Application of the Russig-Laatsch Reaction to Synthesize a Bis[5]helicene Chiral Pocket for Asymmetric Catalysis

Article abstract of DOI:10.1021/jo991498u

The enantiomers of a bis[5]helicenediol ligand ([5]HELOL) have been synthesized in appreciable amounts by a procedure in which key steps are the union of p-benzoquinone with an enol ether of 3-acetylphenanthrene and the displacement of phenol and phenol ether functions by alcohols (the Russig-Laatsch reaction). This diol catalyzes the addition of diethylzinc to aldehydes and gives nonracemic alcohols with enantiomeric excesses as high as 81%. The stereoselectivities and yields are much greater than when the catalyzing diol is BINOL. The enantioselectivities are greater also than those of other reactions catalyzed by helicene ligands.

Full text of DOI:10.1021/jo991498u

J . Org. Chem. 2000, 65, 815-822
815
Ap p lica tion of th e Ru ssig-La a tsch Rea ction to Syn th esize a
Bis[5]h elicen e Ch ir a l P ock et for Asym m etr ic Ca ta lysis
†
,†
‡
‡
Spencer D. Dreher, Thomas J . Katz,* Kin-Chung Lam, and Arnold L. Rheingold
Department of Chemistry, Columbia University, New York, New York 10027, and Department of
Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716
Received September 22, 1999
The enantiomers of a bis[5]helicenediol ligand ([5]HELOL) have been synthesized in appreciable
amounts by a procedure in which key steps are the union of p-benzoquinone with an enol ether of
3
-acetylphenanthrene and the displacement of phenol and phenol ether functions by alcohols (the
Russig-Laatsch reaction). This diol catalyzes the addition of diethylzinc to aldehydes and gives
nonracemic alcohols with enantiomeric excesses as high as 81%. The stereoselectivities and yields
are much greater than when the catalyzing diol is BINOL. The enantioselectivities are greater
also than those of other reactions catalyzed by helicene ligands.
In tr od u ction
yields achieved with it are often excellent, they are not
always. Because the large distance between the center
8
Despite their apparent potential, helicenes have been
used as ligands in asymmetric catalysis only once. The
of chirality and the metal may be to blame, ways have
been studied to move the chirality closer to the metal.
One way has been to prepare derivatives of BINOL in
which the 3 and 3′ positions are substituted, and these
ligands have proven effective in, for example, Diels-
1
probable reason, that they have been difficult to obtain
in large amounts, should be obviated by the procedure
recently developed to make them from p-benzoquinone
and the enol ethers of aryl methyl ketones. This proce-
dure has already made available substantial quantities
9
10
11
Alder, ene, and olefin-metathesis reactions, Claisen
12
rearrangements, and dialkylzinc additions to alde-
2
3
of [5] -, [6]-, and [7]carbo- and heterohelicenes. These
in turn have been used to prepare the first examples of
8b-d,13
hydes.
Another way, developed by Wulff, uses a
14
“
vaulted biaryl” ligand (2), “VAPOL”, to extend the ring
system in the direction of the phenolic hydroxyls, forming
a chiral pocket
4
helical conjugated polymers, helical columnar meso-
5
6
phases, and chiral aggregates. The aggregates exhibit
large optical rotations and large nonlinear optical re-
sponses, and they have been used to demonstrate that
quasi-phase matched second harmonics can be produced
by making use of a material’s chirality.6f
14a
around the metal. The enantioselec-
tivities attained with this ligand in Diels-Alder reactions
14
are excellent. In view of the successes achieved with
substituted BINOLS and VAPOL, we wondered whether
a bis-helicene diphenolic ligand such as 3, because it
should form a larger chiral pocket around the catalytic
metal center, would be an even better asymmetric
catalyst.
1
,1′-Binaphthalene-2,2′-diol (BINOL) (1, R ) H) is
probably the ligand most frequently used for asymmetric
catalysis,7 and although the enantioselectivities and
A significant feature of 3’s structure is that the
stereochemistry of the single bond uniting the two
helicene moieties is constrained, but not by the bulk of
†
Columbia University.
University of Delaware.
‡
(1) (a) Reetz, M. T.; Beuttenmuller, E. W.; Goddard, R. Tetrahedron
Lett. 1997, 38, 3211. See also: (b) Terfort, A.; G o¨ rls, H.; Brunner, H.
Synthesis 1997, 79.
(8) (a) Prasad, K. R. K.; J oshi, N. N. Tetrahederon: Asymmetry 1996,
7, 1957. (b) Hu, Q.-S.; Huang, W.-S.; Vitharana, D.; Zheng, X.-F.; Pu,
L. J . Am. Chem. Soc. 1997, 119, 12454. (c) Ding, K.; Ishii, A.; Mikami,
K. Angew. Chem., Int. Ed. Engl. 1999, 38, 497 and references therein.
(d) Hu, Q.-S.; Pugh, V.; Sabat, M.; Pu, L. J . Org. Chem. 1999, 64, 7528.
(9) (a) Kelly, T. R.; Whiting, A.; Chandrakumar, N. S. J . Am. Chem.
Soc. 1986, 108, 3510. (b) Ishihara, K.; Kurihara, H.; Matsumoto, M.;
Yamamoto, H. J . Am. Chem. Soc. 1998, 120, 6920 and references
therein. (c) Kagan, H. B.; Riant, O. Chem. Rev. 1992, 92, 1007.
(10) (a) Maruoka, K.; Hoshino, Y.; Tadashi, T.; Yamamoto, H.
Tetrahedron Lett. 1988, 29, 3967. (b) Carriera, E. M.; Lee, W.; Singer,
R. A. J . Am. Chem. Soc. 1995, 117, 3649. (c) Berrisford, D. J .; Bolm,
C. Angew. Chem., Int. Ed. Engl. 1995, 34, 1717.
(11) Zhu, S. S.; Cefalo, D. R.; La, D. S.; J amieson, J . Y.; Davis, W.
M.; Hoveyda, A. H.; Schrock, R. R. J . Am. Chem. Soc. 1999, 121, 8251
and references therein.
(12) Maruoka, K.; Banno, H.; Yamamoto, H. J . Am. Chem. Soc. 1990,
112, 7791.
(13) (a) Kitajima, H.; Ito, K.; Aoki, Y.; Katsuki, T. Bull. Chem. Soc.
J pn. 1997, 70, 207. (b) Huang, W.-S.; Hu, Q.-S.; Pu, L. J . Org. Chem.
1998, 63, 1364.
(14) (a) Bao, J .; Wulff, W. D. J . Am. Chem. Soc. 1993, 115, 3814. (b)
Bao, J .; Wulff, W. D. Tetrahedron Lett. 1995, 36, 3321. (c) Bao, J .; Wulff,
W. D.; Dominy, J . B.; Fumo, M. J .; Grant, E. B.; Rob, A. C.; Whitcomb,
M. C.; Yeung, S.-M.; Ostrander, R. L.; Rheingold, A. L. J . Am. Chem.
Soc. 1996, 118, 3392. (d) Heller, D. P.; Goldberg, D. R.; Wulff, W. D. J .
Am. Chem. Soc. 1997, 119, 10551 and references therein.
(2) (a) Katz, T. J .; Liu, L.; Willmore, N. D.; Fox, J . M.; Rheingold,
A. L.; Shi, S.; Nuckolls, C.; Rickman, B. H. J . Am. Chem. Soc. 1997,
1
1
19, 10054. (b) Fox, J . M.; Goldberg, N. R.; Katz, T. J . J . Org. Chem.
998, 63, 7456.
(
3) (a) Dreher, S. D.; Weix, D. J .; Katz, T. J . J . Org. Chem. 1999,
6
4, 3671. (b) Phillips, K. E. S.; Katz, T. J . Unpublished results.
(
(
(
4) Dai, Y.; Katz, T. J . J . Org. Chem. 1997, 62, 1274.
5) Nuckolls, C.; Katz, T. J . J . Am. Chem. Soc. 1998, 120, 9541.
6) (a) Nuckolls, C.; Katz, T. J .: Castellanos, L. J . Am. Chem. Soc.
1
996, 118, 3767. (b) Nuckolls, C.; Katz, T. J .; Katz, G.; Collings, P. J .;
Castellanos, L. J . Am. Chem. Soc. 1999, 121, 79. (c) Nuckolls, C.; Katz,
T. J .; Verbiest, T.; Van Elshocht, S.; Kuball, H.-G.; Kiesewalter, S.;
Lovinger, A. J .; Persoons, A. J . Am. Chem. Soc. 1998, 120, 8656. (d)
Verbiest, T.; Van Elshocht, S.; Kauranen, M.; Hellemans, L.; Snau-
waert, J .; Nuckolls, C.; Katz, T. J .; Persoons, A. Science (Washington,
D.C.) 1998, 282, 913. (e) Fox, J . M.; Katz, T. J .; Van Elshocht, S.;
Verbiest, T.; Kauranen, M.; Persoons, A.; Thongpanchang, T.; Krauss,
T.; Brus, L. J . Am. Chem. Soc. 1999, 121, 3453. (f) Busson, B.;
Kauranen, M.; Nuckolls, C.; Katz, T. J .; Persoons, A. Phys. Rev. Lett.
2
000, 84, 79.
7) Reviews: (a) Rosini, C.; Franzini, L.; Rafaelli, A.; Salvadori, P.
Synthesis 1992, 503. (b) Zimmer, R.; Suhrbier, J . J . Prakt. Chem. 1997,
39, 758. (c) Mikami, K.; Motoyama, Y. In Encyclopedia of Reagents
for Organic Synthesis, Vol 1; Paquette, L. A., Ed.; Wiley: New York,
(
3
1
1
995; pp 397-407. (d) Pu, L. Chem. Rev. 1998, 98, 2405. Also see ref
4c and references therein.
1
0.1021/jo991498u CCC: $19.00 © 2000 American Chemical Society
Published on Web 01/16/2000

8
16 J . Org. Chem., Vol. 65, No. 3, 2000
Dreher et al.
Sch em e 1a
adjacent substituents, as in biaryls such as 1 and 2, but
by the bulk of the distant helicene rings. Adjacent
substituents are not required. If the bond linking the two
aryls were rotated to the (R)-stereochemistry, the heli-
cenes in structure 3 would collide. Thus, when the
structure comprises two (P)-helicenes, the stereochem-
istry about the biaryl single bond must be (S).
a
Key: (a) triisopropylsilyl triflate, Et N, CH Cl , 0 to 25 °C, 2
3
2
2
This paper describes the synthesis of both enantiomers
of this ligand, which we call [5]HELOL, on a significant
scale and in eight steps from commercially available
materials. The [5]HELOL ligand is stable to racemization
at room temperature. Moreover, it acts as a catalyst for
h, 99% yield; (b) p-benzoquinone, toluene, reflux, 24 h; (c) n-BuI,
K2CO3, DMF, 60 °C, 2 h, 81% yield (two steps).
was poured into water, only the helicene precipitated.
After the three steps, the yield of [5]helicenequinone 7
from 3-acetylphenanthrene was 81%.
the addition of diethylzinc to aldehydes,8
,15
and when it
To complete the synthesis of the ligand, the following
transformations were needed: (1) the quinone had to be
reduced; (2) of the resulting phenolic hydroxyls, the one
on the outer periphery had to be converted selectively to
an ether, so only the other, on the inside of the ring
system, could participate in the oxidative dimerization
that would yield the desired bisphenol; and (3) the
enantiomers had to be resolved. Since preliminary stud-
ies showed that dimethyl ether 8 is particularly easy to
resolve (see below), we attempted to synthesize it by
following steps similar to those that led to 7, but with
methyl iodide in place of n-butyl iodide. However, this
approach was unsatisfactory, for the methylation, unlike
the butylation, did not separate the ring system derived
from 6a from that derived from 6b. Remarkably, the
procedure discovered by Russig17 and Laatsch to convert
naphthoquinones into monoalkylated ethers of naphtho-
hydroquinonessreduction with sodium dithionite fol-
lowed by reaction with an alcohol saturated with HCls
achieves the desired result. It converts 7 into 8 in two
simple steps (eq 1). This means it not only transforms
the quinone into the monomethyl ether of the hydro-
quinone, but at the same time it replaces the butoxyl by
a methoxyl. It also proved useful for attaching side-chains
to [6]- and [7]carbohelicenes, and a paper describing these
experiments provides details of the procedure’s applica-
tion.19 The best reaction conditions we could finds
methanol saturated with HCl, with added 1,2-dichloro-
ethane cosolvent, 30 min at 60 °Csgave a 75% yield of 8
(22.0 g). No trace of isomers was found, and the butoxy
side chain was cleanly replaced.
does, the enantioselectivities and yields are larger than
those produced by BINOL.
Resu lts a n d Discu ssion
In designing a helicene ligand for asymmetric catalysis
the following points were considered. First, to be useful,
the ligand must be cheaply and easily preparable on a
large scale. Second, it must be stable both to racemization
and decomposition under the conditions used for cata-
lyzed transformations. Third, a dimeric helicene is likely
to be more effective than a monomer because it can
surround a metal with a chiral pocket. The idea then was
that an appropriate bis-helicenediol might be made easily
from a [5]helicenequinone such as 7, which in turn should
be preparable in just two steps from 3-acetylphenan-
threne (4a ).
18
3
-Acetylphenanthrene containing approximately 10%
-acetylphenanthrene (4b) is available commercially, but
we obtained it more cheaply by acetylating phenanthrene
2
1
6
on a 100 g scale. 2-Acetylphenanthrene is a side
product, and it is difficult to remove.1 Accordingly, an
6c
1
1:1 mixture was used as the starting material for the
synthesis (Scheme 1). This mixture, when treated with
triisopropylsilyl triflate and triethylamine, gave in 99%
yield a mixture of silyl enol ethers 5a and 5b. These,
when combined for 24 h with p-benzoquinone in refluxing
toluene, gave a mixture of [5]helicenequinone 6a (from
3
-acetylphenanthrene) and its isomer 6b (from 2-acetyl-
phenanthrene). They seemed inseparable. However, the
structures separated unexpectedly when the mixture was
treated with n-butyl iodide and K
2
CO
3
in DMF at 60 °C,
Ambient temperatures sufficed to introduce the meth-
oxy group at the 4-position, but to replace the butoxy side
for helicene 6a readily alkylates, while the planar
molecule 6b does not. Thus, when the reaction mixture
1
9
chain required a temperature that was higher. At this
higher temperature, a significant amount of a white solid
side product (20% yield) was also produced. Analyses of
(15) Reviews: (a) Soai, K.; Niwa, S. Chem. Rev. 1992, 92, 833. (b)
Noyori, R. Asymmetric Catalysis in Organic Synthesis; Wiley: New
York, 1994; Chapter 5.
(
16) (a) Mosettig, E.; van de Camp, J . J . Am. Chem. Soc. 1930, 52,
(17) Russig, F. J . Prakt. Chem. 1900, 33.
(18) Laatsch, H. Liebigs Ann. Chem. 1980, 140.
(19) Dreher, S. D.; Paruch, K.; Katz, T. J . J . Org. Chem. 2000, 65,
806.
3
1
704. (b) Kloetzel, M. C.; Pandit, U. K. J . Am. Chem. Soc. 1956, 78,
412. (c) Fern a´ ndez, F.; G o´ mez, G.; L o´ pez, C.; Santos, A. J . Prakt.
Chem. 1989, 331, 15.

Synthesis of a Bis[5]Helicene Chiral Pocket
J . Org. Chem., Vol. 65, No. 3, 2000 817
Sch em e 2a
1
its mass and the NOEs in its H NMR spectrum sug-
gested it to have structure 9.20 As more of this product
was formed when the reaction was run in the light,
further studies were carried out in the dark. But even in
the dark, heating for 24 h at 60 °C converted 8 in
methanolic HCl completely into 9. Fortunately, 9 is
2 2
insoluble in CH Cl , and filtration therefore easily sepa-
rated it from the soluble 8.
a
Key: (a) (1S)-(-)-camphanoyl chloride, DMAP, Et3N, 1,2-
dichloroethane, 2 h, reflux, 80%yield of each diastereomer; (b)
KOH, EtOH, 0 °C, 2 h; (c) Ag2O, Et3N, CH2Cl2, 0 °C, 2 h, 90%
yield (two steps); (d) Zn, AcOH, acetone, 0 °C, 2 h, 100% yield of
each enantiomer.
Resolved (+)- and (-)-8, obtained by de-esterifying (+)-
and (-)-10, appeared to racemize appreciably at room
temperature, so they and the reaction mixtures in which
they were used were maintained at 0 °C. Accordingly, to
obtain stereochemically pure 11, (+)- and (-)-10 were
combined with cold KOH in EtOH-THF in the absence
of air and quenched with acetic acid. Freshly prepared
The [5]HELOL ligand, unlike BINOL, has two inde-
pendently chiral domains. It therefore has both dl and
meso forms. Indeed, when racemic 8 is dimerized with
silver(I) oxide, a mixture of the two is obtained. This
means that to obtain one enantiomer of the dl-dimer, the
precursor 8 must be resolved before it is dimerized. The
resolution was easy to accomplish on a 10 mmol scale
silver(I) oxide24 and triethylamine in CH
mediately added to the resulting (+)- and (-)-8. The
2
2
Cl were im-
2
5
intensely blue-green quinoidal (P,P)-11 and (M,M)-11
were formed in 91% and 89% yields. Because the enan-
tiomers of 11 absorb light strongly, their specific rotations
at the sodium-D wavelength could not be measured.
Therefore they are designated according to their absolute
(
(
Scheme 2). Thus, 8 was combined for 1 h at 80 °C with
S)-(-)-camphanoyl chloride,6b triethylamine, and a cata-
lytic amount of DMAP, and after acidic and basic workup
and trituration with diethyl ether, the [5]helicene cam-
phanate (+)-10 was obtained in 80% yield (9.6 g). Its
1
configurations, analyzed below. H NMR analysis showed
that the samples of 11 each contained less than 2% of
the meso diastereomer.26 The implication is that the
helicene configuration inverted immeasurably during the
de-esterification and dimerization steps.
The [5]HELOL ligands, (+)- and (-)-3, obtained in high
purity and quantitative yields when (P,P)- and (M,M)-
1
diastereomeric excess was >98% according to H NMR
analysis.21 The mother liquor contained (-)-10 of >90%
diastereomeric purity, which was purified to >98% de,
2 2
also in 80% yield, by eluting it with CH Cl from a short
plug of silica gel. The small amount of (+)-10 it contained
remained on the silica. That the (+)-diastereomer is more
polar than the (-)-diastereomer is implied by the former
1
1 were reduced by zinc dust and acetic acid in acetone,
were analyzed to be >98% enantiopure. This analysis was
f
exhibiting the lower R on silica gel and the lower
3
1
carried out by measuring the intensities of the P NMRs
solubility in ether. These observations accord with both
analogies and conformational analyses.22 The camphan-
ate esters (+)- and (-)-10, which are 1-substituted
of the diastereomers formed when (+)- and (-)-3 were
combined with PCl and then with enantiopure (S)-(-)-
3
1
-phenylethylamine.27 That kinetic effects did not bias
[
5]helicenes,23 showed no signs of racemizing during six
months at room temperature.
(24) (a) Tanabe, M.; Peters, R. H. Organic Syntheses; Wiley: New
York, 1990; Collect. Vol. VII, p 388. (b) Goltner, C.; Laatsch, H. Liebigs
Ann. Chem. 1991, 1085.
(25) The double bond between the helicenes is probably trans as in
other 2,2′-diphenoquinones, a preference attributed to repulsion of C-O
dipoles. Hewgil, F. R.; La Greca, B.; Legge, F.; Roga, P. E. J . Chem.
Soc., Perkin Trans. 1 1983, 131.
(
20) Evidence for the assignment of structure 9 includes the follow-
ing: (a) HRMS determined the mass to be 322 (corresponding to the
loss of water from 8); (b) H NMR analysis showed 1 methoxyl (a singlet
at 4.26 ppm) and 1 phenolic hydroxyl (a singlet at 9.37 ppm that is
removed when
1
2
D O is added); and (c) irradiating the hydroxyl
resonance gave rise to NOEs for the singlet at 7.78 ppm (proton c)
and the doublet at 7.57 ppm (proton b), while irradiating the methoxyl’s
resonance gave an NOE only for proton c.
(26) The intensities were analyzed of a doublet at 9.31 ppm in the
H NMR spectrum of (P,P)-11 and a doublet at 9.1 ppm expected in
the spectrum of meso-11. While we did not isolate meso-11, we did
1
(
21) The diastereoselectivity was determined by comparing the
isolate the same structure with butyls in place of methyls. It exhibited
1
1
intensities of the singlet at 0.60 ppm in the H NMR spectrum of (P)-
+)-10 and the singlet at 0.65 ppm in the spectrum of (M)-(-)-10. Had
% of the one diastereomer been present in the other, the analysis
would have detected it.
22) Thongpanchang, T.; Paruch, K.; Katz, T. J .; Rheingold, A. L.;
Lam, K.-C.; Liable-Sands, L. J . Org. Chem., in press.
23) The introduction of a methyl group at the 1-position of
5]helicene raises the racemization barrier to that of [6]helicene. The
a doublet H NMR at 9.1 ppm. This compound was isomeric with the
(
2
analogue of (P,P)-11 that had butyls in place of methyls. It showed no
1
circular dichroisms. The H NMR spectra of these compounds are
displayed in the Supporting Information.
(
(27) The analysis is related to others (a) in which chiral 1,2-diols
were distinguished by menthyldichlorophosphate (Garner, C. M.;
McWhorter, C.; Goerke, A. R. Tetrahedron Lett. 1997, 38, 7717), (b) in
which the phosphonamide formed from its chlorophosphate and (S)-
1-phenylethylamine was used to determine VAPOLs absolute stereo-
chemistry (ref 14c), and (c) [5]HELOLs chlorophosphite was used to
distinguish chiral alcohols, amines, and carboxylic acids (Weix, D.;
Dreher, S. D.; Katz, T. J . Unpublished results).
(
[
q
-1
figures are ∆G ) 24.1 kcal mol for [5]helicene, 38.7 for 1-methyl-
[
5]helicene, and 37.0 for [6]helicene. (a) Goedicke, C.; Stegemeyer, H.
Tetrahedron Lett. 1970, 937. (b) Scher u¨ bl, H.; Fritzche, U.; Mannschreck,
A. Chem. Ber. 1984, 117, 336.

8
18 J . Org. Chem., Vol. 65, No. 3, 2000
Dreher et al.
this analysis was proven by analyzing the phosphon-
to 1.0 M. The ee was 72%. Accompanying this product
3
6
amides that formed when (()-1-phenylethylamine was
was a 25% yield of 2-chlorobenzyl alcohol. For p-
methoxybenzaldehyde to give a good yield (91%), its
concentration had to be doubled to 1.0 M and the time
allowed for it to react had to be doubled as well (to 48 h).
The ee was only 46%.
3
added to (+)-[5]HELOL and PCl . The intensities of the
two 31P resonances in this case were in the ratio 1:1.
The absolute configurations of the [5]HELOL ligands
were determined by analyzing the crystal structure of
(
-)-10. This camphanate was found to have the (M) or
The stereochemistries of the predominant 1-arylpro-
panols produced by (P,P)-(+)-[5]HELOL and (R)-(+)-
BINOL are opposite, (S) by the former, (R) by the latter,
which is reasonable bacause, as discussed in the Intro-
duction and as seen in structure 12, the biaryl link in
the (P,P)-([5]HELOL must have the (S)-stereochemistry,
the opposite of that in (R)-BINOL. Moreover, the favored
way in which benzaldehyde should coordinate to the zinc
is the one pictured in structure 12, not the one in which
the phenyl and hydrogen are interchanged, for the latter
would cause the phenyl to bump against the upper
helicene. Since a zinc atom that is coordinated to one of
left-handed configuration. Since it gave 11 and (-)-3, the
ring systems in these compounds also have the (M)-
configurations.
Asym m etr ic Ca ta lysis. To investigate the catalytic
activity of [5]HELOL, the additions of diethylzinc to aryl
aldehydes15 were studied because in these reactions
BINOL itself and other simple chiral diols (those without
additional heteroatoms at the active site ) are usually
2
8
8
,30
not very effective. In fact, only one simple diol has been
found that is.8
a,31
The optimized experimental conditions used to study
5]HELOL in the diethylzinc additions were to combine
.70 mmol of benzaldehyde, 1.1 mmol of diethylzinc, and
.035 mmol of [5]HELOL in 1.4 mL of toluene at room
the oxygens is believed to transfer the ethyl group to the
[
0
0
aldehyde carbon,3
7,38
and only the oxygen at the top of
structure 12 is positioned to allow such a transfer, the
ethyl must add to the aldehyde’s Si-face. This agrees with
the observations made above as well as those made
previously, that under catalysis by derivatives of (R)-(+)-
temperature for 24 h. 1-Phenylpropanol was formed in
9
the solvents were methylene chloride or THF, or when
the temperature was lower, the reactions were very slow.
When additional diethylzinc was added, or when the
reaction mixtures were more concentrated, the enantio-
selectivities decreased. If BuLi, Et N, or pyridine was
3
added to the reaction mixtures, the products were es-
sentially racemic. As a basis for comparison, BINOL was
3% yield32 and 81% enantiomeric excess (ee). When
33
BINOL, diethylzinc adds to benzaldehydes to give (R)-
3,29
1
-arylpropanols.1
However, [5]HELOL amplifies this
stereospecificity, and the greater hindrance around the
zinc atoms, evident in 12, would account for why. The
additional crowding might also prevent zinc-oxygen
aggregates from forming and therefore account for why
[5]HELOL is more reactive than BINOL. The latter forms
substituted for [5]HELOL. The yield of 1-phenylpropanol
then dropped to 56% and the ee to 34%.34
such aggregates, and their inertness has explained why
it reacts so sluggishly.8
b-d,13a,15b,37
Several other aromatic aldehydes were substituted for
benzaldehyde. p-Chlorobenzaldehyde gave a 90% yield
and a 69% ee. o-Chlorobenzaldehyde and p-methoxy-
benzaldehyde reacted only slowly,35 the former perhaps
because it is sterically hindered. Thus, for o-chlorobenz-
aldehyde to give the expected 1-(2-chlorophenyl)propanol
in good yield (75%), its concentration had to be doubled
(
28) BINOLs that have, besides the ligating diols, other heteroatoms
at the active sites do catalyze the reactions of diethylzinc and aldehydes
13
1
5b
and, like chiral amino alcohols, give excellent enantioselectivities.
The addition of chiral amines to BINOL derivatives also achieves good
8
c
asymmetric catalysis as does replacing one of BINOL’s phenol
Potential problems that might restrict the usefulness
of the [5]HELOL ligand are oxidizability and stereo-
mutability. However, the bright yellow [5]HELOL (3) was
found to be only moderately sensitive to oxidation. When
exposed to air for long periods, especially in solution, it
oxidized to blue-green 11, a transformation that is
probably promoted by the electron-rich alkoxy side chains
and, when compared to 1 and 2, the low barrier to
rotation about the 2,2′-bond. However, this oxidation did
not affect [5]HELOL’s ability to function as a catalyst.
The ligand was generally stored under nitrogen at 4 °C
and was used in air-free solutions. The color of the
reaction mixtures remained yellow throughout.
2
9
functions with an amino group.
29) Vysko cˇ il, S.; J aracz, S.; Smr cˇ ina, M.; Sˇ t ´ı cha, M.; Hanus, V.;
Pol a´ sˇ ek, M.; Ko cˇ ovsky, P. J . Org. Chem. 1998, 63, 7727.
30) In the presence of excess titanium tetraisopropoxide, a titanate
of BINOL, like titanates of other chiral diols, is effective [Zhang, F.-
Y.; Yip, C.-W.; Cao, R.; Chan, A. S. C. Tetrahedron: Asymmetry 1997,
, 585].
31) Rosini, C.; Franzini, L.; Pini, D.; Salvadori, P. Tetrahederon:
Asymmetry 1990, 1, 587.
32) The yields in this reaction were determined by analyzing the
H NMR spectra in CDCl , with 3 serving as an internal standard.
The intensities of a multiplet at 4.55 ppm (1 H) in the spectrum of
-phenylpropanol, a singlet at 9.98 ppm (1 H) in the spectrum of
benzaldehyde, and a singlet at 4.64 ppm (2 H) in the spectrum of benzyl
alcohol were compared with the intensity of a doublet at 8.4 ppm (2
H) in the spectrum of 3. The other reactions were analyzed similarly.
(
(
8
(
(
1
3
1
1
(
33) Determined by analyzing the H NMRs of R-methoxy-R-tri-
fluoromethylphenylacetic acid esters. (a) Dale, J . A.; Mosher, H. S. J .
Am. Chem. Soc. 1973, 95, 512. (b) Parker, D. Chem. Rev. 1991, 91,
441.
34) BINOL at room temperature with diethylzinc and benzaldehyde
The low barrier to [5]helicenes racemizing39 also turned
out not to be as great a problem as at first seemed
possible. To study the stereomutation of nonracemic 11,
1
(
in toluene was reported in ref 8a to give no appreciable reaction and
in ref 8b to transform 37% of the benzaldehyde and give 1-phenylpro-
panol with an ee of 32.9%. To provide a uniform comparison, the
reaction parameters that had been optimized for [5]HELOL were
applied to both it and to BINOL.
(36) Benzyl alcohols are common side products of such reactions.
Examples are in ref 8a,b.
(37) Yamakawa, M.; Noyori, R.J . Am. Chem. Soc. 1995, 117, 6327.
(38) In structure 12, a group attached to the Zn that delivers the
ethyl has been omitted to make the drawing clearer.
(39) Laarhoven, W. H.; Prinsen, W. J . C Top. Curr. Chem. 1984,
125, 63.
(35) The p-methoxy group lowers reactivity in similar reactions.
See: ref 31 and Corey, E. J .; Yuen, P.-W.; Hannon, F. J .; Wierda, D.
A. J . Org. Chem. 1990, 55, 784.

Synthesis of a Bis[5]Helicene Chiral Pocket
J . Org. Chem., Vol. 65, No. 3, 2000 819
it was necessary to monitor the isomerizations until they
had taken place only to a small extent, <15%, because
one enantiomer almost surely converts into the meso
isomer before it converts into the other. Thus, (P,P)-11
was heated at three temperatures, and the conversion
Exp er im en ta l Section
THF was distilled from Na/benzophenone, toluene from Na,
and CH
Aldrich), triisopropylsilyl triflate (GFS, 98%), sodium dithionite
Aldrich, ca. 85%), (dimethylamino)pyridine (Aldrich, 99%),
2 2 2 3 2
Cl , (CHCl) , and Et N from CaH . 1-Iodobutane (99%
(
1
to meso-11 was analyzed by H NMR spectroscopy. For
diethylzinc (Aldrich), benzaldehyde (Aldrich, 99+%), o-chlo-
robenzaldehyde (Aldrich, 99%), p-chlorobenzaldehyde (Aldrich,
97%), and p-methoxybenzaldehyde (Aldrich, 98%) were used
without further purification. p-Benzoquinone (Aldrich, 98%)
was purified by slurrying it in CH Cl with four times its
2 2
weight of basic alumina, filtering through Celite, and drying
under vacuum. 3-Acetylphenanthrene was prepared on a 100
q
-1
the first-order rate constants, ∆H ) 106 ( 3 kJ mol
q
-1
-1
and ∆S ) -10.0 ( 0.1 J mol
K
. The half-life at 25
°
C should therefore be 19 days and at 4 °C 1.4 years. As
a solid at ca. 4 °C, (P,P)-11 seems at least this stable. It
did not convert to detectable amounts of the meso form
during 6 months. The figures show that the barrier to
stereomutation is higher for (P,P)-11 than for [5]helicene
15
44
g scale and (1S)-(-)-camphanic acid on a 100 g scale. (S)-
+)-R-Methoxy-R-(trifluoromethyl)phenylacetyl chloride was
prepared by refluxing (R)-(+)-R-methoxy-R-(trifluoromethyl)-
(
q
-1
q
-1
-1 40
(
∆H ) 95.8 kJ mol and ∆S ) -17.2 J mol
K
),
q
-1
q
45
smaller than for [6]helicene (∆H ) 148 kJ mol and ∆S
phenylacetic acid with SOCl
2
for 1 h. Zn dust (Aldrich, <10
-
1
-1 41
µm, 98%) was activated prior to use.46 Glassware was flame-
)
-7.9 J mol
K ), but similar to that for a [5]heli-
q
-1
q
dried under vacuum and cooled under N . Reactions were
carried out under N . Additions by syringe were through
2
cenebisquinone (∆H ) 116.2 ( 0.4 kJ mol and ∆S )
-1
-1 42
2
-
16.4 ( 4 J mol
K ). Even though the latter has one
rubber septa. Whatman 60 A silica plates were used for TLC
analyses.
more blocking carbonyl than the [5]helicenequinone
moiety of 11, when two [5]helicenes are united, the ring
system of each half of the molecule should interfere with
the stereomutation of the ring system of the other half.
This is indicated in structure 13.
The thermal racemization of [5]HELOL (3) was studied
by comparing the CD spectra of samples of the solid
dextrorotatory enantiomer that had and had not been
heated at 45 °C for 72 h.43 The intensities of the CDs did
not decrease. The CDs of a degassed acetonitrile solution
kept at room temperature for 6 days also did not
decrease. Thus, 3 should not racemize significantly when
it is stored for prolonged periods as a solid at 4 °C under
3
-[1-(Tr iisop r op ylsiloxy)eth en yl]p h en a n th r en e (5a ).
Triisopropylsilyl triflate (37.9 mL, 141 mmol, 43.1 g) was added
to a solution, cooled in an ice bath, of an 11:1 mixture of 3-
and 2-acetylphenanthrenes (28.2 g, 128 mmol)15 and Et
N (53.4
(380 mL), and after the mixture had been
warmed to room temperature it was stirred for 2 h. The
reaction mixture was then diluted with CH Cl , washed three
times with saturated aqueous NaHCO , and dried (Na SO ).
3
2 2
mL) in CH Cl
2
2
3
2
4
The solvent was evaporated. The oily product was purified by
dissolving it in 1:1 hexanes/benzene (+ 3% Et N) and quickly
3
flushing it with more of the same solvents down a short column
(3 in × 3 in) of neutral alumina. This gave 47.6 g (99%) of a
1
light yellow oil, a mixture of 5a and 5b. The H NMR spectrum
of the mixture is in the Supporting Information.
2
N (to minimize oxidation to 11) or when it is kept in
Helicen equ in on e 7. A flask containing a mixture of 5a
and 5b (47.6 g, 128 mmol) and p-benzoquinone (111 g, 1.0 mol)
was fitted with a reflux condenser and then evacuated and
solution under the conditions usually used for catalyzed
reactions. Further evidence of its stability was the
observation that in a diethylzinc addition to benzalde-
hyde catalyzed by [5]HELOL at 25 °C, the enantiopuri-
ties of the 1-phenylpropanol found in aliquots removed
after 1 h and after 24 h were identical within 1%.
Moreover, the stability to stereomutation should be
enhanced when, during a catalytic process, the ring
system is attached to a metal ion, for the complexation
should make 3’s helicene moieties much more rigid.
filled with N
and the mixture under N
reaction mixture was poured from the flask, and CH
2
three times. Toluene (510 mL) was syringed in,
was refluxed for 14 h. The hot
Cl was
2
2
2
added. The mixture was filtered through Celite, and the
solvent was evaporated. After it had been shaken vigorously
with hexanes (500 mL) and filtered, the insoluble portion,
which was mostly benzoquinone, was washed with hexanes
500 mL). Solvent was stripped from the filtrate. Heating at
20 °C under a vacuum of ca. 1 mmHg for ca. 1 h sublimed
the remaining benzoquinone into a cold trap. H NMR analysis
in the Supporting Information) showed the residual dark red-
brown glassy solid to be a 11:1 mixture of 6a and 6b. Since
we could not separate these, the mixture was used for the next
reaction.
(
1
1
(
Con clu sion
Both enantiomers of the [5]HELOL (3) can be prepared
in nonracemic form on a large scale. They are stable to
racemization and decomposition under the experimental
conditions typically used to bring about catalyzed trans-
formations, and they catalyze the additions of diethylzinc
to aldehydes. [5]HELOL is only the second simple al-
kanediol (one that does not have an additional hetero-
atom at the active site) to be catalytically active in this
reaction, and the enantioselectivities it achieves (up to
It was dissolved in DMF (750 mL), and after a flask
containing the solution had been evacuated and filled with N
2
three times, 1-iodobutane (51.0 mL, 128 mmol) was added,
followed by K CO (216 g, 1.56 mol). The mixture was heated
2
3
at 60 °C for 2 h. It turned green and then slowly orange-brown.
The precipitates, which formed when the mixture was cooled
to 25 °C and poured into water (2 L), were filtered through
Celite and washed with water. An orange-brown solid was
2 2
washed from the Celite with CH Cl , and the solvent was
8
1%) are the highest yet obtained in any processes
stripped. The residue was heated and stirred with EtOH (400
mL), and the solids were filtered and dried, giving 36.0 g of 7,
a brick-red solid (an 81% yield if the starting material was
brought about by a helicene-containing catalyst.
9
1% 4a ). A sample recrystallized from EtOAc was character-
(
(
(
(
40) Goedicke, C.; Stegemeyer, H. Tetrahedron Lett. 1970, 937.
41) See footnote 56 in ref 6b.
42) Liu, L.; Katz, T. J . Tetrahedron Lett. 1990, 31, 3983.
43) A more precise analysis of the stereomutation of 3 was difficult
-
1
ized. Mp: 187-187.5 °C. IR (CCl
4
): 2962, 1666, 1298 cm
.
1
3
H NMR (CDCl , 300 MHz): δ 8.44 (d, 1 H, 7.8 Hz), 8.38 (d, 1
because 3 oxidizes to 11, which absorbs strongly at 589 nm, making it
hard to determine specific rotations. An H NMR analysis similar to
that carried out for (P,P)-11 was impossible because no cleanly
resolvable peaks could be found in the spectra of (P,P)- and meso-3.
Their resonances were too broad. CD spectroscopy proved too imprecise
to measure small changes when less than 10-20% stereomutation had
taken place.
1
(44) Gerlach, H.; Kappes, D.; Boeckmann, R. K.; Maw, G. N. Organic
Syntheses; Wiley: New York, 1990; Vol. 71, p 48.
(45) Dale, J . A.; Dull, D. L.; Moser, H. S. J . Org. Chem. 1969, 34,
2543.
(46) Shriner, R. L.; Neumann, F. Organic Syntheses; Wiley: New
York, 1990; Collect. Vol. III, p 73.

8
20 J . Org. Chem., Vol. 65, No. 3, 2000
Dreher et al.
H, 8.7 Hz), 7.94 (m, 3 H), 7.84 (d, 1 H, 8.6 Hz), 7.49 (m, 2 H,
(-3), 356 (sh, 15), 336 (sh, 44), 321 (87), 304 (sh, 32), 280 (-19),
4
2
.3 Hz), 6.95 (d, 1 H, 10.0 Hz), 6.86 (d, 1 H, 10.3 Hz), 4.39 (m,
H), 2.01 (m, 2 H), 1.65 (m, 2 H), 1.07 ppm (t, 3 H, 7.4 Hz).
30 6
272 (-19). Anal. Calcd for C34H O : C, 76.39; H, 5.65.
Found: C, 76.18; H, 5.56.
1
3
C NMR (CDCl
3
, 75 MHz): δ 185.7, 185.3, 158.8, 140.0, 135.9,
The ether solutions were evaporated, leaving the diastere-
1
1
32.9, 132.4, 132.2, 129.7, 129.2, 128.5, 128.2, 128.0, 126.1,
omer with higher R
f
. Its de was >90%.21 It was chromato-
25.9, 125.7, 120.0, 101.2, 69.0, 31.1, 19.4, 13.9 ppm. UV-vis
graphed under aspirator vacuum in two equal batches on silica
gel contained in a 600 mL coarse fritted funnel that was
wrapped with aluminum foil. The product was eluted with
-
5
(CH
3
CN, c ) 7.5 × 10 M): λmax (log ꢀ) 427 (3.71), 385 (3.58),
3
55 (sh, 3.00), 315 (sh, 4.18), 287 nm (4.21). Anal. Calcd for
, C 82.08; H, 5.30. Found: C, 81.79; H, 5.14.
Helicen e 8. Sodium dithionite (172 g, 0.99 mol), EtOAc (1
L), and water (1.5 L) were added to a solution of 7 (30 g, 79.0
mmol) in CH Cl (500 mL) contained in a 4 L separatory
C
26
H
20
O
3
2 2 2 2
solvents ranging from 1:1 hexanes-CH Cl to CH Cl . A small
amount of the slow-moving diastereomer remained on the
column. Isolated was 9.6 g (80%) of (M)-(-)-10 and its de was
1
21
2
2
D
found by H NMR analysis to be >98%. Mp: >250 °C. [R] :
-
funnel. The mixture was shaken for 0.5 h. The organic
materials were separated and washed with water, and the
-1110 (c 0.010, CH
H NMR (CDCl
2
Cl
2 4
). IR (CCl
): 2967, 1796, 1750 cm 1.
, 400 MHz): δ 8.40 (d, 1 H, 8.4 Hz), 8.12 (d, 1
1
3
yellow-brown solution was dried (Na
2
SO
4
). The solvent was
H, 8.4 Hz), 7.92 (m, 3 H), 7.82 (d, 1 H, 8.4 Hz), 7.66 (s, 1 H),
7.52 (t, 1 H, 7.9 Hz), 7.25 (t, 1 H), 6.96 (d, 1 H, 8.4 Hz), 6.91
(d, 1 H, 8.5 Hz), 4.18 (s, 3 H), 4.13 (s, 3 H), 1.52 (m, 1 H), 1.35
(m, 1 H), 1.25 (m, 1 H), 0.92 (s, 3 H), 0.84 (m, 1 H), 0.65 (s, 3
stripped. Methanol saturated with HCl (1 L) was added in the
dark to the solid residue dissolved in 1,2-dichloroethane (250
mL), and the reaction mixture, contained in a flask equipped
with a reflux condenser and wrapped in aluminum foil, was
heated at 60 °C for 30 min. The reaction mixture, cooled to 25
H), 0.50 ppm (s, 3 H). 13C NMR (CDCl
, 75 MHz): δ 177.8,
3
165.4, 153.9, 152.5, 141.0, 132.0, 131.8, 131.3, 128.1, 127.9,
127.7, 127.4, 126.3, 125.6, 125.5, 125.2, 123.5, 121.2, 120.0,
116.0, 105.1, 96.9, 89.8, 55.9, 54.2, 54.0, 29.0, 28.5, 16.3, 16.1,
°
C, was poured slowly into saturated aqueous NaHCO
organic materials were extracted into CH Cl (2 L), which was
then dried (Na SO ) and evaporated. When 500 mL of the
solvent remained, the suspension was filtered, giving 5.0 g (a
0% yield) of 9. To remove small amounts of impurities, the
3
, and
2
2
-
5
2
4
9.5 ppm. UV-vis (CH
3
CN, c ) 4.2 × 10 M): λmax (log ꢀ) 406
(3.22), 384 (3.27), 358 (3.63), 338(3.96), 300 (4.14); 280 nm
-
5
2
(4.15). CD (c ) 4.2 × 10 M, CH
3
CN), nm (∆ꢀ): 409 (5), 396
filtrate was then quickly chromatographed under reduced
pressure on silica gel contained in a 600 mL coarse fritted
(2), 388, (3), 356 (sh, -15), 336 (sh, -44), 321 (-90), 304 (sh,
-31), 280 (19), 272 (18).
funnel that was wrapped in aluminum foil. The eluent was
Cr ysta llogr a p h ic An a lysis of (M)-(-)-10. X-ray quality
crystals, light yellow prisms, were grown by dissolving (M)-
1
CH
2
Cl
2
. The yield of 8, a tan foam, was 20.2 g (72%). H NMR
analysis showed it to contain <5% of 9. A sample purified by
silica gel chromatography, eluting with CH Cl2, was character-
): 3566, 2937, 1606 cm . H NMR (CDCl , 300
MHz): δ 8.48 (d, 1 H, 8.4 Hz), 8.23 (d, 1 H, 8.4 Hz), 7.93 (m,
(-)-10 in CH
cooling the solution to 4 °C for several days. The crystal data
were as follows: M ) 534.58, orthorhombic P2 , a )
8.5136(2) Å, b ) 14.0322(2) Å, c ) 22.3887(4) Å, V )
2 2
Cl , adding this to hexanes at 25 °C, and then
2
-
1 1
ized. IR (CCl
4
3
1 1 1
2 2
3
-3
4
6
3
1
1
1
H), 7.68 (s, 1 H), 7.54 (t, 1 H, 7.2 Hz), 7.28 (t, 1 H, 7.3 Hz),
.96 (d, 1 H, 8.4 Hz), 6.69 (t, 1 H, 8.4 Hz), 4.60 (s, 1H), 4.18 (s,
2674.64(11) Å , Z ) 4, T ) 173(2) K, Dcalc ) 1.328 g cm , µ )
1 2
-
0.90 cm , R(F) ) 5.24%, R (F ) ) 14.90, with a GOF of 1.259
w
1
3
H), 4.08 ppm (s, 3 H). C NMR (CDCl
3
, 75 MHz): δ 153.2,
for 5176 observed independent reflections (4° e 2θ e 52°).
Intensity data were collected using a standard P4 X-ray
diffractometer equipped with a SMART CCD area detector and
a graphite monochromator (λ ) 0.710 73 Å).
48.4, 146.5, 131.6, 131.5, 130.3, 127.9, 127.7, 127.3, 127.1,
26.8, 126.4, 125.6, 125.3, 125.1, 123.5, 120.8, 117.8, 109.6,
07.3, 97.5, 56.1, 55.9 ppm. HRMS (FAB): m/z calcd for
C
24
H
18
O
3
354.1256, found 354.1268.
Properties of 9, an off-white solid, were as follows. H NMR
acetone-d , 300 MHz): δ 9.37 (s, 1 H), 8.95 (d, 1 H, 7.3 Hz),
.82 (d, 1 H, 8.5 Hz), 8.72 (d, 1 H, 8.5 Hz), 8.41 (d, 1 H, 8.5
The systematic absences in the diffraction data were
uniquely consistent with the orthorhombic space group P2 2 2 .
1 1 1
1
(
8
6
The assumption that this is the space group gave results on
refinement that were chemically reasonable and computation-
ally stable. The structure was solved using direct methods,
completed by subsequent difference Fourier syntheses, and
refined by full-matrix least-squares procedures. Empirical
SADABS absorption corrections were applied to the data set.
All non-hydrogen atoms were refined with anisotropic dis-
placement coefficients. Hydrogen atoms were treated as ideal-
ized contributions. The absolute configuration could be as-
signed because that of the camphanate moiety was known. The
software and sources of scattering factors were those in the
SHELXTL (5.10) program library (G. Sheldrick, Siemens XRD,
Madison, WI).
Helicen e Dim er 11. A flask containing (M)-(-)-10 (8.0 g,
2
15.0 mmol) was evacuated and filled with N three times. THF
(90 mL) was added, and the solution was cooled to 0 °C. A
solution of KOH (82 g, 55 mmol) in absolute ethanol (230 mL),
which previously had been degassed by boiling under N , was
2
added by cannula during 15 min, and the reaction mixture
was stirred for 2 h at 0 °C. Acetic acid (115 mL), which had
previously been degassed by boiling under N , was added
2
sufficiently slowly by cannula so the temperature remained
at 0 °C. After it had stirred at 0 °C for 10 min, the mixture
was poured into a 2 L separatory funnel containing water (1
L) and ice (500 g). The organic materials were extracted with
Hz), 8.14 (dd, 2 H, 8.9, 6.6 Hz), 7.98 (t, 1 H, 7.8 Hz), 7.78 (s, 1
1
3
H), 7.57 (d, 1 H, 8.5 Hz), 4.26 ppm (s, 3 H). C NMR (acetone-
d
1
1
6
, 75 MHz): δ 154.3, 152.7, 133.3, 131.8, 130.2, 128.5, 128.1,
27.4, 126.3, 125.9, 125.8, 125.7, 124.8, 124.6, 123.9, 123.4,
22.2, 120.5, 120.4, 120.1, 113.3, 97.7, 56.0 ppm. HRMS
(
FAB): m/z calcd for C23
Mon oca m p h a n a tes (P )-(+)- a n d (M)-(-)-10. 1,2-Dichlo-
roethane (1 L) and Et N (125 mL) were added to a flask
containing 8 (16.0 g, 45.2 mmol), (1S)-(-)-camphanoyl chloride
14.6 g, 67.5 mmol), and 4-(dimethylamino)pyridine (DMAP,
.7 g, 22.1 mmol), and the mixture was heated in an oil bath
at 60 °C for 1 h, then cooled to room temperature and diluted
with CH Cl . After the solution had been washed with aqueous
HCl (1 M), water, twice with saturated aqueous NaHCO , and
again with water, it was dried (Na SO ), and the solvent was
evaporated, giving an off-white solid. This was triturated four
14 2
H O 322.0994, found 322.1004.
3
(
2
2
2
3
2
4
1
times with Et
P)-(+)-10 (9.6 g, 80% yield) showed the de to be >98%.21 This
diastereomer exhibited the lower R when analyzed by TLC
: +970 (c 0.0079, CH Cl ). IR
): 1782, 1760 cm . H NMR (CDCl , 400 MHz): δ 8.38
d, 1 H, 8.5 Hz), 8.17 (d, 1 H, 8.5 Hz), 7.90 (m, 4 H), 7.65 (s, 1
2
O (800 mL each). H NMR analysis of the solid
(
f
on silica. Mp: >250 ° C. [R]
D
2
2
-
1 1
(
(
CCl
4
3
H), 7.47 (t, 1 H, 7.0 Hz), 7.24 (t, 1 H, 7.0 Hz), 6.96 (d, 1 H, 8.4
Hz), 6.78 (d, 1 H, 8.4 Hz), 4.17 (s, 3 H), 4.13 (s, 3 H), 1.54 (m,
CH
Drying (Na
CH Cl (50 mL) was immediately added, the solution was
cooled to 0 °C, and freshly prepared silver(I) oxide (5.3 g, 22.5
2
Cl
2
and then washed three more times with ice water.
1
0
1
1
H), 1.37 (m, 1 H), 1.06 (m, 1 H), 0.94 (s, 4 H), 0.60 (s, 3 H),
2
SO ), and evaporation gave a yellow-brown solid.
4
.47 ppm (s, 3 H). 13C NMR (CDCl
, 75 MHz): δ 177.5, 165.8,
3
2
2
54.1. 152.8, 141.0, 132.1, 132.0, 131.6, 128.2, 127.9, 127.8,
27.6, 126.2, 126.1, 125.9, 125.3, 123.3, 121.5, 120.2, 116.6,
05.1, 97.1, 90.6, 56.0, 54.4, 53.9, 28.7, 16.4, 16.2, 9.5 ppm.
mmol)2 and Et
4a
N (1.3 mL) were added. The color immediately
3
1
turned deep blue-green. Stirring was continued for 2 h at 0
°C. Filtration through Celite, which was then washed with
-
5
UV-vis (CH
3
3
CN, c ) 4.2 × 10 M): λmax (log ꢀ) 406 (3.22),
84 (3.27), 358 (3.63), 338(3.96), 300 (4.14); 280 nm (4.15). CD
CH
2
Cl
2
, evaporation of the solvents, and chromatography on
Cl to CH Cl ), gave 4.7
-
5
(c ) 4.2 × 10 M, CH
3
CN), nm (∆ꢀ): 409 (-5), 396 (-2), 388,
silica gel (eluents: 1:1 hexanes-CH
2
2
2
2

Synthesis of a Bis[5]Helicene Chiral Pocket
J . Org. Chem., Vol. 65, No. 3, 2000 821
g (89%) of (M,M)-11, a blue-green solid. Mp: >250 ° C. IR
Dieth ylzin c Ad d ition s. (P,P)-(+)-3 (0.025 g, 0.035 mmol),
-
1
1
(CCl
4
): 1598 cm . H NMR (CDCl
3
, 300 MHz): δ 9.31 (d, 2
followed by toluene (1.4 mL), was added under N
2
to a small
Zn was added, and the
2
was stirred for 1 h. A yellow
H, 7.6 Hz), 8.28 (d, 2 H, 8.6 Hz), 7.92 (m, 4 H), 7.82 (d, 4 H,
flamed-dried air-free flask. Neat Et
reaction mixture under N
2
8
3
1
1
.8 Hz), 7.51 (m, 4 H), 7.30 (s, 2 H), 7.10 (s, 2 H), 4.18 (s, 6 H),
.52 ppm (s, 6 H). 13C NMR (CDCl
, 75 MHz): δ 158.6, 154.5,
precipitate formed. Benzaldehyde (0.071 mL, 0.070 g, 0.70
mmol) was added, whereupon the precipitate immediately
3
33.4, 132.9, 132.4, 132.0, 128.5, 128.0, 127.7, 127.5, 127.2,
26.3, 124.9, 120.2, 102.6, 98.4, 56.2, 55.7 ppm. UV-vis
dissolved. After the mixture, sealed under N
24 h, it was poured into aqueous NH Cl, and the reaction flask
was washed with CH Cl . The organic solution was dried
(Na SO ), and the solvent was evaporated. The H NMR
spectrum of the entire sample dissolved in CDCl was analyzed
2
, had stirred for
-
5
(CH
3
CN, c ) 5.6 × 10 M): λmax (log ꢀ) 630 (3.92), 600 (sh,
4
3
.87), 470 (3.77), 390 (3.99), 319 (sh, 4.21), 283 nm (4.27). CD
2
2
-
5
1
(
(
(
c ) 5.6 × 10 M, CH
sh, 26), 315 (-117), 337 (-80), 354 (-140), 400 (-15), 430
-23), 630 (21). HRMS (FAB): m/z calcd for C48 704.2199,
3
CN), nm (∆ꢀ): 263 (-13), 283 (57), 296
2
4
3
H
32
O
6
to determine the yield, the (P,P)-(+)-3 acting as an internal
standard. The yield of 1-phenylpropanol47 was 93%, and, as
analyzed by the procedure below, it was predominantly (S)
with 81% ee. Also present were 3% of unreacted benzalde-
hyde47 and 4% of benzyl alcohol. The same reaction was
performed but with (R)-(+)-BINOL in place of the HELOL. The
yield of 1-phenylpropanol was 56%, and it was predominantly
found 704.2200.
(P,P)-11 was prepared similarly. (P)-(+)-10 (7.0 g, 13.1
mmol) in THF (80 mL) was combined with KOH (46.9 g) in
EtOH (200 mL) and, after 2 h at 0 °C, quenched with acetic
acid (100 mL). Oxidation using silver(I) oxide (4.6 g, 19.7
47
mmol) and Et
3 2 2
N (1.1 mL) in CH Cl (44 mL) at 0 °C, followed
(
R) with 34% ee. Also present were 37% of unreacted benz-
aldehyde and 7% of benzyl alcohol.
by chromatography, gave 4.2 g (91%) of (P,P)-11, mp >250 °
1
13
C. Its H NMR, C NMR, and UV-vis spectra were identical
-
5
In similar experiments with [5]HELOL, p-chlorobenzalde-
to those of (M,M)-11. CD (c ) 5.6 × 10 M, CH
63 (11), 283 (-59), 296 (sh, -28), 315 (120), 337 (81), 354
143), 400 (16); 430 (24), 630 (-21).
P ,P )-(+)-[5]HELOL (3). Acetone (20 mL) and then acetic
3
CN), nm (∆ꢀ):
hyde gave (S)-1-(4-chlorophenyl)propanol8a in 90% yield. Its
2
(
4
7
ee was 69%. Also present were 5% of 4-chlorobenzyl alcohol
and 5% of p-chlorobenzaldehyde.47
(
o-Chlorobenzaldehyde under conditions that were identical,
acid (0.80 mL) were added to a flask containing (P,P)-11 (1.0
g, 1.4 mmol) and zinc dust (4.6 g, 70 mmol) at 0 °C. The
mixture turned yellow after several minutes. After it had
stirred for 2 h at 0 °C, the reaction mixture was quickly filtered
except that only 0.7 mL of toluene was used, gave a 75% yield
4
8
of (S)-1-(2-chlorophenyl)propanol (ee 72%) and a 25% yield
4
7
of 2-chlorobenzyl alcohol.
p-Methoxybenzaldehyde49 under these last conditions gave,
after 48 h, (S)-1-(4-methoxyphenyl)propanol in 91% yield. Its
ee was 46%. Also present was a 3% yield of 4-methoxybenzyl
alcohol44 and 6% of residual p-methoxybenzaldehyde.
through Celite, which was then washed with CH
organic solutions were washed three times with water, dried
MgSO ), and filtered, and the solvent was evaporated. When
2 2
Cl . The
(
4
47
most of the solvent had been removed, a small amount of
pentane was added to precipitate a brilliant yellow solid, which
was subjected to a vacuum for 24 h at 25 °C. Obtained was
En a n tioselectivity Deter m in a tion s. The reaction mix-
tures were distilled under reduced pressure (ca. 0.1 mm),
mixed for 12 h with excess (S)-(+)-R-methoxy-R-(trifluoro-
1
N
2
2
2
.0 g (100%) of (P,P)-(+)-3, dec 210 °C, which was stored under
at 4 °C. [R] : +2580 (c 0.010, CH CN). IR (CCl ): 3527,
, 400 MHz): δ 8.34 (br s,
H), 8.28 (d, 2H, 8.4 Hz), 8.00 (m, 8H), 7.55 (s, 2H), 7.49 (br s,
H), 7.23 (s, 2H), 7.04 (s, 2H), 4.21 (s, 6H), 3.92 (s, 6H), 3.33
methyl)phenylacetyl chloride,45 DMAP, Et
N, and CH
3
2 2
Cl , and
2
D
-
3
4
1 1
the enantioselectivities were measured by analyzing the
955, 1607 cm . H NMR (DMSO-d
6
3
3
resonances of the protons on the carbons next to the oxygens.
The chemical shifts for the substituted 1-phenylpropanol
1
3
derivatives were as follows: H, δ 5.89, 5.82 ppm; p-OCH , δ
ppm (s, 2H). C NMR (DMSO-d
6
, 75 MHz): δ 152.6, 146.7,
45.2, 132.9, 130.5, 130.4, 128.5, 127.8, 127.4, 126.9, 126.0,
25.6, 125.0, 124.9, 124.6, 124.2, 120.3, 119.1, 110.6, 96.8, 56.2,
3
5.85, 5.77 ppm; p-Cl, δ 5.85, 5.78 ppm; o-Cl, 6.37, 6.27 ppm.
In each case in which (P,P)-(+)-[5]HELOL was the asymmetric
catalyst, the diastereomer with the resonance at higher field
1
1
5
4
2
-
5
5.9 ppm. UV-vis (CH
67 (3.33), 419 (sh, 3.96), 300 (4.46), 256 (4.41), 243 (4.43),
27 nm (4.31). CD (c ) 4.6 × 10 M, CH
3
CN, c ) 4.6 × 10 M): λmax (log ꢀ)
5
0
predominated. This is the ester of the (S)-1-arylpropanol.
-
5
When (R)-(+)-BINOL was the asymmetric catalyst, the other
diastereomer predominated.
Ra tes of Ra cem iza tion . Samples of (P,P)-11 (ca. 5 mg) in
3
CN), nm (∆ꢀ): 216
(
-122), 229 (-5), 238 (-48), 262 (61), 285 (sh, -91), 296
(-112), 328 (152), 359 (sh, 104). HRMS (FAB): m/z calcd for
C
48
H
34
O
6
706.2355, found 706.2379.
0.75 mL of CDCl
3.0, 53.5, and 59.6 °C, cooled to 25 °C, and their H NMR
3
contained in an NMR tube were heated at
1
4
(
M,M)-(-)-[5]HELOL (3). Under identical conditions (M,M)-
spectra were analyzed quickly. The first-order rate constants
1
-
D
1 (1 g) gave (M,M)-(-)-3, dec 210 °C, also in 100% yield. [R] :
2470 (c 0.016, CH
of its solutions in DMSO-d
CH
of solutions of the two materials in CH
of one another.
-
6
-1
1
13
for the stereomutation were: at 43.0 °C, 5.2 × 10
s ; at 53.5
2
Cl
2
). The H NMR and C NMR spectra
and its UV-vis spectrum in
CN were identical to those of (P,P)-(+)-3. The CD spectra
-
5
-1
-5 -1
°
C, 2.02 × 10
s
; and at 59.6 °C, 4.07 × 10
s
.
6
(P,P)-(+)-3 (1.4 mg) was dissolved in CH
3
CN (50 mL), and
3
the yellow solution was degassed by freeze-pump-thawing. The
CD spectra of aliquots were the same whether they were
removed when the solution was first prepared or after the
3
CN were mirror images
Deter m in ation of th e Ster eoch em ical P u r ities of (P ,P )-
+)- a n d (M,M)-(-)-3. DMAP (a small single crystal) was
solution had been kept, sealed under N
2
, for 6 days at ambient
(
temperature (ca. 25 °C). In a similar experiment in which solid
P,P)-(+)-3 was heated at 45 °C for 2 days, the CD spectra
before and after the heating were essentially identical.
added to (P,P)-(+)-3 (25 mg, 0.035 mmol) in a small vial, which
(
was then evacuated and purged with N
of a 0.10 M solution in CH Cl , 0.039 mmol) was added,
followed by Et N. After the mixture had stirred for 15 min at
5 °C, (S)-(-)-1-phenylethylamine (42 mg, 0.035 mmol, 4.5 µL)
was added, and stirring was continued for 2 h. CDCl was
added, and with a capillary containing 85 wt % H PO as an
2 3
. Then PCl (0.39 mL
2
2
3
Ack n ow led gm en t. We acknowledge with special
thanks Dr. J oseph M. Fox, who suggested that 3-acetyl-
phenanthrene be used as a precursor to helicenes. We
2
3
3
4
3
1
external reference, the P NMR spectrum at 121.5 MHz was
determined. A single resonance was observed, at 150.44 ppm.
When the same analysis was performed starting with (M,M)-
(
47) These compounds were identified by their 1H NMR resonances
13 1
(
The Aldrich Library of C and H FT NMR Spectra; Pouchert, C. J .,
(
-)-3, again only a single resonance was observed, but it was
at 149.67 ppm. When the analysis was performed with (P,P)-
+)-3 and (()-1-phenylethylamine, both resonances were seen,
Behnke, J ., Eds.; Aldrich Chemical Co.: Milwaukee, 1993; Vol. 2).
1
(48) A multiplet in the H NMR spectrum at 4.99 ppm, similar to
resonances in the spectra of 1-phenylpropanol (4.52 ppm), 1-(4-
chlorophenyl)propanol (4.40 ppm), and 1-(4-methoxyphenyl)propanol
4.47 ppm) was used to analyze for 1-(2-chlorophenyl)propanol.
(
and their intensities were equal within 1%. The data are
displayed in the Supporting Information. The conclusion is
that the amount of (M,M)-(-)-3 in (P,P)-(+)-3 and the amount
of (P,P)-(+)-3 in (M,M)-(-)-3 must both be <2%.
(
(
(
49) Alonso, E.; Ramon, D.; Yus, M. Tetrahedron 1996, 52, 14341.
50) Bolm, C.; Schlingloff, G.; Harms, K. Chem. Ber. 1992, I125,
1191.

8
22 J . Org. Chem., Vol. 65, No. 3, 2000
Dreher et al.
thank the NSF for grant support (CHE98-02316) and
Daniel J . Weix for help with the experiments.
spectra of (P,P)- and (M,M)-11 and of (P,P)-(+)- and (M,M)-
(-)-3; 31P NMR spectra of adducts of (P,P)-(+)- and (M,M)-
(
-)-3, PCl , and both (S)-(-)- and (()-1-phenylethylamine;
3
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
1
NOEs in the H NMR spectrum of 9; X-ray diffraction data
for (M)-(-)-10. This material is available free of charge via
the Internet at http://pubs.acs.org.
5
a plus 5b, 6a plus 6b, and meso- and (P,P)-11 (but with butyls
1
13
in place of methyls); IR, H NMR, and C NMR spectra of 7,
, 9, (P)-(+)- and (M)-(-)-10, (P,P)-11, and (P,P)-(+)-3; UV-
vis spectra of 7 and (P)-(+)- and (M)-(-)-10; CD and UV-vis
8
J O991498U