Photoswitching of dextro/levo rotation with axially chiral binaphthyls linked to an azobenzene

Article abstract of DOI:10.1021/jo901030s

(Figure Presented) To examine the reversible photoisomerization and subsequent change of asymmetric field, we synthesized optically active 3,3′-disubstituted-1,1′-binaphthyls with an azobenzene moiety. Reflecting the structural change, the specific rotati

Full text of DOI:10.1021/jo901030s

pubs.acs.org/joc
been selected as objects in computational chemistry.4
Photoswitching of Dextro/Levo Rotation with
Axially Chiral Binaphthyls Linked to an Azobenzene
Moreover, reflecting their chirality, some axially chiral
binaphthyls exhibit relatively large optical rotations and
circular dichroism (CD).5 The large positive-negative-rever-
sible changes of such optical properties have been evaluated
with promising results in the development of novel photo-
switching materials. Moreover, optical rotation can be de-
tected at an unabsorbed wavelength, so target compounds
do not degrade during measurement. Hence, a switch for
dextro (positive)/levo (negative) rotation will lead to the
development of noise-cancellation, nondestructive reading
of memory devices.6c Although practical recording-repro-
duction systems have been designed by using optical rota-
tion, practical applications have yet to emerge due to the lack
of appropriate compounds.6d,e Irie, Branda, Wang, and
Yokoyama et al. have independently reported changes using
helicenoids via photochroism,7 but the CD variations are not
large or unreported and the change in optical rotations at the
sodium D-line, the universal wavelength (589 nm), does not
involve a sign inversion. Our research strives to change the
CD variations and optical rotations using axially chiral
binaphthyl derivatives.
Kazuto Takaishi,*,† Masuki Kawamoto,† Kazunori
Tsubaki,‡ and Tatsuo Wada*,†
†Supramolecular Science Laboratory, RIKEN, Hirosawa 2-1,
Wako, Saitama 351-0198, Japan, and ‡Graduate School of
Life and Environmental Science, Kyoto Prefectural
University, Shimogamo Hangi-cho, Sakyo, Kyoto 606-8522,
Japan
ktakaishi@riken.jp; tatsuow@riken.jp
Received May 18, 2009
We have reported basic compound (R)-1 (Figure 1), which
is composed of binaphthyl and azobenzene skeletons.8 Azo-
benzene skeletons are heavily used photochromic parts,9 and
show a significant change in length between the cis and trans
forms. The azobenzene moiety of (R)-1 can be reasonably
cis-trans photoisomerized. Additionally, isomerization can
be detected by the change in CD intensity, which is likely due
to the change in the dihedral angle formed by the two
naphthalene rings because the dihedral angle is related to
the CD intensity. This change is reversible, but the change is
small and sign reversal is not detected. Although we have
attempted to shorten the linkers so that the structural change
of the azobenzene moiety can be more directly transmitted to
the binaphthyl, our attempts have been unsuccessful. As an
alternative, herein we investigate whether the 3,30-substitu-
ents of the binaphthyl moiety can produce large confor-
mation changes between the cis and trans forms as well as
initiate subsequent changes in optical rotation and CD.
To examine the reversible photoisomerization and sub-
sequent change of asymmetric field, we synthesized opti-
cally active 3,30-disubstituted-1,10-binaphthyls with an
azobenzene moiety. Reflecting the structural change,
the specific rotation and circular dichroism underwent
significant variations. Under certain conditions, the po-
sitive-negative signals were reversible. Furthermore, the
magnitude of these changes showed a 3,30-substituent
dependency. Dibenzyloxy or bis(diphenylmethyloxy) de-
rivatives were better suited for sign interconversion of the
optical properties. In contrast, the hydroxy group(s)
lacked both optical signals and durability.
(5) (a) Minatti, A.; Doetz, K. H. Tetrahedron: Asymmetry 2005, 16, 3256–
3267. (b) Tsubaki, K.; Miura, M.; Morikawa, H.; Tanaka, H.; Kawabata, T.;
Furuta, T.; Tanaka, K.; Fuji, K. J. Am. Chem. Soc. 2003, 125, 16200–162001.
(c) Higuchi, H.; Ohta, E.; Kawai, H.; Fujiwara, K.; Tsuji, T.; Suzuki, T.
J. Org. Chem. 2003, 68, 6605–6610.
(6) (a) Murguly, E.; Norsten, T. B.; Brand, N. R. Angew. Chem., Int.
Ed. 2001, 40, 1752. (b) Feringa, B. L. Molecular Switches; Wiley-VCH:
Weinheim, Germany, 2001. (c) Kawata, Y.; Kawata, S. Chem. Rev. 2000,
100, 1777. (d) Yokoyama, Y.; Ubukata, T.; Saito, M. JP Patent 224834, 2008.
(e) Wada, H.; Nishino, K. JP Patent 50317, 2003.
(7) (a) Okumura, T.; Tani, Y.; Miyake, K.; Yokoyama, Y. J. Org. Chem.
2007, 72, 1634–1638. (b) Wang, Z. Y.; Todd, E. K.; Meng, X. S.; Gao, J. P. J.
Am. Chem. Soc. 2005, 127, 11552–11553. (c) Wigglesworth, T. J.; Sud, D.;
Norsten, T. B.; Lekhi, V. S.; Branda, N. R. J. Am. Chem. Soc. 2005, 127,
7272–7273. (d) Irie, M. Chem. Rev. 2000, 100, 1685-1716 and references
cited therein.
Axially chiral binaphthyls have a wide, flexible asym-
metric field. Hence, they have been extensively used in
catalytic asymmetric syntheses,1 specific molecular recogni-
tion,2 and helical twisting of liquid crystallines,3 and have
(1) (a) Fraile, J. M.; Garcia, J. I.; Mayoral, J. A. Chem. Rev. 2009, 109,
360–417. (b) Yamamoto, H.; Abel, J. P. J. Am. Chem. Soc. 2008, 130,
10521–10523. (c) Rabalakos, C.; Wulff, W. D. J. Am. Chem. Soc. 2008, 130,
13524–13525. (d) Brunel, J. M. Chem. Rev. 2005, 105, 857–898.
(2) (a) Ema, T.; Tanida, D.; Hamada, K.; Sakai, T. J. Org. Chem. 2008,
73, 9129–9132. (b) Park, H.; Nandhakumar, R.; Hong, J.; Ham, S.; Chin, J.;
Kim, K. M. Chem.;Eur. J. 2008, 14, 9935–9942. (c) Wang, Q.; Chen, X.;
Tao, L.; Wang, L.; Xiao, D.; Yu, X.-Q.; Pu, L. J. Org. Chem. 2007, 72,
97–101.
(3) (a) Yoshizawa, A.; Kobayashi, K.; Sato, M. Chem. Commun. 2007,
257–259. (b) Akagi, K.; Guo, S.; Mori, T.; Goh, M.; Piao, G.; Kyotan, M. J.
Am. Chem. Soc. 2005, 127, 14647–14654.
(4) (a) Kamberaj, H.; Low, R. J.; Neal, M. P. Mol. Phys. 2006, 104,
335–357. (b) Kranz, M.; Clark, T.; Schleyer, P. v. R. J. Org. Chem. 1993, 58,
3317–3325.
(8) Kawamoto, M.; Aoki, T.; Wada, T. Chem. Commun. 2007, 930–932.
(9) (a) Tamaoki, N.; Mathews, M. J. Am. Chem. Soc. 2008, 130, 11409–
11416. (b) Dri, C.; Peters, M. V.; Schwarz, J.; Hecht, S.; Grill, L. Nature
Nanotechnol. 2008, 3, 649–653. (c) Muraoka, T.; Kinbara, K.; Aida, T.
Nature 2006, 440, 512–515. (d) an Delden, R. A.; Mecca, T.; Rosini, C.;
Feringa, B. L. Chem.;Eur. J. 2004, 10, 61–70.
DOI: 10.1021/jo901030s
r
Published on Web 06/09/2009
J. Org. Chem. 2009, 74, 5723–5726 5723
2009 American Chemical Society

Takaishi et al.
JOCNote
FIGURE 1. (R)-1 and its photoisomerization.
SCHEME 1. Synthetic Route to (R)-4-8a
FIGURE 2. (a, b) CD spectra of (R)-4 and (R)-1. (c, d) Absorption
spectra of (R)-4 and (R)-1 (horizontal scales are common). (R)-4 after
365 nm irradiation (blue solid line), (R)-4 after 436 nm irradiation (red
solid line), (R)-1 after 365 nm irradiation (blue dotted line), and (R)-1
after 436 nm irradiation (red dotted line). Conditions: 1,4-dioxane
(1.0 ꢀ 10-5 M), 20 °C, light path length = 10 mm, irradiation wave-
length= 365nm(10 mW/cm2, 100s)and 436nm(10 mW/cm2, 100s).
aConditions: (a) K2CO3, 36%, (b) TiCl4, 91%, (c) RX, base, 35-64%.
As shown in Scheme 1, 3,30-disubstituted binaphthyl-
azobenzene diads (R)-4-8 were synthesized starting from
optically active (R)-2.10 Williamson synthesis of diol (R)-2
and dibromide 3 afforded cyclic compound (R)-4 in 36%
yield. Then debenzylation of (R)-4 with titanium tetrachlor-
ide for 5 min gave (R)-5 in high yield (91%). Using a higher
temperature, longer reaction time, or another Lewis acid
(e.g., aluminum trichloride or niobium pentachloride) resulted
in an extremely low yield (<5%) because some of the other
O-alkyl bonds of (R)-4 were also disconnected, and subse-
quent polymerization was also detected. Additionally,
generalized deprotection with palladium-activated carbon
(Pd/C) and hydrogen was unsuccessful under the reduction
of the azobenzene moiety. (R)-6-8 were prepared in mod-
erate yields by coupling (R)-5 with the appropriate halide. In
the synthesis of (R)-6 and -7, excess methyl iodide (MeI) and
R-bromodiphenylmethane were used, respectively. Although
in the synthesis of (R)-8, 1.1 equiv (based on (R)-5) of
benzyl bromide was used for preferential monobenzylation,
mixtures of desired (R)-8, unreacted (R)-5, and dibenzylated
(R)-4 were obtained; they were easily separated by gel
permeation chromatography (GPC).
FIGURE 3. 1H NMR spectra of the benzyl protons of (R)-4 after
(A) 365 nm irradiation and (B) 436 nm irradiation. Conditions:
CDCl3 (1 ꢀ 10-3 M), 300 MHz, 22 °C, irradiation 365 nm (10 mW/cm2,
500 s) and 436 nm (10 mW/cm2, 500 s) in an NMR test tube (φ= 5 mm).
cantly, (2) a positive-negative-reversible region was observed
on the long-wavelength side (around 450-500 nm) for the
trans form, which unlike (R)-1, exhibited a positive Cotton
effect, and (3) the Cotton effect at a short wavelength
1
(around 230 nm), which was attributed mainly to the Bb
transition moment and reflected twisting of two naphthalene
rings,12 did not resemble that of (R)-1.13 Meca, et al. have
calculated the racemization barriers of select 2,20-dis-
ubsituted-1,10-binaphthyl derivatives; all the values are
greater than 38 kcal/mol. Hence, racemizations of these
binaphthyls are atypical.14 Racemization or epimerization
Cis-trans isomerization11 was confirmed by the change in
absorption near 360 nm, which indicated a π-π* transition of
the trans form. On the whole, the absorption spectra of (R)-4
and (R)-1 were similar (Figure 2). On the other hand, the CD
spectra differed; the CD spectral shape of (R)-4 was atypical
in three ways: (1) the cis and trans forms differed signifi-
(12) Harada, N.; Nakanishi, K. Acc. Chem. Res. 1972, 5, 257–263.
(13) Although Bari et al. used calculations to determine that the Cotton
effect of the CD was inverted at a naphthalene dihedral angle of 110°, no
exceptions have yet to be reported for this empirical rule: (a) Bari, L. D.;
Pescitelli, G.; Marchetti, F.; Salvadori, P. J. Am. Chem. Soc. 2000, 122, 6395–
6398. (b) Bari, L. D.; Pescitelli, G.; Salvadori, P. J. Am. Chem. Soc. 1999, 121,
7998–8004.
(10) Tsubaki, K.; Morikawa, H.; Tanaka, H.; Fuji, K. Tetrahedron:
Asymmetry 2003, 14, 1393–1396.
(11) ΔHq and ΔSq for cis-trans isomerization of (R)-4 and 7 have been
experimentally determined to be 22 kcal/mol and -13 cal/(mol K) for (R)-4
3
and 20 kcal/mol and -17 cal/(mol K) for (R)-7.
(14) Meca, L.; Reha, D.; Havlas, Z. J. Org. Chem. 2003, 68, 5677–5680.
3
5724 J. Org. Chem. Vol. 74, No. 15, 2009

Takaishi et al.
JOCNote
TABLE 1. CD Data and Cis-Trans Ratio of (R)-4-8a
CDλ after 365 nm
irradiation (nm)
Δε after 365 nm cis:trans after 365 nm CDλ after 436 nm
Δε after 436 nm cis:trans after 436 nm
irradiation
entry
1
compd
irradiation
irradiationb
71:29
irradiation (nm)
irradiationb
31:69
(R)-4
235.0
434.0
230.5
436.5
233.5
436.5
234.0
435.5
233.0
435.0
+50.7
-19.3
+50.0
-16.7
+53.3
-14.5
+37.0
-19.8
+48.2
-19.4
237.0
482.0
232.0
426.5
234.5
485.5
238.0
486.5
235.0
488.5
+119.3
+14.4
+51.7
-10.5
+87.3
+7.2
2
3
4
5
(R)-5
(R)-6
(R)-7
(R)-8
65:35
72:28
69:31
68:32
31:69
22:78
35:65
29:71
+93.3
+14.2
+86.7
+9.4
aConditions: 1,4-dioxane (1.0ꢀ10-5 M), 20 °C, light path length=10 mm, irradiation wavelength = 365 nm (10 mW/cm2, 100 s) and 436 nm
(10 mW/cm2, 100 s). b Determined by HPLC (column, Inertsil SIL 100-A (GL science, Inc.); eluent, CHCl3:n-hexane = 1:1 for (R)-4, 6, 7, 8, CHCl3 for (R)-5).
TABLE 2. [R]D and [Φ]D of (R)-4-8 after Photoirradiationa
[R]D ([Φ]Db) after 365 nm [R]D ([Φ]D) after 436 nm
entry compd
irradiation (deg)
irradiation (deg)
1
2
3
4
5
(R)-4
(R)-5
(R)-6
(R)-7
(R)-8
-314 (-2500)
-427 (-2600)
-370 (-2400)
-252 (-2400)
-415 (-2900)
+642 (+5100)
-490 (-3000)
+248 (+1600)
+278 (+2600)
-285 (-2000)
aConditions: chloroform, c = 0.10g/dL, 20 °C, light path length = 10
cm, irradiation wavelength = 365 nm (10 mW/cm2, 500 s) and 436 nm
(10 mW/cm2, 500 s). b [Φ]D = [R]D/100 ꢀ (molecular weight).
does not have to be aromatic because the CD movement of
(R)-6, which possesses methoxy groups, was similar to that of
(R)-4, but with a smaller variation. Thus, we speculated that
although bulkier is not necessarily better, the steric effect is
important. Dihydroxy groups were too small to exibit unique
properties (Figure 4).16 This photoisomerization can be
cycled numerous times without decomposition by using
(R)-4, -6, and -7, but (R)-5 and -8, which possess hydroxy
group(s), degraded slightly after 20 irradiation cycles.
Table 2 shows the specific optical rotations ([R]D) and
molar optical rotations ([Φ]D) of (R)-4-8 after photoirra-
diation until the values were constant. Absorption at the
sodium D-line (589 nm) did not occur despite analysis at high
concentrations and long path lengths. The cis-trans ratios
were the same as those shown in Table 1. The absolute values
of [R]D were lower than those of helicenes and other chiral
metallic compounds,17 but were greater than general axially
chiral compounds. Furthermore, [R]D of (R)-4, -6, and -7
resulted in a sign inversion; (R)-4 showed the largest change
(ca. 1000°). The absolute values of [R]D of compound (R)-7
remained nearly constant, but the sign was reversed. From
the result of (R)-8, both 3- and 30-substitents, which are
bulkier than the hydroxy group, are necessary to realize a
sufficient sign inversion. [Φ]D, which is the molecular weight-
adjusted value, of (R)-4-8 after 365 nm irradiation was
similar. Thus, the conformation around the azobenzene
moiety about the cis form has a negligible difference, and
consequently, the 3,30-substituent is key to the conformation
of the trans form.
FIGURE 4. (a) CD spectra of (R)-5 and (b) absorption spectra of
(R)-5; after 365 nm irradiation (blue line) and after 436 nm irradia-
tion (red line). Conditions: 1,4-dioxane (1.0 ꢀ 10-5 M), 20 °C, light
path length= 10mm, irradiation wavelength = 365nm(10mW/cm2,
100 s) and 436 nm (10 mW/cm2, 100 s).
required strong acids or metallic catalysts.15 Therefore, (R)-4
maintains its original configuration. As a precautionary
measure, (R)-4 was derivatized to a known compound to
verify that inversion did not occur (see the Supporting
Information). Hence, the approximate inversion of the Cot-
ton effect near 230 nm likely caused the defferences between
the dihedral angle of naphthalene rings or the wavelength of
maximum absorption for (R)-4 and that of (R)-1.
The chemical shift and signal splitting of the benzyl
protons far from the light-driven part of (R)-4 in the
1H NMR vastly differed between the cis and trans forms
because the signal of (R,cis)-4 appeared as a singlet, whereas
that of (R,trans)-4 was an AB quartet with Δδ of 0.3 ppm
(Figure 3). Similar to the examples above, the 3,30-dibenzyl
groups played an important role in the entire conformation
of these compounds.
Table 1 shows the CD data and isomer ratio at the points
of (R)-4-8. The three tendencies described above were true
for (R)-6-8, but the intensities varied. The 3-substituent
(16) Because the CD spectral shape of (R)-5 in methanol and that in 1,4-
dioxane are the same, it is likely that hydrogen bonds do not contribute much
to its conformation.
(17) (a) Graule, S.; Rudolph, M.; Vanthuyne, N.; Autschbach, J.;
ꢀ
(15) Tsubaki, K.; Tanaka, H.; Takaishi, K.; Miura, M.; Morikawa, H.;
Furuta, T.; Tanaka, K.; Fuji, K.; Sasamori, T.; Tokitoh, N.; Kawabata, T.
J. Org. Chem. 2006, 71, 6579–6587.
Roussel, C.; Crassous, J.; Reau, R. J. Am. Chem. Soc. 2009, 131, 3183–
3185. (b) Katzenelson, O.; Edelstein, J.; Avnir, D. Terahedron: Asymmetry
2000, 11, 2695–2704.
J. Org. Chem. Vol. 74, No. 15, 2009 5725

Takaishi et al.
JOCNote
In summary, we synthesized a series of 3,30-disubstituted-
1,10-binaphthyls with an azobenzene moiety, and the varia-
tion of their CD spectra and optical rotations were detected
before and after photoisomerization. The optical properties
of the axially chiral compounds changed greatly. Moreover,
novel, reversible positive-negative signals were detected.
This result may be useful in designing molecular switches,
binaphthyl-type catalysts, and their leads. Studies on the
substituent effect are currently underway.
Synthesis of (R)-5. A suspension of (R)-4 (400 mg, 0.504mmol)
and TiCl4 (111 μL,1.01 mmol, 2.0 equiv) in CH2Cl2 (20 mL) was
stirred for 5 min at 0 °C. Then the reaction mixture was poured
into a mixed solvent of chloroform, ethyl acetate, and water. The
organic layer was separated and washed successively with water
(twice) and brine. After drying over sodium sulfate, the solvent
was evaporated in vacuo to give a residue. The residue was
purified by column chromatography (SiO2; n-hexane/chloro-
form/ethyl acetate=2/3/1) to afford (R)-5 (281 mg, 0.459 mmol,
91%): red amorphous;IR (KBr) 3374, 1439, 1283, 1250, 752 cm-1
;
1H NMR (400 MHz, CDCl3) δ 1.4-1.9 (m, 4H), 3.5-4.0 (m, 8H),
6.54 (br s, 2H), 6.7-7.5 (m, 16H), 7.62, 7.73 (two d, J = 8.4,
8.4 Hz, 2H); 13C NMR (75 MHz, CDCl3) δ 29.0, 29.5, 67.2, 67.4,
69.8, 71.1, 111.1, 111.5, 114.4, 115.4, 119.6, 120.7, 120.8, 121.1,
123.6, 123.9, 124.5, 124.7, 125.2, 125.3, 125.7, 125.8, 126.6, 128.2,
128.5, 128.6, 129.2, 130.9, 131.6, 143.4, 144.9, 145.2, 145.7, 147.9,
148.3, 148.3, 149.0, 153.2; HRMS (ESI+) calcd for C38H32N2O6-
Na (M + Na)+ 635.2158, found 635.2147; LR MS (FAB+) 613.2
(M + H)+, 635.2 (M + Na)+.
Experimental Section
Synthesis of (R)-4. A suspension of (R)-2 (2.00 g, 4.02 mmol),
K2CO3 (1.22 g, 8.83 mmol, 2.2 equiv) and 3 (1.92 g, 4.02 mmol,
1.0 equiv) in DMF (300 mL) was stirred for 12 h at 50 °C. Then
the reaction mixture was poured into a mixed solvent of chloro-
form and water. The organic layer was separated and washed
successively with 0.1 N hydrochloric acid solution, water
(twice), and brine. After drying over sodium sulfate, the solvent
was evaporated in vacuo to give a residue, which was purified by
column chromatography (SiO2; n-hexane/chloroform = 1/9)
and GPC to afford (R)-4 (1.15 g, 1.45 mmol, 36%): red
Acknowledgment. This work was supported in part by
funding for a special doctoral fellow (61101) to K.T. from
RIKEN. We are indebted to members of the molecular
characterization team at RIKEN for the NMR, MS spectral,
R, and CD measurements. We also acknowledge Professor
Takeo Kawabata and Mr. Daisuke Sue of Kyoto University
for furnishing (R)-2.
amorphous; IR (KBr) 2959, 1439, 1281, 1250, 1017, 748 cm-1
;
1H NMR (400 MHz, CDCl3) δ 1.2-2.1 (m, 4H), 3.2-4.3 (m,
8H), 4.88, 5.19 (ABq, νAB = 61.7 Hz, JAB = 11.2 Hz, s, 4H),
6.4-7.5 (m, 26H), 7.73, 7.78 (two d, J = 8.4, 8.4 Hz, 2H); 13
C
NMR (75 MHz, CDCl3) δ 29.5, 29.7, 64.8, 65.1, 68.9, 70.1, 70.5,
70.6, 108.4, 108.4, 113.1, 113.9, 119.4, 120.2, 124.0, 124.2, 125.0,
125.6, 125.7, 125.2, 126.3, 127.8, 127.9, 128.2, 128.5, 128.8,
128.8, 130.5, 130.6, 130.8, 136.4, 136.5, 142.6, 143.7, 145.9,
146.4, 151.1, 151.3, 152.8 (some peaks overlapped); HRMS
(ESI+) calcd for C52H45N2O6 (M + H)+ 793.3278, found
793.3303; LR MS (FAB+) 793.4 (M + H)+.
Supporting Information Available: CD and UV-vis spectra
of 4-8, including select overlapping displays, full experimental
details, and characterization data of all new compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
5726 J. Org. Chem. Vol. 74, No. 15, 2009