Effect of chain length on radical to carbanion cyclo-coupling of bromoaryl alkyl-linked oxazolines: 1,3-Areneotropic migration of oxazolines

Article abstract of DOI:10.1021/jo0620720

(Chemical Equation Presented) 2-Halophenylalkyl-2-oxazolines with alkyl chain spacers of two to six C atoms (n = 0-4) were prepared and their S _RN1-type reactions with several base systems examined. The best conditions to promote cyclocoupling to the corresponding benzocycloalkane derivatives involved use of LDA in THF. The precursors with 3-C-atom and 4-C-atom spacers gave good yields of 2-(1′-phenylindan-1′-yl)-2- oxazolines and 2-(1-phenyl-1,2,3,4-tetrahydronaphthalen-1-yl)-2-oxazoline, respectively. The major products from the precursor with a 5-C-atom spacer were derivatives of benzocycloheptane in which the oxazoline group had undergone a novel areneotropic migration from the end of the spacer to the benzo ring. The product from reaction of the corresponding 2-C-atom precursor was a 9-oxazolinophenanthrene derivative. EPR spectroscopy showed the intermediates of the LDA-promoted reactions to be radical anions of the product benzocycloalkanes. This supported an S_RN1-type chain mechanism involving initial production of aryl radicals connected to azaenolate ions via the spacer groups. Intramolecular radical to carbanion coupling then generated ring-closed benzocycloalkane radical anions that transferred an electron to more precursor. Diastereoselective radical to carbanion cyclo-coupling reactions were carried out with 2-bromophenylpropyl precursors containing chiral 2-oxazolines. The diastereoselectivity achievable was modest, but the product diastereoisomeric Indane derivatives were easily separable by chromatography.

Full text of DOI:10.1021/jo0620720

Effect of Chain Length on Radical to Carbanion Cyclo-Coupling of
Bromoaryl Alkyl-Linked Oxazolines: 1,3-Areneotropic Migration of
Oxazolines
Laura J. Marshall, Mark D. Roydhouse, Alexandra M. Z. Slawin, and John C. Walton*
UniVersity of Saint Andrews, EastChem, School of Chemistry, Saint Andrews,
Fife, KY16 9ST, United Kingdom
jcw@st-andrews.ac.uk
ReceiVed October 6, 2006
2-Halophenylalkyl-2-oxazolines with alkyl chain spacers of two to six C atoms (n ) 0-4) were prepared
and their SRN1-type reactions with several base systems examined. The best conditions to promote cyclo-
coupling to the corresponding benzocycloalkane derivatives involved use of LDA in THF. The precursors
with 3-C-atom and 4-C-atom spacers gave good yields of 2-(1′-phenylindan-1′-yl)-2-oxazolines and 2-(1-
phenyl-1,2,3,4-tetrahydronaphthalen-1-yl)-2-oxazoline, respectively. The major products from the precursor
with a 5-C-atom spacer were derivatives of benzocycloheptane in which the oxazoline group had undergone
a novel areneotropic migration from the end of the spacer to the benzo ring. The product from reaction
of the corresponding 2-C-atom precursor was a 9-oxazolinophenanthrene derivative. EPR spectroscopy
showed the intermediates of the LDA-promoted reactions to be radical anions of the product
benzocycloalkanes. This supported an SRN1-type chain mechanism involving initial production of aryl
radicals connected to azaenolate ions via the spacer groups. Intramolecular radical to carbanion coupling
then generated ring-closed benzocycloalkane radical anions that transferred an electron to more precursor.
Diastereoselective radical to carbanion cyclo-coupling reactions were carried out with 2-bromophenylpropyl
precursors containing chiral 2-oxazolines. The diastereoselectivity achievable was modest, but the product
diastereoisomeric Indane derivatives were easily separable by chromatography.
Introduction
character, its potential to routinely produce a range of ring sizes,
the possibility of smoothly forming quaternary centers, and
because the process works for a different range of functionality
The quest for alternatives to organotin compounds as media-
tors of free-radical-based syntheses1a,b has not only led to the
identification of several new families of reagents2a-g but also
heightened interest in new ring-closure strategies.2e,3a-c Coupling
together of a radical with a carbanion during an SRN1 process
is frequently fast and may even be diffusion controlled.4a,b
Interest in the intramolecular version of this reaction, i.e., cyclo-
coupling, is gathering momentum because of its radical chain
(2) (a) Lang, S.; Corr, M.; Muir, N.; Khan, T. A.; Schonebeck, F.;
Murphy, J. A.; Payne, A. H.; Williams, A. C. Tetrahedron Lett. 2005, 46,
4027-4030. (b) Murphy, J. A.; Khan, T. A.; Zhou, S.-Z.; Thomson, D.
W.; Mahesh, M. Angew. Chem., Int. Ed. 2005, 44, 1356-1360. (c)
Chatgilialoglu, C. Organosilanes in Radical Chemistry; Wiley: Chichester,
2004; Chapters 6 and 7. (d) Studer, A.; Amrein, S.; Schleth, F.; Schulte,
T.; Walton, J. C. J. Am. Chem. Soc. 2003, 125, 5726-5733. (e) In Radicals
in Organic Synthesis; Renaud, P., Sibi, S., Eds.; Wiley-VCH: Weinheim,
2001. (f) Jackson, L. V.; Walton, J. C. Chem. Commun. 2000, 2327-2328.
(g) Binmore, G.; Walton, J. C.; Cardellini, L. Chem. Commun. 1995, 27-
28.
(1) (a) Baguley, P. A.; Walton, J. C. Angew. Chem., Int. Ed. 1998, 37,
3072-3082. (b) Studer, A.; Amrein, S. Synthesis 2002, 835-849.
10.1021/jo0620720 CCC: $37.00 © 2007 American Chemical Society
Published on Web 01/03/2007
898
J. Org. Chem. 2007, 72, 898-911

1,3-Areneotropic Migration of Oxazolines
SCHEME 1. Preparation of 2-Halophenylalkyloxazoline
Precursors
from the classic 5-exo-ring closures onto alkene type acceptors.
Wolfe and Goehring used this strategy to prepare several types
of heterocycles via base-promoted carbanion formation from
aromatic amides. For example, indol-2-one derivatives were
made from N-acyl-o-haloanilines,5 and 2-phenyl-1,3-benzothia-
zoles were prepared by analogous routes.6 Six-membered and
larger rings have been made in good yields using methodology
of this type.7 Several alkaloids such as eupoulauramine, (()-
tortuosamine8 and an ergot-type alkaloid9 have also been
synthesized by routes in which radical anion cyclo-couplings
were the key steps. Rossi et al. reviewed the limited number of
such ring closures in the context of SRN1 reactions in general.10
Many bioactive molecules, such as the anticonvulsant ben-
zocaramiphen,11 contain an Indane, tetralin, or analogous
heterocyclic ring, embodying a quaternary center. The radical
carbanion cyclo-coupling strategy appeared promising as a
synthetic approach to such structures. Our primary objectives,
therefore, were to test the response of the SRN1 ring-closure
process to the size of the forming ring and explore ways of
controlling stereochemistry during this step.
Wolfe and co-workers showed that carbanions derived from
2-alkyl-2-oxazolines took part in intermolecular SRN1 reactions
with aromatic and heteroaromatic radicals, giving coupled
products in moderate to high yields.12 Suitably functionalized
2-oxazolines are useful chiral auxiliaries13 and readily undergo
further functional-group manipulations.14 These factors prompted
us to choose haloaromatics, substituted with alkyl spacers of
different lengths bearing 2-oxazoline units, as our model
precursors. Part of this research was reported in a preliminary
communication.15
required for assembly of our chosen 2-haloaryl-2-oxazoline
precursors. 2-Bromophenethyl iodide 1b was made from com-
mercial 2-bromophenylacetic acid by reduction with LiAlH4
followed by iodination with a mixture of iodine, imidazole, and
triphenylphosphine. To avoid debromination during the reduc-
tion step, the published procedure16 was modified to involve
only 1 equiv of LiAlH4 at 0 °C for 20 min. Chain elongation
was accomplished by a novel sequence that involved coupling
1b with 2,4,4-trimethyl-2-oxazoline, hydrolysis of the resulting
oxazoline 2 to afford the corresponding acid, and reduction of
the latter with LiAlH4. The desired 2-bromophenylbutyl iodide
1c was obtained in 6 steps from 2-bromophenylacetic acid in
41% overall yield (Scheme 1). 2-Bromophenylpropanoic acid
was obtained from 2-bromobenzaldehyde and Meldrum’s acid.17
Conversion to the corresponding iodide was achieved by the
same reduction/iodination process. Chain elongation followed
an oxazoline-mediated sequence similar to that of Scheme 1
and afforded 2-bromophenylpentyl iodide in 45% overall yield
in 7 steps.
The oxazolino precursors 3-7 were obtained by alkylation
of 2,4,4-trimethyl-2-oxazoline or 2-benzyl-4,4-dimethyl-2-ox-
azoline with the appropriate phenalkyl iodide (or bromide)
mediated with n-BuLi. During the alkylations control of solvent
polarity was found to be crucial. By using a 3:2 mixture of
THF:hexane, loss of halogen from the aryl ring was essentially
completely suppressed and the (2-haloaryl)alkyl-2-oxazolines
3-7 were obtained in satisfactory yields (Scheme 1).
Results and Discussion
Synthesis of Haloaromatics with Oxazolino-alkyl Arms.
2-Halophenylalkyl iodides with a range of chain lengths were
(3) (a) Walton, J. C. Top. Curr. Chem. 2006, 264, 163-200. (b) Walton,
J. C. Acc. Chem. Res. 1998, 31, 99-107. (c) Schiesser, C. H.; Wild, L. M.
Tetrahedron 1996, 52, 13265-13314.
(4) (a) Save´ant, J.-M. Tetrahedron, 1994, 50, 10117-10165 and refer-
ences therein. (b) Rossi, R. A.; Pierini, A. B.; Palacios, S. M. AdV. Free
Radical Chem. 1990, 1, 193-252.
(5) Goehring, R. R.; Sachdeva, Y. P.; Pisipati, J. S.; Sleevi, M. C.; Wolfe,
J. F. J. Am. Chem. Soc. 1985, 107, 435-443. Wolfe, J. F.; Sleevi, M. C.;
Goehring, R. R. J. Am. Chem. Soc. 1980, 102, 3646-3647.
(6) Bowman, W. R.; Heaney, H.; Smith, P. H. G. Tetrahedron Lett. 1982,
23, 5093-5096.
(7) Dandekar, S. A.; Greenwood, S. N.; Greenwood, T. D.; Mabic, S.;
Merola, J. S.; Tanko, J. M.; Wolfe, J. F. J. Org. Chem. 1999, 64, 1543-
1553.
(8) Goehring, R. R. Tetrahedron Lett. 1992, 33, 6045-6048. Goehring,
R. R. Tetrahedron Lett. 1994, 35, 8145-8146.
(9) Martin, S. F.; Liras, S. J. Am. Chem. Soc. 1993, 113, 10450-10451.
Liras, S.; Lynch, C. L.; Fryer, A. M.; Vu, B. T.; Martin, S. F. J. Am. Chem.
Soc. 2001, 123, 5918-5924.
(10) Rossi, R. A.; Pierini, A. B.; Pen˜e´n˜ory, A. B. Chem. ReV. 2003, 103,
71-167.
(11) DeHaven-Hudkins, D. L.; Allen, J. T.; Hudkins, R. L.; Stubbins, J.
F.; Tortella, F. C. Life Sci. 1995, 56, 1571-1576. Sommer, M. B.; Begtrup,
M.; Bøgesø, K. P. J. Org. Chem. 1990, 55, 4822-4827.
(12) Wong, J.-W.; Natalie, J. K.; Nwokogu, G. C.; Pisipati, J. S.;
Flaherety, P. T.; Greenwood, T. D.; Wolfe, J. F. J. Org. Chem. 1997, 62,
6152-6159.
Production of Five- and Six-Membered Benzocycloalkanes
by Radical to Carbanion Cyclo-Coupling. The projected
radical to carbanion cyclo-coupling process is shown in Scheme
2. Treatment of each precursor with base was expected to
generate the azaenolate moiety of 8. Transient intermediate 8
will be generated by electron transfer (ET) to the azaenolate
(13) Lutomski, K. A.; Myers, A. I. In Asymmertic Synthesis; Morrison,
J. D., Ed.; Academic Press: Orlando, 1984; Vol. 3, Chapter 3. Aitken, R.
A.; Kile´nyi, S. N. Asymmetric Synthesis; Blackie: London, 1992; Chapter
5.
(14) Ager, D. J.; Prakash, I.; Schaad, D. R. Chem. ReV. 1996, 96, 835-
(16) Faust, R.; Garrett, P. J.; Jones, R.; Yeh, L.-K.; Tsotinis, A.;
Panoussopoulou, M.; Calogeropoulou, T.; The, M.-T.; Sugden, D. J. Med.
Chem. 2000, 43, 1050-1061.
875.
(15) Roydhouse, M. D.; Walton, J. C. Chem. Commun. 2005, 4453-
4455.
(17) Toth, G.; Kover, K. E. Synth. Commun. 1995, 25, 3067-3074.
J. Org. Chem, Vol. 72, No. 3, 2007 899

Marshall et al.
SCHEME 2.
SRN1 Chain Mechanism for Cyclo-Coupling of
appreciable reduced product 12 plus minor styrenes 14 (X )
Br, H) (entry 11). The advantage of using 3 equiv of LDA was
demonstrated by reaction with 1 equiv of LDA, which gave a
mixture of products in lower yields (entry 12). This may be
because the LDA acts as an electron-transfer reagent as well as
a base (see below). The similar yields of 11 (n ) 1, R ) Ph)
and 11 (n ) 1, R ) H) under matching conditions in the LDA-
promoted reactions (cf. entries 11 and 14) demonstrated that
ring closure to give the quaternary center took place just as
readily as ring closure to give the tertiary center. A less polar
solvent system consisting of THF/hexane (2:1) was also tried
(entry 13). The Indane yield (41% after 3 h) was comparable
to that obtained in neat THF (57% after 6 h), so solvent polarity
was not crucial. It has been reported that Fe(II) salts significantly
enhanced the rates of some SRN1 reactions.18 We found,
however, that for 2-oxazoline precursors addition of FeCl2
(0.1-1 equiv) depressed the yield of 11 and made workup more
difficult (entries 15 and 16). Stable complexes are known to
form between Fe(II) salts and oxazolines,19 and this may account
for the deleterious effect. The cleanest and most efficient
conditions, i.e., those of entry 9, were adopted as standard
(method A) for most subsequent SRN1 processes.
Intermediates from (2-Haloaryl)alkyl-2-oxazolines
In many of the reactions reduced compound 12 was an
important byproduct, particularly when photostimulation was
employed. Possibly 12 was formed by hydrogen-atom abstrac-
tion by the aryl radical 9 from either solvent or other reactants
before cyclo-coupling could take place. Alternatively, LDA
might transfer an electron under irradiation to the aryl radical
intermediate, giving an anion which is subsequently protonated
to 12. The reaction mixtures were strongly basic, so the final
protonation step giving 12 probably occurred during workup.
For precursors 4a-d reacting with LDA, 2-bromostyrene and
styrene 14 (X ) Br, H) were also formed as byproducts (Table
1). The bromostyrene may have originated from deprotonation
of the substrate at the benzyl position followed by â-elimination
of the oxazolinyl moiety 15 (Scheme 2). The 2-bromostyrene
can itself undergo electron capture and loss of bromide to afford
the 2-vinylphenyl radical, which will rapidly abstract an H atom
to give the observed styrene. The fact that styrenes were not
observed for precursors with n * 1 fits this pattern because
â-elimination of their nonstabilized oxazolinylalkyl moieties
would be thermodynamically much less favorable.
The analogous coupling reaction to produce a 6-membered
ring was next examined. Treatment of bromophenylbutyl-2-
oxazoline 5 with 3 equiv of LDA for 48 h gave a 55% yield of
the tetralin derivative 11 (n ) 2, R ) Ph) (Table 1, entry 17),
and 50% was obtained with photostimulation in 6 h (entry 18).
No reduced product 12 or styrenes were detected. The slightly
lower yield of tetralin compared to Indane might possibly be
due to a slower rate of six-membered ring closure. In this case
crystals of the tetralin 11 (n ) 2, R ) Ph) were obtained and
X-ray structure analysis showed a single conformer with the
six-membered ring in a half-chair having the oxazoline in the
pseudo-equatorial orientation and the phenyl in the pseudo-axial
orientation. As expected, the C-C bonds to the quaternary C
atom (1.542 Å to benzo, 1.547 Å to ring CH2, 1.547 Å to Ph,
and 1.532 Å to oxazoline) were significantly longer than
unstrained C-C bonds.
and should very rapidly eject bromide to produce the distonic
radical anion 9. Rapid cyclo-coupling of the reactive centers in
9 closes the ring and yields benzocycloalkane radical anion 10.
SET from 10 to another precursor azaenolate then yields a
functionalized benzocycloalkane 11 while also propagating the
chain.
Several sets of conditions for promoting ring closure,
embodying different bases and solvents, were examined. One
of the traditional conditions for SRN1 processes utilizes t-BuOK
as base in DMSO. However, when 4a or 4d was treated with
this base, no cyclized product was obtained, even in the presence
of added Et3N as a photoelectron transfer agent (Table 1, entries
1 and 2). Possibly efficient deprotonation of the precursors did
not take place. A solution of dimsyl sodium was prepared from
sodium hydride and DMSO, and precursor 4a was added. On
photolysis at rt for 4 h mainly the reduced product 12 (n ) 1,
R ) Ph) was obtained, but on photolysis at 100 °C a 66% yield
of the desired Indane 11 (n ) 1, R ) Ph) was observed (Table
1, entries 3 and 4). Alkali metal amides in liquid ammonia were
also tested as bases. The oxazolines in THF were added to the
gray suspensions of metal amide in liquid NH3, and the solutions
were irradiated for 1-2 h in quartz apparatus. When bromide
4a was reacted a poor yield of the Indane (16%) was obtained
together with nearly twice as much of the reduced product (entry
5). The reactants were diluted by a factor of 10, but no Indane
was observed (entry 6). The chloride 4d gave only a trace
amount of reduced product together with mainly unreacted
precursor (entry 7).
LDA in THF has been used as an alternative base/solvent
pair for SRN1 reactions, and the results we obtained with this
combination under various conditions also appear in Table 1.
For precursor 4a the cyclo-coupling proceeded smoothly at room
temperature without photostimulation, but 48 h was needed for
full consumption of the reactant. This was the cleanest and most
efficient reaction and afforded the Indane 11 (n ) 1, R ) Ph)
in 75% yield (compare entries 7-9). The reaction was acceler-
ated by use of UV irradiation but was not so clean, giving
(18) Galli, C.; Gentili, P. J. Chem. Soc., Perkin Trans. 1993, 2, 1135-
1140. Baumgartner, M. T.; Pierini, A. B.; Rossi, R. A. J. Org. Chem. 1999,
64, 6487-6489.
(19) Bart, S. C.; Hawrelak, E. J.; Schmisseur, A. K.; Lobkowsky, E.;
Chirik, P. J. Organometallics 2004, 23, 237-246.
900 J. Org. Chem., Vol. 72, No. 3, 2007

1,3-Areneotropic Migration of Oxazolines
TABLE 1.
S
RN1-Type Reactions of Propyl- (4) and Butyl-2-oxazolines (5) with Various Basesa
cyclized, 11
reduced, 12
styrenes, 14
entry
precursor
conditions
(n ) 1, R ) Ph) %
(n ) 1, R ) Ph) %
(X ) Br, H) %
1
2
3
4
5
6
7
8
4a (X ) Br)
4d (X ) Cl)
4a (X ) Br)
4a (X ) Br)
4a (X ) Br)
4a (X ) Br)
4d (X ) Cl)
4a (X ) Br)
4a (X ) Br)
4a (X ) Br)
4a (X ) Br)
4a (X ) Br)
4a (X ) Br)
4c (X ) Br, R ) H)
4a (X ) Br)
4a (X ) Br)
5
t-BuOK, Et3N, DMSO, UV, rt, 4 h
t-BuOK, Et3N, DMSO, UV, rt, 3 h
NaH, DMSO, UV, rt, 4 h
0
0
0
66
16
0
0
2
0
96
0
31
1
1
0
0
0
0
0
0
0
0
0
0
NaH, DMSO, UV, 100 °C, 4 h
KNH2, liq. NH3, UV, -33 °C, 1 h
KNH2, liq. NH3, UV, -33 °C, 1 h, 1/10 dil.c
LiNH2, liq. NH3, UV, -33 °C, 1 h
LDA (3 equiv), THF, rt, 20 h
50
8
9
LDA (3 equiv), THF, rt, 48 h
75b
2b
5
10
11
12
13
14
15
16
17
18
LDA (3 equiv), THF, reflux, 3h
LDA (3 equiv), THF, UV, rt, 6 h
LDA (1 equiv), THF, rt, 2 h
45
57b
15b
13
12b
0
16
17
0
7b
2
11
LDA(3 equiv), THF/hexane,d rt, 3 h
LDA (3 equiv), THF, UV, rt, 6 h
LDA (3 equiv), THF, FeCl2 (1 equiv), rt, 1 h
LDA (3 equiv), THF/hexane,d FeCl2 (0.1 equiv)
LDA (3 equiv), THF, rt, 48 h
41b
6b
0
58b (R ) H)
23
8
8
0
0
25
55b (n ) 2)
50b (n ) 2)
5
LDA (3 equiv), THF, UV, rt, 6 h
0
a Yields (mol %) determined by NMR except as indicated otherwise. In most experiments the shortfall in product mass was shown to be unreacted
starting material. b Isolated yields. c All reactants diluted by 1/10. d THF:hexane ) 2:1.
SCHEME 3. Products from LDA-Promoted Reactions
(methods A and B) of Pentyl 6 and Hexyl 7 Precursors
spectroscopy to be the saturated analog 17. Closure of the seven-
membered ring and attachment of the oxazoline group to the
1
aromatic ring were confirmed by COSY H-1H and HMBC
1H-13C correlations (see Supporting Information). One of the
minor products was shown to be 19, derived by reductive
dehalogenation. The second minor component crystallized, and
it was shown by X-ray diffraction to be the fluorene derivative
18. It was evident that closure to the seven-membered ring had
taken place but that this was followed, in the case of the major
products, by migration of the oxazoline group onto the aromatic
ring.
The homologous hexyl-precursor 7 was reacted with LDA
using the conditions of both methods A and B. In each case a
complex mixture of products was obtained that could not be
separated by chromatography. Individual fractions from the
1
columns were examined by H NMR spectroscopy and GC-
MS. Comparison of the spectroscopic data with that obtained
from products 16-19 demonstrated that the major products were
most probably the eight-membered ring analogs, i.e., benzocy-
clooctene 20, benzocyclooctane 21 and the analogous fluorene
derivative 22, accompanied by several minor unidentified
byproducts. Thus, a similar areneotropic migration of the
oxazoline unit took place, although the process was less efficient.
Rearrangements Accompanying Seven- and Eight-Mem-
bered Ring Formation. When bromophenylpentyl-oxazoline
6 was treated with LDA and the solution warmed to room
temperature and stirred for 48 h in daylight (method A), the
product mixture contained two major and two minor compo-
nents,20 none of which was the expected 5-phenyl-5-oxazoli-
nobenzocycloheptene 11 (n ) 3, R ) Ph). The reaction was
also carried out over 6 h, in quartz apparatus, with UV irradiation
(method B) and yielded the same products, though in different
proportions (Scheme 3). The four components were separated
by chromatography and characterized by spectroscopic and other
means.
The first eluting major component was found to be benzo-
suberene derivative 16 in which an unprecedented migration
of the oxazoline group from the end of the pentyl chain to the
aromatic ring had taken place. Crystals of this component were
eventually obtained, and X-ray analysis fully confirmed this
structure (see Supporting Information). The second major
component did not crystallize but was shown by 2D NMR
EPR Spectroscopic Study of the Intermediates. To probe
the nature of the intermediates, the reactions of three homolo-
gous bromophenylalkyl dimethyl-2-oxazoline precursors 4a, 5,
and 6, promoted by LDA in THF solution, were examined by
9 GHz EPR spectroscopy. Solutions (2.5 and 1.5 mM) were
prepared with 3 equiv of LDA in the same way as for the
preparative work, and 0.02 mL aliquots were transferred to
quartz capillary tubes (1 mm i.d.) and hence to the resonant
cavity of the EPR spectrometer. When even traces of oxygen
were present the spectrum of a nitroxide (aminoxyl) radical (g
) 2.0064) was obtained in each case. Comparison of the
hyperfine splittings (hfs) a(N) ) 14.6 G and a(2H) ) 4.2 G at
290 K with the literature21 showed that this was due to the di-
isopropyl nitroxide 24. Minor signals from other nitroxides
accompanied this spectrum.
(20) In some experiments a fifth minor component, 2-(1,5-diphenylnonan-
5-yl)4,4-dimethyl-2-oxazoline, derived from reaction of an intermediate with
excess butyl lithium (or degradation product there from) was also obtained.
(21) Hudson, A.; Hussain, H. A. J. Chem. Soc. (B) 1967, 1299-1300.
Flesia, E.; Surzur, J.-M. Tetrahedron Lett. 1974, 123-126.
J. Org. Chem, Vol. 72, No. 3, 2007 901

Marshall et al.
TABLE 2. Comparison of Experimental EPR with DFT Computed EPR Parametersa
hydrogen and nitrogen atom hfs/G
species/method
(basis set)
1
2
3
4
5
6
7
8
9
10
4.2
-3.6
-3.2
3.4
11
1.5
-1.4
-0.1
12
0.1
-0.7
1.5
13
4.2
-4.0
-3.0
3.4
10 (n ) 1)/exptl. EPRb
10 (n ) 1)/DFT(6-31G(d))
10 (n ) 1)/DFT(epr-ii)
25e /exptl EPRc
5.7
-5.6
-4.4
1.5
1.0
0.5
4.2
-4.4
-2.8
3.4
4.2
-3.8
-3.6
3.4
1.5
-1.0
-1.0
1.5
1.5
1.0
1.4
-0.7
-0.6
0.1
0.4
0.2
-0.6
1.5
1.5
25e /DFT(6-31G(d))
25e /DFT(epr-ii)
-0.2
0.0
1.8
-1.4
-1.0
-6.4
-5.1
-4.0
3.3
1.9
0.5
0.4
-0.2
-0.3
-0.2
0.0
-0.5
-0.3
3.4
-3.1
-2.7
1.8
-0.2
-0.3
5.1
-0.5
-0.4
1.8
3.1
-2.7
3.4
1.7
25a /exptl EPRc
1.8
25a /DFT(6-31G(d))
25a /DFT(epr-ii)
-1.4
-1.2
N
-0.3
-0.1
2
0.0
-0.1
3
-1.2
-1.2
4
0.3
0.3
5
0.7
0.4
6
0.0
0.0
7
0.0
0.0
8
-1.4
-1.1
9
-0.4
-0.5
10
-7.1
-6.4
11
0.7
0.6
12
-3.7
-3.2
13
27/DFT(6-31G(d))
27/DFT(epr-ii)
4.9
3.8
-3.4
-2.3
0.3
-0.4
-8.1
-7.1
-0.5
-0.5
-1.6
-1.5
0.2
0.3
-0.2
-0.2
-0.1
-0.1
0.1
0.1
0.0
0.0
0.1
0.1
0.0
0.1
a Numbering shown in structure 26; DFT computations using UB3LYP with a 6-31G(d) or epr-ii basis set. Tentative assignments of the experimental hfs
to individual H atoms are shown. b g ) 2.0030. c g ) 2.0028.
SCHEME 4. Formation of Di-isopropyl Nitroxide
Detection of di-isopropyl nitroxide in the presence of oxygen
is evidence in favor of the radical-mediated chain. The nitroxide
is probably derived from the di-isopropylaminyl radical gener-
ated from LDA (Scheme 4). The i-Pr2N- anion could transfer
an electron to a precursor molecule, thus initiating a new chain
and generating the i-Pr2N• radical together with species 8.22 This
amounts to an initiation step and would account for the fact
that excess LDA was always found to be advantageous. The
i-Pr2N• radical will capture dioxygen to give the corresponding
peroxyl radical, and coupling of two peroxyls will produce the
observed persistent di-isopropyl nitroxide together with oxygen.
Radical 24 is a secondary nitroxide and will decay slowly by
disproportionation to the corresponding oxime and nitrone.
However, it will also inhibit the SRN1 chain, and this is consistent
with the observed sensitivity of the reactions to oxygen.
When oxygen was rigorously excluded from the reaction of
FIGURE 1. EPR spectrum obtained on treatment of 6 with LDA: (top)
the propyl precursor 4a, a longer lived species having g )
experimental at 195 K in THF (second derivative); (bottom) computer
2.0030 was observed. The isotropic EPR signal was broad at
simulation including both species.
300 K, and its intensity increased with UV photolysis. On
lowering the temperature the intensity increased again, and best
well resolved, and the second-derivative presentation (Figure
resolution was obtained at temperatures of about 205 K. This
1) suggested that two similar paramagnetic species were present.
A satisfactory simulation including both species (R ) 0.922)
was obtained, with the parameters shown in Table 2, and this
is also displayed in Figure 1.
The g factors and magnitudes of the hfs of these species
suggested they were aromatic radical anions. Cyclo-coupling
of the aromatic radical carbanions derived from 4a, 5, and 6
spectrum was observable for many hours, although its intensity
gradually weakened. Broadly similar long-lived EPR spectra,
with essentially identical g factors, were obtained on treatment
of precursors 5 and 6 with LDA in THF and illumination with
UV light. A good computer simulation (R ) 0.991) was obtained
for the spectrum from 4a with the hfs listed in Table 2. The
spectrum obtained from precursor 5 was considerably broader
will give radical anions that contain a benzocycloalkane unit
than the others with asymmetry caused by the presence of an
substituted by phenyl and oxazoline rings, i.e., 10. EPR data
additional signal originating from the quartz tube; a reliable
simulation could not be derived. However, the g factor (2.0031),
spectral width, and lifetime all suggested the species responsible
from the literature for the model radical anions of diphenyl-
methane23 and Indane24 (which had been prepared by treatment
of the corresponding hydrocarbons with K metal) were similar
for this spectrum was analogous to that derived from the other
in terms of g factor and hfs magnitudes. The experimental hfs
precursors. The spectrum obtained from pentyl precursor 6 was
(22) Alternatively, i-Pr2N• could be formed by some other radical (or
radicals) in the system (X•) abstracting an H atom from adventitious i-Pr2-
NH.
(23) Gerson, F.; Martin, W. B., Jr. J. Am. Chem. Soc. 1969, 91, 1883-
1891.
(24) Bauld, N. L.; Farr, F. R. J. Am. Chem. Soc. 1974, 96, 5633-5634.
902 J. Org. Chem., Vol. 72, No. 3, 2007

1,3-Areneotropic Migration of Oxazolines
SCHEME 5. Radical Anions Derived from Precursor 6
SCHEME 6. Mechanism for Oxazoline Migration
were tentatively assigned to specific H atoms of 10 with the
aid of DFT computations (see below) and are compared in Table
2. A noteworthy feature is that the spin density is equally
distributed to both phenyl rings in the radical anion of
diphenylmethane. In radical anions 10 spin density is also
distributed to both rings but the two rings are not equivalent,
so the spin is not equally partitioned.
The radical anion concentrations, measured by double
integration of the EPR signals at 200 K, ranged from 4 × 10-6
for 10 (n ) 3) to 3 × 10-4 M for 10 (n ) 1). The concentrations
decreased at higher temperatures, but at 290 K (temperature of
the preparative experiments) the ratios of the radical anion to
initial precursor were found by extrapolation to be 1.5%, 0.8%,
and 0.6% for 10 (n ) 1)/4a, 10 (n ) 2)/5, and 10 (n ) 3)/6,
respectively. These values represent a significant proportion of
the initial concentration of the precursor molecules, and hence,
the EPR spectra are not from minor stray species but important
intermediates.
The data are consistent with the SRN1 chain process of
Scheme 2 (Scheme 6 for 6) with a modest chain length. The
observed increase in radical concentration at lower temperatures
is consistent with the expected slower chain propagation at lower
temperatures. In particular, Scheme 2 proposes that the radical
anions 10 decay by bimolecular electron transfer to more
precursor (4, 5, or 6). This step would slow down as temperature
was lowered, and hence, if this is the rate-limiting step, the
concentration of the radical anions would be expected to build
up, exactly as observed.
The computer fits for the spectra obtained with precursor 6
were “soft” because of the limited S/N but indicated a major
species (60%) and a minor species (40%). In this case rapid
cyclo-coupling within 9 will give the ring-closed radical anion
as a pair of enantiomers, each of which can exist in two
conformations, 25e and 25a, having the oxazoline group pseudo-
equatorial and -axial, respectively (Scheme 5). DFT computa-
tions,25 carried out on 25e and 25a at the UB3LYP/epr-ii//
UB3LYP/6-31G(d) level of theory,26 predicted that 25a was 3.9
kcal/mol lower in energy than 25e. It may be expected, therefore,
that ring inversion will be slow at 195 K, and hence, the two
EPR spectra correspond to the two conformers. The DFT
computations showed that the SOMO of 25a was associated
mainly with the two aromatic rings. The SOMO of 25e was
similar but had higher amplitude on the phenyl ring. These
properties were reflected in the computed hfs which correlated
satisfactorily with the experimental EPR data (Table 2).
The equatorial oxazoline moiety is sited comparatively close
to the aromatic ring in conformer 25e so that 1,3-migration of
the oxazoline group toward the “southernmost” C atom (C-1)
of the aromatic ring can be envisaged via a four-membered
transition state such as 28. 1,3-Proton transfer in radical anion
29 will yield the rearranged radical anion 27. DFT computations
(UB3LYP/epr-ii//UB3LYP/6-31G(d)) on 27 predicted the over-
all transformation from 25e to 27 to be strongly exothermic
(by 25 kcal/mol). The computed SOMO for 27 was mainly
confined to the aromatic ring and the oxazoline unit with very
little amplitude on the Ph ring. Consequently, only four
hydrogen hfs >0.1 mT and a comparatively large N-hfs were
obtained (Table 2). The predicted EPR spectrum of 27 is,
therefore, very different from either of the experimentally
observed spectra, in agreement with our assignment of them to
25a and 25e. Transfer of an electron from rearranged radical
anion 27 to more precursor 6 will yield product 17 and
simultaneously propagate the chain (Scheme 6). The most
plausible mechanism for formation of the unusual alkene 16
involves lithiation of product 17 at the bisaryl C atom by the
excess base followed by loss of LiH. Precedents for this kind
of elimination are known,27 and it is supported by the greater
yield of 16 from the longer running experiment (method A)
where more 17 would have been available.
In conformer 25a it is the equatorial phenyl ring that is
comparatively close to the southernmost C-1 of the aromatic
ring. It follows that cyclo-coupling from C-1 to an ortho-C atom
of the phenyl ring would afford a ring-closed radical anion,
containing the fluorenyl unit, that could readily be transformed
to product 18 by electron transfer and rearomatization by
oxidative loss of two H atoms.
(25) Frisch M. J. et al. Gaussian 03, Revision A.1; Gaussian Inc.:
Pittsburgh, PA, 2003 (see Supporting Information for full citation).
(26) Barone, V. In Recent AdVances in Density Functional Theory;
Chong, D. P., Ed.; World Scientific Publishing Co.: Singapore, 1996.
(27) Withnall, R.; Dunkin, I. R.; Snaith, R. J. Chem. Soc., Perkin Trans.
1994, 2, 1973-1977 and references cited therein.
J. Org. Chem, Vol. 72, No. 3, 2007 903

Marshall et al.
SCHEME 8. Preparation of a Dihydroindene-1-one
SCHEME 7. Formation of a Phenanthrene Derivative from
Ethyl Precursor 3
Derivative from Compound 3
SCHEME 9. Stereochemistry of the Cyclo-Coupling Step
A) none of 31 was obtained but instead 9-oxazolino-phenan-
threne derivative 35 (88%) was isolated. A plausible mechanism
to account for this involves initial formation of azaenolate 30
(analogous to 9 proposed for the longer chain precursors). Cyclo-
coupling to give the radical anion of 31 appears to be disfavored
because of the strain in the four-membered ring. Instead cyclo-
coupling with closure to a six-membered ring occurred via the
resonance structure 32 affording radical anion 33. Transfer of
an electron from 33 to more 3 will propagate an SRN1-type chain
and generate dihydrophenanthrene derivative 34. In view of the
strongly basic solution, aromatization to 35 probably occurs as
shown in Scheme 7 by deprotonation and loss of lithium hydride.
Alternatively, the mechanism might involve oxidative loss of
two H atoms, as proposed above for formation of fluorene
derivative 18. A small sample of the reaction mixture was
examined by EPR spectroscopy, which showed a strong and
long-lived spectrum (g ) 2.003). The signal was not well
enough resolved for definitive analysis, but the observation
supports the idea that radical anions were the main intermediates.
We also carried out an LDA-promoted reaction with com-
pound 3 but with an additional 1 equiv of n-BuLi added to the
THF solution. Interestingly, the reaction took a completely
different course in this case and gave spiro-Indane derivative
36 in good yield. Hydrolysis of the oxazoline moiety in 36 with
dilute HCl yielded dihydroindene-1-one derivative 37, and the
latter was fully characterized. A possible mechanism for
formation of 36 involves rapid butylation of the azaenolate anion
(possibly by attack on BuBr formed from the n-BuLi). Subse-
quent electron transfer and loss of bromide would generate an
aryl-type radical that could undergo a normal 5-exo-cyclization
onto the CdN bond of the oxazoline moiety. Final H-atom
abstraction from the THF by the resulting cyclic aminyl radical
would then yield 36.29
The emerging pattern is that cyclo-coupling of aryl radicals
with azaenolate-type carbanions readily takes place for ring sizes
from five- to eight-membered. This is in good accord with the
known rapidity of intermolecular radical anion coupling.28 For
benzocycloalkane radical anions containing seven- and eight-
membered rings, R-substituents take part in subsequent unique
rearrangements in which the 1,3-oxazoline group migrates to
the benzo moiety. An intriguing question is why the seven- and
eight-membered benzocycloalkane radical anions rearrange
whereas the five-and six-membered analogs do not. Stereoelec-
tronic factors are likely to be important. In the ring-closed radical
anions 10 much of the unpaired electron density resides in the
benzo ring. Increasing the adjacent cycloalkane ring size brings
the pseudo-equatorial substituent closer to the benzo ring. For
example, the DFT-computed, direct, through-space, distances
from C(1) of the benzo ring to the C(dN) atom of the oxazoline
unit were 3.41, 3.16, 2.79, and 2.91 Å in the equatorial
conformers of the five-, six-, seven-, and eight-membered ring
anions, respectively. The considerably closer approach for the
seven- and eight-membered intermediates would mean less strain
in the transition states for migration (28), and hence, their
energies would be lowered.
Attempted Ring Closure to a Benzocyclobutane Deriva-
tive. The precursor 3 with the shorter ethyl-type spacer was
also examined to see if radical anion cyclo-coupling could take
place to form benzocyclobutene 31 containing a four-membered
ring. However, when precursor 3 was treated with LDA (method
Diastereoselective Radical Carbanion Cyclo-Coupling
Reactions. The stereochemistry of the ring-closure step is an
important consideration. It seemed possible that diastereo-
selectivity could be achieved in our intramolecular SRN1
reactions using precursors containing chiral 2-oxazolines. The
(28) Alternative mechanisms involving initial formation of arynes are
possible but are less likely because the EPR evidence proves the participation
by radical anions because of the sensitivity of the reactions to oxygen and
(29) Alternatively, after the butylation step, debrominative lithiation of
the aryl ring would lead to an aryllithium that could undergo nucleophilic
attack at the CdN bond of the oxazoline.
UV radiation and because of literature precedents for analogous SRN
processes mediated by LDA (see refs 7, 12, and 15).
1
904 J. Org. Chem., Vol. 72, No. 3, 2007

1,3-Areneotropic Migration of Oxazolines
TABLE 3. Yields and Diastereoselectivities in Cyclo-Coupling
Reactionsa of Chiral Oxazolines
SCHEME 10. Cyclo-Coupling with Chiral Oxazolines
entry
precursor (R1)
method/conditions
yield 47% de%
1
2
3
4
5
6
7
8
(S)-40 (i-Pr)
(S)-40 (i-Pr)
(S)-40 (i-Pr)
(S)-44 (Chm)e
(S)-44 (Chm)e
(S)-41 (t-Bu)
(S)-41 (t-Bu)
A
72
50
55
ndf
ndf
60
85
nd
nd
nd
nd
nd
42b
48c
18b
16
A at 0 °C
B
A
B
B
A
16
33c
16d
2
(S)-41 (0.036M) A + (-)-50 (3.5 equiv)
9
(S)-41 (0.1 M)
(S)-41 (0.1 M)
(R)-41 (0.1 M)
(R)-41 (0.1 M)
A + (-)-50 (3.5 equiv)
A + (-)-50 (10 equiv)
A + (-)-50 (3.5 equiv)
A + (-)-50 (10 equiv)
-4
-14
21
10
11
12
14
a Method A: LDA (3 equiv), THF, rt, 48 h. Method B: LDA (3 equiv),
THF, rt, UV, 6 h. b From the ratio of isolated products. c Obtained by GC-
MS. d Obtained from NMR spectra. e Chm ) cyclohexylmethyl. f nd ) not
determined.
stereocenter becomes sp2 on deprotonation. The SRN1 reactions
of the chiral precursors were then examined using the previously
developed methods A and B and Indane product de’s were
determined using 1H NMR spectroscopy and/or chiral GC-MS
(Table 3).
The SRN1 reaction of isopropyl precursor (S)-40 without
photostimulation (entry 1) gave a 72% yield of the two Indane
diastereoisomers, which was comparable to that obtained for
the dimethyl analogue under these conditions (Table 1, entry
9). A moderate selectivity (42% de) was obtained, and this
increased to 48% (entry 2) when the reaction was carried out
at 0 °C. An attempt to carry out the reaction at -33 °C resulted
in an intractable mixture in which no product was evident. When
the reaction mixture was UV irradiated for 6 h (entry 3) both
the yield and selectivity decreased. The two diastereoisomers
were easily separable by column chromatography. Neither
oxazolino-Indane diastereoisomer crystallized, but mild hy-
drolysis of the major isomer with dilute HCl gave the â-hy-
droxyamide (S,S)-48. Crystals of this compound were obtained,
and an X-ray structure revealed the configuration to be S,S (see
Supporting Information), and thus the absolute configuration
of the major oxazolino-Indane was also (S,S). This suggested
that the major conformer of the azaenolate was (S,S)-39. Chiral
oxazoline (S)-41 containing the larger tert-butyl group was also
reacted under similar conditions. The 85% yield obtained in
the unstimulated experiment (entry 7) was the highest achieved
in our intramolecular SRN1 series. However, the de’s obtained
were actually poorer than those with the isopropyl substituent
(presumably the proportion of 38E was lower in this case).
An attempt was made to carry out the reaction at -33 °C
but as before, this resulted in an intractable mixture. The
cyclohexylmethyl-substituted oxazoline (S)-44 was also exam-
ined, but under both sets of conditions only about one-third of
the precursor was consumed and the stereoselectivity was poor
(entries 4 and 5). A possible reason might be poor solubility in
the reaction medium.
ring-closure step involves coupling of an aryl radical to the
azaenolate moiety, and clearly any diastereoselectivity will
depend on one of the two possible azaenolate anion conforma-
tions 38E and 38Z predominating. It was anticipated that
approach from below azaenolate 38E (shown) would be
sterically favored by bulky oxazoline substituents R and would
give the (S,S)-39-diastereomer
However, approach from aboVe in azaenolate 38Z (shown)
would be sterically favored by bulky substituents in the
oxazoline and would lead to the (R,S)-39-diastereomer. Thus,
any stereoselectivity will depend on the relative concentrations
of the azaenolates 38E and 38Z.
To test this possibility, a set of chiral 3-(2-bromophenyl)-
propyl-2-oxazolines 40-46 was prepared. Chiral 2-benzyl-2-
oxazolines were made by condensing benzyl acetimidate ethyl
ester hydrochloride (BnC(OEt)dNH‚HCl) with the appropriate
chiral 1,2-aminoalcohol.30 The desired precursors were then
obtained in yields ranging from 40% to 88% by n-BuLi-
promoted alkylation of each oxazoline with 2-bromophenylethyl
iodide (1b). Each precursor was obtained as a mixture of
diastereoisomers (at the 3 position), but this should not affect
the outcome of the SRN1 cyclo-coupling because the C(3)
When the phenyl-substituted oxazoline (S)-42 was treated
with LDA (method A) the only product (apart from unreacted
starting material) was probably oxazole derivative 49 (60%).
Formation of 49 may be a consequence of the acidity of the
proton at C(4) of the oxazoline ring in (S)-42. Removal of this
proton by the base would give a delocalized anion that could
undergo lithiation and subsequent loss of LiH to yield 49.
Alternatively, SET from the anion could give the corresponding
benzyl-type radical, which could convert to 49 by oxidative
(30) Gant, T. G.; Meyers, A. I. Tetrahedron 1994, 50, 2297-2360.
J. Org. Chem, Vol. 72, No. 3, 2007 905

Marshall et al.
removal of an H atom. Chiral GC-MS was employed in an
attempt to determine any selectivity at the Indane site, but only
one peak of the required mass was observed, so the result was
inconclusive. Precursor (S,S)-45, derived from the classic
Meyers’ oxazoline, was reacted with LDA (method A), but only
an intractable mixture of products was obtained. The spectra
of the mixture indicated many aromatic environments and
complete loss of the signals of the oxazoline moiety. This
suggested that the 2-oxazoline ring proton(s) had been lost.
Similar behavior was observed for the diphenylmethanol deriva-
tive (S)-46. An extra equivalent of LDA was used to ensure
deprotonation of the hydroxyl group, but only a complex mixture
resulted. The benzyl-substituted oxazoline (S)-43 also gave a
multitude of products with either method A or method B. When
only 1 equiv of LDA was employed a 2:1 mixture of starting
material and the reduced product (analogous to 12) was obtained.
An excess of the weaker base t-BuOK in THF gave no reaction
at rt for 48 h and gave only reduced product on UV photo-
stimulation for 6 h. The base LiHMDS in THF was also tried
with a short irradiation period (1 h) but to no avail, yielding
only an intractable mixture. We concluded that oxazolines with
substituents that increase the acidity of the ring protons degrade
under the SRN1 reaction conditions, resulting in a range of
products and loss of stereoinduction.
(-)-Spartiene 50 has been used as a bidentate ligand for
lithium cations in several stereoselective syntheses with varying
success.31 When a chiral lithium ligating agent is used in
conjunction with a chiral auxiliary in the precursor there is the
possibility of either enhancement (matched pair) or decrease/
reversal in selectivity (mismatched pair). In an attempt to
improve the diastereoselectivity in our SRN1 ring closures several
experiments were carried out using different amounts of 50.
The (S)-tert-butyl substrate (S)-41 was chosen for study because
it delivered the cleanest SRN1 ring closure. The reaction was
carried out using method A conditions with 3.5 equiv of 50,
but this gave a reduced selectivity (2% de, Table 3, entry 8)
compared to the result without spartiene (16%, entry 7). When
the substrate concentration was increased (0.1 M) a reversal of
selectivity (-4%) was obtained, and on increasing the amount
of spartiene to 10 equiv, the reverse selectivity increased to
-14% (Table 3, entries 9 and 10). Clearly this data indicated
that (S)-41 and (-)-sparteine formed a mismatched pair. The
opposite precursor diastereoisomer (R)-41 was prepared from
(R)-tert-leucinol, and the SRN1 reactions were carried out with
3.5 and 10 equiv of 50. A slightly increased selectivity was
observed when 3.5 equiv of 50 was employed (21%, entry 11),
but a poorer selectivity was obtained in the case of 10 equiv
(entry 12). The final experiment had to be carried out on a very
small scale due to the expense of the substrate, and serious
solvent loss occurred during the reaction; hence, the result may
not represent the true selectivity. However, entries 11 and 12
showed that although (R)-41 and 50 formed a matched pair,
the enhancement in selectivity was very modest.
as ring closure forming a tertiary center. The reactions lent
themselves to study by EPR spectroscopy, which supported an
SRN1-type chain mechanism involving a radical to carbanion
cyclo-coupling step producing a 1-phenyl-benzocycloalkane
radical anion as the main intermediate. Interestingly, although
the phenyl and benzo groups in individual radical anions were
separated by sp3-hybridized C atoms, the unpaired electron was
distributed to both aromatic rings. For benzocycloalkane radical
anions containing seven- and eight-membered rings (pentyl and
hexyl spacers), R-oxazoline substituents took part in subsequent
unique rearrangements toward the adjacent benzo unit, i.e., 1,3-
oxazoline migration (cyclo-coupling for phenyl substituents).
This can be attributed to the closer approach of the 2-oxazoline
substituent to the benzo ring in the pseudo-equatorial conformers
of the seven-membered rings. Overall the alkyl arms served to
escort the comparatively remote oxazoline units into the orbit
of the arene radical anions. With the bromophenylethyl-
oxazoline precursor (ethyl spacer) cyclo-coupling to produce a
four-membered ring did not compete. Instead, coupling occurred
with the phenyl ring to form the six-membered ring of a
phenanthrene derivative. Diastereoselective cyclo-coupling of
2-bromophenyl)propyl precursors containing chiral oxazolines
yielded diastereoisomeric Indanes. However, the maximum
selectivity achieved with a 4-isopropyl substituent in the
oxazoline (48% de) was modest. A redeeming feature was the
separability of the two diastereoisomers by conventional chro-
matography. It was found that substituents that enhanced the
acidity of the oxazoline protons were not tolerated.
Experimental Section
General Method for the Preparation of 2-(Arylalkyl)-2-
oxazolines. n-BuLi (2.5 M) in hexanes (1.05 equiv) was added
dropwise to a stirring solution of the 2-oxazoline (1 equiv) in dry
THF (0.75 mL/mmol) and dry hexane (0.38 mL/mmol) under
nitrogen at -78 °C. After stirring for 5 min o-halophenethyl halide
(1.05 equiv) in dry THF (0.38 mL/mmol) and dry hexane (0.2 mL/
mmol) was added over a 10 min period. The solution was then
allowed to warm to room temperature slowly over 3 h; then water
(3 × 2 mL/mmol) was added. The aqueous layer was extracted
with ether (3 × 3.2 mL/mmol), and the combined organic layers
were washed with water (3.2 mL/mmol), dried (MgSO4), and
concentrated to give the crude substituted 2-oxazoline. Purification
was achieved by column chromatography using SiO2 or Al2O3 and
ethyl acetate/hexanes mixtures as eluant.
2-[2-(2-Bromophenyl)-1-phenylethyl]-4,4-dimethyl-2-oxazo-
line (3). Yellow/orange oil which crystallized as orange crystals
(52%), mp 62 °C; Rf (SiO2, 8:2 hexane:EtOAc) 0.35; IR νmax(film)/
1
cm-1 1664 (CdN); H NMR δ 1.20 (3H, s, CH3), 1.26 (3H, s,
CH3), 3.19 (1H, dd, J ) 6.6 and 13.5 Hz, CHAHB), 3.49 (1H, dd,
J ) 9.0 and 13.5 Hz, CH), 3.86 (2H, s, CH2O), 4.06 (1H, dd, J )
6.6 and 9.0 Hz, CHAHB), 7.00-7.14 (3H, m, ArH), 7.21-7.37 (5H,
m, ArH), and 7.49-7.55 (1H, m, HCCBr); 13C NMR δ 28.6 (CH3),
28.8 (CH3), 41.0 (CH2), 45.3 (CH), 67.4 ((CH3)2C), 79.3 (CH2O),
125.2 (CBr), 127.3 (CHAr), 127.6 (CHAr), 128.1 (CHAr), 128.5
(CHAr), 129.0 (CHAr), 132.2 (CHAr), 133.2 (CHAr), 138.8 (Cq), 140.0
(Cq), and 166.1 (CdN); m/z (CI) 358 [99%, (MH)+)]. Calcd for
C19H21ON79Br (MH)+: 358.0807. Found: 358.0812.
Conclusions
2-[3-(2-Chlorophenyl)propyl]-4,4-dimethyl-2-oxazoline (4d).
2-Halophenylalkyl-2-oxazolines containing propyl and butyl
spacers cyclized with reasonable efficiency on treatment with
LDA to give Indane and tetralin derivatives, respectively. Ring
closure forming a quaternary center took place just as readily
Clear oil (54%); Rf (Al2O3, 9:1, hexane:EtAc) 0.3; νmax(film)/cm-1
1
1648 (CdN); H NMR δ 1.27 (6H, s, 2 × CH3), 1.91-2.01 (2H,
m, CH2(CH2)2), 2.31 (2H, t, J ) 7.5 Hz, CH2), 2.79 (2H, t, J ) 7.7
Hz, ArCH2), 3.91 (2H, s, CH2O), 7.10-7.24 (3H, m, ArH), and
7.31-7.34 (1H, m, HCCCl); 13C NMR δ 26.1 (CH2), 27.6 (CH2),
28.4 (2CH3), 32.8 (CH2), 66.9 ((CH3)2CH), 78.9 (CH2O), 126.7
(CHAr), 127.4 (CHAr), 129.4 (CHAr), 130.5 (CHAr), 133.9 (CCl),
(31) Hoppe, D.; Hense, T. Angew. Chem., Int. Ed. 1997, 36, 2282-
2316.
906 J. Org. Chem., Vol. 72, No. 3, 2007

1,3-Areneotropic Migration of Oxazolines
139.1 (Cq), and 165.6 (CdN); m/z (CI) 252 [100%, (MH)+]. Calcd
for C14H19ON35Cl (MH)+: 252.1155. Found: 252.1147.
isomer), 34.2 (ArCH2, major isomer), 45.1 (PhCH, major isomer),
45.2 (PhCH, minor isomer), 69.8 (CH2O), 71.7 (CHN), 124.3 (CBr),
127.0 (CHAr, major isomer), 127.1 (CHAr, minor isomer), 127.3
(CHAr), 127.6 (CHAr), 127.8 (CHAr, major isomer), 128.0 (CHAr,
minor isomer), 128.5 (CHAr), 130.3 (CHAr), 132.7 (CHAr), 139.6
(Cq, minor isomer), 139.9 (Cq, major isomer), 140.8 (Cq, minor
isomer), 140.9 (Cq, major isomer), and 167.8 (CdN), m/z (ES) 386
[100%, (MH+)]. Calcd for C21H25NO79Br (MH+): 386.1120.
Found: 386.1122.
2-[3-(2-Bromophenyl)propyl]-4,4-dimethyl-2-oxazoline (4c).
Clear oil (80%); Rf (Al2O3, 9:1, hexanes:EtAc) 0.25; νmax(film)/
cm-1 1668 (CdN); 1H NMR δH 1.27 (6H, s, 2 × CH3), 1.88-2.00
(2H, m, CH2CH2CH2), 2.32 (2H, t, J ) 7.7 Hz, CH2), 2.79 (2H, t,
J ) 7.8 Hz, ArCH2), 3.91 (2H, s, CH2O), 7.01-7.07 (1H, m, ArH),
7.21-7.28 (2H, m, ArH), and 7.50-7.53 (1H, m, HCCBr); 13C
NMR δ 26.2 (CH2), 27.6 (CH2), 28.4 (2 × CH3), 35.3 (CH2), 66.9
((CH3)2CH), 78.9 (CH2O), 124.4 (CBr), 127.3 (CHAr), 127.6 (CHAr),
130.4 (CHAr), 132.7 (CHAr), 140.8 (Cq), and 165.5 (CdN); m/z (CI)
296 [100%, (MH)+]. Anal. Calcd for C14H18ONBr: C, 56.8; H,
6.1; N, 4.7. Found: C, 56.3; H, 6.6; N, 4.7.
(4R)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-isopropyl-2-ox-
azoline [(R)-40]. Yield (68%): see data for (S)-40.
(4S)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-tert-butyl-2-ox-
azoline [(S)-41]. White solid; mixture of diastereoisomers (72%);
1
Rf (9:1 hexane:EtOAc) 0.2; νmax/cm-1 1669 (CdN); H NMR δ
2-[3-(2-Chlorophenyl)-1-phenylpropyl]-4,4-dimethyl-2-oxazo-
0.86 (9H, s, C(CH3)3, minor isomer), 0.90 (9H, s, C(CH3)3, major
isomer), 2.06-2.21 (1H, m, CHAHB), 2.29-2.46 (1H, m, CHAHB),
2.61-2.86 (2H, m, ArCH2), 3.62-3.75 (1H, m, PhCH), 3.81-3.89
(1H, m, CHN), 3.99-4.10 (1H, m, CHCHDO), 4.10-4.18 (1H, m,
CHCHDO), 7.00-7.08 (1H, m, ArH), 7.16-7.41 (7H, m, ArH), and
7.49-7.51 (1H, m, BrCCH); 13C NMR δ 25.8 (tBu, minor isomer),
26.0 (tBu, major isomer), 33.7 (CH2, major isomer), 33.7 (CH2,
minor isomer), 34.2 (CH2, major isomer), 34.4 (CH2, minor isomer),
45.3 (PhCH, minor isomer), 45.4 (PhCH, major isomer), 68.6
(CH2O, minor isomer), 68.6 (CH2O, major isomer), 75.4 (CHN,
minor isomer), 75.6 (CHN, major isomer), 124.4 (CBr), 127.0
(CHAr), 127.1 (CHAr), 127.4 (CHAr), 127.6 (CHAr), 127.9 (CHAr),
128.1 (CHAr), 128.5 (CHAr), 130.4 (CHAr), 132.8 (CHAr), 139.7 (Cq,
major isomer), 140.0 (Cq, minor isomer), 141.0 (Cq, major isomer),
141.1 (Cq, minor isomer), 167.7 (CdN, minor isomer), and 167.8
(CdN, major isomer); m/z (ES) 400 [100%, (MH+]. Anal. Calcd
for C22H26ONBr: C, 66.0; H, 6.5; N, 3.5. Found: C, 65.8; H, 6.6;
N, 3.3. Calcd for C22H27NO79Br (MH)+: 400.1276. Found: 400.1279.
(4R)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-tert-butyl-2-ox-
azoline [(R)-41]. Yield (80%): see data for (S)-41.
line (4b). Clear oil (48%); Rf (Al2O3, 19:1, hexane:EtAc) 0.2; νmax
-
(film)/cm-1 1635 (CdN); 1H NMR δ 1.27 (3H, s, CH3), 1.28 (3H,
s, CH3), 2.07-2.19 (1H, m, CHCHAHBCH2), 2.32-2.44 (1H, m,
CHCHAHBCH2), 2.63-2.80 (2H, m, ArCH2), 3.60 (1H, t, J ) 7.8
Hz, CH), 3.87 (2H, s, CH2O), 7.07-7.20 (3H, m, ArH), and 7.22-
7.37 (6H, m, ArH); 13C NMR δ 28.1 (CH3), 28.3 (CH3), 31.4 (CH2),
33.5 (CH2), 44.9 (CH), 66.8 ((CH3)2CH), 78.9 (CH2O), 126.6
(CHAr), 127.0 (CHAr), 127.3 (CHAr), 127.7 (CHAr), 128.5 (CHAr),
129.3 (CHAr), 130.4 (CHAr), 133.8 (CCl), 139.0 (Cq), 139.6 (Cq),
and 166.5 (CdN); m/z 328 (CI) [100%, (MH)+]. Calcd for C20H23-
ON35Cl (MH)+: 328.1468. Found: 328.1478.
2-[3-(2-Bromophenyl)-1-phenylpropyl]-4,4-dimethyl-2-oxazo-
line (4a). White solid (66%); Rf (SiO2, 9:1 hexanes:EtOAc) 0.25;
mp 69.5-71.5 (EtOAc:hexanes); νmax(nujol)/cm-1 1668 (CdN);
1H NMR δ 1.28 (3H, s, CH3), 1.30 (3H, s, CH3), 2.05-2.19 (1H,
m, CHCHACHBCH2), 2.31-2.43 (1H, m, CHCHACHBCH2), 2.60-
2.81 (2H, m, ArCH2), 3.61 (1H, t, J ) 7.8 Hz, CH), 3.90 (2H, s,
CH2O), 7.00-7.06 (1H, m, ArH), 7.16-7.38 (7H, m, ArH), and
7.48-7.51 (1H, m, HCCBr); 13C NMR δ 28.1 (CH3), 28.3 (CH3),
33.6 (CH2), 34.0 (CH2), 44.9 (CH), 67.0 ((CH3)2CH), 79.0 (CH2O),
124.3 (CBr), 127.1 (CHAr), 127.3 (CHAr), 127.6 (CHAr), 127.8
(CHAr), 128.5 (CHArH), 130.4 (CHAr), 132.7 (CHAr), 139.7 (Cq),
140.8 (Cq), and 166.6 (CdN); m/z (CI) 372 [95%, (MH)+]. Calcd
for C20H23ON79Br (MH)+: 372.0963. Found: 372.0945. Anal.
Calcd for C20H22ONBr: C, 64.5; H, 6.0; N, 3.8. Found: C, 64.7;
H, 5.75; N, 3.6.
(4S)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-phenyl-2-ox-
azoline [(S)-42]. White solid; mixture of diastereoisomers (40%);
1
Rf (9:1 hexanes:EtOAc) 0.1; νmax/cm-1 1656 (CdN); H NMR δ
2.16-2.30 (1H, m, CHAHB), 2.41-2.55 (1H, m, CHAHB), 2.68-
2.89 (2H, m, ArCH2), 3.73-3.79 (1H, m, PhCH), 4.03-4.11 (1H,
m, CHCHDO), 4.53-4.63 (1H, m, CHCHDO), 4.53-4.63 (1H, m,
CHN), 7.01-7.08 (1H, m, ArH), 7.18-7.45 (12H, m, ArH), and
7.49-7.52 (1H, m, BrCCH); m/z (CI) 420 [100%, (MH+)]. Calcd
for C24H23NO79Br (MH)+: 420.0963. Found: 420.0957.
(4S)-2-[3-(2-Bromophenyl)propyl]-4-isopropyl-2-oxazoline. Yield
(50%); Rf (SiO2, 4:1 hexanes:EtOAc) 0.25; νmax/cm-1 (film) 1670
1
(CdN); H NMR δ 0.88 (3H, d, J ) 6.9, CH3), 0.96 (3H, d, J )
(4S)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-benzyl-2-oxazo-
line [(S)-43]. White solid; 1:1 mixture of diastereoisomers (88%);
Rf (9:1 hexanes:EtOAc) 0.25; νmax/cm-1 1645 (CdN); 1H NMR δ
2.04-2.20 (1H, m, PhCHCHAHB), 2.28-2.46 (1H, m, PhCH-
CHAHB), 2.58-2.78 (3H, m, ArCH2 and PhCHCHD), 3.07-3.15
(1H, m, PhCHCHD), 3.56-3.66 (1H, m, PhCH), 3.88-3.96 (1H,
m, CHEHFO), 4.06-4.16 (1H, m CHEHFO), 4.34-4.46 (1H, m,
CHN), 7.00-7.05 (1H, m, Ar), 7.14-7.36 (12H, m, Ph and Ar),
and 7.48-7.51 (1H, m, Ar); 13C NMR32 δ 33.5 (CH2, isomer A),
33.7 (CH2, isomer B), 34.0 (CH2, isomer A), 34.1 (CH2, isomer
B), 41.5 (CH2), 44.9 (PhCH, isomer A), 45.0 (PhCH, isomer B),
66.9 (CHN, isomer A), 67.0 (CHN, isomer B), 71.5 (CH2O), 124.4
(CBr), 126.4 (CHAr), 127.1 (CHAr), 127.2 (CHAr), 127.3 (CHAr),
127.6 (CHAr), 127.88 (CHAr), 127.9 (CHAr), 128.38 (CHAr), 128.42
(CHAr), 128.5 (CHAr), 129.2 (CHAr), 129.3 (CHAr), 130.4 (CHAr),
132. 7 (CHAr), 137.6 (Cq, isomer A), 137.7 (Cq, isomer B), 139.5
(Cq, isomer A), 139.6 (Cq, isomer B), 140.77 (Cq, isomer A), 140.8
(Cq, isomer B), 168.55 (CdN, isomer A), and 168.57 (CdN, isomer
B); m/z (CI) 434 [100%, (MH+)]. Calcd for C25H25NO79Br
(MH)+: 434.1120. Found: 434.1109.
6.7 Hz, CH3), 1.74 (1H, dqq, J ) 6.9, 6.9 and 6.7, (CH3)2CH),
1.75 (2H, qu, J ) 7.8 Hz, CH2), 2.35 (2H, t, J ) 6.8, CH2CdN),
2.80 (2H, t, J ) 6.8 Hz, ArCH2), 3.84-3.95 (2H, m, CHN and
CHAHBO), 4.14-4.23 (1H, m, CHAHBO), 7.00-7.10 (1H, m, ArH),
7.16-7.30 (2H, m, ArH), and 7.50-7.54 (1H, m, HCCBr); 13C
NMR δ 18.1 (CH3), 18.8 (CH3), 26.3 (CH2), 27.6 (CH2CdN), 32.5
((CH3)2CH), 35.4 (ArCH2), 69.8 (CH2O), 72.0 (CHN), 124.4 (CBr),
127.3 (CHAr), 127.6 (CHAr), 130.4 (CHAr), 132.7 (CHAr), 140.8 (Cq),
and 166.6 (CdN); m/z (CI) 310 [100%, (MH)+]. Calcd for C15H21-
NO79Br (MH)+: 310.0807. Found: 310.0807.
(4S)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-isopropyl-2-ox-
azoline [(S)-40]. White solid (71%); 2:1 mixture of diastereoisomers
1
Rf (SiO2, 2:1 hexanes:EtOAc) 0.38; νmax/cm-1 1664 (CdN); H
NMR δ (500 MHz, CDCl3) 0.85 (3H, d, J ) 6.8 Hz, CH3, major
isomer), 0.87 (3H, d, J ) 6.8 Hz, CH3, minor isomer), 0.93 (3H,
d, J ) 6.8 Hz, CH3, major isomer), 0.96 (3H, d, J ) 6.8 Hz, CH3,
minor isomer), 1.71-1.82 (1H, m, CH(CH3)2), 2.10-2.20 (1H, m,
CHAHB), 2.33-2.43 (1H, m, CHAHB), 2.64-2.82 (2H, m, ArCH2),
3.66 (1H, t, J ) 7.8 Hz, PhCH, major isomer), 3.69 (1H, t, J ) 8.1
Hz, PhCH, minor isomer), 3.88-3.95 (2H, m, CHCHDO, CHN),
4.01-4.21 (1H, m, CHCHD), 6.99-7.06 (1H, m, Ar), 7.15-7.27
(3H, m, Ar), 7.30-7.38 (4H, m, ArH), and 7.48-7.50 (1H, m,
ArH); 13C NMR δ 17.9 (CH3, major isomer), 18.0 (CH3, minor
isomer), 18.7 (CH3, major isomer), 18.8 (CH3, minor isomer), 32.4
((CH3)2CH, minor isomer), 32.5 ((CH3)2CH, major isomer), 33.5
(CH2, minor isomer), 33.7 (CH2, major isomer), 34.1 (ArCH2, minor
(4S)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-cyclohexylmethyl-
2-oxazoline [(S)-44]. White solid; 3:2 mixture of diastereoisomers
1
(49%); Rf (9:1 hexanes:EtOAc) 0.25; νmax/cm-1 1647 (CdN); H
NMR δH 0.83-1.01 (2H, m, C6H11), 1.08-1.32 (4H, m, C6H11-
(32) Assignment of isomer A or B is arbitrary.
J. Org. Chem, Vol. 72, No. 3, 2007 907

Marshall et al.
CHAHB and C6H11), 1.34-1.48 (1H, m, C6H11), 1.58-1.75 (6H,
m, C6H11CHAHB and C6H11), 2.08-2.20 (1H, m, PhCHCHAHB),
2.30-2.44 (1H, m, PhCHCHAHB), 2.62-2.81 (2H, m, ArCH2),
3.60-3.68 (1H, m, PhCH), 3.76-3.81 (1H, m, CHCHDO), 4.09-
4.30 (1H, m, CHN and CHCHDO), 7.00-7.08 (1H, m, ArH), 7.16-
7.38 (1H, m, ArH), and 7.48-7.51 (1H, m, HCCBr); 13C NMR δ
26.2 (CH2), 26.5 (CH2), 33.4 (CH2), 33.57 (CH2, minor isomer),
33.61 (CH2, major isomer), 33.8 (CH2), 34.1 (CH2, minor isomer),
34.2 (CH2, major isomer), 34.7 (CH, minor isomer), 34.8 (CH,
major isomer), 44.2 (CH2, major isomer), 44.3 (CH2, minor isomer),
45.1 (PhCH), 63.8 (CHN, minor isomer), 63.9 (CHN, major isomer),
73.0 (CH2O, major isomer), 73.1 (CH2O, minor isomer), 124.4
(CBr), 127.1 (CHAr, major isomer), 127.2 (CHAr, minor isomer),
127.4 (CHAr), 127.9 (CHAr), 128.6 (CHAr), 130.4 (CHAr), 132.8
(CHAr), 139.7 (Cq, minor isomer), 139.8 (Cq, major isomer), 140.9
(Cq, minor isomer), 141.0 (Cq, major isomer), and 167.7 (CdN);
m/z (CI) 440 [100%, (MH+)]. Calcd for C25H31NO79Br (MH)+:
440.1589. Found: 440.1570.
(nujol)/cm-1 1660 (CdN); 1H NMR δ 1.26 (3H, s, CH3), 1.27 (3H,
s, CH3), 1.46-1.74 (2H, m, ArCH2CH2), 1.84-1.96 (1H, m,
PhCHCHACHB), 2.11-2.23 (1H, m, PhCHCHACHB), 2.66-2.82
(2H, m, ArCH2), 3.58 (1H, t, J ) 7.8 Hz, PhCH), 3.86 (2H, s,
CH2O), 6.98-7.04 (1H, m, ArH), 7.12-7.36 (7H, m, ArH), and
7.48 (1H, d, J ) 7.8 Hz, HCCBr); 13C NMR δ 27.7 (CH2), 28.2
(CH3), 28.3 (CH3), 33.5 (CH2), 35.9 (CH2), 45.2 (PhCH), 66.8
((CH3)2C), 78.9 (CH2O), 124.4 (CBr), 127.0 (CHAr), 127.3 (CHAr),
127.4 (CHAr), 127.7 (CHAr), 128.5 (CHAr), 130.2 (CHAr), 132.7
(CHAr), 140.0 (Cq), 141.4 (Cq), and 166.8 (CdN); m/z (ES) 386
[100%, (MH+)]. Anal. Calcd for C21H24ONBr: C, 65.3; H, 6.3; N,
3.6. Found: C, 65.3; H, 6.8; N, 3.8. Calcd for C21H25NO79Br
(MH)+: 386.1120. Found: 386.1108.
2-[3-(2-Bromophenyl)-1-phenylpentyl]-4,4-dimethyl-2-oxazo-
line (6). Clear oil (60%); Rf (SiO2, 9:1 hexanes:EtOAc) 0.2; νmax
-
(nujol)/cm-1 1661 (CdN); 1H NMR δ 1.25 (3H, s, CH3), 1.26 (3H,
s, CH3), 1.15-1.47 (2H, m, CH2), 1.54-1.74 (2H, m, ArCH2CH2),
1.77-1.91 (1H, m, PhCHCHACHB), 2.03-2.17 (1H, m, PhCH-
CHACHB), 2.68 (2H, t, J ) 7.7 Hz, ArCH2), 3.53 (1H, t, J ) 7.8
Hz, PhCH), 3.85 (2H, s, CH2O), 6.98-7.04 (1H, m, ArH), 7.11-
7.33 (7H, m, ArH), and 7.49 (1H, dd, J ) 8.0 and 1.0 Hz, HCCBr);
13C NMR δ 27.1 (CH2), 28.2 (CH3), 28.3 (CH3), 29.4 (CH2), 33.6
(CH2), 35.8 (CH2), 45.3 (PhCH), 66.7 ((CH3)2C), 78.8 (CH2O),
124.4 (CBr), 126.9 (CHAr), 127.2 (CHAr), 127.3 (CHAr), 127.7
(CHAr), 128.5 (CHAr), 130.2 (CHAr), 132.6 (CHAr), 140.2 (Cq), 141.7
(Cq), and 166.8 (CdN); m/z (ES) 400 [100%, (MH+)]. Calcd for
C22H27NO79Br (MH)+: 400.1276. Found: 400.1292.
(4S,5S)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-methoxy-5-
phenyl-2-oxazoline [(S,S)-45]. White solid; 3:2 mixture of dias-
tereoisomers (55%); Rf (4:1 hexanes:EtOAc) 0.25; νmax/cm-1 1640
1
(CdN); H NMR δH 2.15-2.39 (1H, m, PhCHCHAHB), 2.38-
2.53 (1H, m, PhCHCHAHB), 2.70-2.88 (ArCH2), 3.39 (3H, s, MeO,
minor isomer), 3.41 (3H, s, MeO, major isomer), 3.46-3.54 (1H,
m, CHCHDOMe), 3.65 (1H, dd, J ) 9.6, 4.3 Hz, CHCHDOMe, major
isomer), 3.66 (1H, dd, J ) 9.7, 4.4 Hz, CHCHDOMe, minor isomer),
3.73 (1H, t, J ) 7.7 Hz, PhCHCHAHB, minor isomer), 3.80 (1H, t,
J ) 7.8 Hz, PhCHCHAHB, major isomer), 4.08-4.17 (1H, m,
CHN), 5.28 (1H, d, J ) 6.6 Hz, PhCHO), 7.00-7.12 (3H, m, ArH),
7.17-7.38 (8H, m, ArH), 7.40-7.44 (2H, m, ArH), and 7.49-
7.52 (1H, m, ArH); 13C NMR δ 33.2 (CH2, major isomer), 33.5
(CH2, minor isomer), 34.1 (CH2), 45.1 (PhCH, minor isomer), 45.3
(PhCH, major isomer), 59.2 (CHN), 74.26 (MeO, major isomer),
74.3 (CH2O, major isomer), 74.39 (MeO, minor isomer), 74.42
(CH2O, major isomer), 83.2 (PhCHO, minor isomer), 83.5 (PhCHO,
major isomer), 124.4 (CBr), 125.2 (CHAr), 125.4 (CHAr), 127.3
(CHAr), 127.4 (CHAr), 127.7 (CHAr), 127.9 (CHAr), 127.96 (CHAr),
128.0 (CHAr), 128.1 (CHAr), 128.6 (CHAr), 130.5 (CHAr), 132.8
(CHAr), 139.4 (Cq, major isomer), 139.5 (Cq, minor isomer), 140.86
(Cq, major isomer), 140.88 ((Cq, minor isomer) and (Cq, major
isomer)), 140.9 (Cq, minor isomer), 168.9 (CdN, major isomer),
and 169.0 (CdN, minor isomer); m/z (CI) 464 [100%, (MH+)].
Calcd for C26H27NO279Br (MH)+: 464.1225. Found: 464.1219.
2-[3-(2-Bromophenyl)-1-phenylhexyl]-4,4-dimethyl-2-oxazo-
line (7). Clear oil (61%); Rf (SiO2, 9:1 hexanes:EtOAc) 0.2; νmax
-
(nujol)/cm-1 1661 (CdN); 1H NMR δ 1.27 (3H, s, CH3), 1.28 (3H,
s, CH3), 1.20-1.45 (4H, m, 2 × CH2), 1.53-1.63 (2H, m,
ArCH2CH2), 1.75-1.87 (1H, m, PhCHCHACHB), 2.00-2.13 (1H,
m, PhCHCHACHB), 2.68 (2H, t, J ) 7.7, ArCH2), 3.52 (1H, t, J )
7.8 Hz, PhCH), 3.87 (2H, s, CH2O), 6.99-7.05 (1H, m, ArH),
7.13-7.35 (7H, m, ArH), and 7.50 (1H, dd, J ) 8.0 and 0.9 Hz,
HCCBr); 13C NMR δ 27.3 (CH2), 28.2 (CH3), 28.4 (CH3), 29.0
(CH2), 29.7 (CH2), 33.7 (CH2), 36.0 (CH2), 45.4 (PhCH), 66.8
((CH3)2C), 78.9 (CH2O), 124.4 (CBr), 126.9 (CHAr), 127.2 (CHAr),
127.3 (CHAr), 127.7 (CHAr), 128.5 (CHAr), 130.2 (CHAr), 132.7
(CHAr), 140.3 (Cq), 141.9 (Cq), and 166.9 (CdN); m/z (ES) 414
[100%, (MH+)]. Calcd for C23H29NO79Br (MH)+: 414.1433.
Found: 414.1439.
General Procedure for Reaction of 2-Haloarylalkyl-2-oxazo-
lines with tert-BuOK/DMSO. Freshly sublimed potassium tert-
butoxide (1.1 equiv) was dissolved in dry distilled DMSO (6 mL/
mmol). The substrate oxazoline (1 equiv) in DMSO (6 mL/mmol)
was then added and stirred for 5 min. Triethylamine (1 equiv) was
added, and the resulting mixture was irradiated for 3-4 h. Water
(6 mL/mmol) was added, and the aqueous phase was extracted with
dichloromethane (3 × 12 mL/mmol). The combined organic extracts
were washed (water (24 mL/mmol)), dried (MgSO4), and concen-
trated to give the crude product.
(4S)-{2-[3-(2-Bromophenyl)-1-phenylpropyl]-2-oxazolin-4-yl}-
diphenylmethanol [(S)-46]. White solid; 2:1 mixture of diastere-
oisomers (41%); Rf (9:1 hexanes:EtOAc) 0.2; νmax/cm-1 1654 (Cd
N) and 3547 (OH); 1H NMR δ 2.05-2.17 (1H, m, PhCHCHAHB),
2.24-2.40 (2H, m, OH and PhCHCHAHB), 2.61-2.78 (2H, m,
ArCH2), 3.73 (1H, t, J ) 7.7 Hz, PhCH), 4.02-4.13 (2H, m, CH2O),
5.28 (1H, t, J ) 9.6 Hz, CHN, minor isomer), 5.32 (1H, t, J ) 9.4
Hz, CHN, major isomer), 6.99-7.06 (1H, m, ArH), 7.09-7.40
(16H, m, ArH), 7.46-7.51 (1H, m, ArH), and 7.55-7.60 (2H. m,
ArH); 13C NMR δ 33.7 (CH2, minor isomer), 33.8 (CH2, major
isomer), 34.2 (CH2, minor isomer), 34.3 (CH2, major isomer), 45.38
(PhCH, major isomer), 45.44 (PhCH, minor isomer), 69.1 (CH2O,
major isomer), 69.2 (CH2O, minor isomer), 72.4 (CHN), 78.0 (Ph2C,
minor isomer), 78.2 (Ph2C, major isomer), 124.3 (CBr, major
isomer), 124.4 (CBr, minor isomer), 125.7 (CHAr), 126.7 (CHAr),
126.8 (CHAr), 127.0 (CHAr), 127.1 (CHAr), 127.3 (CHAr), 127.7
(CHAr), 128.0 (CHAr), 128.2 (CHAr), 128.6 (CHAr), 130.3 (CHAr),
130.4 (CHAr), 132.8 (CHAr), 139.4 (Cq, minor isomer), 139.7 (Cq,
major isomer), 140.79 (Cq, major isomer), 140.84 (Cq, minor
isomer), 144.0 (Cq, major isomer), 144.1 (Cq, minor isomer), 145.7
(Cq), 171.5 (CdN, minor isomer), and 171.6 (Cq, major isomer);
m/z (CI) 526 [100%, (MH+)]. Calcd for C31H29NO279Br (MH)+:
526.1354. Found: 526.1372.
Reaction of 2-[3-(2-Chlorophenyl)propyl]-4,4-dimethyl-2-ox-
azoline (4d). General procedure for t-BuOK with 3 h irradiation.
After work up the 1H-NMR spectrum showed the presence of only
starting material.
Reaction of 2-[3-(2-Bromophenyl)-1-phenylpropyl]-4,4-di-
methyl-2-oxazoline (4a) with t-BuOK. After the general procedure
with 4 h irradiation, 1H NMR and GCMS revealed the presence of
a 1:2 mixture of the starting material 4a and the dehalogenated
product 2-[1,3-diphenylpropyl]-4,4-dimethyl-2-oxazoline 12 (n )
1
1, R ) Ph); Rf (Al2O3, 19:1, hexane:EtAc) 0.3, H NMR δ 1.25
(3H, s, CH3), 1.26 (3H, s, CH3) 2.09-2.18 (1H, m, CHACHB),
2.37-2.44 (1H, m, CHACHB), 2.59 (2H, t, J ) 7.7 Hz, ArCH2),
3.54 (1H, t, J ) 7.7 Hz, PhCH), 3.81 (1H, s, CHCHDO), 3.82 (1H,
s, CHCHDO), and 7.11-7.34 (10H, m, ArH); 13C NMR δ 28.6
(CH3), 28.8 (CH3), 34.0, 35.9, 45.1, 67.3, 79.3, 126.3, 127.5, 128.3,
128.6, 128.8, 129.0, 129.1, 140.5, 141.9, and 167.1; m/z (EI) 293
(1%, M+), 189 (100, M-C8H8), 91 (16).
2-[3-(2-Bromophenyl)-1-phenylbutyl]-4,4-dimethyl-2-oxazo-
line (5). Clear oil (69%); Rf (SiO2, 9:1 hexanes:EtOAc) 0.2; νmax
-
908 J. Org. Chem., Vol. 72, No. 3, 2007

1,3-Areneotropic Migration of Oxazolines
General Procedure for Reaction of 2-Haloarylalkyl-2-oxazo-
lines with MNH2/NH3 . To a solution of ferric nitrate (0.03 equiv)
(CH2O), 124.3 (CHAr), 124.6 (CHAr), 126.4 (CHAr), 127.3 (CHAr),
141.3 (Cq), 144.0 (Cq), and 167 (CdN); m/z (CI) 216 [100%,
(MH)+]. Calcd for C20H22ON (MH)+: 216.1388. Found: 216.1396.
A mixture of styrene and 2-bromostyrene (14) (10%) with 1H NMR
spectra in agreement with the literature was also obtained. Yields
for the same reaction carried out in Pyrex glassware with method
A are shown in Table 1.
-
in liquid ammonia (15 mL/mmol) at -78 °C in a quartz flask was
added with caution either lithium or potassium metal (3 equiv).
The resulting mixture was warmed to -33 °C and stirred until the
blue color faded to a gray suspension (ca. 30 min). The substrate
oxazoline (1 equiv) in dry THF (1 mL/mmol) was added dropwise
over 5 min and stirred for a further 10 min. After allowing the
mixture to warm to reflux (-33 °C), it was irradiated with UV for
either 1 or 2 h, after which solid NH4Cl was added with extreme
caution! The liquid ammonia was allowed to evaporate while being
slowly replaced by ether (20 mL/mmol). The ethereal solution was
decanted, and the solids were washed with ether (2 × 50 mL/mmol).
The combined portions were dried (MgSO4) and concentrated.
Reaction of 2-[3-(2-Chlorophenyl)propyl]-4,4-dimethyl-2-ox-
azoline (4d) with MNH2/NH3. Following the general procedure
above 4d was reacted with LiNH2 in liquid ammonia with
irradiation for 2 h to yield after work up a 5:1 mixture of starting
material 4d and the reduced product 12(n ) 1, R ) H).
Reaction of 2-[3-(2-Bromophenyl)-1-phenylpropyl]-4,4-di-
methyl-2-oxazoline (4a) with MNH2/NH3. Precursor 4a was added
to a stirring solution of potassium amide in liquid ammonia as in
the general procedure with 1 h irradiation. Dibromomethane as used
as an internal NMR standard. The 1H NMR revealed three separate
species; these were starting material 4a (51%), diphenylpropyl-2-
oxazoline (31%) and the desired 4,4-dimethyl-2-(1′-phenylindan-
1′-yl)-2-oxazoline 11 (n ) 1, R ) Ph) (16%); Rf (SiO2, 9:1 hexanes:
EtOAc) 0.35; νmax(film)/cm-1 1649 (CdN); 1H NMR δ 1.24 (3H,
s, CH3), 1.40 (3H, s, CH3), 2.30 (1H, ddd, J ) 12.6, 7.7, 5.1 Hz,
CHAHB), 2.82 (1H, ddd, J ) 15.6, 7.7, 7.4 Hz, ArCHCHD), 3.00
(1H, ddd, J ) 15.6, 7.9, 5.1 Hz, ArCHCHD), 3.18 (1H, ddd, J )
12.6, 7.9, 7.4 Hz, CHAHB), 3.90 (1H, AB doublet J ≈ 7.9 Hz,
CHEHFO), 3.97 (1H, AB doublet J ≈ 7.9 Hz, CHEHFO), 7.06-
7.11 (2H, m, ArH), 7.18-7.31 (6H, m, ArH), and 7.48-7.51 (1H,
m, ArH); 13C NMR δ 28.0 (CH3), 28.3 (CH3), 30.3 (CH2), 41.3
(ArCH2), 58.2 ((CH3)2C), 66.8 (PhC), 79.4 (CH2O), 124.7 (CHAr),
126.4 (CHAr), 126.5 (CHAr), 126.6 (CHAr), 126.7 (CHAr), 127.7
(CHAr), 128.3 (CHAr), 143.6 (Cq), 144.3 (Cq), 144.8 (Cq), and 167.6
(CdN); m/z (CI) 292 [100%, (MH)+]. Calcd for C20H22ON
(MH)+: 292.1701. Found: 292.1708. The reaction was repeated
under identical conditions but at a dilution of one-tenth; the yields
are shown in Table 1.
General Procedure for Reaction of 2-Haloarylalkyl-2-oxazo-
lines with LDA in THF. To a solution of LDA (3 equiv) in either
THF or 2:1 hexane:THF mixture (6 mL/mmol) at -78 °C under
nitrogen was added a solution of the substrate (1 equiv) in the
reaction solvent (3 mL/mmol). After stirring for 10 min the solution
was allowed to warm to room temperature over 30 min, at which
time more solvent was added (18 mL/mmol). In Method A the
mixture was then stirred for between 0 and 48 h at between -30
°C and reflux. In Method B a quartz flask was used, and irradiation
by a 400 W medium-pressure Hg arc placed ca. 15 cm from the
flask was continued usually for 6 h. In both methods, after reaction,
a saturated solution of ammonium chloride (10 mL/mmol) was
added and the aqueous layer was extracted with three portions of
ether (3 × 5 mL/mmol). The combined organic layers were washed
with water (10 mL/mmol) and dried (MgSO4) to give the crude
product.
Reaction of 2-[3-(2-Bromophenylpropyl]-4,4-dimethyl-2-ox-
azoline (4d) with LDA. Using method B 4d was reacted under
UV radiation in quartz glassware for 4 h to yield after column
chromatography starting material 4d (29%), the desired 4,4-
dimethyl-2-indan-1′-yl-2-oxazoline 11 (n ) 1, R ) H) (58%); Rf
(SiO2, 9:1 hexanes:EtOAc) 0.2; νmax/cm-1 1653 (CdN); 1H NMR
δ 1.28 (3H, s, CH3), 1.30 (3H, s, CH3), 2.29-2.47 (2H, m, CH2),
2.87-2.98 (1H, m, ArCHAHB), 3.03-3.13 (1H, m, ArCHAHB), 3.93
(2H, s, CH2O), 4.09, (1H, t, J ) 7.9 Hz, CH), 7.08-7.27 (3H, m,
ArH), and 7.30-7.34 (1H, m, ArH); 13C NMR δ 28.2 (CH3), 25.5
(CH3), 29.2 (CH2), 31.8 (ArCH2), 44.4 (CH), 66.8 (C(CH3)2), 79.2
Reaction of 2-[3-(2-Bromophenyl)-1-phenylpropyl]-4,4-di-
methyl-2-oxazoline (4a) with LDA. Precursor 4a was reacted using
method A for 48 h. Column chromatography, using 9:1 hexanes:
EtOAc as elutant, afforded Indane 11 (n ) 1, R ) Ph) (75%),
starting material 4a (<1%), and the mixture of styrenes (2%). Yields
for the same reaction in the dark and under irradiation for between
0 and 20 h with 1-3 equiv of LDA are detailed in Table 1.
Reaction of 2-[3-(2-Bromophenyl)-1-phenylbutyl]-4,4-di-
methyl-2-oxazoline (5) with LDA. Method A yielded 55% of the
desired 4,4-dimethyl-2-(1-phenyl-1,2,3,4-tetrahydronaphthalen-1-
yl)-2-oxazoline 11 (n ) 2, R ) Ph); Rf (SiO2, 9:1 hexanes:EtOAc)
0.25; νmax(film)/cm-1 1638 (CdN); 1H NMR δ 1.28 (3H, s, CH3),
1.38 (3H, s, CH3), 1.54-1.69 (1H, m, CHAHB), 1.80-1.93 (1H,
m, CHAHB), 2.03-2.12 (1H, m, CHCHD), 2.70-2.79 (1H, m,
CHCHD), 2.83-2.89 (2H, m, ArCH2), 3.93 (1H, AB doublet J∼8.2,
CHEHFO), 3.96 (1H, AB doublet J∼8.2, CHEHFO), 7.00-7.04 (2H,
m, ArH), 7.07-7.27 (6H, m, ArH), and 7.31-7.34 (1H, m, ArH);
13C NMR δ 19.2 (CH2), 28.1 (CH3), 28.3 (CH3), 29.8 (CH2), 37.8
(ArCH2), 50.5 (PhC), 66.8 ((CH3)2C), 79.1(CH2O), 125.4 (CHAr),
126.3 (CHAr), 126.8 (CHAr), 127.8 (CHAr), 127.9 (CHAr), 129.2
(CHAr), 130.9 (CHAr), 137.3 (Cq), 137.9 (Cq), 146.6 (Cq), and 168.7
(CdN); m/z (CI) 306 [100%, (MH)+]. Anal. Calcd for C21H23NO:
C, 82.6; H, 7.6; N, 4.6. Found: C, 81.8; H, 8.0; N, 4.7. Calcd for
C21H24ON (MH)+: 306.1858. Found: 306.1853. An X-ray structure
for 11 (n ) 2, R ) Ph) confirmed the structure (see Supporting
Information).
Reaction of oxazoline 5 using method B gave tetralin 11 (n )
2, R ) Ph) in a 50% yield.
Reaction of 2-[3-(2-Bromophenyl)-1-phenylpentyl]-4,4-di-
methyl-2-oxazoline (6) with LDA. Using Method A, several
components were isolated by column chromatography.
2-(1-n-Butyl-1,5-diphenylpentyl)-4,4-dimethyl-2-oxazoline
(13%): Rf (SiO2, 4:1 hexanes:EtOAc) 0.4; νmax(film)/cm-1 1655
1
(CdN); H NMR δ 0.81-0.88 (3H, m, CH3), 0.93-1.34 (6H, m,
3 × CH2), 1.27 (3H, s, (CH3)2C), 1.30 (3H, s (CH3)2C), 1.62 (2H,
q, J ) 7.6 Hz, CH2), 1.89-2.15 (4H, m, 2 × CH2), 2.49-2.65
(2H, m, CH2), 3.73 (1H, d, J ) 8.1 Hz, CHAHBO), 1H, d, J ) 8.1
Hz, CHAHBO), and 7.10-7.34 (10H, m, 2 × Ph); 13C NMR δ 14.0
(CH3), 23.0 (CH2), 23.1 (CH2), 25.9 (CH2), 28.2 ((CH3)2C), 31.5
(CH2), 33.8 (CH2), 34.0 (CH2), 35.5 (CH2), 47.2 (PhC), 66.8
((CH3)2C), 78.7 (CH2O), 125.5 (CHAr), 126.2 (CHAr), 126.4 (CHAr),
128.1 (CHAr), 128.3 (CHAr), 142.4 (Cq), 143.8 (Cq), and 169.5 (Cd
O); m/z (ES) 378 [100%, (MH+)]. Calcd for C26H36NO (MH)+:
378.2797. Found: 378.2811.
4,4-Dimethyl-2-(9-phenyl-6,7-dihydro-5H-benzocyclohepten-
1
1-yl)-2-oxazoline (16): colorless needles (25%); H NMR δ 0.72
(3H, s, CH3), 1.07 (3H, m, CH3), 1.22-1.32 (2H, m, CH2), 2.01-
2.20 (2H, m, CH2), 2.56-2.69 (2H, m, CH2), 3.23 (1H, d, J ) 7.8
Hz, CHAHBO), 3.71 (1H, d, J ) 7.8 Hz, CHAHBO), 6.54 (1H, t, J
) 7.4 Hz, vinyl), 7.13-7.35 (6H, m, CHAr), 7.39 (1H, dd, J ) 7.6
and 1.5 Hz), and 7.65 (1H, dd, J ) 7.6 and 1.5 Hz); 13C NMR δ
25.0 (CH2), 27.7 (CH3), 28.1 (CH3), 32.4 (CH2), 34.1 (CH2), 66.6
(Cq), 79.0 (CH2O), 126.5 (CHAr), 126.7 (CHAr), 127.2 (CHAr), 128.1
(CHAr), 128.4 (vinyl), 128.7 (CHAr), 131.1 (CHAr), 138.3 (Cq), 141.5
(Cq), 141.7 (Cq), 142.1 (Cq), and 163.4 (CdN); m/z 318 [100%,
(MH)+]. Calcd for C22H24ON (MH)+: 318.1858. Found: 318.1865.
X-ray crystal diffraction data confirmed the structure (see Sup-
porting Information).
4,4-Dimethyl-2-(4,5,6,7-tetrahydro-cyclohepta[jk]fluoren-7a-
1
yl)-2-oxazoline (18): colorless needles (6%); H NMR δ 1.24-
1.30 (2H, m, CH2), 1.26 (3H, m, CH3), 1.27 (3H, m, CH3), 1.95-
2.18 (2H, m, CH2), 2.25-2.42 (1H, m, CHAHB), 2.72-2.85 (2H,
J. Org. Chem, Vol. 72, No. 3, 2007 909

Marshall et al.
m, CH2), 2.94-3.01 (1H, m, CHAHB), 3.76 (1H, d, J∼8.1,
CHCHDO), 3.81 (1H, d, J ≈ 8 Hz, CHCHDO), 7.06 (1H, d, J ) 7.4
Hz, CHAr), 7.30 (1H, d, J ) 7.4 Hz, CHAr), 7.31 (1H, dd, J ) 7.4
and 1.3 Hz, CHAr), 7.37 (1H, dt, J ) 7.4 and 1.3 Hz, CHAr), 7.57-
to give a yellow oil, which was purified by column chromatography
(SiO2 9:1 hexanes:EtOAc) to afford 2-butyl-2-phenyl-indan-1-one
37 (15%); H NMR δ 0.84 (3H, t, J ) 7.4 Hz, CH3), 1.07-1.34
1
(2H, m, CH2), 1.23-1.34 (2H, m, CH2), 1.93-2.06 (1H, m,
CHAHB), 2.08-2.20 (1H, m, CHAHB), 3.60 (1H, AB, CHAHB), 3.39
(1H, AB, CHAHB), 7.16-7.24 (1H, m, ArH), 7.24-7.33 (1H, m,
ArH), 7.37-7.42 (1H, m, ArH), 7.34-7.45 (1H, m, ArH), 7.47-
7.52 (1H, m, ArH), 7.58-7.65 (2H, m, ArH), and 7.75-7.80 (2H,
m, ArH); 13C NMR δ 13.9 (CH3), 23.2 (CH2), 26.9 (CH2), 38.3
(CH2), 41.0 (CH2), 57.1 (Cq), 124.6 (CHAr), 126.1 (CHAr), 126.5
(CHAr), 126.6 (2 × CHAr), 127.6 (CHAr), 128.5 (2 × CHAr), 135.0
(CHAr), 136.4 (Cq), 142.5 (Cq), 152.8 (Cq), and 208.1 (CdO); m/z
(GC-MS) 264; m/z (CI) 265 [99%, (MH)+]. Calcd for C19H21O
(MH)+: 265.1592. Found: 265.1594. EPR (9.5 GHz, with 100 kHz
modulation, modulation amplitude 0.2 Gpp, power 1 mW) in THF
solution at 320 K, complex multiplet at g 2.0034.
Reaction of (4S)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-
isopropyl-2-oxazoline (S)-40 with LDA. (S)-40 was reacted
according to method A. After column chromatography using 9:1
hexanes:EtOAc as eluant, diastereomeric indanes (S,R)-47 and (S,S)-
47(R1 ) i-Pr, R2 ) H) were obtained in a combined yield of 72%.
The de was 42%, and (S,S)-47 was found to be the major isomer
after hydrolysis to the corresponding â-hydroxyamide and subse-
quent X-ray analysis.
7.62 (2H, m (∼t), CHAr), and 7.69-7.71 (1H, m (∼d), CHAr); 13
C
NMR unfortunately, there was insufficient material to obtain a clear
spectrum; m/z (EI) 317 [100%, (M+)], 274 (65), 203 (70), 100 (84).
X-ray crystallographic data confirmed the structure (see Supporting
Information).
4,4-Dimethyl-2-(9-phenyl-6,7,8,9-tetrahydro-5H-benzocyclo-
1
hepten-1-yl)-2-oxazoline (17): clear oil (19%); H NMR δ 1.24
(3H, s, CH3), 1.28 (3H, s, CH3), 1.30-1.40 (1H, m, C6HAHB),
1.68-1.89 (3H, m, C6HAHB and C7H2), 1.95-2.06 (1H, m,
C8HCHD), 2.47-2.66 (3H, m, C8HCHD and C5H2), 3.89 (1H, d, AB
doublet, J ≈ 8.1 Hz, CHEHFO), 3.94 (1H, d, AB doublet, J ≈ 8.1
Hz, CHEHFO), 5.01 (1H, dd, J ) 5.9 and 2.7, C9H), 7.02 (2H, d,
J ) 8. Hz 1, C10H), 7.10-7.31 (5H, m, C3H, C4H, C11H and C12H),
7.44 (1H, dd, J ) 5.8 and 3.4 Hz, C2H); 13C NMR δ 25.1 (C7),
27.4 (C6), 28.0 (CH3), 28.1 (CH3), 30.7 (C8), 35.9 (C5), 44.4 (C9),
67.5 ((CH3)2C), 78.9 (CH2O), 125.1 (C12), 126.3 (C3), 127.4 (C10),
127.8 (C2), 128.0 (C11), 130.2 (C1), 132.5 (C4), 142.2 (C9′ or C9′′),
142.3 (C9′ or C9′′), 144.2 (C4′), and 163.9 (CdN) (see Supporting
Information for 2D NMR data); m/z (EI) 319 [100%, (M+)], 291
(27), 290 (20), 247 (20), 228 (16), 219 (24), 202 (25), 189 (24).
Calcd for C22H25NO: 319.1936. Found: 319.1933.
(4S,1′R)-4-Isopropyl-2-(1′-phenylindan-1′-yl)-2-oxazoline (S,R)-
47 (R1 ) i-Pr, R2 ) H). Rf (SiO2, 9:1 hexanes:EtOAc) 0.2; [R]D
25
2-[1,5-Diphenylpentyl]-4,4-dimethyl-2-oxazoline (19) (7%):
m/z (EI) 321 [20%, (M+)], 202 (100), 188 (82), 91 (38). Yields
were determined by a combination of 1H NMR and GCMS analysis.
Alkene 16 and fluorene 18 were obtained pure by crystallization
from the fractions obtained from column chromatography. Reaction
with method B gave the yields shown in Table 1.
96.8 (c ) 1.02 in MeOH); νmax/cm-1 1655(CdN); 1H NMR δ 0.94
(3H, d, J ) 6.7 Hz, CH3), 1.02 (3H, d, J ) 6.7 Hz, CH3), 1.90
(1H, sept, J ) 6.7 Hz, CH), 2.26-2.34 (1H, m, CHAHBCHCHD),
2.79-2.89 (1H, m, CHAHBCHCHD), 2.97-3.07 (1H, m, CHAHB
-
CHCHD), 3.19-3.28 (1H, m, CHAHBCHCHD), 3.87-4.04 (2H, m,
CHEHF and CHN), 4.19-4.25 (1H, m, CHEHF), 7.09-7.38 (8H,
m, ArH), and 7.46-7.49 (1H, m, ArH); 13C NMR δ 17.9 (CH3),
19.2 (CH3), 30.5 (CH2), 32.4 ((CH3)2CH), 41.2 (ArCH2), 58.6
(PhC), 70.1 (CH2O), 71.8 (CHN), 124.6 (CH), 126.3 (CH), 126.6
(CH), 126.7 (CH), 126.8 (CH), 127.7 (CH), 128.2 (CH), 143.9 (C),
144.4 (C), 144.7 (C), and 168.8 (CdN); m/z (CI) 306 [100%,
(MH)+]. Calcd for C21H24NO (MH)+: 306.1858. Found: 306.1862.
(4S,1′S)-4-Isopropyl-2-(1′-phenylindan-1′-yl)-2-oxazoline (S,S)-
Reaction of 2-[3-(2-Bromophenyl)-1-phenylhexyl]-4,4-di-
methyl-2-oxazoline (7) with LDA. Reactions were carried out with
both methods A and B. 1H NMR spectroscopy revealed both
reactions produced a multitude of products. Due to the small scale
of these reactions (ca. 0.24 mmol), it was decided to combine the
mixtures from both methods to enable column chromatography to
be carried out with hope of identifying reaction products. The
fractions (ca. 8) obtained from the column were mixtures of
1
25
47 (R1 ) i-Pr, R2 ) H). Rf (SiO2, 9:1 hexanes:EtOAc) 0.15; [R]D
components which were analyzed by H NMR spectroscopy and
106.9 (c ) 0.32 in CHCl3); νmax/cm-1 1654 (CdN); H NMR δ
1
GCMS, which indicated the main products were analogous to those
from 6, i.e., compounds 20-23 (see Supporting Information).
Reaction of 2-[2-(2-Bromophenyl)-1-phenylethyl]-4,4-di-
methyl-2-oxazoline (3) with LDA. Method A gave 2-phenanthren-
9-yl-4,4-dimethyl-2-oxazoline (35) (88%); 1H NMR δ 1.19 (3H, s,
CH3), 1.21 (3H, s, CH3), 3.41 (2H, AB, CH2O), 7.60 (4H, m, CHAr),
7.87 (1H, m, CHAr), 8.28 (1H, m, CHAr), 8.64 (2H, m, CHAr), and
9.02 (1H, m, CHAr); 13C NMR δ 28.54 (CH3), 30.94 (CH3), 65.86
(C(CH3)2), 78.52 (CH2O), 122.58 (CHAr), 122.76 (CHAr), 126.79
(CHAr), 126.91 (CHAr), 127.09 (CHAr), 127.23 (CHAr), 128.28
(CHAr), 128.81 (CHAr), and 129.60 (CHAr); m/z (GC-MS) 275. m/z
(EI) 276 [99%, (MH)+]. Calcd for C19H18NO (MH)+: 276.1388.
Found: 276.1391.
Reaction of 3a with LDA and BuLi. To an LDA solution was
added a solution of substrate 3a (1 equiv, 500 mg) in dry THF
(4.19 mL). After stirring for 10 min, the solution was warmed to
rt over 30 min, at which time more dry THF (25 mL) was added
along with n-BuLi (1 equiv, 0.6 mL). A sample was taken for EPR
analysis, and the mixture was stirred for 48 h at rt. After 48 h, the
reaction was quenched with a saturated solution of NH4Cl (20 mL).
The aqueous layer was extracted with ether (4 × 10 mL), the
combined organic layers were dried (Na2SO4), and the solvent was
evaporated. Chromatography (SiO2, 8:2 hexanes:EtOAc) gave spiro-
compound 36 (0.37 g, 81%). The sample was hydrolyzed overnight
in 2% HCl solution (50 mL). The reaction mixture was extracted
with ether (50 mL), and the aqueous layer was treated with 10%
sodium bicarbonate solution. The resulting mixture was extracted
with DCM (3 × 50 mL). The combined organic layers were washed
with brine and dried over (Na2SO4), and solvents were evaporated
0.74 (3H, d, J ) 6.8 Hz, CH3), 0.85 (3H, d, J ) 6.8 Hz, CH3),
1.70 (1H, sept, J ) 6.8 Hz, CH), 2.26-2.35 (1H, m, CHAHB-
CHCHD), 2.80-2.90 (1H, m, CHAHBCHCHD), 2.98-3.08 (1H, m,
CHAHBCHCHD), 3.15-3.24 (1H, m, CHAHBCHCHD), 3.95-4.06
(2H, m, CHEHF and CHN), 4.19-4.26 (1H, m, CHEHF), 7.09-
7.13 (1H, m, ArH), 7.17-7.30 (7H, m, ArH), and 7.41-7.48 (1H,
m, ArH); 13C NMR δ 17.6 (CH3), 18.5 (CH3), 30.5 (CH2), 32.6
(CH(CH3)2, 41.3 (CH2), 70.3 (CH2O), 71.4 (CHN), 124.6 (CHAr),
126.3 (CHAr), 126.5 (CHAr), 126.6 (CHAr), 126.7 (CHAr), 127.7
(CHAr), 128.2 (CHAr), 144.0 (Cq), 144.2 (Cq), 144.6 (Cq), and 168.9
(CdN); m/z (CI) 306 [100%, (MH)+]. Calcd for C21H24NO
(MH)+: 306.1858. Found: 306.1865. Results for reactions under
UV irradiation for 6 h and at lower temperatures in THF are shown
in Table 3.
(1S,1′S)-1′-Phenylindan-1′-carboxylic Acid (2-Hydroxy-1-iso-
propylethyl)-amide (S,S)-48. (4S,1′S)-4-Isopropyl-2-(1′-phenylin-
dan-1′-yl)-2-oxazoline (0.5 g, 1.6 mmol) in 2% hydrochloric acid
solution (50 mL) was stirred at room temperature overnight. The
reaction mixture was extracted with ether (50 mL), and the aqueous
layer was treated with 10% sodium hydrogen carbonate solution.
The resulting mixture was extracted with dichloromethane (3 ×
50 mL). The combined organic layers were washed with brine, dried
(MgSO4), and evaporated to give a yellow oil. The oil was left to
stand until it had all crystallized and was then recrystallized from
EtOAc/hexanes to give the amide (S,S)-48 (0.5 g, 95%) as white
25
needles; mp 115.5-117.0; [R]D 8.9 (c ) 0.96 in MeOH); νmax
/
cm-1 1638 (CdO), 3290 and 3357; H NMR δ 0.65 (3H, d, J )
6.9 Hz, CH3), 0.77 (3H, d, J ) 6.9 Hz, CH3), 1.66-1.82 (1H, m,
1
910 J. Org. Chem., Vol. 72, No. 3, 2007

1,3-Areneotropic Migration of Oxazolines
CH(CH3)2), 2.17 (1H, br s, OH), 2.36-2.45 (1H, m, CHAHB-
CHCHD), 2.84-2.93 (1H, m, CHAHBCHCHD), 2.96-3.07 (1H, m,
CHAHBCHCHD), 3.18-3.27 (1H, m, CHAHBCHCHD), 3.57-3.63
(1H, m, OCHEHF), 3.68-3.78 (2H, m, CHN and OCHEHF), 5.68
(1H, d, J ) 7.4 Hz, NH), and 7.18-7.37 (9H, m, ArH); 13C NMR
δ 18.0, 19.5, 28.8, 30.8, 41.0, 57.9, 64.8, 66.0, 124.9 (CHAr), 125.4
(CHAr), 126.8 (CHAr), 127.1 (CHAr), 127.7 (CHAr), 128.1 (CHAr),
128.6 (CHAr), 143.2, 145.0, 145.1, and 175.4 (CdO); m/z (CI) 324
[100%, (MH)+]. Anal. Calcd for C21H25NO2: C, 78.0; H, 7.7; N,
4.3. Found: C, 78.2; H, 8.0; N, 4.4. Calcd for C21H26NO2 (MH)+:
324.1964. Found: 324.1964. X-ray diffraction data fully confirmed
the structure (see Supporting Information) and indicated the
stereochemistry was (S,S).
Reaction of (4S)-2-[3-(2-Bromophenyl)-1-phenyl-propyl]-4-
phenyl-2-oxazoline (S)-42 with LDA. Method A gave 4-phenyl-
2-(1′-phenylindan-1′-yl)-oxazole (49) (60%); H NMR δ 2.30-
2.40 (1H, m, CHAHB), 2.78-2.88 (1H, m, CHAHB), 2.96-3.06 (1H,
m, CHAHB), 3.15-3.24 (1H, m, CHAHB), and 6.98-7.70 (15H,
m, vinyl and CHAr).
1
Reaction of (4S)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-
cyclohexylmethyl-2-oxazoline (S)-44. Methods A and B led to poor
conversion but afforded impure 4-cyclohexylmethyl-2-(1′-phenylin-
dan-1′-yl)-2-oxazoline (47) m/z (EI) 359 [100%, (M+)], 193 (90),
115 (28). Analysis of the mixture by GC-MS gave the diastereo-
meric excesses, 16% (method A) and 16% (method B).
General Procedure for Reactions with LDA in the Presence
of Iron(II) Chloride. To a solution of LDA (3 equiv) in either
THF or 2:1 hexane:THF mixture (6 mL/mmol) at -78 °C under
nitrogen was added a solution of the substrate in the reaction solvent
(3 mL/mmol). After stirring for 10 min the solution was allowed
to warm to room temperature over 30 min, after which time more
solvent was added (18 mL/mmol) followed by addition of either
10 mol % or 1 equiv of anhydrous iron(II) chloride. The mixture
was then stirred for 1 h, after which a saturated solution of
ammonium chloride (10 mL/mmol) was added. The aqueous layer
was extracted with three portions of ether (3 × 5 mL/mmol), and
the combined organic layers were washed with water (10 mL/mmol)
and dried (MgSO4) to give the crude product.
Oxazoline 4a was reacted in both THF and 2:1 THF:hexanes
with addition of 0.1 or 1 equiv of FeCl2. 1H NMR analysis revealed
the formation of 4,4-dimethyl-2-(1′-phenylindan-1′-yl)-2-oxazoline
11 (n ) 1, R ) Ph) (for data see above), 2-[1,3-diphenylpropyl]-
4,4-dimethyl-2-oxazoline 12 (n ) 1, R ) Ph), and styrenes together
with starting material. The product distributions are shown in Table
1.
Reaction of 2-Haloarylalkyl-2-oxazolines with NaH/DMSO.
To NaH (60% dispersion in mineral oil) (32.4 mg, 0.81 mmol),
under nitrogen, was added dry THF (1 mL) with stirring. NaH was
allowed to settle, and THF was withdrawn using a syringe. This
was repeated. Dry, distilled DMSO (4 mL) was then added slowly
with stirring. After 10 min, oxazoline 4a (1 mL, 0.27 M solution
in dry DMSO) was added dropwise with stirring. Then with either
heating and/or UV irradiation and/or FeCl2 (1 equiv) the mixture
was stirred for between 10 min and 4 h. Finely ground NH4Cl (0.5
g) was added and stirred for 10 min. Water (5 mL) and ether (10
mL) were added, and the aqueous layer was extracted with ether
(2 × 5 mL). The combined organic extracts were washed (water
(5 mL)), dried (MgSO4), and concentrated to give the crude reaction
mixture. NMR spectroscopy showed the following products: 4,4-
dimethyl-2-(1′-phenylindan-1′-yl)-2-oxazoline 11 (n ) 1, R ) Ph)
and 2-[1,3-diphenylpropyl]-4,4-dimethyl-2-oxazoline 12 (yields in
Table 1).
Reaction of (4S)-2-[3-(2-Bromophenyl)-1-phenylpropyl]-4-
tert-butyl-2-oxazoline (S)-41 with LDA. After work up the
1H NMR spectrum showed only diastereomeric indanes (S,R)- and
(S,S)-47 (R1 ) t-Bu, R2 ) H). Yields and selectivities were
determined by both 1H NMR (integration of tBu signals) and GCMS
(peak areas). An attempt was made to separate the two isomers
from each other by use of column chromatography but gave an
isomeric mixture of â-hydroxyamides as a result of hydrolysis.
(4S)-4-tert-Butyl-2-(1′-phenylindan-1′-yl)-2-oxazoline [(S,S)-
1
and (S,R)-47 (R1 ) t-Bu, R2 ) H)]: H NMR δ 0.63 (9H, s,
(CH3)3, major isomer), 0.82 (9H, s, (CH3)3, minor isomer), 2.10-
2.20 (1H, m, CHAHBCHCHD), 2.63-2.78 (1H, m, CHAHBCHCHD),
2.82-2.97 (1H, m, CHAHBCHCHD), 3.00-3.18 (1H, m, CHAHB-
CHCHD), 3.69 (1H, dd, J ) 10.4 and 8.2 Hz, CHN, minor isomer),
3.79 (1H, dd, J ) 10.0 and 6.7 Hz, CHN, major isomer), 3.90-
3.95 (1H, m, CHAHBO), 4.01-4.07 (1H, m, CHAHBO), 6.95-7.02
(2H, m, ArH), 7.02-7.16 (6H, m, ArH) and 7.28-7.37 (1H, m,
ArH). Hydrolysis gave (1′S)-1-phenylindan-1-carboxylic acid (1′-
hydroxymethyl-2′,2′-dimethyl-propyl)amide as a 3:2 mixture of
isomers (80%), Rf (9:1 hexane:EtOAc) 0.1; νmax(film)/cm-1 1629
1
(CdN) and 3344 (NH and OH); H NMR δ 0.68 (9H, s, (CH3)3,
major isomer), 0.81 (9H, s, (CH3)3, minor isomer), 2.38-2.48 (1H,
m, CHAHBCHCHD), 2.79-3.07 (2H, m, CHAHBCHCHD and CHAH-
BCHCHD), 3.18-3.30 (1H, m, CHAHBCHCHD), 3.35-3.49 (1H, m,
CHN), 3.74-3.88 (2H, m, CH2O), 5.67 (1H, d, J ) 8.0 Hz, NH,
minor isomer), 5.73 (1H, d, J ) 8.0 Hz, NH, major isomer), and
7.16-7.37 (9H, m, ArH); 13C NMR δ 26.6 ((CH3)3C, major isomer),
26.8 ((CH3)3C, minor isomer), 30.7 (CH2, minor isomer), 30.8 (CH2,
major isomer), 33.0 ((CH3)3C, major isomer), 33.3 ((CH3)3C, minor
isomer), 40.4 (CH2, minor isomer), 41.2 (CH2, major iomer), 60.5
(CHN, minor isomer), 60.6 (CHN, major isomer), 63.8 (CH2O,
minor isomer), 63.9 (CH2O, major isomer), 65.9 (PhC, minor
isomer), 66.1 (PhC, major isomer), 124.9 (CHAr), 125.38 (CHAr,
minor isomer), 125.44 (CHAr, major isomer), 126.7 (CHAr, minor
isomer), 126.8 (CHAr, major isomer), 127.1 (CHAr, major isomer),
127.2 (CHAr, minor isomer), 127.6 (CHAr), 127.9 (CHAr, minor
isomer), 128.1 (CHAr, major isomer), 128.5 (CHAr, minor isomer),
128.6 (CHAr, major isomer), 142.9 (Cq, minor isomer), 143.0 (Cq,
major isomer), 144.8 (Cq, major isomer), 144.9 (Cq, minor isomer),
145.0 (Cq, major isomer), 145.3 (Cq, minor isomer), 175.5 (CdO,
major isomer), and 175.8 (CdO, minor isomer); m/z (ES+) 360
[100%, (MNa)+]. Calcd for C22H27NO2Na (MNa)+: 360.1939.
Found: 360.1935.
Acknowledgment. We thank the EPSRC for financial
support.
Supporting Information Available: General experimental
procedures; preparations of halophenylalkyl halides and chiral
oxazolines; NMR and MS data for reaction of 7; DFT Cartesian
coordinates and total energies for radical anions 10, 25e, 25a, and
27; views of the SOMOs for 25e, 25a, and 27; EPR spectra obtained
from 4a, 5, and 6; NMR spectra for precursor oxazoline derivatives,
for 11(n ) 1, R ) Ph) and 47(R1)i-Pr, R2)H); 2D NMR spectra
for 17; X-ray diffraction data for 11(n ) 2, R ) Ph), (S,S)-48, 16
and 18. This material is available free of charge via the Internet at
http://pubs.acs.org.
Reactions of (S)- and (R)-2-[3-(2-Bromophenyl)-1-phenyl-
propyl]-4-tert-butyl-2-oxazoline (S)- and (R)-41 with LDA and
(-)-Spartiene. Reactions were as described in the general proce-
dures except that 3.5-10 equiv of (-)-spartiene (50) was added
prior to addition of the oxazoline. The amount of THF added on
reaching rt was adjusted such that the oxazoline concentration was
equal to either 0.036 or 0.1 M. After a standard workup a small
amount of product mixture was analyzed via GCMS to obtain the
diastereomeric excess of the formed Indane (see Table 3).
JO0620720
J. Org. Chem, Vol. 72, No. 3, 2007 911