The Unique Behavior of a Chiral Binaphthyl Oxazoline in the Presence of Cu(I) and Its Role as a Chiral Catalyst

Full text of DOI:10.1021/jo971761w

412
J . Org. Chem. 1998, 63, 412-413
Communications
configuration (Scheme 1, 4A) preferred to have both naph-
thalene rings at 68° and the oxazolines almost orthoganol
to the naphthalenes. The oxazoline rings did not appear to
show any rotation about the connecting C-C bond (Scheme
1A, bond a), for this would place the tert-butyl groups into
the π-face of the naphthalene rings. Since the 3:1 mixture
of 4:5 returns to >98% 4 at room temperature, this supports
the earlier report5 that the activation barrier to atropi-
somerization of many 1,1′,8,8′-binaphthyls is rather low.
This is especially so when sp2 substituents are present at
the 8,8′ position. The barrier is much higher with sp3
substituents OH and OMe, since they present greater
volume during the rotation, and these derivatives have
indeed been resolved into enantiomers.6-8 Fuji4 also re-
ported facile atropisomerization in similar systems contain-
ing phosphorus substituents and attributes the behavior to
distortion from planarity in the naphthalene rings, which
was confirmed by X-ray data. Careful examination of the
X-ray structure of 42 showed no such distortion in the
naphthalene rings so the stability of 4 as the only atropi-
somer at room temperature must be due to other factors
(vide supra).
Th e Un iqu e Beh a vior of a Ch ir a l Bin a p h th yl
Oxa zolin e in th e P r esen ce of Cu (I) a n d Its
Role a s a Ch ir a l Ca ta lyst
A. I. Meyers* and Alan Price
Department of Chemistry, Colorado State University,
Fort Collins, Colorado 80523-1872
Received September 22, 1997
We have been engaged for several years in the acquisition
of chiral binaphthyls from the Ullmann coupling of 1-bromo-
2-oxazolinylnaphthalenes 11 and 8-bromo-1-oxazolinylnaph-
thalenes 3.2 In both cases, the binaphthyl coupling products
2 and 4, respectively, were obtained in high diastereoselec-
tivity (>99%).1,2 In view of the large effort expended by a
Although the low barrier to rotation in 4 to 5 may preclude
its use as a chiral catalyst, we still felt further study was
necessary. We were surprised to learn that when the
Ullmann coupling was performed in DMF in place of
pyridine, th e sole p r od u ct obtained was the aR-atropiso-
mer 5, with none of the aS-derivative detectable by NMR.2
When the copper salts were removed on workup, the aR-
atropisomer was stable as a crystalline solid at room
temperature for several hours and indefinitely at -20 °C.
However, when dissolved in various solvents at room tem-
perature, it was rapidly and completely transformed back
to the aS-isomer 4. Of most interest was the ability to
transform 4 back to 5 only after heating a DMF solution to
90-100 °C containing 1.0 equiv of CuBr. A solution of the
copper complex of 5 in CDCl3 at room temperature was also
found to be stable to atropisomerization indefinitely.
The reasons for this behavior are now more clearly
revealed in light of these data. For example, when 1 was
transformed into 2, only the aS-atropisomer was formed, and
this was explained earlier1 by assuming the Cu was holding
the two ligands (oxazoline nitrogen) in a manner that
minimized all nonbonded interactions. Furthermore, the
barrier for aryl-aryl rotation in 2 is very high so the
kinetically controlled product was the only one formed and
isolated. In the present case of naphthyloxazoline 3, when
the coupling is carried out in pyridine, we were surprised
to find that 4 and not 5 was the product, even though the
latter is capable of readily forming a bidentate oxazoline
complex to copper. We, therefore, attributed the formation
of 4 to buttressing of the tert-butyl groups.1 In light of the
present results, pyridine may compete with the oxazolines
for the copper ion (R-configuration). In the absence of copper
ion, 5 can suffer lone-pair repulsion from both oxazoline
nitrogens, resulting in atropisomerization to 4A. On the
other hand, using DMF to couple 3, the kinetically formed
number of laboratories to utilize chiral binaphthyls, mainly
of the type depicted by 2, as chiral catalysts, we have focused
our recent attention on the more rarely studied4 1,1′,8,8′-
systems, 4. In our recent preliminary report,2 we showed
that the chiral bromo derivative 3 gave, upon heating in
refluxing pyridine containing activated copper powder, the
single diastereomer (aS,S)-4 which was quite stable to
atropisomerism, even after heating for prolonged periods.
Surprisingly, heating 4 for 24 h at 145 °C gave a mixture of
4 and 5 in a 75:25 ratio, only to return to 4 upon standing
at room temperature. An X-ray structure2 of 4 showed that
(1) Nelson, T. D.; Meyers, A. I. J . Org. Chem. 1994, 59, 2655.
(2) Meyers, A. I.; McKennon, M. J . Tetrahedron Lett. 1995, 36, 5869.
(3) For recent reviews see: (a) Rossini, C.; Fannzini, L.; Raffaelli, A.;
Salvadori, P. Synthesis 1992, 503. (b) Narasaka, K. Synthesis 1991, 1. (c)
Tomioka, K. Synthesis 1990, 541.
(4) Fuji, K.; Sakurai, M.; Tohkai, N.; Kuroda, A.; Kawabata, T.; Fuka-
zawa, Y.; Kinoshita, T.; Tada, T. J . Chem. Soc., Chem. Commun. 1996, 1609.
(5) Harris, M. M.; Patel, P. K.; Korp, J . D.; Bernal, I. J . Chem. Soc., Perkin
Trans. 2 1981, 12, 1621 and earlier references cited.
(6) Fabbri, D.; Delogu, G.; DeLucchi, O. J . Org. Chem. 1995, 60, 6599.
(7) Fuji, K.; Kawabata, T.; Kuroda, A.; Taga, T. J . Org. Chem. 1995, 60,
1914.
(8) Artz, S. P.; deGrandpre, M. P.; Cram, D. J . J . Org. Chem. 1985, 50,
1486.
S0022-3263(97)01761-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/14/1998

Communications
J . Org. Chem., Vol. 63, No. 3, 1998 413
Sch em e 1. (A) a S-Bin a p h th yloxa zolin e w ith ou t Cu Com p lex (4) a n d a R-“Con figu r a tion ” w ith Cu -In ser ted (5).
(B) Sa m e, bu t Sim p lified Lin e Sch em a tic
cyclopropane products as a 1:1 mixture of 6 and 7 both
5A is stable as a copper complex since the DMF cannot
showing virtually no enantiomeric excess. However, when
compete as a ligand with the oxazolines for the copper ion
5 was complexed11 with CuOTf, an unoptimized yield of 76%
of cyclopropane products (6, 7) was obtained in a 2:1 ratio.
In this instance, (1R,2S)-7 was formed in 90% ee and
(Scheme 1). In summary, the atropisomerization of 4 to 5
and its reverse is not due to any specific steric buttressing
by the tert-butyl groups, but purely to a stereoelectronic
(1R,2R)-6 in 67% ee. This result suggests that 5 may form
effect, namely the lone-pair repulsions in 5 that drive the
a complex that may be more organized (e.g., bidentate) to
molecule back to 4 or the copper complexation, which drives
4 to 5. Therefore, one may safely assume that the metal-
free thermodynamic product in this process is 4, but in DMF
solvent, the kinetic product, due to Cu-complexation, is
enter into the transition state of the cyclopropanation than
the copper complex of 4. Clearly, a 1:1 complex of 4 with
Cu(I) would be highly unlikely due to a severe tert-butyl-
naphthalene interaction. In fact, any attempt to complex
the Cu(I) to the two oxazolines in 4 would drive the tert-
butyl group directly into the face of the naphthalene rings.
actually 5.
The question was next posed to determine whether either
or both binaphthyl bis-oxazolines would serve as a chiral
catalyst. Thus, 4 and 5 were employed as chiral ligands for
Because of this, the copper complex of 4 is probably oligo-
meric or exists as nonspecific aggregates.11,12
the well-known9,10 copper-catalyzed cyclopropanation of
In summary, this difference in behavior is undoubtedly
styrene using ethyl diazoacetate. The cyclopropanation was
due to the low barrier to rotation, which will allow a single
carried out in CH2Cl2 in the presence of 2 mol % copper(I)
diastereomer (4) to be transformed by heating into an
alternate configuration by simple metal coordination. The
fact that the two copper forms are separated by a significant
energy barrier, such that no interconversion occurs at room
temperature, also adds to their utility. Further studies and
use of other metals are under investigation, as well as
catalyst formed at room temperature by in situ mixing of
CuOTf‚1/2C6H6 and either 4 or 5. The ratio of trans:cis
products where monitored by 1H NMR and the enantiomeric
excess of 6 and 7, were determined by chiral HPLC (Chiracel
OB). The complex of 4 and CuOTf gave a 21% yield of
structural information on the copper complexes.
Ack n ow led gm en t. The authors are grateful to the
National Science Foundation and the National Institutes
of Health for support of this study.
J O971761W
(9) (a) Ojima, I. Catalytic Asymmetric Synthesis; VCH: New York, 1993.
(b) Pfaltz, A. Acc. Chem. Res. 1993, 26, 339. See also: Uozumi, Y.; Kyota,
H.; Kishi, E.; Kitayama, K.; Hayashi, T. Tetrahedron: Asymmetry 1996, 7,
1603.
(10) A recent report (Gant, T. G.; Noe, M. C.; Corey, E. J . Tetrahedron
Lett. 1995, 36, 8745) described the use of chiral biphenyloxazolines as
effective Cu-ligands for cyclopropanation.
(11) It is important to recall, as mentioned earlier, that 4 cannot be
transformed into 5 unless it is heated to 90-100 °C in the presence of CuBr
in DMF. Thus, at room temperature, only a “vague complex” or aggregate
is formed between Cu(I) and 4.
(12) Aggregations of Cu(I)‚oxazoline complexes have been reported
(Evans, D. A.; Woerpel, K. A.; Scott, M. J . Angew. Chem., Int. Ed. Engl.
1992, 31, 430).