Synthesis and aggregation of a conjugated helical molecule

Article abstract of DOI:10.1021/ja983248l

Described are an improved method to synthesize helicenebisquinone 1 and an efficient method to resolve its enantiomers and those of other helicenebisquinones by separating derived camphanate esters. The union of p- benzoquinone and the bis(triisopropylsilylenol ether) of a 2,7-diacetyl-4,5- dialkoxynaphthalene gives the helicene skeleton in 56% yield. Molecules of nonracemic 1 aggregate when their dodecane solutions are concentrated to 5 x 10^-4 M or more. Chloroform dissociates the aggregates. Aggregation is characterized by enhanced circular dichroisms and g values, red-shifted UV- vis absorptions, increased specific rotations, increased light scattering at an absorption frequency, and, as reported previously, fluorescence emissions that shift to the red, ¹H NMR resonances that shift upfield, and organization of the pure material into fibers. The CD and UV-vis absorption spectra of spin-coated films are similar to those of concentrated solutions in dodecane, but the g values are higher. The specific rotation of a cast film, [α](D), is 158 000 (deg cm²)/10 g. Unlike the nonracemic material, the racemic in bulk does not form visible fibers and is too insoluble to give concentrated solutions in dodecane. Also nonracemic helicene 3 does not show the characteristics of aggregation either in solution or in bulk.

Full text of DOI:10.1021/ja983248l

J. Am. Chem. Soc. 1999, 121, 79-88
79
Synthesis and Aggregation of a Conjugated Helical Molecule
Colin Nuckolls,† Thomas J. Katz*,† Gail Katz,† Peter J. Collings,‡ and Louis Castellanos†
Contribution from The Department of Chemistry, Columbia UniVersity, New York, New York 10027, and
The Department of Physics and Astronomy, Swarthmore College, Swarthmore, PennsylVania 19081
ReceiVed September 10, 1998
Abstract: Described are an improved method to synthesize helicenebisquinone 1 and an efficient method to
resolve its enantiomers and those of other helicenebisquinones by separating derived camphanate esters. The
union of p-benzoquinone and the bis(triisopropylsilylenol ether) of a 2,7-diacetyl-4,5-dialkoxynaphthalene gives
the helicene skeleton in 56% yield. Molecules of nonracemic 1 aggregate when their dodecane solutions are
concentrated to 5 × 10-4 M or more. Chloroform dissociates the aggregates. Aggregation is characterized by
enhanced circular dichroisms and g values, red-shifted UV-vis absorptions, increased specific rotations,
increased light scattering at an absorption frequency, and, as reported previously, fluorescence emissions that
1
shift to the red, H NMR resonances that shift upfield, and organization of the pure material into fibers. The
CD and UV-vis absorption spectra of spin-coated films are similar to those of concentrated solutions in
dodecane, but the g values are higher. The specific rotation of a cast film, [R]D, is 158 000 (deg cm2)/10 g.
Unlike the nonracemic material, the racemic in bulk does not form visible fibers and is too insoluble to give
concentrated solutions in dodecane. Also nonracemic helicene 3 does not show the characteristics of aggregation
either in solution or in bulk.
Introduction
spectra and specific rotations resembling those of the Lang-
muir-Blodgett films, but differing from those of solutions that
are more dilute.3
Nonracemic helicene 1 as the pure material,1,2 in Langmuir-
Blodgett films,3 and when dissolved in saturated hydrocarbon
solvents1,3 assembles into helical columns, within which the
molecules appear to be stacked along their helix axes as shown
schematically in structure 2.2 In the pure material, the columnar
Organizing into columns significantly alters 1’s optical
properties. Its second-harmonic-generating ability,4 circular
dichroisms, and optical rotations1,3 increase greatly, and the
UV-vis absorptions and fluorescence emissions shift.1,3 More
light at an absorption wavelength is scattered.5,6 Not all helicenes
surrounded by alkoxy chains behave in this way, and it appears
that in structures similar to those of 1 the quinones are essential
for the stacking to occur. Thus, a derivative of 1 that has only
one of the termini reduced and acylated also aggregates into
helical columns,7 but none in which both quinones are reduced
has been found by polarized light microscopy to form an
anisotropic liquid phase.1,8
Because the optical and structural properties of 1 and its
derivatives7 are unique, it is unfortunate that the previously
reported synthesis of 1 affords only very small amounts of
material (<10 mg).1 Accordingly, we tried to improve the
synthesis, and in this paper, we show how to achieve such an
improved preparation. In addition, the larger amounts available
made it possible to more fully study the CD and UV-vis
absorption spectra, optical rotation, light scattering, and thermal
racemization of nonracemic 1. The results of these studies are
also recorded herein, as are similar studies of nonracemic 3,
which provide a standard to which the properties of 1 can be
assemblies are further ordered into very long, micrometer-wide
lamellar fibers.1,2 Analyses of the fibers by electron and X-ray
diffraction and by electron and polarized light microscopy show
the columns to be ca. 42 Å in diameter and hexagonally packed
collinearly with the fiber axes.2 In Langmuir-Blodgett films,
the organization in parallel columns can be seen directly by
atomic force microscopy4 and inferred from the similarity
between the interlayer spacings of multilayers3 and the diameters
of the columns in the fibers of pure material,2 both analyzed
by X-ray diffraction. In dodecane solutions of sufficient
concentration, the molecules aggregate,1 and that they assemble
into columns is implied by the CD and UV-vis absorption
† Columbia University.
(5) This was first shown by Peter Collings (see ref 1).
‡ Swarthmore College.
(6) Pasternack, R. F.; Collings, P. J. Science 1995, 269, 935.
(7) A derivative of 1 was recently found to be the first core-chiral
columnar liquid crystal: Nuckolls, C.; Katz, T. J. J. Am. Chem. Soc. 1998,
120, 9541.
(8) Besides 3, the following, investigated preliminarily, appeared isotropic
under the polarizing microscope: (a) analogues of 3 with pentyl, heptyl,
nonyl, and heptadecyl side chains instead of dodecyls; (b) an analogue of
3 with methoxyls in place of acetates; (c) analogues of 14 with eight lauryls,
4-heptyloxybenzoyls, 3-cholesteryloxycarbonyls, and I-pyr+(CH2)11CO
groups in place of the acetyls, the last compound in water (Colin Nuckolls,
Thomas J. Katz, and Louis Castellanos, unpublished results).
(1) Nuckolls, C.; Katz, T. J.; Castellanos, L. J. Am. Chem. Soc. 1996,
118, 3767.
(2) Lovinger, A. J.; Nuckolls, C.; Katz, T. J. J. Am. Chem. Soc. 1998,
120, 264.
(3) Nuckolls, C.; Katz, T. J.; Verbiest, T.; Van Elshocht, S.; Kauranen,
M.; Kuball, H.-G.; Kiesewalter, S.; Lovinger, A. J.; Persoons, A. J. Am.
Chem. Soc. 1998, 120, 8656.
(4) Verbiest, T.; Van Elshocht, S.; Kauranen, M.; Hellemans, L.;
Snauwaert, J.; Nuckolls, C.; Katz, T. J.; Persoons, A. Science 1998, 282,
913.
10.1021/ja983248l CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/19/1998

80 J. Am. Chem. Soc., Vol. 121, No. 1, 1999
Nuckolls et al.
Scheme 1a,b
compared. Additionally, the new synthesis proceeds by way of
racemic 1, some of whose properties are compared with those
of the enantiomerically pure material.
Results and Discussion
Helicene Syntheses. Scheme 1 shows how a key step in the
original synthesis (top equation) was improved (bottom equa-
tion).9 The essential difference is that the diene component
originally had the two phenols derivatized as pivalate esters,
whereas in the newer synthesis they are derivatized as dodecyl
ethers. Although a variety of ways were tried to improve the
yield of helicene 5,10 the best that could be obtained was only
16%. Accompanying it as the major product seems to be the
aromatized compound in which only one of the silyl enol ethers
of 4 has combined with the p-benzoquinone and the other has
converted into the methyl ketone.13 Initially there was an
additional difficulty because the silyl ether functions of helicene
5 were too labile to survive chromatography.14 When they were
removed, the phenolic product was rapidly destroyed, presum-
ably by oxidation. While isolating 5 by chromatography was
therefore precluded, the problem could be overcome, because
everything in the reaction mixture except helicene 5 is soluble
in acetone. This means that the desired product could be isolated,
analytically pure, simply by filtering the acetone suspension.
Nevertheless, the reaction of 4 with p-benzoquinone gives 5
in yields that are dismal. However, when 6 was heated with an
excess of p-benzoquinone, the only product the aromatic region
of the 1H NMR spectrum identified in the crude reaction mixture
was the desired helicene.15 Once again, the lability of the silyl
ethers made it impossible to purify the product by chromatog-
a Abbreviations: TBDMS ) t-BuMe2Si, TIPS ) i-Pr3Si, Piv )
Me3CCO. b Reagents and conditions: (a) p-benzoquinone, PhMe; (b)
(i) p-benzoquinone, heptane; (ii) Ac2O, Et3N‚3HF, Et3N.
Scheme 2a,b
(9) The reactions of bis(enol ether)s of arylmethyl ketones with p-
benzoquinone have given a number of helicenebisquinones: (a) Willmore,
N. D.; Liu, L.; Katz, T. J. Angew. Chem., Int. Ed. Engl. 1992, 31, 1093. (b)
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, 119, 10054. (c)
Fox, J. M.; Goldberg, N. R.; Katz, T. J. J. Org. Chem. 1998, 63, 7456.
(10) The following attempts to improve the yield failed: changing the
solvent to octane, benzene, or acetonitrile; running the reaction without
solvent; adding trichloroacetic acid in catalytic11 or stoichiometric amounts;
adding more or adding less benzoquinone; purifying benzoquinone by
recrystallyzing it from acetone;12 adding 2,6-di-tert-butylpyridine; adding
pyridinium dichromate; adding bis(2,6-di-tert-butyl-4-hydroxyphenyl)-
methane; adding Lewis acids (titanium tetraisopropoxide, tert-butyldi-
methylsilyl triflate, or europium tris(6,6,7,7,8,8,8)heptafluoro-2,2-dimethyl-
3,5-octanedionate); adding 4 Å moleular sieves or K2CO3.
a Abbreviations: TBDMS ) t-BuMe2Si, TIPS ) i-Pr3Si, Piv )
t-BuCO. b Reagents and conditions: (a) (i) MeMgBr, THF; (ii)
pyridinium dichromate, CH2Cl2, 73% yield. (b) TBDMSOTf, Et3N,
CH2Cl2, 100% yield. (c) (i) KOH, EtOH, 97% yield; (ii) C12H25I,
K2CO3, DMF, 98% yield. (d) (i) MeLi, THF, 96% yield; (ii) tetraprop-
ylammonium perruthenate, N-methylmorpholine N-oxide, CH2Cl2, 4
Å molecular sieves, 85% yield. (e) TIPSOTf, Et3N, CH2Cl2, 95% yield.
raphy, and since the helicene formed in this reaction, unlike 5,
is soluble in acetone, the silyl ethers were converted into acetate
esters,16 which allowed 7 to be obtained from 6 in a remarkable
56% yield. The important point is that when the substituents
meta to the enol ethers are ethers rather than esters, the yields
of helicenebisquinone are very high. We suppose that the ethers
exert their effect by donating electrons, thereby increasing either
the reactivity of the diene or the rate at which the diene-
benzoquinone adduct oxidizes.
(11) Liu, L.; Katz, T. J. Tetrahedron Lett. 1990, 31, 3983.
(12) Willmore, N. D. Ph.D. Dissertation, Columbia University, 1994.
1
(13) The evidence is the observation in the H NMR spectrum of two
meta-coupled aromatic hydrogens, one at very low fields, and the acetyl
group. The spectrum (400 MHz) displays the former at δ 8.39 and 7.66
ppm (both 1 H, d, J ) 1.6 Hz) and the latter at 2.67 ppm (3 H, s). In
addition, there are the following resonances: two sharp singlets, at 7.76
and 7.57 ppm (both 1 H), expected from the other aromatic hydrogens; an
AB quartet at 7.07 and 6.98 ppm (J ) 10 Hz), attributable to the quinone
hydrogens; a broad singlet at 1.48 ppm (18 H), attributable to the pivalates;
and the two singlets expected from the TBDMS group, at 1.08 and 0.39
ppm (9 and 6 H, respectively).
(14) The adsorbants tried were silica gel, silica gel doped with triethyl-
amine, basic alumina, neutral alumina, and florisil. The lability is presumed
to be associated with conjugation between the siloxyls and the quinone’s
carbonyls. Other helical derivatives of 6-siloxy-1,4-naphthoquinone prepared
in our lab proved less sensitive. They decomposed only slightly when
chromatographed (ref 9b,c and S. Dreher, unpublished results).
(15) The aromatic region of the 1H NMR spectrum of the desired product
is easily identified because it consists of two very sharp singlets and an
AB quartet.
As summarized in Scheme 2, both 4 and 6 were synthesized
from dipivaloxydialdehyde 8,17 itself prepared on a multigram
scale from the known 3,6-dimethyl-1,8-naphthalenediol.18 Thus,
(16) We could not find a previous example in which Ac2O, Et3N‚3HF,
and Et3N had been used to replace a silyl ether by an acetate, but similar
combinations have been used, such as ZnCl2-AcCl (Kim, S.; Lee, W. J.
Synth. Commun. 1986, 16, 659) and FeCl3-Ac2O (Ganem, B.; Small, V.
R., Jr. J. Org. Chem. 1974, 39, 3928).
(17) The three-step synthesis is described in the Supporting Information
to ref 1.
(18) Synthesized in three steps from dehydroacetic acid: (a) Bethell, J.
R.; Maitland, P. J. Chem. Soc. 1962, 3751. (b) Overeem, J. C.; Van der
Kerk, G. J. M. Recl. TraV. Chim. Pays-Bas 1964, 83, 1005.

Synthesis of a Conjugated Helical Molecule
J. Am. Chem. Soc., Vol. 121, No. 1, 1999 81
methylmagnesium bromide adds selectively to 8’s aldehyde
groups, oxidation with pyridinium dichromate then gives
diacetylnaphthalene 9,19 and this, on reaction with tert-butyldi-
methylsilyl triflate and triethylamine, converts into enol ether
4.20 The yield of 4 from 8 was 73%. For the last transformation
in this series and a number of similar ones, we found that silyl
triflates give higher yields and more easily purified products
than other reagents that convert ketones into silyl enol ethers.21
To synthesize 6, the esters of 8 were saponified and the resulting
phenols then alkylated. Addition of methyllithium and oxidation
by Ley’s method, using tetrapropylammonium perruthenate plus
N-methylmorpholine N-oxide,22 gave very clean diacetylnaph-
thalene 11. Once again, it was reaction with the silyl triflate
(triisopropylsilyl triflate in this case) that gave the corresponding
enol ether, 6.23
Resolution of Helicene Enantiomers. Prior to this work, a
compound similar to 1, but with only two alkoxyl groups, had
been resolved into its enantiomers by a reduction and esterifi-
cation that produced a separable mixture of two diastereomeric
tetra-N-toluenesulfonyl-L-proline esters, which were then each
reconverted into the bisquinone, now optically active.9b Also
previously resolved into their enantiomers by way of diesters
of a nonracemic acid, in this case (-)-menthoxyacetic acid,24
were some helicenediols that were subsequently transformed
into [6]- and [7]helicenebisquinones.25 However, for the resolu-
tion of helicenebisquinones derived from those synthesized
according to Scheme 1, tetraesters, formed by reduction and
esterification with N-toluenesulfonyl-L-proline, (-)-menthoxy-
acetic acid, and a variety of other nonracemic acids,26 all proved
unsatisfactory. What does work superbly, not only for these
examples, but also for 10 other helicenequinones synthesized
subsequently in the course of this and related studies,27 is
reduction and esterification with (1S)-(-)-camphanoyl chlo-
ride.28
Scheme 3a
a Reagents and conditions: (a) (i) Na2S2O4, H2O, THF; (ii)
(1S)-(-)-camphanoyl chloride, Et3N, CH2Cl2. Yields: 76%, (-)-
diastereomer; 96%, (+)-diastereomer. (b) C12H25I, K2CO3, DMF, 93%
yield. (c) (i) Na2S2O4, H2O, THF; (ii) (1S)-(-)-camphanoyl chloride,
Et3N, CH2Cl2. Yields: 85%, (-)-diastereomer; 82%, (+)-diastereomer.
and (1S)-(-)-camphanoyl chloride in boiling toluene, but it gave
yields that were slightly lower: a 78% yield of the (-)-
diastereomer and an 86% yield of the (+)-diastereomer.
Chromatography on an ordinary silica gel column easily
separates the diastereomers of 12 and 13 in yields of 76-96%.
In both cases the separation is excellent. In the case of 13 the
difference in Rf values was measured to be 0.3, which is large,30
and the separation was carried out on a gram scale. In fact, the
separations are so good that, according to 1H NMR, 13C NMR,
and thin layer chromatographic analysis, no contamination by
diastereomeric material could be detected in any of the four
samples. The ratio of the signal to noise in the 1H NMR spectra
was such that the diastereomeric purities of the samples prepared
from 12 were g98% and from 13 g99.5%.
Conversion of 12 and 13 into Nonracemic 1. To complete
the synthesis of nonracemic 1, the camphanates of the separated
diastereomers of 12 and 13 had to be removed and the resulting
helicenes oxidized. In addition, for the synthesis from 12, the
silyl ethers and pivalate esters had to be converted into long-
chain ethers. Notably, when combined with methyllithium in
1,2-dimethoxyethane and then acetic anhydride, the silyl,
pivalate, and camphanate groups of 12 were all replaced by
acetates in only one step. The yield of 14 was 95%.
A strategy that had been used to convert only some of the
ester groups of hexaacetoxytriphenylenes into ethers, heating
with potassium carbonate, 18-crown-6, and dodecyl iodide in
isobutyl methyl ketone,31 removed the six least hindered esters
of octaacetate 14, replacing them with dodecyls in 30% yield.32
The structure 3 was assigned to the product on the basis of the
NMR chemical shift of the acetate protons. Of the acetate
resonances of 14, at δ 1.22, 2.48, 2.50, and 2.53 ppm, the ones
that are absent in 3 are the latter three, those at the normal
position for aryl acetate NMRs.33 The one that remains (in 3 it
resonates at δ 1.12 ppm), like the related resonances of
previously synthesized helicenes that have acetoxy groups
occupying the positions assigned to those in 3,9b,12,34 is at ca.
1.5 ppm higher field than normal. In addition, the structure is
But before (()-1 could be resolved, it had to be prepared
from 7, and as Scheme 3 shows, this was done in a way that
avoided the isolation of the phenolic quinone. The combination
of potassium carbonate and dodecyl iodide in warm DMF
transformed the acetates into dodecyl ethers in a yield of 93%.
The resolution was then effected in the following way. Sodium
dithionite29 reduced helicenequinones 5 and (()-1 to the
bishydroquinones, which, because they were easily oxidized in
the air, were esterified directly with (1S)-(-)-camphanoyl
chloride and triethylamine. An alternative was to use Fox’s
modification9c of Willmore’s procedure,9b using Zn, TMEDA,
(19) The method of Corey, E. J.; Schmidt, G. Tetrahedron Lett. 1979,
399.
(20) The method of Mander, L. N.; Sethi, S. P. Tetrahedron Lett. 1984,
25, 5953.
(21) (a) TBDMSCl, NaI, Et3N: see ref 9b. (b) TBDMSCl and KH:
Orban, J.; Turner, J. V.; Twitchin, B. Tetrahedron Lett. 1984, 25, 5099.
(22) Ley, S. V.; Norman, J.; Griffith, W. P.; Marsden, S. P. Synthesis
1994, 639.
(23) By the method of Corey, E. J.; Cho, H.; Ru¨cker, C.; Hua, D. H.
Tetrahedron Lett. 1981, 22, 3455.
(24) Yang, B. V. Ph.D. Dissertation, Columbia University, 1987.
(25) Yang, B.; Liu, L.; Katz, T. J.; Liberko, C. A.; Miller, L. L. J. Am.
Chem. Soc. 1991, 113, 8993.
(26) (-)-Menthyl carbonic, cholesteryl carbonic, (R)-(-)-O-acetylman-
delic, and camphorsulfonic. Either the yields of the diastereomers were poor
or the diastereomers could not be separated.
(27) Reference 9c and unpublished results by a number of co-workers
in our lab.
(28) Commercially available and easily prepared (Gerlach, H.; Kappes,
D.; Boeckman, R. K., Jr.; Maw, G. N. Organic Syntheses; Wiley: New
York, 1998; Collect. Vol. IX, p 151).
(29) (a) Ulrich, H.; Richter, R. In Methoden der Organischen Chemie
(Houben-Weyl); Grundmann, C., Ed.; Georg Thieme Verlag: Stuttgart, 1977;
Vol. VII/3a, pp 649-650. (b) Liu, L. Ph.D. Dissertation, Columbia
University, 1991, applied sodium dithionite to reduce helicene bisquinones.
(30) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923.
(31) (a) Kranig, W.; Hu¨ser, B.; Spiess, H. W.; Kreuder, W.; Ringsdorf,
H.; Zimmermann, H. AdV. Mater. 1990, 2, 36. (b) Kreuder, W.; Ringsdorf,
H. Makromol. Chem., Rapid Commun. 1983, 4, 807.
(32) Also formed in this reaction was a triacetate, presumably a material
with the two acetoxy groups of 3 and one other.
(33) 2-Naphthyl acetate and phenyl acetate have resonances for the acetate
methyls at 2.31 and 2.27 ppm, respectively, Pouchert, C. J.; Behnke, J.
The Aldrich Library of 13C and 1H FT-NMR Spectra, Edition 1, Vol. 2;
Aldrich Chemical Co.: Milwaukee, 1993; pp 1298C and 1285B.
(34) Dai, Y.; Katz, T. J. J. Org. Chem. 1997, 62, 1274. In addition, this
reference shows a similar effect applies to isobutyrate esters.

82 J. Am. Chem. Soc., Vol. 121, No. 1, 1999
Nuckolls et al.
Scheme 4a
a Reagents and conditions: (a) (i) MeLi, DME; (ii) Ac2O, 95% yield.
(b) C12H25I, K2CO3, 18-crown-6, i-BuCOMe, 30% yield. (c) (i) KOEt,
THF; (ii) (NH4)2Ce(NO3)6, H2O, dioxane, 94% yield. (d) (i) MeLi, THF;
(ii) chloranil, 95% yield.
Figure 2. (a, top) CD spectra (ordinate on the left, displaying molar
circular dichroism) and (b, bottom) UV-vis absorption spectra (ordinate
on the right, displaying molar extinction coefficient) of 2.5 × 10-2
M
solutions of (-)-1: (s) in dodecane; (- - -) in 10% CHCl3-90%
dodecane; (- - -) in 50% CHCl3-50% dodecane. The path lengths are
all 10 µm.
× 10-2 M, circular dichroisms at ca. 251, 342, and 511 nm
increase greatly. In particular, the circular dichroism at 251 nm
becomes very large, whereas in the dilute solutions it is tiny.
These changes are one indication that the molecules of non-
racemic 1, which at 1.0 × 10-5 M behave independently of
one another, aggregate when their solutions in dodecane are
concentrated. Similar changes in CD spectra are seen when a
related helicene that organizes into a columnar mesophase is
dissolved in dodecane and similarly concentrated7 and when
planar chromophores aggregate that have chiral groups at-
tached.37 Figure 1 also shows that in the most concentrated
solution in dodecane (the one that is 2.5 × 10-2 M), the UV
absorption maximum at 315 nm decreases in intensity and an
additional maximum appears at 332 nm.
Figure 1. (a, top) CD spectra (ordinate on the left, displaying molar
circular dichroism) and (b, bottom) UV-vis absorption spectra (ordinate
on the right, displaying molar extinction coefficient) of (-)-1 in
dodecane: (s) when the concentration is 2.5 × 10-2 M and the path
length is 10 µm; (- - -) when the concentration is 1.0 × 10-4 M and
the path length is 2 mm; (- - -) when the concentration is 1.0 × 10-5
M and the path length is 1 cm.
Figure 2 shows that when chloroform is added to the
dodecane, the CD and UV-vis absorption spectra of a 2.5 ×
10-2 M solution of (-)-1, which in the absence of chloroform
are those attributed to the aggregates, change to those attributed
to the unassociated molecules.38 The spectra of a solution in
1:1 CHCl3-dodecane are essentially those of the 1.0 × 10-5
M solution in dodecane displayed in Figure 1. When the solvent
is 10% CHCl3-90% dodecane, the spectra are partway between
the two extremes.
supported by the demonstration that removing the acetates of 3
and then oxidizing gives 1. Thus, (Scheme 4) treatment with
potassium ethoxide in THF and then with cerric ammonium
nitrate35 gives the helicenebisquinone in 94% yield.
It is because yields are so poor in the Diels-Alder oxidation
step producing 5 (Scheme 1) and in the reaction converting
octaacetate 14 into the hexaether diacetate 3 that only small
amounts of 1 could be obtained following the procedure reported
originally.1 The synthesis that proceeds by way of 13 does not
have these shortcomings. Not only is the yield good in the
reaction that gives 7, but, as seen in Scheme 4, the removal of
the camphanate esters of 13, by treatment with methyllithium,
followed by the immediate oxidation of the resulting bishy-
droquinone, by treatment with chloranil,36 gives 1 in 95% yield.
Aggregation of Nonracemic 1 in Solution. When solutions
of nonracemic 1 in dodecane are concentrated, a number of their
properties change.
(37) For examples in which the CDs of aggregating molecules are
enhanced by concentration see footnote 4 in ref 1 and (a) Gottarelli, G.;
Mezzina, E.; Spada, G. P.; Carsughi, F.; Di Nicola, G.; Mariani, P.;
Sabatucci, A.; Bonazzi, S. HelV. Chim. Acta 1996, 79, 220 and references
therein. For examples in which solvents and temperature are the agents of
change, see: (b) Palmans, A. R. A.; Vekemans, J. A. J. M.; Havinga, E.
E.; Meijer, E. W. Angew. Chem., Int. Ed. Engl. 1997, 36, 2648. (c) Matile,
S.; Berova, N.; Nakanishi, K. Chem. Biol. 1996, 3, 379. (d) Huang, D.;
Matile, S.; Berova, N.; Nakanishi, K. Heterocycles 1996, 42, 723 and
references therein. (e) Schnur, J. M.; Ratna, B. R.; Selinger, J. V.; Singh,
A.; Jyothi, G.; Easwaran, K. R. K. Science 1994, 264, 945. For examples
in which organization in films results in enhanced CD spectra, see ref 37b.
Also see: (f) van Nostrum, C. F.; Bosman, A. W.; Gelinck, G. H.; Schouten,
P. G.; Warman, J. M.; Kentgens, A. P. M.; Devillers, M. A. C.; Meijerink,
A.; Picken, S. J.; Sohling, U.; Schouten, A.-J.; Nolte, R. J. M. Chem. Eur.
J. 1995, 1, 171. (g) Barbera´, J.; Iglesias, R.; Serrano, J. L.; Sierra, T.; de la
Fuente, M. R.; Palacios, B.; Pe´rez-Jubindo, M. A.; Va´zquez, J. T. J. Am.
Chem. Soc. 1998, 120, 2908.
(a) CD and UV-Vis Absorption Spectra and Specific
Rotations. Figure 1 shows that while the CD spectra of 1.0 ×
10-5 and 1.0 × 10-4 M solutions of (-)-1 in dodecane are
essentially identical, when the concentration is increased to 2.5
(35) By the method of Brimble et al. and Jacob et al.: (a) Brimble, M.
A.; Hodges, R.; Stuart, S. J. Tetrahedron Lett. 1988, 29, 5987. (b) Jacob,
P., III; Callery, P. S.; Shulgin, A. T.; Castagnoli, N., Jr. J. Org. Chem.
1976, 41, 3627.
(38) Similar effects of CHCl3 on aggregates of other molecules are in
ref 37b. For those of ClCH2CH2Cl see: (a) Markovitsi, D.; Bengs, H.;
Ringsdorf, H. J. Chem. Soc., Faraday Trans. 1992, 88, 1275. (b) Gallivan,
J. P.; Schuster, G. B. J. Org. Chem. 1995, 60, 2423.
(36) By the method described in ref 9b.

Synthesis of a Conjugated Helical Molecule
J. Am. Chem. Soc., Vol. 121, No. 1, 1999 83
Figure 4. Light scattering from solutions of (-)-1: (s) in dodecane
at a concentration of 2.0 × 10-2 M; (- - -) in dodecane at a
concentration of 5.0 × 10-3 M; ( - - - ) in dodecane at a
Figure 3. (a, top) CD spectra (ordinate on the left, displaying molar
circular dichroism) and (b, bottom) UV-vis absorption spectra (ordinate
on the right, displaying molar extinction coefficient) of (-)-1 in
dodecane: (s) when the concentration is 2.5 × 10-2 M and the path
length is 10 µm; (- - -) when the concentration is 1.0 × 10-3 M and
the path length 105 is µm; (- - -) when the concentration is 5.0 × 10-4
M and the path length is 1 mm.
s
s
concentration of 1.0 × 10-3 M; (- - -) in 50% CHCl3-50% dodecane
at a concentration of 2.0 × 10-2 M. The solutions were in 2 mm path
length cells.
by light scattering data (Figure 4). There are two significant
observations. One is that while a 2.5 × 10-2 M solution of (-)-1
in 1:1 CHCl3-dodecane, in which the helicene molecules appear
to be unassociated (Figure 2), scatters no amount of measurable
light between 500 and 700 nm, a solution in dodecane that is 1
× 10-3 M does scatter detectable amounts of light, maximally
at ca. 590 nm. The second is that, as the concentration in
dodecane increases, the scattered light increases in intensity.6,41
Enhanced light scattering is expected near absorption bands of
aggregates whose chromophores couple, an effect called reso-
nance light scattering,6 and as illustrated by the curves in Figure
4, the scattering can be perturbed by intense absorption.42 The
peaks are at wavelengths much greater than those of the lowest
energy absorption bands (ca. 490 nm, Figures 1 and 3),
presumably because strong absorption of light at lower wave-
lengths shifts the observed scattering peaks to higher wave-
lengths. Moreover, it shifts the peaks further when the concen-
tration, and therefore the absorption, increases. However, the
telling observation is that, at wavelengths above 625 nm, the
intensity of scattered light overcomes the effects of absorption.
It increases with the sample’s concentration. This is possible
only if the aggregates increase appreciably in size.
Table 1. Specific Rotation and g Values (∆ꢀ/ꢀ) at Three
Wavelengths of (-)-1 in Dodecane and Neat
g × 103
concn (M)
252 nm
350 nm
530 nm
[R]D × 103 a
1 × 10-5
5 × 10-4
2.5 × 10-2
1.0 × 10-1
neatb
-0.19
4.5
4.5
-1.7
-4.5
-5.2
-0.36
-2.2
-4.1
-0.60
-5.4
-7.1
-16
12
-9.8
-8.1
-160
a (deg cm2)/10 g. b The g values were measured using a film on a
quartz plate that had been spin coated from a 1 × 10-3 M octane
solution. To determine the [R]D of such a film, the thickness (0.41 µm)
was measured by comparing the absorbance with that of a sample of
known thickness (2 µm). The density was assumed to be 0.85 g/mL.
Figure 3 shows that concentrations much less than 2.5 × 10-2
M are sufficient for aggregation to be recognizable in the
electronic spectra. Thus, even a solution in dodecane as dilute
as 5.0 × 10-4 M displays CD and UV-vis absorption spectra
essentially identical to those of a 2.5 × 10-2 M solution in
dodecane. Measures of the aggregation are the values of ∆ꢀ/ꢀ,
Kuhn’s dissymmetry factors g,39 which Table 1 summarizes at
three wavelengths for the dodecane solutions of varying
concentration.
The table also displays how the specific rotations of the
samples vary with concentration. It shows that the g values are
larger for the more concentrated solutions than for the one that
is most dilute and that the [R]D values rise from -600 (deg
cm2)/10 g when the sample is most dilute, to -5400 (deg cm2)/
10 g when aggregation is first detected spectroscopically (at
5.0 × 10-4 M), to -16 000 (deg cm2)/10 g when the solution
is 0.1 M.40 The g values suggest that more concentrated samples
have higher specific rotations because the aggregates they
contain are larger.
There is a more evident manifestation of aggregation. Just
like solutions of other molecules that aggregate, those of
nonracemic 1 in dodecane, which are not notably viscous when
their concentrations are e1 × 10-3 M, are highly viscous when
the concentration is 2.5 × 10-2 M.43 In sum, the increasing g
(41) (a) Pasternack, R. F.; Bustamante, C.; Collings, P. J.; Gianetto, A.;
Gibbs, E. J.J. Am. Chem. Soc. 1993, 115, 5393. (b) Pasternack, R. F.;
Schaefer, K. F. Inorg. Chem. 1994, 33, 2062. (c) de Paula, J. C.; Robblee,
J. H.; Pasternack, R. F. Biophys. J. 1995, 68, 335. (d) Parkash, J.; Robblee,
J. H.; Agnew, J.; Gibbs, E.; Collings, P.; Pasternack, R. F.; de Paula, J. C.
Biophys. J. 1998, 74, 2089. (e) Purrello, R.; Scolaro, L. M.; Bellacchio, E.;
Gurrieri, S.; Romeo, A. Inorg. Chem. 1998, 37, 3647.
(42) (a) Pasternack, R. F.; Goldsmith, J. I.; Sze´p, S.; Gibbs, E. J. Biophys.
J. 1998, 75, 1024. (b) Collings, P. J.; Gibbs, E. J.; Pasternack, R. F.
unpublished results.
(43) Molecules whose solutions are either viscous or gel include, besides
many that hydrogen-bond to one another, a number that are united by
π-stacking forces: (a) Scheibe, G. Kolloid-Z. 1938, 82, 1. (b) Lin, Y.;
Kachar, B.; Weiss, R. G. J. Am. Chem. Soc. 1989, 111, 5542. (c) Snijder,
C. S.; de Jong, J. C.; van Bolhuis, F.; Feringa, B. L. Chem. Eur. J. 1995,
1, 594. (d) van Nostrum, C. F.; Picken, S. J.; Schouten, A.-J.; Nolte, R. J.
M. J. Am. Chem. Soc. 1995, 117, 9957. (e) Brotin, T.; Utermo¨hlen, R.;
Fages, F.; Bouas-Laurent, H.; Desvergne, J.-P. J. Chem. Soc., Chem.
Commun. 1991, 416.
(b) Light Scattering and Viscosity. Evidence that aggrega-
tion increases when the solutions are concentrated is provided
(39) (a) Kuhn, W. Trans. Faraday. Soc. 1930, 26, 293. (b) Eliel, E. L.;
Wilen, S. H.; Mander, L. N. Stereochemistry of Organic Compounds;
Wiley: New York, 1994; Chapter 13.
(40) There was no sign of anisotropy. The specific rotations did not
change when the samples were rotated about the light beam, and no features
were seen when the samples, sandwiched between crossed polarizers, were
observed with a microscope.

84 J. Am. Chem. Soc., Vol. 121, No. 1, 1999
Nuckolls et al.
The films of nonracemic 1 have unparalleled ability to rotate
the plane of polarization of plane-polarized light. Initially we
reported1 that 10 µm thick samples rotated plane-polarized light
at 589 nm (the sodium D-line) by 1400°/mm, which corresponds
to [R]D 170 000 (deg cm2)/10 g.48 After further study, we found
that 633 nm light from a low-power HeNe laser was appreciably
scattered by these films. We therefore experimented with making
films that are thinner and found that, by casting solutions in
octane onto quartz plates, we could prepare thin films that
scattered less. The thickness of one such film was determined
to be 0.41 µm by comparing its absorbance of 550 nm light
with that of a film known to be 2 µm thick.49 The specific
rotation of 589 nm light was measured each time this film,
perpendicular to the light beam, was rotated about the beam
through six successive 30° increments. The values diverged little
and averaged to 0.534 ( 0.03°. This means that R ) -1300
deg/mm and [R]D -158 000 (deg cm2)/10 g.50
The specific rotation of 1 appears to be the highest recorded.
Other materials with high specific rotations include a metal-
locene oligomer ([R]D 34 000 (deg cm2)/10 g)51 and a ferro-
electric liquid crystal (R ) 100°/mm).52 Cholesteric liquid
crystals exhibit comparable rotations, but only for light at the
reflecting wavelength (which is the product of their pitch and
refractive index).53
Figure 5. (a, top) CD spectra (ordinate on the left, displaying ellipticity
in mdeg) and (b, bottom) UV-vis absorption spectra (ordinate on the
right, displaying absorbance) of a spin-coated film of (-)-1.
values, specific rotations, and light scattering all imply that as
the concentrations of nonracemic 1 in dodecane increase, so
do the sizes of the aggregates the molecules form.
Chiroptical Properties of Thin Film and Solvent-Free
Samples of Nonracemic 1. Displayed in Figure 5 are the CD
and UV-vis absorption spectra of a thin film that was prepared
by spin coating from a 1 × 10-3 M solution of (-)-1 in octane
onto a quartz slide. The CD spectrum did not change appreciably
when the sample was rotated in the plane perpendicular to the
beam of light,44 but to eliminate any effects of linear birefrin-
gence and linear dichroism, a spectrum was measured each time
the sample had been rotated through six successive 30°
increments.45 Figure 5 is the average of these spectra. The
significant observation is that the film has essentially the same
CD and UV-vis absorption spectra as the solutions of ag-
gregated 1 (that is the 2.5 × 10-2 M solution in Figure 1), with
one important difference. The g values (Table 1) are larger for
the film than for the solutions (e.g., at 252 nm, 3 times larger).
They are similar to one of the highest recorded, that of an
optically active polythiophene.46 That the value of g is larger
for the film than for the most concentrated solution (Table 1)
suggests that the aggregates in the bulk material are larger too.
This notion is supported by the organization of the molecules
in the pure material into very long fibers that are ca. 3 µm
wide.1,2
The spectra of the spin-coated films are similar to those of
previously prepared Langmuir-Blodgett (LB) films.3 Both are
characteristic of the aggregates in dodecane solution. Notable
differences are that the spectra of the spin-coated sample are
sharper and their g values are higher (g ) 1.2 × 10-2 at 252
nm, compared to g ) 7.5 × 10-3 at 240 nm for the LB film).3
Possibly the differences originate from the LB film having been
spread from CHCl3 and the spin-coated film from octane, in
which, according to Figures 1 and 2 (if octane is like dodecane),
the molecules should be more organized than in CHCl3.47
Properties of Racemic 1. Racemic 1 is much less soluble
than nonracemic 1 in dodecane. Accordingly, its UV-vis
absorption spectra in this solvent could be determined only if
the concentration was less than ca. 5.4 × 10-4 M. These spectra
were the same as those of the most dilute solutions of the
nonracemic material. Unlike the nonracemic material, racemic
1 does not form fibers visible under the microscope when it is
treated in the same way as the nonracemic material, that is, when
it is heated to a temperature (ca. 210 °C) at which it flows freely
and then is cooled. This accords with the observation that AFM
images of Langmuir-Blodgett films of the racemic material
do not show the long columnar structures that are seen in
comparable films of the nonracemic material.4 The implication
is that long columns form and assemble into large fibrous
structures only if the molecules have the same handedness.2
Properties of Nonracemic 3. To provide a reference to which
the properties of nonracemic 1 could be compared, the optical
properties of (+)-3 were examined. The thought was that the
cores of 1 and 3 have the same number of rings, the side chains
projecting from the cores are similar, and 3 even has two more
dodecyl side chains than 1, which might make stacking more
favorable. Yet as Figure 6 shows, the CD, UV-vis absorption,
and fluorescence emission spectra of solutions of nonracemic
3 in dodecane are essentially the same when the concentration
(47) Similar effects of solvents on the films of phthalocyanines have
been attributed to the formation of a lyotropic phase (Fujiki, M.; Tabei, H.;
Kurihara, T. Langmuir 1988, 4, 1123).
(48) In the initial communication,1 the units of the specific rotations were
stated incorrectly as (deg cm2)/g instead of (deg cm2)/10 g. The [R]D for
pure nonracemic 1 should have been listed as 170 000 (deg cm2)/10 g instead
of 170 000 (deg cm2)/g.
(49) This sample was made by sandwiching 1 between quartz plates that
were separated by a 2 µm aluminum spacer from The Goodfellow Corp.,
Berwyn, PA.
(44) The beam was ca. 200 mm2 in area.
(50) The density was assumed, as previously,1 to be 0.85 g/mL.
(51) Katz, T. J.; Sudhakar, A.; Teasley, M. F.; Gilbert, A. M.; Geiger,
W. E.; Robben, M. P.; Wuensch, M.; Ward, M. D. J. Am. Chem. Soc. 1993,
115, 3182.
(45) (a) Kuball, H.-G.; Scho¨nhofer, A. In Circular Dichroism: Principles
and Applications; Nakanishi, K., Berova, N., Woody, R., Eds.; VCH: New
York, 1994; Chapter 4. (b) Scho¨nhofer, A.; Kuball, H.-G. Chem. Phys. 1987,
115, 159. (c) Norde´n, B. Acta Chem. Scand. 1972, 26, 1763. (d) Tunis-
Schneider, M. J. B.; Maestre, M. F. J. Mol. Biol. 1970, 52, 521. (e) Cornell,
D. G. J. Colloid Interface Sci. 1979, 70, 167.
(46) For optically active polythiophenes in dodecanol g values are said
to be typically 2 × 10-2 (Langeveld-Voss, B. M. W.; Janssen, R. A. J.;
Christiaans, M. P. T.; Meskers, S. C. J.; Dekkers, H. P. J. M.; Meijer, E.
W. J. Am. Chem. Soc. 1996, 118, 4908).
(52) Musˇevicˇ, I.; Sˇkarabot, M.; Kityk, A. V.; Blinc, R.; Moro, D.;
Heppke, G. Ferroelectrics 1997, 203, 133.
(53) (a) de Vries, H. Acta Crystallogr. 1951, 4, 219. (b) Chandrasekhar,
S.; Prasad, J. S. Mol. Cryst. Liq. Cryst. 1971, 14, 115. (c) Isaert, N.;
Berthault, J.-P.; Billard, J. J. Opt. 1980, 11, 17. (d) Cano, R.; Chatelain, P.
Compt. Rend. 1964, 259, 352. (e) Baessler, H.; Laronge, T. M.; Labes, M.
M. J. Chem. Phys. 1969, 51, 3213.

Synthesis of a Conjugated Helical Molecule
J. Am. Chem. Soc., Vol. 121, No. 1, 1999 85
Table 2. DSC Transition Temperatures and Enthalpies for
Nonracemic and Racemic 1a
first transition T, °C
(∆H, kJ/mol)
second transition T, °C
(∆H, kJ/mol)
cycle
Nonracemic 1
heating
cooling
heating
cooling
heating
cooling
230.7 (45.0)
212.7 (-18.2)
222.8 (15.7)
195.6 (-7.9)
206.4 (23.6)
197.4 (-21.0)
209.4 (23.6)
192.8 (-9.2)
Racemic 1
206.7 (36.2)
191.5 (-30.2)
heating
cooling
a The temperature was scanned between 30 and 250 °C at a rate of
10 °C/min.
This unusual behavior would be understandable if the helicene
molecules invert configuration during the analyses. In accord
with this explanation, racemic 1 exhibits the same DSC
characteristics as nonracemic 1 that has been heated and cooled
a number of times (Table 2) and nonracemic 1 does racemize
sufficiently rapidly at the temperatures in Table 2. Thus, in 1,2,4-
trichlorobenzene the half-times for racemization were measured
as 49 min at 198.0 °C, 28 min at 203.0 °C, and 17 min at 207.5
°C. The temperature at which the half-time for racemization is
ca. 1 h, 196 °C, is thus essentially the same as that for a similar
[6]helicenebisquinone that lacks the alkoxyls in the peri
positions, 190 °C,9b and for the parent [6]helicene, 201 °C.56
However, the enthalpy of activation, ∆Hq ) 48.1 kcal/mol, is
much larger than for the racemization of the parent [6]helicene
(∆Hq ) 35.4 kcal/mol).56 This might relate to the 3.9 kcal/mol
greater enthalpy of activation for racemization of a [5]heli-
cenebisquinone11 compared to its parent, [5]helicene.57 More
remarkable, however, is the enormously larger entropy of
Figure 6. (a, top) CD spectra (ordinate on the left, displaying molar
circular dichroism) and (b, bottom) UV-vis absorption spectra (ordinate
on the right, displaying molar extinction coefficient) of 3 in dodecane:
(s) when the concentration is 2.0 × 10-2 M and the path length is 10
µm and (- - -) when the concentration is 2.0 × 10-5 M and the path
length is 10 mm. The inset (c) displays emission spectra (excitation at
325 nm) of dodecane solutions of 3 in a 10 µm cell: (s) when the
concentration is 2.1 × 10-2 M and (- - -) when the concentration is 2.1
× 10-5 M solutions.
is 2 × 10-2 M as when it is 2 × 10-5 M, and as reported
previously,1 so are the specific rotations ([R]D 410 and 413 (deg
cm2)/10 g). Furthermore, the pure material in a 10 µm cell
rotates plane-polarized light at 589 nm by 0.11°, corresponding
to [R]D 1300 (deg cm2)/10 g.1,48,50 When viewed under a
microscope between crossed polarizers, it is isotropic. These
properties contrast with those of nonracemic 1, whose CD and
UV-vis absorption spectra (Figure 1), fluoresence emission
spectra,1 and specific rotations (Table 1) change greatly when
its solutions in dodecane are concentrated and which under a
polarizing microscope is seen to be highly anisotropic.1,2 We
suppose the reason that 1 aggregates, whereas 3 does not, is
that donor-acceptor interactions between electron-rich rings of
one molecule and electron-poor rings of the adjacent molecules
stabilize columnar stacks,54 and 3, unlike 1, has only electron-
rich rings. An alternative explanation might be that stacking is
inhibited in 3 by the bulk of the additional acetate groups, but
a helicene that has ester groups at one end does organize into
columns.7 In any case, an essential point is that, for helicenes
to stack, it seems that many side chains alone do not suffice.
activation for the racemization of 1 (∆Sq ) 26 cal mol-1 K-1
)
compared to that for the racemization of the parent [6]helicene
(∆Sq ) -1.9 cal mol-1 K-1).56 This could be due to the relief
of conformational restraints at the peri positions, but we do not
have data to address the point. The 45 kJ/mol enthalpy for the
initial DSC transition of nonracemic 1 is 3 times greater than
that reported for a helicene liquid crystal7 and also greater than
the 1-20 kJ/mol for various discotic liquid crystals,58 which
we associate with the assembly of 1 into fibrous structures that
the other materials do not form.
Conclusions
An improved synthesis of helicene 1 is described. Notable
are the 56% yield in the helicene-forming step and the method
developed to resolve helicenequinones. Nonracemic helicene
1, unlike nonracemic 3 and also unlike any previously known
helicene, aggregates in solution and in thin films. In dodecane,
aggregation occurs even in solutions as dilute as 5 × 10-4 M.
The aggregation is characterized by enhanced circular dichro-
isms and g values, red-shifted UV-vis absorptions, increased
specific rotations, increased light scattering at an absorption
frequency, and, as reported previously, fluorescence emissions
that shift to the red and 1H NMR resonances that shift upfield.
The chiroptical properties of concentrated solutions in dodecane
to which CHCl3 has been added are similar to those of the most
Differential Scanning Calorimetry and Kinetics of Race-
mization. Table 2 summarizes the differential scanning calo-
rimetric (DSC) analyses of nonracemic and racemic 1.55 Only
one transition is seen in the first heating cycle, at 230.7 °C, but
in subsequent cooling and heating cycles, two transitions are
seen initially and then only one, at a temperature 21 °C lower.
(54) The notion that π-donor-acceptor interactions stabilize aggregating
columnar systems was also proposed by others: ref 38a,b. Also see: (a)
Green, M. M.; Ringsdorf, H.; Wagner, J.; Wu¨stefeld, R. Angew. Chem.,
Int. Ed. Engl. 1990, 29, 1478 and references therein. (b) Mohr, B.; Wegner,
G.; Ohta, K. J. Chem. Soc., Chem. Commun. 1995, 995.
(56) (a) Martin, R. H.; Marchant, M.-J. Tetrahedron Lett. 1972, 3707.
(b) Martin, R. H.; Marchant, M. J. Tetrahedron 1974, 30, 347. The values
of ∆Hq (35.4 kcal/mol) and ∆Sq (-1.9 cal mol-1 K-1) were recalculated
using the kinetic data in the first of these references.
(57) Goedicke, C.; Stegemeyer, H. Tetrahedron Lett. 1970, 937.
(58) See footnote 15 in ref 7.
(55) In the initial publication,1 the DSC was scanned to only 210 °C
and therefore just missed the phase transitions recorded here.

86 J. Am. Chem. Soc., Vol. 121, No. 1, 1999
Nuckolls et al.
gave 3.0 g of pure diol (a 96% yield), an off-white oil that slowly
dilute solutions in dodecane alone, implying that CHCl3 causes
the aggregates to dissociate. Cast films of the pure material show
high specific rotations, [R]D -158 000 (deg cm2)/10 g. The CD
and UV-vis absorption spectra of spin-coated films are similar
to those of the concentrated solutions in dodecane, but the g
values are higher. Unlike the nonracemic material, the racemic
material in bulk, after it has been heated and cooled, does not
form visible fibers, and it is too insoluble in dodecane to give
concentrated solutions.
1
crystallized: IR (CCl4, KBr) 3616, 3500 cm-1; H NMR (400 MHz,
CDCl3) δ 0.92 (t, 6 H, J ) 7.0 Hz), 1.30-1.65 (m, 36 H), 1.47 (d, 6
H, J ) 6.4 Hz), 1.93 (quint, 4 H, 8.1 Hz) 3.92 (t, 4 H, 6.5 Hz), 4.75
(q, 2 H, 6.4 Hz), 6.87 (s, 2 H), 7.27 (s, 2 H); 13C NMR (75 MHz,
CDCl3) δ 14.1, 22.7, 24.9, 26.3, 29.4-29.7 (m), 31.9, 69.3, 70.5, 104.2,
116.3, 116.5, 137.1, 144.1, 157.1; HRMS (FAB) calcd for C38H64O4
584.4805, found 584.4818.
To 3.0 g of this diol (5.13 mmol) were added 2.5 g of 4 Å molecular
sieves, 25 mL of dry CH2Cl2, 95 mg of tetrapropylammonium
perruthenate (0.27 mmol), and 1.58 g of N-methylmorpholine N-oxide
(13.5 mmol). The deep green-black mixture was stirred at room
temperature for ca. 1 h and then poured onto a silica gel column (2 in.
× 6 in.). CH2Cl2 eluted 2.5 g (an 85% yield) of 11, a slightly yellow,
waxy solid: mp (from CH2Cl2) 98-101 °C; IR (NaCl) 1683 cm-1; 1H
NMR (400 MHz, CDCl3) δ 0.88 (t, 6 H, J ) 6.9 Hz), 1.20-1.70 (m,
36 H), 1.93 (quint, 4 H, J ) 8.0 Hz), 2.72 (s, 6 H), 4.14 (t, 4 H, J )
6.6 Hz), 7.46 (d, 2 H, J ) 1.2 Hz), 8.06 (d, 2 H, J ) 1.3 Hz); 13C
NMR (75 MHz, CDCl3) δ 14.5, 23.1, 26.7, 26.8, 29.8-30.1 (m), 69.8,
106.2, 122.5, 124.5, 136.1, 136.3, 157.9, 198.0. Anal. Calcd for
C38H60O4: C, 78.57; H, 10.41. Found: C, 78.48; H, 10.42.
Bis(enol ether) 6. CH2Cl2 (14 mL) and Et3N (3.6 mL) were syringed
into 11 (1.5 g, 2.6 mmol) under N2 in a flame-dried round-bottomed
flask containing a magnetic stir bar. The flask was cooled to 0 °C, and
triisopropylsilyl triflate (1.43 mL, 5.3 mmol) was added in drops from
a syringe. The reaction mixture was allowed to warm to room
temperature (ca. 10 min) and then recooled to 0 °C. Saturated aqueous
NaHCO3 was added, and the mixture was extracted with benzene (2×).
The combined benzene extracts were washed with saturated aqueous
NaHCO3, 1 N NaOH, and water, dried over 4 Å molecular sieves, and
filtered through Celite. Removal of solvent and chromatography on a
1 in. × 3 in. column of neutral alumina in 1:1 benzene-hexane plus
3% Et3N (eluent: the same solvent) gave, after the product had been
heated at 100 °C in a vacuum, 2.2 g of 6 (a 95% yield), a clear, brightly
fluorescent oil that slowly crystallized at room temperature: 1H NMR
(400 MHz, CDCl3) δ 0.89 (t, 6 H, J ) 7.0 Hz), 1.16 (d, 36 H, J ) 7.3
Hz), 1.25-1.60 (m, 42 H), 1.92 (quint, 4 H, J ) 7.2 Hz), 4.07 (t, 4 H,
6.7 Hz), 4.51 (d, 2 H, 1.7 Hz), 4.97 (d, 2 H, 1.7 Hz), 7.06 (d, 2 H, 1.3
Hz), 7.67 (d, 2 H, 1.4 Hz); 13C NMR (75 MHz, CDCl3) δ 13.1, 14.5,
18.4, 23.1, 26.8, 29.6-30.2 (m), 32.2, 69.7, 91.3, 105.1, 117.9, 118.2,
136.0, 139.8, 156.4, 157.0; HRMS (FAB, M + 1) calcd for C56H101O4-
Si2 893.7238, found 893.7237.
Helicenebisquinone 7. p-Benzoquinone (5.3 g, 49.3 mmol) was
added to a solution of 6 (2.2 g, 2.45 mmol) in anhydrous heptane (35
mL) under N2, contained in a flame-dried 100 mL round-bottomed flask,
outfitted with a magnetic stir bar and reflux condenser, that had been
evacuated and flushed with N2 (5×). The reaction mixture was
evacuated, flushed with N2, and gently refluxed for 4 d. During the
course of the reaction, p-benzoquinone that had sublimed into the reflux
condenser was scraped back into the reaction flask. The solvent was
distilled, and the residue under vacuum was heated at 135 °C to sublime
away p-benzoquinone and hydroquinone. THF (30 mL), Ac2O (9.35
mL, 99 mmol), Et3N (13 mL), and Et3N‚HF (1.19 g, 9.9 mmol) were
added to the residual glassy solid, and the mixture was stirred for 2 h
at room temperature. CH2Cl2 was added, washed with H2O (3×), dried
(MgSO4), filtered, and removed under reduced pressure. Chromatog-
raphy on a 2 in. × 6 in. column of silica gel (eluent: benzene containing
amounts of EtOAc that increased from 5 to 10%) yielded 1.2 g of 7 (a
56% yield), a brick-red powder: mp (from PhH) 164-166 °C; 1H NMR
(400 MHz, CDCl3) δ 0.88 (t, 6 H, J ) 6.9 Hz), 1.20-1.70 (m, 36),
1.97 (m, 4 H), 2.54 (s, 6 H), 4.22 (m, 2 H) 4.24 (m, 2 H), 6.60 (d, 2
H, J ) 9.8 Hz), 6.79 (d, 2 H, J ) 9.8 Hz), 7.23 (s, 2 H), 7.95 (s, 2 H);
13C NMR (75 MHz, CDCl3) δ 14.5, 21.6, 23.1, 26.8, 29.8-30.1 (m),
32.3, 70.0, 99.5, 116.7, 121.2, 127.3, 131.0, 132.4, 132.7, 136.9, 139.8,
148.5, 159.0, 168.7, 184.1, 185.9; HRMS (FAB, M + 1) calcd for
C54H65O10 873.4573, found 873.4578. Anal. Calcd for C54H64O10: C,
74.29; H, 7.39. Found: C, 74.30; H, 7.45.
Experimental Section
THF and toluene were distilled from Na plus benzophenone, CH2-
Cl2 and Et3N from CaH2. Benzoquinone was purified by mixing it with
4 times its weight of basic alumina, a small amount of 4 Å molecular
sieves, and CH2Cl2. The slurry was filtered through Celite and the
solvent removed under reduced pressure.
Methyllithium (1.6 M in Et2O, low halide content) was purchased
from Acros. The following were purchased from Aldrich and used
without further purification: tetrapropylammonium perruthenate (97%),
N-methylmorpholine N-oxide (97%), 4 Å powdered molecular sieves
(activated), triisopropylsilyl trifluoromethanesulfonate (97%), heptane
(anhydrous, 99%), 1-iodododecane (98%), DMF (anhydrous, 99.8%),
(1S)-(-)-camphanoyl chloride (98%), and chloranil (99%).
Details of the syntheses of compounds 3, 4, 5, 8, 9, and 14 and of
the previous synthesis of 1 can be found in the Supporting Information
to ref 1.
4,5-Dihydroxy-2,7-naphthalenedicarboxaldehyde. EtOH (76 mL,
100%) was injected into a mixture of powdered KOH (2.92 g, 52.0
mmol) and 8 (2 g, 5.2 mmol) that had been evacuated three times and
flushed with N2 in a flame-dried round-bottomed flask containing a
magnetic stir bar. The mixture was stirred for 35 min, by which time
a heavy yellow precipitate had formed. Cooling in an ice-water bath
and stirring with 1 N HCl and then cold water precipitated the product,
which was filtered and washed with much water to remove pivalic acid.
Drying, first with air in the funnel for 30 min and then under high
vacuum overnight, yielded 1.09 g (a 97% yield) of 4,5-dihydroxy-2,7-
naphthalenedicarboxaldehyde, a yellow powder: mp (from EtOH/
H2O) > 225 °C; IR (NaCl) 1680 cm-1; 1H NMR (400 MHz, acetone-
d6) δ 7.35 (d, 2 H, J ) 1.3 Hz), 8.20 (d, 2 H, J ) 1.3 Hz), 10.11 (s,
2 H), 10.65 (br s, 2 H); 13C NMR (75 MHz, acetone-d6) δ 107.9, 120.9,
128.0, 136.7, 137.1, 156.0 192.2; HRMS (FAB) calcd for C12H8O4
216.0423, found 216.0418.
4,5-Didodecyloxy-2,7-naphthalenedicarboxaldehyde (10). 1-Io-
dododecane (6.85 mL, 27.8 mmol) and then dry DMF (30 mL) were
added to 4,5-dihydroxy-2,7-naphthalenedicarboxaldehyde (1.00 g, 4.63
mmol) and powdered K2CO3 (3.20 g, 23.1 mmol) that had been
evacuated and flushed with N2 several times in a flame-dried round-
bottomed flask containing a magnetic stir bar. The mixture was stirred
at 50 °C overnight, cooled to room temperature, and poured into ice-
water. Extraction with CH2Cl2 (3×), washing with water, drying
(MgSO4), filtration, removal of solvent under reduced pressure, and
silica gel chromatography (eluent: first benzene and then 2% EtOAc
in benzene) gave 2.50 g of 10 (a 98% yield), a buff-colored, waxy
1
solid: mp (from PhH) 105-107 °C; IR (NaCl) 1699 cm-1; H NMR
(400 MHz, CDCl3) δ 0.90 (t, 6 H, J ) 6.9 Hz), 1.20-1.70 (m, 36 H),
1.96 (quint, 4 H, J ) 8.0 Hz), 4.16 (t, 4 H, J ) 6.6 Hz), 7.39 (d, 2 H,
J ) 1.1 Hz), 7.98 (d, 2 H, J ) 1.1 Hz), 10.11 (s, 2H); 13C NMR (75
MHz, CDCl3) δ 14.1, 22.7, 26.3, 29.3-29.7 (m), 31.9, 69.5, 104.4,
123.6, 128.3, 135.5, 136.0, 158.2, 191.7; HRMS (FAB) calcd for
C36H56O4 552.4179, found 552.4185.
2,7-Diacetyl-4,5-didodecyloxynaphthalene (11). THF (50 mL) was
added to 3.0 g of 10 (5.4 mmol) in a flame-dried 100 mL round-
bottomed flask containing a magnetic stir bar. Although cooling to 0
°C gave some precipitate, 10.2 mL of 1.6 M MeLi in Et2O (16.2 mmol)
was syringed in drops into the stirred slurry. After the mixture had
stirred at room temperature for 30 min, it was recooled to 0 °C, treated
with saturated aqueous NH4Cl to destroy excess MeLi, and combined
with H2O (20 mL). Extraction (EtOAc, 3×), drying (powdered activated
4 Å molecular sieves), filtration through Celite, and removal of solvent
(()-1. Anhydrous DMF (10 mL) was syringed into 7 (0.4 g, 0.41
mmol), K2CO3 (1.26 g, 9.2 mmol), and 1-iodododecane (5.7 mL, 23
mmol) contained in a flame-dried round-bottomed flask with a magnetic
stir bar that had been evacuated and flushed with N2. After the reaction

Synthesis of a Conjugated Helical Molecule
J. Am. Chem. Soc., Vol. 121, No. 1, 1999 87
mixture had again been repeatedly evacuated and flushed with N2, it
was heated to 65 °C and stirred vigorously for 2.5 h. The reaction
mixture was cooled to room temperature, and EtOH (10 mL) was added
to precipitate the product, which was collected in a fritted funnel on
top of a pad of Celite and washed with large amounts of EtOH. CH2-
Cl2 was used to wash the product from the Celite, and after the solvent
had been removed under reduced pressure, the residue was chromato-
graphed on a 1 in. × 4 in. column of silica gel, which yielded 0.48 g
solution turned from lemon yellow to deep red. Filtration and rotary
evaporation gave a solid, which was chromatographed on a 1 in. × 3
in. silica gel column. Elution with 2.5% EtOAc in benzene, concentra-
tion at reduced pressure, dissolution in a small amount of CH2Cl2, and
precipitation by the addition of ethanol yielded 0.22 g (a 95% yield)
of (-)-1, a red wax: IR (KBr) 1656 cm-1; 1H NMR (400 MHz, CDCl3)
δ 0.87 (m, 6H), 1.28 (m, 32H), 1.60 (m, 4H), 1.99 (m, 4H), 4.25 (m,
4H), 6.53 (d, J ) 11.1 Hz, 1H), 6.70 (d, J ) 10.8 Hz, 1H), 7.41 (s,
1H), 7.59 (s, 1H); 13C NMR (75 MHz, CDCl3) δ 14.11, 22.70, 26.23,
26.45, 29.06, 29.39, 29.44, 29.63, 29.68, 29.70, 29.73, 29.76, 29.79,
31.94, 69.27, 69.42, 98.81, 102.07, 120.57, 126.46, 128.00, 129.11,
131.78, 132.14, 135.92, 139.73, 156.81, 184.93, 185.26; HRMS calcd
for C74H110O8 1126.8200, found 1126.8230.
1
of (()-1 (a 93% yield), a red wax whose H and 13C NMR, IR, and
UV spectra were identical to those of samples of 1 prepared according
to the published procedure.59
Preparation and Resolution of the Enantiomers of 13. THF (30
mL) and then water (15 mL) that had been boiled to remove O2 and
cooled under N2 were added to (()-1 (1.0 g, 0.89 mmol) and Na2S2O4
(2 g, 11.5 mmol) in a flame-dried 100 mL round-bottomed flask with
a magnetic stir bar. After the mixture had stirred for 30 min, the color
had changed from deep red to lemon yellow, indicating that the
bisquinone had been reduced. Distilled CH2Cl2 (30 mL) and H2O (10
mL) were added, solvent was removed from the organic layer under
reduced pressure, and the residual yellow solid was heated overnight
at 100 °C and then cooled to room temperature. (1S)-(-)-Camphanoyl
chloride (2.3 g, 10.7 mmol) was added, and the flask containing the
two solids was evacuated and flushed with N2. After THF (25 mL)
had been added and the mixture cooled to 0 °C, Et3N (4 mL) was added
in drops from a syringe, and the reaction mixture was refluxed and
stirred for 24 h. Benzene (20 mL) was added to the cooled mixture
and then washed with saturated aqueous NaHCO3, dilute aqueous NH3,
1 N HCl, H2O, and finally saturated aqueous NaHCO3, dried (MgSO4),
filtered, and removed under reduced pressure. Chromatography on a 2
in. × 8 in. silica gel column (eluent: 25% EtOAc-75% hexanes)
yielded (-)-13 (0.70 g, an 85% yield) and then (eluent: 33% EtOAc-
67% hexanes) (+)-13 (0.68 g, an 83% yield), both yellow glassy
substances.
Data for (-)-13: IR (CCl4, KBr) 1798, 1752 cm-1; 1H NMR (400
MHz, CDCl3) δ 0.52 (s, 6 H), 0.61 (s, 6 H), 0.85-1.00 (m, 20 H),
1.05-1.65 (m, 112 H), 1.70-2.10 (m, 12 H), 2.35-2.45 (m, 2 H),
2.72-2.82 (m, 2 H), 4.15-4.32 (m, 6H), 4.52-4.58 (m, 2 H), 6.25 (d,
2 H, J ) 8.3 Hz), 6.85 (d, 2 H, J ) 8.3 Hz), 7.17 (s, 2 H), 7.78 (s, 2
H); 13C NMR (75 MHz, CDCl3) δ 9.6, 9.8, 14.1, 16.1, 17.0, 22.7, 26.4,
26.6, 28.9, 29.1, 29.4-29.8 (m), 31.1, 31.9, 54.2, 54.4, 54.7, 55.0, 68.2,
69.4, 89.5, 91.4, 97.7, 100.3, 114.8, 117.5, 117.8, 120.2, 121.1, 126.4,
127.1, 128.4, 142.9, 143.5, 153.7, 156.0, 164.9, 166.2, 177.8, 178.1;
[R]D -431 (c 0.17, CH2Cl2); HRMS (FAB) calcd for C114H160O20
1849.1503, found 1849.1501.
A similar procedure converted 410 mg of (+)-13 into 245 mg of
(+)-1 (a 98% yield).
Preparation and Resolution of (()-12. Dry oxygen-free THF (50
mL) and then deoxygenated water (20 mL) were added to 5 (0.5 g, 0.6
mmol) and Na2S2O4 (1 g, 6 mol) in a dry round-bottomed flask fitted
with a reflux condenser. The suspension was refluxed for 30 min and
then cooled. Dry CH2Cl2 (50 mL) and 10 mL of H2O were syringed
in, and the organic layer was cannulated into another dry oxygen-free
flask. Two 10 mL portions of CH2Cl2 washed the remainder over.
Stripping yielded a bright yellow solid, which was heated for 3-4 h at
100 °C. Powdered 4 Å molecular sieves, followed by 10 mL of CH2-
Cl2, were added to the air-sensitive bishydroquinone, and after it had
cooled to 0 °C, the solution was transferred by cannula to another flask
that contained (-)-camphanoyl chloride (1.53 g, 7.1 mmol) at 0 °C.
CH2Cl2 (5 mL) was used to transfer residual bishydroquinone. Et3N (1
mL, 7.1 mmol) was added in drops from a syringe, and the yellow
solution was stirred at 0 °C for 2 h. The cold solution and then repeated
CH2Cl2 washings were filtered into 1 M HCl, and the organic solution
was separated and washed with 1 M HCl, aqueous NH3, 1 M HCl, and
saturated aqueous NaHCO3. After it had been dried (MgSO4) and
filtered and the solvent had been stripped, the crude waxy product was
chromatographed on a 3.5 in. × 14 in. silica gel column and eluted
first with hexanes-EtOAc-CH2Cl2 (4:2:1) and then with hexanes-
EtOAc-CH2Cl2 (3:1.75:1). The levorotatory diastereomer, (-)-12 (351
mg, 76%), eluted first, followed by a small amount (<50 mg) of the
mixed diastereomers and then the dextrorotatory diastereomer (445 mg,
96%).
Data for the (-)-isomer: 1H NMR (400 MHz, CDCl3) 0.33 (s, 6
H), 0.48 (s, 6 H), 0.58 (s, 12 H), 0.92 (s, 6 H), 1.05-1.30 (m, 40 H),
1.37 (m, 2 H), 1.48 (s, 18 H) 1.62 (m, 2H), 1.86 (m, 2H) 2.07 (m, 2H),
2.43 (m, 2H), 2.75 (m, 2H), 6.43 (d, J ) 8.4 Hz, 2 H), 7.02 (d, J ) 8.4
Hz, 2 H), 7.39 (s, 2 H), 7.96 (s, 2 H); 13C NMR (75 MHz, CDCl3) d
9.55,9.79, 16.01, 16.96, 17.01, 18.30, 25.71, 27.38, 28.75, 29.01, 29.48,
31.06, 39.67, 54.06, 54.39, 54.93, 89.54, 91.16, 105.30, 114.59, 115.20,
118.20, 120.64, 121.77, 123.95, 127.38, 128.01, 142.77, 143.87, 146.43,
150.06, 164.74, 165.96, 177.54, 177.98; HRMS calcd for C88H108O22-
Si2 1572.6870, found 1572.6820.
Data for the (+)-isomer: 1H NMR (400 MHz, CDCl3) δ 0.32 (s,
6 H), 0.45 (s, 6 H), 0.47 (s, 6 H), 0.74 (s, 6 H), 1.13 (s, 18 H), 1.22 (s,
6 H), 1.24 (s, 6 H), 1.26 (s, 6 H), 1.30-1.50 (m, 24 H), 1.64 (m, 2 H),
1.87 (m, 2 H) 2.10 (m, 2H), 2.37 (m, 2H), 2.71 (m, 2H), 6.34 (d, J )
8.4 Hz, 2 H), 6.91 (d, J ) 8.4 Hz, 2 H), 7.12 (s, 2 H), 7.93 (s, 2 H);
13C NMR (75 MHz, CDCl3) δ 9.47,9.59, 16.15, 16.71, 16.91, 18.18,
25.55, 27.27, 28.70, 28.82, 29.90, 30.63, 31.13, 39.60, 53.88, 54.12,
54.57, 54.73, 89.99, 90.62, 90.75, 104.43, 114.8, 116.8, 118.6, 121.62,
122.07, 122.28, 122.49, 122.92, 126.60, 128.17, 128.73, 128.85, 142.53,
144.012, 147.06, 150.47, 165.37, 166.01, 176.75, 177.47.
Among peaks in the 1H NMR spectra that are isolated and prominent
are those at 7.39 ppm in the (-)-diastereomer and at 7.12 ppm in the
(+)-diastereomer. The former resonance could not be found in the
latter’s spectrum and vice versa, and the ratio of signal to noise implied
that each diasteriomer contained not more than 2% of the other.
CD, UV-Vis Absorption, Fluorescence Emission, Polarimetry,
and Light Scattering Analyses. UV-vis absorption spectra were
recorded using a Perkin-Elmer Lambda 19 UV-vis spectrometer,
circular dichroism spectra using a JASCO 720 spectrometer, fluores-
cence spectra using a Spex Fluorolog 86 fluorimeter, and specific
rotations using a JASCO Dip 1000 polarimeter. Dodecane was either
Data for (+)-13: IR (CCl4, KBr) 1799, 1750 cm-1; 1H NMR (400
MHz, CDCl3) δ 0.47 (s, 6 H), 0.77 (s, 6 H), 0.87-1.05 (m, 20 H),
1.20-1.75 (m, 110 H), 1.87-1.95 (m, 2 H), 2.00-2.18 (m, 10 H),
2.37-2.45 (m, 2 H), 2.72-2.82 (m, 2 H), 4.20-4.38 (m, 6 H), 6.26
(d, 2 H, J ) 8.3 Hz), 6.86 (d, 2 H, J ) 8.3 Hz), 7.01 (s, 2 H), 7.73 (s,
2 H); 13C NMR (75 MHz, CDCl3) δ 9.6, 9.7, 14.1, 16.3, 16.7, 17.1,
22.7, 26.5, 26.6, 28.6, 28.9, 29.1-29.8, 31.2, 31.9, 54.1, 54.3, 54.4,
54.9, 68.6, 69.6, 90.3, 91.0, 96.6, 100.3, 115.8, 117.6, 118.5, 119.3,
122.4, 125.5, 127.9, 128.9, 142.7, 144.0, 153.9, 156.9, 165.0, 166.2,
177.3, 177.7; [R]D +310 (c 0.17, CH2Cl2); HRMS (FAB) calcd for
C
114H160O20 1849.1503, found 1849.1519.
Among peaks in the 1H NMR spectra that are isolated and prominent
are those at 7.17 ppm in the (-)-diastereomer and at 7.01 ppm in the
(+)-diastereomer. The former resonance could not be found in the
latter’s spectrum and vice versa, and the excellent ratio of signal to
noise implied that each diasteriomer contained 0.5% or less of the other.
(-)-(1). THF (15 mL) and then, while stirring, MeLi (3.95 mL 6.32
mmol) in drops were added to (-)-13 (0.390 g, 0.021 mmol) under N2
in a flame-dried round-bottomed flask that was cooled to -78 °C.
Stirring was continued while the mixture was allowed to warm to room
temperature and for an additional 30 min. Saturated aqueous NH4Cl
was added in drops to the mixture, the mixture was recooled to 0 °C,
until excess MeLi had reacted, and then 1 N HCl and benzene were
added. The organic layer was washed with water and dried (MgSO4).
Before the MgSO4 was filtered, chloranil was added, whereupon the
(59) See the Supporting Information to ref 1.

88 J. Am. Chem. Soc., Vol. 121, No. 1, 1999
Nuckolls et al.
olefin-free material from Fluka or was freed of olefin by shaking with
concentrated H2SO4, washing, drying, and distilling. The cells used to
measure the spectra were as follows: a 10 µm quartz cell from Starna
Cells, used for ca. 10-2 M solutions; a 105 µm quartz cell, used for ca.
0.500; 120, 0.591; 150, 0.510. The average is 0.534 ( 0.03. The
absorbance at 550 nm of this film was 4.9 times less than that of a
film prepared by crushing the sample between quartz plates separated
by a 2 µm aluminum spacer from the Goodfellow Corp., Berwyn, PA.
10-3 M solutions; a 1 mm and a 2 mm quartz cell, used for ca. 10-4
M
solutions; and a 1 cm quartz cell used for ca. 10-5 M solutions. The
fluorescence and light scattering spectra measured emission from the
front faces of 10 µm and 2 mm quartz cells, respectively.
Acknowledgment. We thank the NSF (Grant CHE9802316)
and The Kanagawa Academy of Science and Technology for
support and numerous colleagues for assistance, especially
Professors Timothy Swager and Scott Trzaska of MIT for the
DSC analyses, for letting us use their polarizing microscope,
and for advice, Professor Louis Brus (Columbia) for the use of
his spin-coater, Dr. Steffen Jockusch (Columbia) for help with
the light-scattering measurements, Professor Carl Gryte (Co-
lumbia) for the use of his microscope, and Evan A. Hendrix
for help with the graphics.
The neat samples of 1 used for the measurements in Figure 5 were
prepared by spin coating at 6000 rpm from a 1 × 10-3 M octane
solution onto a quartz plate, which was then mounted in a circular holder
that allowed the plate to be rotated through 360° at 30° increments.
The film used to measure the specific rotation recorded in Table 1 was
prepared by placing a 1 × 10-3 M octane solution into the well formed
by sandwiching a ca. 100 µm Teflon washer (Ace Glass Inc., Vineland,
NJ) and a quartz flat (NSG Precision Cells Inc., Farmingdale, NY) in
a metal clamp. The well was loosely covered with a glass plate, which
allowed the octane solution to evaporate slowly during the course of a
day. To remove any remaining solvent, the quartz disk was placed under
vacuum overnight. The resulting circular film was centered in the holder
described above. A HeNe laser beam that passed through the film had
a well-defined center spot, which meant that little light was scattered
by the sample. The rotation (R) of a plane-polarized light beam 8 mm
in diameter was measured as a function of the angle (φ) through which
the sample was rotated in the plane perpendicular to the beam, as
follows (φ, R, both in degrees): 0, 0.525; 30, 0.550; 60, 0.530; 90,
Supporting Information Available: 1H NMR, 13C NMR,
and IR spectra of 4,5-dihydroxy-2,7-naphthalenedicarboxalde-
hyde, (-)-1, 6, 7, 10, the diol formed when MeLi is added to
10, 11, (+)-12, (-)-12, (+)-13, and (-)-13, and DSC traces
for (-)-1 and (()-1 (39 pages, print/PDF). See any current
masthead page for ordering information and Web access
instructions.
JA983248L