Optically Active Oxo(phthalocyaninato)vanadium(IV)
FULL PAPER
collected, washed with water, and after drying it was purified by silica gel
chromatography with chloroform as eluent, then recrystallized from a
chloroform/hexane mixture, to give 263 mg (41.6%) of the dark blue
shiny solid, H2Pc. Elemental analysis (%): calcd for C56H66N8O4: C 73.50,
H 7.27, N 12.24; found: C 73.86, H 7.29, N 12.15; 1H NMR (CDCl3): d=
8.88 (d, 4H, J=6.85 Hz), 7.99 (t, 4H, J=7.33 Hz), 7.54 (d, 4H, J=
7.33 Hz), 4.66 (t, 8H, J=6.36 Hz), 2.46 (m, 8H), 2.08 (m, 8H), 1.72 (m,
8H), 1.57 (m, 8H), 1.06 (t, 12H, J=7.33 Hz), ꢀ0.66 ppm (s, 2H); UV/Vis
(CHCl3): lmax (10ꢀ4e)=728 (18.2), 694 (16.3), 678 (5.00), 628 (3.56), 353
(4.76), 322 nm (6.72 molꢀ1 dm3 cmꢀ1). The above NMR data clearly indi-
cate that the hexyloxy groups are attached to the phthalocyanine periph-
ery in a single-handed rotation.[5,6]
chiral SiPcs coordinated by axial OH and hexyloxy ligands
did not exhibit a meaningful CD sign.
Comparison of the experimental data with those calculat-
ed by using TDDFT indicates that the first-eluted VOPc1,
which shows a negative CD sign associated with the Q00
band, is a right-handed conformer (VOPc-a). Thus, our pres-
ent data suggest that, in alkoxy-substituted Pcs with a
domed p skeleton, the right-handed conformer shows a neg-
ative CD sign in the Q band and a positive CD sign in the
n–p* transition band. The left-handed conformer exhibits
the opposite CD sign in both the Q and n–p* bands. It has
been further demonstrated that the calculated CD sign in
the Q region is sensitive to the relative orientation of the
peripheral alkoxy groups. The relationship between the Q-
band CD and the dihedral angle (q) was sigmoidal for the
vanadium complexes, and the averaged CD sign was found
to be negative for the right-handed conformer. In contrast,
the Q-band CD of SiPcs was predicted to be almost silent
due to the cancellation of two Q-band CD signals of oppo-
site sign, regardless of the relative orientation of the axial
group. It should be noted that the Q-band CD sign would
be inverted even in a chiral Pc with single-handed orienta-
tion when the position of peripheral substituents is flipped
with respect to the Pc plane. (In the present system, howev-
er, it appears energetically highly unfavorable for substitu-
ents to tilt significantly from the macrocyclic plane (see Fig-
ures 7 and 9).) Although, at present, it might be difficult to
confirm the above prediction experimentally, we believe
that the correlation found between the CD sign in the Q
region and the relative orientation of the peripheral groups
has implications not only in the present system but also in
many different chiral Pc and porphyrin systems.
1,8,15,22-Tetrahexyloxyphthalocyaninatovanadium(IV) oxide, VOPc:
A
mixture of the single-handed H2Pc (40.7 mg, 0.045 mmol) described
above, vanadium(V) pentoxide (99 mg, 5.44 mmol), and dried urea
(547 mg, 9.1 mmol) was stirred at 1408C for ꢂ1 h, and after addition of
DMF (1 mL) the reaction was continued for further 17.5 h. After remov-
al of the solvent by an evaporator, the residue was washed with water,
dried, and purified by column chromatography over alumina with carbon
tetrachloride/methanol (10:1 v/v) as the eluting solvent. Recrystallization
from chloroform/methanol gave the desired VOPc (29.2 mg (66%). Ele-
mental analysis (%): calcd for C56H64N8O5V: C 68.62, H 6.58, N 11.43;
found: C 69.98, H 6.75, N 11.73; UV/Vis (CHCl3): lmax (10ꢀ4e): 733
(24.4), 700 (3.82), 657 (4.39), 434 (1.48), 353 (4.96), 324 nm
(5.91 molꢀ1 dm3 cmꢀ1).
1,8,15,22-Tetrahexyloxyphthalocyaninatosilicon hydroxide hexyl oxide,
SiPc. In freshly distilled dry quinoline (9.3 mL), 4-n-hexyloxy-1H-isoin-
dole-1,3ACHTUNTRGNEU(GN 2H)-diimine (1.51 g, 6.16 mmol), prepared from 3-hexyloxyph-
thalonitrile and ammonia gas by a general procedure,[25] and silicon tetra-
chloride (2.3 mL, 20 mmol) were stirred at 1808C under nitrogen for
18 h. After the system had been cooled, chloroform and water were
added to it, the chloroform layer was collected, and the solvents were re-
moved as much as possible by an evaporator. The residue was subjected
to alumina chromatography (Act. IV) with first hexane (to remove the
residual quinoline) and then chloroform as the eluting solvents. The mix-
tures of Si(OH)2Pc isomers obtained were separated on an HPLC
(Ultron VX-ODS) column using methanol/chloroform (7:2 v/v) as the
eluting solvent (flow rate=9 mLminꢀ1, detection wavelengths=220, 350,
600, 700 nm).
A portion of the separated 1,8,15,22-tetrahexyloxyphthalocyaninatosili-
con dihydroxide, Si(OH)2Pc (6 mg, 6.15ꢁ10ꢀ3 mmol) and hexanol
(400 mL, 3.20ꢁ10ꢀ3 mmol) were heated at reflux in toluene (75 mL) for
1.5 h in the presence of dried calcium dichloride (2.5 g).[26] The solvent
was removed by a vacuum evaporator and the residue was placed on an
alumina column (Act. V) with first hexane and then chloroform as elu-
ents, and further on an alumina TLC plate with chloroform as eluent, to
give the desired SiPc (ca. 3.4 mg, 52% from Si(OH)2Pc) as a green
powder; 1H NMR (CDCl3): d=9.16 (d, 4H), 8.17 (t, 4H), 7.70 (d, 4H),
4.75 (m, 8H), 2.52 (m, 8H), 2.15 (m, 8H), 1.75 (m. 8H), 1.21 (m, 8H),
1.05 (t, 12H), 0.27 (m, 2H), 0.23 (t, 3H), ꢀ0.39 (m, 2H), ꢀ1.32 (m, 2H),
ꢀ1.57 (m, 2H), ꢀ1.94 (m, 2H), ꢀ5.50 ppm (s, 1H); UV/Vis (CHCl3): lmax
(10ꢀ4e) 712 (18.0), 678 (2.73), 636 (3.03), 322 nm (4.38 molꢀ1 dm3 cmꢀ1);
FABMS [M]+: calcd for C62H78N8O6Si: 1058.6; found: 1058.6. These
NMR data confirm that SiPc is indeed a Pc with a single-handed rota-
tion.
Experimental Section
General: Electronic absorption spectra were measured by using a Hitachi
U-3410 spectrophotometer. CD and MCD measurements were per-
formed by using a JASCO J-720 spectrodichrometer with a chloroform
solution (10ꢀ6 molLꢀ1), without and with a JASCO electromagnet that
produced magnetic fields of up to 1.09 T. CD curves shown here were ob-
tained by accumulation of more than 8 h.
Synthesis
3-Hexyloxyphthalonitrile: In a minor modification of Leznoffꢂs proce-
dure,[5,24] 3-nitrophthalonitrile (2.0 g, 11.6 mmol), hexanol (2.5 mL,
20.0 mmol), and anhydrous potassium carbonate (1.2 g, 8.7 mmol) were
stirred at room temperature under a nitrogen atmosphere in dry DMF
(12 mL). Potassium carbonate (1.2 g in each portion) was added to the
system after 24 and 48 h and the stirring was continued for three days in
total. The crude reaction mixture was poured into water (400 mL). The
precipitate was collected by filtration, washed thoroughly with water, and
dried. The crude product was purified by silica gel column chromatogra-
phy with chloroform as the eluting solvent. Off-white needles of 3-hex-
Separation of VOPc and SiPc into two enantiomers: Single-handed VOPc
and SiPc were separated into two enantiomers on a reversed-phase silica
gel HPLC column (0.46fꢁ25 cm) coated with cellulose tris(3,5-dimethyl-
phenylcarbamate) by using mixtures of hexane/chloroform (6:1 v/v) for
VOPc (flow rate=1 mLminꢀ1
, detection wavelength=350 nm) and
carbon tetrachloride/chloroform (6:1 v/v) for SiPc (flow rate=
0.7 mLminꢀ1, detection wavelengths=210, 350, 600, 700 nm). In the case
of VOPc, two fractions were eluted at 9 min (VOPc-1) and 19 min
(VOPc-2) in areal ratio 49:51 (Figure 2a); for SiPc two fractions were
eluted at 7 min (SiPc-1) and 9 min (SiPc-2) in areal ratio 49:51 (Fig-
ure 2b). The electronic absorption spectra of VOPc-1 and VOPc-2 were
exactly the same, and for SiPc-1 and SiPc-2 not only the electronic spec-
ACHTUNGTRENNUNGyloxyphthalonitrile (2.00 g, 76%), mp 110–1118C, were recrystallized
from a chloroform/hexane mixture.[6]
1,8,15,22-Tetrahexyloxyphthalocyanine, H2Pc:[6] Lithium (1.1 g, 158 mmol)
was dissolved in anhydrous hexanol under reflux and a nitrogen atmos-
phere. After cooling, 3-hexyloxyphthalonitrile (630 mg, 2.76 mmol) was
added and the mixture was stirred at room temperature for three days.[5]
The solvent was removed in a rotatory evaporator and the residue was
1
tra but also the H NMR spectra were found to be the same.
Chem. Eur. J. 2009, 15, 10173 – 10181
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10179