 4:1) to give 4a in 34%.
2-Benzoyl-3-methyl-pentanoic acid ethyl ester (4b) was synthesized
likewise in 41% yield (Table 4).
N-(2-Hydroxy-3,4-dimethyl-2-p-tolyl-cyclobutyl)-acetamide (2g).
79%, mp 204-205 °C. H NMR (CDCl3, 300 MHz): δ 1.18 (d, 3H,
1
J 6.1), 1.19 (d, 3H, J 6.1), 1.89 (m, 1H), 2.03 (s, 3H), 2.02-2.05 (m,
1H), 2.37 (s, 3H), 4.19 (d, 1H, J 8.8), 6.32 (d, 1H, J 8.8), 7.18 (d, 2H,
J 7.9), 7.34 (d, 2H, J 7.9). 13C NMR (CDCl3, 75.5 MHz): δ 12.1 (CH3),
18.0 (CH3), 21.4 (CH3), 23.0 (CH3), 43.1 (CH), 44.8 (CH), 57.7 (CH),
80.6 (Cq), 126.6 (CH), 130.1 (CH), 138.0 (Cq), 144.4 (Cq), 173.1 (Cq).
Anal. Calcd for C15H21NO2 (247.3): C, 72.84; H, 8.56; N, 5.66.
Found: C, 72.99; H, 8.40; N, 5.47.
Photolyses of 1a-1l and 4a,b. General Procedure. A solution of
0.2 mmol of the starting material in 50 mL of benzene was irradiated
under a nitrogen atmosphere at 15 °C for 5 h (TLC- and GC-control)
with phosphor-coated mercury low-pressure lamps emitting at 350 (
20 nm. After evaporation of the solvent, the crude reaction mixture
was separated and purified by column chromatography (silica gel, PE/
EE 2:1). The Norrish I cleavage products are all literature-known and
were compared with authentic samples or spectral data from the
literature. As a Norrish II cleavage product, N-(4-methyl-benzoyl)-
acetamide (3) was formed during the photolyses of 1a-1g, and the
corresponding amides were formed from 1h-1l.
N-(2-Hydroxy-3-methyl-2-p-tolyl-cyclobutyl)-acetamide (2b). 22%,
colorless oil. 1H NMR (CDCl3, 300 MHz): δ 1.11 (d, 3H, J 6.6), 1.66
(m, 1H), 1.91 (s, 3H), 2.29 (s, 3H), 2.30 (m, 1H), 2.39 (m, 1H), 4.95
(m, 1H), 6.47 (d, 1H, J 8.6), 7.09 (d, 2H, J 7.9), 7.24 (d, 2H, J 7.9).
13C NMR (CDCl3, 75.5 MHz): δ 12.6 (CH3), 20.9 (CH3), 23.1 (CH3),
35.0 (CH), 39.3 (CH2), 55.7 (CH), 74.0 (Cq), 81.4 (Cq), 124.9 (CH),
129.2 (CH), 137.3 (Cq), 141.9 (Cq), 169.7 (Cq). Anal. Calcd for C14H19-
NO2 (233.3): C, 72.07; H, 8.21; N, 6.00. Found: C, 71.83; H, 8.16;
N, 5.75.
N-(2-Hydroxy-3-ethyl-2-p-tolyl-cyclobutyl)-acetamide (2c). 41%,
colorless oil. 1H NMR (CDCl3, 300 MHz): δ 0.82 (t, 3H, J 7.3), 1.60
(m, 1H), 1.91 (s, 3H), 2.21 (m, 1H), 2.29 (s, 3H), 2.41 (m, 1H), 4.56
(m, 1H), 6.49 (d, 1H, J 8.8), 7.09 (d, 2H, J 7.9), 7.28 (d, 2H, J 7.9).
13C NMR (CDCl3, 75.5 MHz): δ 11.5 (CH3), 21.0 (CH2), 21.3 (CH3),
23.2 (CH3), 32.4 (CH2), 42.1 (CH), 48.4 (CH), 81.4 (Cq), 124.8 (CH),
129.1 (CH), 137.1 (CH), 141.7 (Cq), 169.7 (Cq). Anal. Calcd for C15H21-
NO2 (247.3): C, 72.84; H, 8.56; N, 5.66. Found: C, 71.72; H, 8.46;
N, 5.56.
N-(2-Hydroxy-4-methyl-2-p-tolyl-cyclobutyl)-benzamide (2h). ca.
1
65% (from crude NMR), colorless oil. H NMR (CDCl3, 300 MHz):
δ 1.14 (d, 3H, J 6.8), 1.78 (dd, 1H, 1H, J 9.8, 11.9), 2.32 (s, 3H), 2.31
(m, 2H), 4.49 (t, 1H, J 8.8), 7.02-7.85 (m, 9H).
1-(2-Hydroxy-4-methyl-2-p-tolyl-cyclobutyl)-pyrrolidine-2,5-di-
1
one (2i). ca. 65% (from crude NMR). H NMR (CDCl3, 300 MHz) δ
1.15 (d, 3H, J 6.6), 1.96 (dd, 1H, J 9.2, 12.3), 2.30 (s, 3H), 2.57 (dd,
1H, J 12.3), 2.74 (s, 4H), 2.78 (m, 1H), 4.58 (d, 1H, J 9.4), 7.11 (d,
2H, J 7.9), 7.29 (d, 2H, J 7.9). 13C NMR (CDCl3, 75.5 MHz): δ 19.6
(CH3), 20.9 (CH3), 27.8 (CH2), 41.7 (CH2), 44.6 (CH), 60.7 (Cq), 80.5
(Cq), 124.8 (CH), 129.1 (CH), 136.9 (Cq), 141.1 (Cq), 179.2 (Cq).
1-(6-Hydroxy-6-phenyl-5-aza-bicyclo[2.1.1]hex-5-yl)-ethanone (2l).
1H NMR (CDCl3, 300 MHz, mixture with N-acetylpyrrolidine): δ 1.88
(m, 4H), 2.00 (s, 3H), 2.31 (s, 3H), 5.22 (m, 2H), 7.14 (d, 2H, J 7.9),
7.28 (d, 2H, J 7.9). 13C NMR (CDCl3, 75.5 MHz, mixture with
N-acetylpyrrolidine): δ 20.6 (CH3), 22.1 (CH2), 24.3 (CH3), 63.0 (CH),
77.2 (Cq), 125.4 (CH), 129.2 (CH), 137.6 (Cq), 144.0 (Cq), 169.1 (Cq).
N-(4-Methyl-benzoyl)-acetamide (3). mp 142 °C (143 °C).51
Ethyl 2-Hydroxy-4-methyl-2-phenyl-cyclobutanoate (5a). ca. 26%
1
(from crude NMR, mixture with the Norrish I cleavage product). H
NMR (CDCl3, 300 MHz): δ 1.21 (d, 3H, J 6.9), 1.22 (t, 3H, J 6.8),
1.99 (dd, 1H, J 9.1, 13.1), 2.48 (dd, 1H, J 8.2, 13.1), 2.92 (m, 1H),
3.21 (d, 1H, J 9.2), 4.21 (q, 2H, J 6.8), 7.23-7.49 (m, 5H). 13C NMR
(CDCl3, 75.5 MHz): δ 14.2 (CH3), 20.6 (CH3), 29.4 (CH), 41.3 (CH2),
54.1 (CH), 75.6 (Cq), 124.9 (CH), 127.3 (CH), 128.2 (CH), 146.1 (Cq),
173.2 (Cq).
N-(2-Hydroxy-3,3-dimethyl-2-p-tolyl-cyclobutyl)-acetamide (2d).
1
47%, mp 192-194 °C. H NMR (CDCl3, 300 MHz): δ 0.76 (s, 3H),
1.21 (s, 3H), 1.83 (m, 1H), 1.90 (s, 3H), 2.02 (m, 1H), 2.28 (s, 3H),
5.06 (m, 1H), 6.45 (d, 1H, J 8.3), 7.07 (d, 2H, J 7.9), 7.21 (d, 2H, J
7.9). 13C NMR (CDCl3, 75.5 MHz): δ 21.1 (CH3), 21.2 (CH3), 21.7
(CH3), 23.4 (CH3), 38.9 (CH2), 42.6 (Cq), 52.6 (CH), 77.3 (Cq), 124.9
(CH), 129.1 (CH), 137.4 (Cq), 141.7 (Cq), 169.7 (Cq). Anal. Calcd for
C15H21NO2 (247.3): C, 72.84; H, 8.56; N, 5.66. Found: C, 72.16; H,
8.49; N, 5.98.
N-(2-Hydroxy-4-methyl-2-p-tolyl-cyclobutyl)-acetamide (2e). 52%,
mp 172-173 °C. 1H NMR (CDCl3, 300 MHz): δ 1.19 (d, 3H, J 6.6),
1.78 (dd, 1H, J 9.7, 11.8), 2.02 (s, 3H), 2.31 (s, 3H), 2.38 (m, 2H),
4.36 (t, 3H, J 8.9), 6.30 (d, 1H, J 8.4), 7.14 (d, 2H, J 7.9), 7.28 (d, 2H,
J 7.9). 13C NMR (CDCl3, 75.5 MHz): δ 19.1 (CH3), 21.0 (CH3), 23.4
(CH3), 35.2 (CH), 38.5 (CH2), 56.9 (CH), 78.0 (Cq), 124.9 (CH), 129.1
(CH), 137.4 (Cq), 141.8 (Cq), 169.6 (Cq). Anal. Calcd for C14H19NO2
(233.3): C, 72.07; H, 8.21; N, 6.00. Found: C, 71.09; H, 8.16; N,
5.79.
N-(2-Hydroxy-4,4-dimethyl-2-p-tolyl-cyclobutyl)-acetamide (2f).
43%, colorless oil. 1H NMR (CDCl3, 300 MHz): δ 1.19 (s, 3H), 1.21
(s, 3H), 2.01 (s, 3H), 2.06 (d, 1H, J 9.5), 2.33 (s, 3H), 3.04 (d, 1H, J
9.5), 4.49 (d, 1H, J 9.0), 6.22 (d, 1H, J 9.0), 7.15 (d, 2H, J 7.9), 7.25
(d, 2H, J 7.9). 13C NMR (CDCl3, 75.5 MHz): δ 21.0 (CH3), 23.3 (CH3),
23.4 (CH3), 29.8 (CH3), 36.6 (CH2), 38.1 (Cq), 58.1 (CH), 75.9 (Cq),
124.8 (CH), 129.1 (CH), 130.1 (Cq), 137.2 (Cq), 169.9 (CHq). Anal.
Calcd for C15H21NO2 (247.3): C, 72.84; H, 8.56; N, 5.66. Found: C,
72.34; H, 8.34; N, 5.39.
Ethyl 2-Hydroxy-3,4-dimethyl-2-phenyl-cyclobutanoate (5b). ca.
15% (from crude NMR, mixture with the Norrish I cleavage product).
1. diastereoisomer, 1H NMR (CDCl3, 300 MHz): δ 0.55 (d, 3H, J 7.7),
1.01 (d, 3H, J 6.8), 1.22 (t, 3H, J 6.8), 2.45 (m, 1H), 3.02 (m, 1H),
3.20 (d, 1H, J 10.1), 4.21 (q, 2H, J 6.8), 7.23-7.49 (m, 5H). 2.
1
diastereoisomer, H NMR (CDCl3, 300 MHz): δ 1.06 (d, 3H, J 6.8),
1.12 (d, 3H, J 6.8), 1.22 (t, 3H, J 6.8), 2.05 (m, 1H), 2.34 (m, 1H),
2.81 (d, 1H, J 9.4), 4.23 (q, 2H, J 6.8), 7.23-7.49 (m, 5H).
Acknowledgment. Financial support by the Deutsche Fors-
chungsgemeinschaft (GR-881/7-1,2,3) and the Fonds der Chem-
ischen Industrie is greatly acknowledged. For generous gifts of
amino acids we thank the Degussa AG. We are grateful to
Andrei Kutateladze for informing us about his work on the
conformational dependence of SOC in 2-oxatetramethylenes.44
JA0111941
(51) Knittel, D. Monatsh. Chem. 1986, 117, 679-688.
(52) Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.; Lipton, M.;
Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput. Chem.
1990, 11, 440-467.
9
J. AM. CHEM. SOC. VOL. 124, NO. 3, 2002 403