Photochromism in 1,1'-bis-2-naphthols

Full text of DOI:10.1021/ja960516b

9990
J. Am. Chem. Soc. 1996, 118, 9990-9991
Photochromism in 1,1′-Bi-2-naphthols
Marino Cavazza and Maurizio Zandomeneghi*
Dipartimento di Chimica e Chimica Industriale
UniVersita’ di Pisa, Via Risorgimento 35
I-56126 Pisa, Italy
Akihiko Ouchi and Yoshinori Koga
National Institute of Materials
and Chemical Research
1-1 Higashi, Tsukuba, Ibaraki 305, Japan
ReceiVed February 19, 1996
For a long time the events following electronic excitation of
naphthols have been studied;1 excited-state proton transfer is a
typical process in such cases.1b We report here on the
photochemistry of both racemic and optically active 1,1′-bi-2-
naphthol (1). A novel photochromic cycle involving the main
photoproduct has been observed: the photoequilibrium was
strongly dependent on both the wavelength of the irradiating
light and the solvent. In addition, the photoreactions discovered
were highly selective, stereospecific, and of synthetic value for
compounds with complex structure and stereochemistry.
When 1,1′-bi-2-naphthol (5 × 10-4 M in methanol) was
irradiated with an immersion medium-pressure Hg lamp (100
W), a sluggish decrease of the concentration of 1 was observed
with concomitant formation of four aromatic products, which
were detected by HPLC analysis. We were able to isolate the
most relevant compound (2), which, however, was produced
with an almost negligible quantum yield. The structure best
Figure 1. (a) UV spectra of 1 (continuous line) and 2 (dashed line) in
MeOH. (b) Circular dichroism spectra of (-)-1 (continuous line) and
of (-)-2 (dashed line) in n-hexane.
1
fitting our experimental data (mass spectrometry data, H and
Scheme 1
13C NMR, UV, CD, and IR spectroscopies)2 is reported in
Scheme 1.
In Figure 1a, the UV absorption spectra of 1 and 2 are
displayed together for comparison. In Figure 1b, the CD spectra
of (-)-1 and of (-)-2, the compound obtained from irradiation
of (-)-1 (Vide infra), are also reported.
Photoaddition of an OH group to a C-C double bond is a
uncommon event reported, for example, in the intramolecular
cyclization of o-allylphenol to give dihydrobenzofuran.3
The absence of phenolic OH function in 2 is confirmed by
unreactivity of 2 with CH2N2 in diethyl ether solution, while,
in identical experimental conditions, quantitative formation of
the dimethoxy derivative of 1 is obtained. In contrast with the
Hg lamp photochemical experiment, irradiation with the 334
nm emission of a CW Ar+ laser (100 mW) of a methanol
solution of 1 (5 × 10-4 M) resulted in the production of 2 with
a reasonable quantum yield (ca. 0.038) and ca. 50% conversion
(Vide infra). However, 2 was not formed by the other UV Ar+
laser emissions at 351 and 364 nm. In addition, photochemical
reversion of 2 to 1 was observed with all three emissions of the
Ar+ laser. Scheme 1 shows the main features of the reactions
involving 1 and 2.
From the UV spectra in Figure 1a, we see the strong
absorption of 2 at λ > 350 nm, likely due to the o-xylylene
moiety in 2,4 in contrast with the quasi-transparency of 1 in the
same region. Thus, we can understand why 2 is not formed
during irradiation of 1 at the other UV Ar+ laser emissions of
351 and 365 nm. From Scheme 1 it is clear that irradiation
must result in a photostationary state in which the [2]/[1] ratio
depends on the extinction coefficients ꢀ1 and ꢀ2 and on the ratio
of the quantum yields, φ1f2 and φ2f1, at the irradiation
wavelength according to the formula:
(1) (a) Griffiths, J.; Chu, K. Y.; Hawkins, C. J. Chem. Soc., Chem.
Commun. 1976, 676-677. Akhtar, I. A.; McCullough, J. J. J. Org. Chem.
1981, 46, 1447-1450. Chow, Y. L.; Buono-Core, G. E.; Liu, X. Y.; Ito,
K.; Qiou, P. J. Chem. Soc., Chem. Commun. 1987, 12, 913-915. (b) Weller,
A. Z. Elektrochem. Ber. Bunsen-Ges. Physik. Chem. 1952, 56, 662-668.
Shizuka, H. Acc. Chem. Res. 1985, 18, 141-147. Van Stam, J.; Lofroth,
J. E. J. Chem. Educ. 1986, 63, 181-184. Jankowski, A.; Stefanowicz, P.
Studia Biophysica 1989, 131, 55-66. Iwanek, W.; Mattay, J. J. Photochem.
Photobiol., A 1992, 67, 209-226. Arnaut, L. G.; Formosinho, S. J. J.
Photochem. Photobiol., A 1993, 75, 1-20; 1993, 75, 21-48.
[2]/[1] ) ꢀ1φ1f2/ꢀ2φ2f1
(2) 1H NMR (C6D6) δ: 3.00 (2H, m, H-3), 4.87 (1H, br s, OH), 4.97
(1H, m, H-4), 5.85 (1H, dddd, J3,4′ ) 2.0 and 2.68, J4,4′ ) 3.0, J3′,4′
)
10.26, H-4′), 6.25 (1H, dddd, J3.3′ ) 1.5 and 2.36, J4,3′ ) 1.9, J3′,4′ ) 10.26,
H-3′), 6.86 and 7.00 (2H, AB system, J ) 8.3, H-2′ and H-1′), 7.10-8.00
(series of d, naphthalene protons). 13C NMR (CDCl3) spectrum displays,
in addition to the aromatic carbons, two signals at δ 29.1 (t) and 71.7 (d)
ppm. These signals were unambiguously assigned to C-3 and C-4 by a
HETCOR experiment. The IR spectrum presents a series of bands not
present in the 1 spectrum. The new C-H out of plane bending vibrations
The ratio [2]/[1], at λ ) 334 nm, was calculated to be 0.8
(i.e., 45% 2 and 55% 1 in the photostationary mixture). The
observed proportion of 2 and 1 was 49 and 51%, respectively,
(4) Kolc, J.; Michl, J. J. Am. Chem. Soc. 1973, 95, 7391-7401. Flynn,
C. R.; Michl, J. J. Am. Chem. Soc. 1974, 96, 3280-3288. Migirdicyan,
E.; Baudet, J. J. Am. Chem. Soc. 1975, 97, 7400-7404. Tseng, K. L.;
Michl, J. J. Am. Chem. Soc. 1977, 99, 4840-4842.
occur at 639 (m), 701 (s), 791 (s), 859 (m) cm-1
.
(3) Miranda, M. A.; Tormos, R. J. Org. Chem. 1993, 58, 3304-3307.
S0002-7863(96)00516-1 CCC: $12.00 © 1996 American Chemical Society

Communications to the Editor
J. Am. Chem. Soc., Vol. 118, No. 41, 1996 9991
with (1.5% experimental error. The difference between
calculated and observed concentration ratios can be reasonably
explained by a small amount of side reactions from 1 which
were not included in Scheme 1, in addition to experimental
errors in the quantum yields, evaluated at ca. 15% conversion
of 1 and 2, respectively. Also, the inefficient production of 2
observed with a standard medium-pressure Hg lamp can be
explained. In fact, the photoequilibrium ratio was calculated
to be 0.03 taking into account the energy distribution of the
emissions of the 100 W medium-pressure Hg lamp used
(essentially 366 and 334 nm), the pyrex transmittance, and the
absorbance of 1 and 2 at these emissions. Probably, this small
photoequilibrium ratio explains why 2, obtained in a very small
amount, was not detected and reported until now.
We note that atropisomerism of 1 is converted into the more
usual asymmetric carbon dependent chirality in 2, which
stereospecifically reverts to the original atropisomerism by the
2 f 1 photo(or thermal, Vide infra)transformation. Thus,
experimental findings show that the chirality in 2, derived from
a covalent structure, is more photostable than the chirality in 1,
derived from conformational barriers. Finally, the asymmetry
of 2 appears higher than that of 1, when the maximum values
of the dissymmetry factor, the Kuhn g-factor,8 in the 200-400
nm spectral region is considered; gmax ) (∆ꢀ/ꢀ)max ) 0.0067 at
260 nm for 2, compared to gmax ) 0.0034 of 1 at 240 nm.
The effect of the solvent in the 1 f 2 photorearrangement is
particularly noteworthy. In fact, this reaction was strongly
depressed in an aprotic solvent such as CH3CN (a 20-fold
reduction), compared with that in CH3OH. The inverse 2 f 1
photoreaction was only lowered by half. Thus, the photoequi-
librium [2]/[1] ratio in CH3CN became almost zero at 334 nm:
in practice it was difficult to observe any photoreaction of 1 in
CH3CN. Interestingly, in a 80:20 (v/v) CH3CN/H2O mixture,
the photoreaction proceeded as well as when carried out in CH3-
OH. Finally, we were unable to observe the 1 f 2 phototrans-
formation when 1 was dissolved in n-hexane.
Let us now turn to the properties and structure of the novel
molecule 2. Observing the o-xylylene moiety present in 2 and
remembering its intrinsic lability also in substituted xylylenes,4
one is inclined to tribute some thermal instability to 2. In reality,
this fact, showing 2 as a valence isomer of 1, was the first to
be noticed. The thermal back reaction of 2 revealed some
interesting features:
(1) It was strongly dependent on the solvent. At 85 °C we
found no reaction in CH3CN, but the monomolecular kinetic
constants k ) 6.2 × 10-6 s-1 and k ) 1.8 × 10-5 s-1 were
measured for the thermal back reaction in CH3CN-H2O (80:20,
v/v) and in CH3OH, respectively. In other words, CH3CN
protects 1 from phototransformation into 2 and protects 2 from
thermal return to 1.
Subsequently, some stereochemical aspects of the 1 f 2
transformation have to be reported. It is noted that 334 nm
irradiation of 99.5% optically pure (-)-1 in MeOH gave 2 with
(-)-2 as the prevailing stereoisomer.5 The optical purity of both
1 and 2 recovered from the irradiation mixtures was lower than
that of the starting material; the optical purity decreased with
the amount of the energy absorbed. When (-)-1 (99.5%
enantiomeric excess (ee)) in MeOH was photolyzed up to 15%
photoconversion, (-)-2 with 99% ee (analysis by chiral HPLC)
was obtained. At 28% phototransformation, the ee2 was found
to have lowered to 98.5% ee2, at 36% conversion 97.8% ee2
was measured, and finally, at 44.2% conversion 95.3% ee2 was
observed. The corresponding enantiomeric excesses of 1 were
98.1, 97.3, 95.9, and 92.3%, respectively. Thus, ee1 < ee2; this
fact could be the result of photoracemization of 1 or connected
to the 2 f 1 phototransformation. The racemization of 1
through its direct electronic excitation is probable because such
processes have been observed in other atropisomeric diaryls6
and also in the photolysis of the (-)-1 antipode complexed to
bovine serum albumin.7 In contrast with the above photocon-
versions starting from (-)-1, selective photolysis of chemically
pure (-)-2 resulted in the production of optically active (-)-1,
whose ee was exactly equal to that of 2. During the photolysis
ee2 remained constant. Selective photolysis of 2 was obtained
by using the 364 nm laser emission (this line is not absorbed
by formed 1, cf., Figure 1a). Moreover, the constancy of ee2
proved that excitation of 2 gave no racemization of 2, taking
into account that ca. 95% of absorbed 364 nm photons
electronically excite 2 without converting it, as revealed by the
quantum yield value 0.048 of the 2 f 1 conversion. Then,
photoracemization of 1 explains the lowering of ee1 and ee2 as
well as the result of ee1 < ee2 during the irradiations of optically
active 1. It seems very likely that inversion of the chirality
during the 1 f 2 transformation, if any, is a minor occurrence
with respect to the photoracemization of 1.
(2) Thermal racemization in the conversion 2 f 1 was not
detected at 85 °C in CH3OH.
(3) An activation energy of 106 kJ/mol was measured in the
temperature range of 65-85 °C in CH3CN/H2O solutions (80:
20, v/v). The activation energy relates to the thermal reversion
in other photochromic systems being of the same order of
magnitude (e.g., for cis-/trans-azobenzene and azonaphthalene
it is ca. 96 kJ/mol).9
Finally, an extension of the above findings to other binaph-
thols was attempted. We found formation of the methyl ether
derivative of 2 starting from the monomethyl ether of 1. The
UV and CD absorption properties of this compound were very
similar to those of 2. On the contrary, no phototransformation
of the dimethyl ether of 1 was found. Clearly, the presence of
acidic hydrogen(s) is required by the photoreaction described.
(5) For the pure (-)-2 antipode, obtained by semipreparative chiral
HPLC, [R] ) -571, -615, and -871 at 578, 546, and 446 nm, respectively,
in CH3OH and 0.135 g/100 mL. Chiral HPLC: tR ) 35 and 41 min for
the (+)-2 and (-)-2 antipodes, respectively; 0.5 mL/min flux; mobile phase
n-hexane/isopropyl alcohol 85:15 (v/v); Chiralcel OJ.
(6) Zimmermann, H. E.; Crumrine, D. S. J. Am. Chem. Soc. 1972, 94,
498-506. Irie, M.; Yoshida, K.; Hayashi, K. J. Phys. Chem. 1977, 81,
969-972. Tetreau, C.; Lavalette, D.; Cabaret, D.; Geraghty, N.; Welvart,
Z. NouV. J. Chim. 1982, 6, 461-465.
Acknowledgment. The MURST project “Ecocompatibilita’ dei
Processi Chimici: Aspetti Metodologici ed Applicativi” is gratefully
acknowledged. A particular aknowledgement goes to one of the
referees for his sound criticism.
(7) Levi-Minzi, N.; Zandomeneghi, M. J. Am. Chem. Soc. 1992, 114,
9300-9304.
(8) Kuhn, W. Ann. ReV. Phys. Chem. 1958, 9, 417-438.
(9) Fisher, E. EPA Newsletter 1995, 54, 16-32.
JA960516B