Effect of Peripheral Substitution on the Electronic Absorption and Fluorescence Spectra of Metal-Free and Zinc Phthalocyanines

Article abstract of DOI:10.1002/chem.200304834

The effect of substituents on the position and intensity of the electronic absorption and fluorescence spectra of phthalocyanines (Pcs) was examined for 35 Pc compounds. When electron-releasing groups are bound to four β-benzo positions of the Pc skeleton

Full text of DOI:10.1002/chem.200304834

FULL PAPER
Effect of Peripheral Substitution on the Electronic Absorption and
Fluorescence Spectra of Metal-Free and Zinc Phthalocyanines
Nagao Kobayashi,*[a] Hiroshi Ogata,[a] Naokazu Nonaka,[b] and Eugene A. Luk×yanets[c]
Abstract: The effect of substituents on
the position and intensity of the elec-
tronic absorption and fluorescence spec-
tra of phthalocyanines (Pcs) was exam-
ined for 35 Pc compounds. When elec-
tron-releasing groups are bound to four
a-benzo positions of the Pc skeleton, the
B and Q bands shift to longer wave-
length. Relative to this shift, the effect of
introducing the same electron-releasing
groups at the other four a positions
amounts to about 1.6 2.0. Although the
effect is not always clearly seen, intro-
duction of electron-releasing groups in
the b-benzo positions of the Pc skeleton
generally shifts the Q band to shorter
wavelength. The effect of electron-with-
drawing groups is exactly the opposite
with respect to the a and b positions.
These effects can be reasonably ex-
plained by considering the magnitude
of the atomic orbital coefficients of the
carbon atoms derived from molecular
orbital (MO) calculations. In addition,
the following intriguing phenomena
were observed in the experiments, al-
though not all were explained theoret-
ically: 1) the splitting of the Q band of
metal-free Pcs decreases with increasing
wavelength of the Q band, 2) the ring
currents of Pcs with Q bands at longer
wavelength are generally smaller, and
3) the absorption coefficients of the Q
band of Pc compounds with 16-electron-
releasing substituents are larger than
those of the corresponding tetra- and
octasubstituted Pcs by several tens of
percent. 4) Our PPP calculations sug-
gested that the absorption coefficient of
the Q band of Pcs with more strongly
electron releasing substituents is larger.
5) The second HOMO of the Pcs with
the Q band at longer wavelength has b1u
symmetry, as opposed to the a2u symme-
try of normal Pcs. 6) Pcs showing S1
emission maxima at wavelengths longer
than about 740 nm generally have quan-
tum yields of less than 0.1.
Keywords: fluorescence spectrosco-
py ¥ molecular orbital calculations ¥
phthalocyanines ¥ substituent effects
¥ UV/Vis spectroscopy
display devices,[8] low-dimensional metals,[9] gas sensors,[10]
liquid crystals,[11] nonlinear optics,[12] photodynamic reagents
for cancer therapy and other medical applications,[13] and
electrocatalytic reagents.[14] In a number of these applications,
the wavelength of the major p p* transitions in the UV/Vis
region is of critical importance. The position and bandwidth of
absorption bands of Pcs can be adjusted in two ways. One is by
changing size and symmetry of the p-conjugated system of
Pcs,[15] and the other is by varying the central metal atom and
the type, number, and positions of peripheral substituents.[16]
It is known, for example, that the Q band can be shifted with
approximate additivity when the same substituents are
introduced at the same position of each benzene ring of the
Pcs.[17] However, the effect of introducing a plurality of
substituents on each benzene unit has not been examined
systematically to date, except for our preliminary communi-
cation.[18] Here we report regiospecific and nonlinear effects
of substituents on both the energy and intensity of the
electronic absorption and fluorescence emission spectra of
metal-free and zinc Pcs, as well as their effect on the splitting
of the Q00 band of the metal-free Pcs. We examined 35 Pcs
with electron-releasing alkoxyl or alkyl(phenyl)thio groups
and electron-withdrawing nitro or sulfonyl groups. Some of
Introduction
Phthalocyanines (Pcs) have traditionally found use as dyes
and pigments, as catalysts for desulfurization processes in the
oil industry, and more recently as photoconducting agents in
photocopiers, deodorants, germicides, optical computer read/
write disks, and as promotion or retardation films for plant
growth in greenhouses, because of their easy synthesis, high
stability, and the presence of p p* transitions in the near-UV/
visible/near-IR region.[1] In recent decades, there has been
renewed interest in the use of Pcs in a variety of high-tech
fields, including semiconductor devices,[2] photovoltaic solar
cells,[3] electrophotography,[4] rectifying devices,[5] molecular
electronics,[6] Langmuir Blodgett films,[7] electrochromism in
[a] Prof. Dr. Dr. N. Kobayashi, H. Ogata
Department of Chemistry, Graduate School of Science
Tohoku University, Sendai 980-8578 (Japan)
Fax : (‡81)-22-217-7719
[b] N. Nonaka
Pharmaceutical Institute, Tohoku University
Sendai 980-8578 (Japan)
[c] Prof. Dr. E. A. Luk×yanets
Organic Intermediates and Dyes Institute
1/4 B. Sadovaya Str., 103787 Moscow (Russia)
Chem. Eur. J. 2003, 9, 5123 5134
DOI: 10.1002/chem.200304834
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5123

N. Kobayashi et al.
FULL PAPER
the experimental observations were reproduced by MO
calculations within the frameworkof the Pariser Parr
Pople (PPP) approximation.
clearly seen). As can be seen in the figure, for Pcs with the
same number of alkoxyl groups, substitution at the a positions
results in a Q band absorption at longer wavelength (lower
energy). An increase in the number of substituents from four
to eight results in opposite behavior for a and b substitution.
Substitution at a positions further shifts the Q band to longer
wavelength, and substitution at b positions to shorter wave-
length. The latter shift suggests that substitution at the b
positions by electron-releasing groups shifts the band to
shorter wavelength, although this shift is smaller than the
longer wavelength shift due to substitution at the a positions.
The shift to shorter wavelength due to b substitution appears
real, however, since the Q band of 16(OCH2CF3)H2Pc is at
shorter wavelength than that of 8a(OBu)H2Pc. In accordance
with previous studies,[16] the shift in the B band region is not
clear, although there is a trend that substitution at the a
positions broadens the band.[20, 21] Interestingly, the shapes of
the B bands of 4b(OBu)- and 8b(OBu)H2Pc are similar, and
this also holds true for 4a(OBu)- and 8a(OBu)H2Pc. The
H2Pcs substituted at the b positions have three peaks in the B
band region, while the a-sub-
Results and Discussion
The structures and abbreviations of Pcs used in this study are
shown in Figure 1. In the abbreviations, the first number
indicates the number of substituents introduced on the Pc
periphery, a or b indicates their positions, and substituent is
identified in parentheses. For example, 4a(SBu)ZnPc denotes
a ZnPc with four butylthio groups in the a positions, while
4a,4b(NO2)H2Pc is a metal-free Pc with four nitro groups
each in the a and b positions.
Absorption and magnetic circular dichroism spectra: Figure 2
shows the absorption and magnetic circular dichroism (MCD)
spectra of alkoxy-substituted H2Pcs (the B[19] and Q band
regions are shown separately so that the peakpositions can be
stituted H2Pcs have the stron-
gest peaks at 315 330 nm.
Comparison of the spectra of
8a(OBu)-, 8b(OBu)-, and
16(OCH2CF3)H2Pc
further
shows that the absorption coef-
ficient of 16(OCH2CF3)H2Pc is
roughly 1.2 times larger than
those of the other two. Further-
more, comparison of the spec-
tra of 4a- and 4b(OBu)H2Pc
suggests that the intensity of the
Q band of the former is slightly
higher.
The MCD spectra are shown
for the three representative
species. In the Q band region,
8a(OBu)- and 16(OCH2CF3)-
H2Pc display dispersion-type,
Faraday A-term-like curves.
Since the compounds have
D2h symmetry, this is a pseudo-
Faraday A term produced by
superimposition of closely ly-
ing Faraday B terms of oppo-
site signs.[22] However, care-
ful analysis of this apparent
dispersion-type
Q
band of
8b(OBu)H2Pc perhaps experi-
mentally substantiates that the
splitting of the
this species is larger than
those of 8a(OBu)- and
Q band of
16(OCH2CF3)H2Pc. The shape
of the B MCD indicates that the
main B band of these three
species lies at about 330
360 nm.
Figure 1. Structures and abbreviations of phthalocyanines in this study.
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5123 5134
5124

Phthalocyanines
5123 5134
Figure 3. Electronic absorption and MCD spectra of 8a(SBu)ZnPc (solid
lines), 4a(SBu)ZnPc (broken lines), and 4b(SBu)ZnPc (dotted broken
lines) in THF in the Q (top) and B (Soret) band (bottom) regions.
Figure 2. Electronic absorption and MCD spectra of some metal-free Pcs
in THF in the Q (top) and B (Soret) band (bottom) regions.
Figure 3 shows the electronic absorption and MCD spectra
of three butylthio-substituted ZnPcs as representative spectra
of metalated Pcs. As for the metal-free species in Figure 2, the
Q band of 4b(SBu)-, 4a(SBu)- and 8a(SBu)ZnPc shifts to
longer wavelength, while, conversely, the most intense peakin
the B band region of these species shifts to shorter wavelength
in this order; this suggests that the configuration interaction
between the B and Q states increases in this order (unfortu-
nately, attempts to prepare 16(OCH2CF3)ZnPc were unsuc-
cessful).[23] In the MCD spectra, since all compounds have D4h
symmetry, at least one Faraday A term is expected for the Q
and B bands, respectively, of each compound. In the Q band
region, all ZnPcs show dipersion-type A terms corresponding
to the Q00 band. In the B band region, the most intense A
terms of 4a- and 4b(SBu)ZnPc are very clear and appear at
340 and 360 nm, respectively. However, the A term of
8a(SBu)ZnPc is small (ca. 290 300 nm region).
Figure 4. Correlation between the splitting of the Q band, Qx00 À Qy00, and
the mid-energy of the Q band (Qx00 ‡ Qy00)/2.
splitting generally decreases with increasing wavelength of the
Q band, and H2Pcs (and metal-free naphthalocyanine) with
the Q band at around 800 nm do not show apparent splitting.
Experimental electronic absorption data of H2Pcs and
ZnPcs are summarized in Tables 1 and 2, respectively. Note
that the absorption coefficients of unsubstituted H2Pc and
8a(OBu)ZnPc may be not correct due to low solubility of the
former and instability of the latter in solution. By subtracting
the Q band energy of unsubstituted Pcs from those of
substituted ones, we can calculate the effect of substituents
quantitatively (in the case of metal-free species, (Qx ‡ Qy)/2
indicates the Q band position). For electron-releasing groups,
While collecting absorption data of H2Pcs substituted with
various groups, we noticed a wavelength-dependent splitting
of the Q band. These data are summarized in Figure 4. The
Chem. Eur. J. 2003, 9, 5123 5134
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5125

N. Kobayashi et al.
FULL PAPER
Table 1. Electronic absorption data of metal-free phthalocyanines in THF.
[a]
Qx [cmÀ1
l [nm]
]
Qy [cmÀ1
l [nm]
]
10À4 e [LmolÀ1 cmÀ1
]
Qx À Qy [103 cmÀ1
]
0.5(Qx ‡ Qy) [103 cmÀ1
]
D [103 cmÀ1
]
[b]
H2Pc (reference)
14316.4
698.5
15048.9
664.5
0.73
14.68
0
electron-withdrawing substituents
[b]
4a(NO2)H2Pc
4b(NO2)H2Pc
4a,4b(NO2)H2Pc
14409.2
694.0
14035.1
712.5
14035.1
712.5
15174.5
659.0
14814.8
675.0
14814.8
675.0
0.77
14.79
14.42
14.42
0.11
À 0.26
À 0.26
[b]
0.78
[b]
0.78
electron-releasing substituents
4a(OBu)H2Pc
8a(OBu)H2Pc
4b(OBu)H2Pc
8b(OBu)H2Pc
16(OCH2CF3)H2Pc
4a(SBu)H2Pc
8a(SBu)H2Pc
4b(SBu)H2Pc
13831.3
723.0
13123.4
762.0
14222.7
703.1
14320.5
698.3
13650.0
732.6
13633.3
733.5
12430.1
804.5
14015.4
713.5
14499.1
689.7
13440.8
744.0
15008.3
666.3
15146.9
660.2
14170.3
705.7
14084.5
710.0
12430.1
804.5
14673.5
681.5
16.57
11.20
8.90
0.67
0.32
0.79
0.83
0.52
0.45
0
14.17
13.28
14.62
14.73
13.91
13.85
12.43
14.34
14.01
14.43
13.81
À 0.51
À 1.40
À 0.06
0.05
11.25
14.35
10.14
13.56
6.34
À 0.77
À 0.83
À 2.25
À 0.34
À 0.67
À 0.25
À 0.87
0.66
0.58
0.66
0.53
8b(SBu)H2Pc
13717.4
729.0
14094.4
709.5
13541
738.5
14295.9
699.5
14755.8
677.7
14074.6
710.5
4.11
4b(SPh)H2Pc
10.66
14.35
4a(SPh)4b(tBu)H2Pc
[a] Shift relative to the Q band energy of H2Pc. [b] Reliable data could not be obtained because of low solubility of the compounds.
substitution at the a positions shifts the Q band to longer
wavelength. Tetrasubstitution produces Q band shifts of
À0.51 and À0.83 kcmÀ1 for 4a(OBu)H2Pc and 4a(SBu)H2Pc,
respectively, relative to H2Pc, and shifts of À0.55 and
À0.80 kcmÀ1 for 4a(OBu)ZnPc and 4a(SBu)ZnPc, respec-
tively, relative to ZnPc. These increase to À1.40, À2.25,
À1.72, and À2.09 kcmÀ1 in 8a(OBu)H2Pc, 8a(SBu)H2Pc,
8a(OBu)ZnPc, and 8a(SBu)ZnPc, respectively. These values
indicate, for example, that the effect of a thioether group is
about 1.45 1.62 times larger than that of the alkoxyl groups,
and that the magnitude of the shift due to the second four
substituents is about 1.6 2.0 times larger than that owing to
the first four substituents. According to symmetry-adapted
perturbation theory,[15] which is a first-order perturbation
theory, the shifts in the Q band produced by increasing the
number of substituent groups from zero to four and from four
to eight are expected to be the same. The value of 1.6 2.0
therefore suggests the presence of a higher order perturbation
in the second tetrasubstitution at the a positions.
shifts the Q band to longer wavelength, and the size of the
shift approximately pararells the number of substituent
groups. Since the position of the Q band of 8a(OBu)-,
16(OCH2CF3)-, and 8b(OBu)H2Pc changes from longer to
shorter wavelength, we suggest that substitution by electron-
releasing groups at the b positions essentially moves the Q
band to shorter wavelength. The slight longer wavelength shift
due to alkylthio groups may reflect the increase of the size of
effective p system, since the 3p orbital of sulfur is known to
mix well with the p orbital of tetraazaporphyrins.[24]
The effect of electron-withdrawing groups is the opposite of
that of electron-releasing groups. As can be seen in the
comparison of 4a(NO2)- and 4b(NO2)H2Pcs and -ZnPcs,
substitution at a or b positions shifts the Q band to shorter or
longer wavelength, respectively. However, the magnitude of
the shift for substitution at b positions is not necessarily small
compared with that at a positions [cf., for example, the data of
4a- (0.15 kcmÀ1) and 4b(SO2Ph)ZnPc (À0.14 kcmÀ1)], in
contrast to the above cases of electron-releasing groups.
The substituent effect at the b positions is much smaller
than that in the a positions, and appears different for alkoxyl
and thioether groups. In the case of alkoxyl groups, the shift in
the Q band of 4b(OBu)H2Pc and 8b(OBu)H2Pc relative to the
Q band of H2Pc is less than Æ0.06 kcmÀ1, while those of
4b(SBu)H2Pc and 8b(SBu)H2Pc are À0.34 and À0.67 kcmÀ1,
respectively. Thus, substitution by thioalkyl groups slightly
Fluorescence emission and ring-current strength: Like other
metal-free and ZnPcs,[25] all compounds in this study show
fluorescence emission from the S1 state. Figure 5 displays the
correlation between the wavenumber of fluorescence max-
imum and quantum yield of 35 Pcs. The following facts
emerge from this figure: 1) Pcs showing emission peaks at
5126
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5123 5134

Phthalocyanines
5123 5134
Table 2. Electronic absorption data of zinc phthalocyanines in THF.
[a]
Q0-0 [cmÀ1
l [nm]
]
10À4 e [LmolÀ1 cmÀ1
at Qx
]
D [103cmÀ1
]
ZnPc (reference)
14914.2
670.5
28.00
0
electron-withdrawing substituents
4a(NO2)ZnPc
4b(NO2)ZnPc
14943.2
669.2
14903.1
671.0
14836.7
674.0
15060.2
664.0
14.41
0.03
À 0.01
À 0.08
0.15
[b]
[b]
4a,4b(NO2)ZnPc
4a(SO2Ph)ZnPc
4b(SO2Ph)ZnPc
9.48
17.86
22.22
14777.6
676.7
À 0.14
0.05
Figure 5. Correlation between wavenumber of fluorescence peakmax-
imum and quantum yield FF of S1 fluorescence.
4a(SO2Ph)4b(tBu)ZnPc 14961.1
668.4
between the HOMO and the LUMO, and 2) is due to the
heavy-atom effect of zinc. A relation such as that in point 3)
has not been reported to date. The data for S1 emissions of H2-
and ZnPcs are summarized in Tables 3 and 4.
electron-releasing substituents
4a(OBu)ZnPc
8a(OBu)ZnPc[c]
4b(OBu)ZnPc
8b(OBu)ZnPc
16(OCH2CF3)ZnPc
4a(SBu)ZnPc
8a(SBu)ZnPc
4b(SBu)ZnPc
8b(SBu)ZnPc
4a(SPh)ZnPc
14370.0
696.0
13190.0
758.0
14830.0
674.5
14830.0
674.5
14144.3
707.0
14112.3
708.6
12827.1
779.6
14547.6
687.4
14144.3
707.0
14150.3
706.7
14609.2
684.5
14128.3
707.8
17.26
17.58
13.35
20.42
15.26
18.30
12.45
22.88
13.91
25.81
29.21
24.88
24.20
À 0.55
À 1.72
À 0.09
À 0.09
À 0.77
À 0.80
À 2.09
À 0.37
À 0.77
À 0.76
À 0.31
À 0.79
À 0.82
Figure 6 shows the relationship between the position of the
Q
00 band and pyrrole proton signals of alkoxy- and alkylthio-
substituted H2Pcs in the 1H NMR spectra. In compounds such
as porphyrins and Pcs, pyrrole protons appear with negative
chemical shifts due to the ring current of the macrocycle.
4b(SPh)ZnPc
8b(SPh)ZnPc
4a(SPh)4b(tBu)ZnPc
14094.4
709.5
[a] Shift relative to the Q band energy of H2Pc. [b] Reliable data could not
be obtained because of low solubility of the compounds. [c] Because of
instability of the solution in THF, spectrum was recorded in pyridine.
Figure 6. Correlation between chemical shift [ppm] of the pyrrole proton
and the mid-energy of the Q band (Qx00 ‡ Qy00)/2 [cmÀ1] of metal-free Pcs.
Table 3. S1 fluorescence emission data of metal-free phthalocyanines in THF.
Excitation Fluorescence Fluorescence
wavelength peaklifetime
F
Excitation
Stokes
peak shift
l [nm]
maximum l[nm] [ns]
maximum l [nm] [103 cmÀ1
]
H2Pc (reference)
640
704.5
0.600
shorter wavelength have higher
quantum yields; 2) the quantum
yields of metal-free Pcs are gen-
erally higher, even though their
emission maxima appear at low-
er energy than those of the
corresponding ZnPcs; and 3)
Pcs showing S1 emission maxi-
ma at wavelengths longer than
about 740 nm generally have
quantum yields of less than 0.1.
Point 1) suggests that nonradia-
tive decay becomes more diffi-
cult as the energy gap increases
electron-withdrawing substituents
4a(NO2)H2Pc
4b(NO2)H2Pc
4a,4b(NO2)H2Pc
4a(OBu)H2Pc
8a(OBu)H2Pc
4b(OBu)H2Pc
8b(OBu)H2Pc
640
640
640
633
700
625
633
701.5
718.5
718.5
729.2
809
7.36
5.72
5.62
4.85
3.80
7.78
8.83
0.787
0.270
0.258
0.070
0.189
0.490
0.579
696
715
0.11
0.07
0.06
0.13
1.13
0.11
0.07
715.5
722.6
741
704.5
699.5
710
703
electron-releasing substituents
740 10.8
4.54
16(OCH2CF3)H2Pc
4a(SBu)H2Pc
8a(SBu)H2Pc
4b(SBu)H2Pc
8b(SBu)H2Pc
4b(SPh)H2Pc
633
630
722.5
630
630
640
0.0607 733.8
0.11
0.01
1.19
0.04
0.12
0.02
0.06
737
809
720
733.4
715
744
0.038
0.032
0.138
0.052
0.334
0.073
736.5
738
718
726.8
714
740.8
1.31
6.30
4.34
6.44
4.79
4a(SPh)4b(tBu)H2Pc 640
Chem. Eur. J. 2003, 9, 5123 5134
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5127

N. Kobayashi et al.
FULL PAPER
Table 4. S1 fluorescence emission data of zinc phthalocyanines in THF.
16(O- or S-alkyl)Pcs, were cal-
culated to have split and broad B
bands. In particular, this trend is
more evident for alkylthio-sub-
stituted species, for which one of
the split components is expected
to appear midway between the
Q and B bands, as is indeed
observed for 8a(SBu)ZnPc at
about 500 nm. Another impor-
tant point is that the most in-
tense B band of 8a(S-alkyl)Pc is
expected to lie at the shortest
wavelength among alkoxy- and
alkylthio-substituted Pcs. In-
deed, as seen in the absorption
spectra in Figure 3, the most
intense B band of 8a(SBu)ZnPc
lies at wavelength shorter than
300 nm. Thus, Pcs which have
the Q band at longer wave-
lengths generally have the B
band at shorter wavelengths,
and this indicates that the con-
figuration interaction in these
Excitation Fluorescence
wavelength peaklifetime
Fluorescence
F
Excitation
Stokes
peakshift
l [nm]
maximum l [nm] [ns]
maximum l [nm] [103 cmÀ1
]
ZnPc (reference)
620
686.5 4.36
0.300
electron-withdrawing substituents
4a(NO2)ZnPc
4b(NO2)ZnPc
4a,4b(NO2)ZnPc
4a(SO2Ph)ZnPc
4b(SO2Ph)ZnPc
608
678
695
698
720
686
678.5
706
4.63
3.83
3.36
4.69
4.35
3.98
3.05(81%)
4.53 (19%)
0.57
0.488 672.5
0.159 691
0.088 691
0.038 714.5
0.397 684.5
0.441 672.5
0.143 696
0.12
0.08
0.15
0.11
0.03
0.13
0.20
612.5
612.5
608
608
4a(SO2Ph)4b(tBu)ZnPc 608
4a(OBu)ZnPc
8a(OBu)ZnPc
4b(OBu)ZnPc
633
675
633
807.5
0.062 739
1.15
0.03
electron-releasing substituents
689
5.31 (88%)
1.51 (12%)
4.39
3.93
2.63
1.58
3.93
3.18
2.65
0.422 687.6
8b(OBu)ZnPc
16(OCH2CF3)ZnPc
4a(SBu)ZnPc
8a(SBu)ZnPc
4b(SBu)ZnPc
8b(SBu)ZnPc
4a(SPh)ZnPc
4b(SPh)ZnPc
8b(SPh)ZnPc
4a(SPh)4b(tBu)ZnPc
633
633
630
700
630
630
620
620
620
620
681.5
716
718.2
781
685
733
687.5
687.5
721.2
727.4
0.616 676.4
0.0745 713
0.137 708
0.036 712.5
0.386 683.5
0.119 700
0.276 686.5
0.439 684
0.090 713.2
0.224 721.8
0.11
0.06
0.20
1.23
0.03
0.64
0.02
0.08
0.16
0.11
3.93
3.35
3.05
Interestingly, as seen in this figure, H2Pcs having Q bands at
longer wavelength show lower ring currents. This type of
relationship has not been reported to date.
Pcs is stronger than that in Pcs which have the Q band at
shorter wavelengths. This type of relationship was recently
substantiated for ZnPc, ZnNc, and their intermediate com-
pounds.[28]
Molecular orbital calculations: To enhance our understanding
on the effect of substituents, we carried out molecular orbital
(MO) calculations on several alkylthio- and alkoxy-substitut-
ed (pyrrole-proton-deprotonated) H2Pc species within the
frameworkof the Pariser Parr Pople (PPP) approxima-
tions.[26] The calculated electronic absorption spectra are
shown in Figure 7, and transition energies, oscillator strengths
f, and configurations are summarized in Table 5. The exper-
imental order of the wavelengths of the Q band according to
the type of substituents is well reproduced by the calculation;
the Q band shifts to longer wavelength in the order of
unsubstituted, 8b(O-alkyl)-, 8b(S-alkyl)-, 16(O-alkyl)-, 16(S-
alkyl)-, 8a(O-alkyl)-, and 8a(S-alkyl)-Pcs. For these species,
the estimated wavelengths were 662, 685, 694, 738, 744, 770,
and 799 nm, while the experimental values of zinc Pcs in THF
were, 671, 675, 707, 707, [no datum], 758, and 780 nm. In all
cases the Q band shifts to longer wavelength on introduction
of electron-releasing groups. A comparison of the corre-
sponding alkoxy- and alkylthio-substituted species shows that
the Q band of the latter appears at longer wavelength, and its
intensity is higher than that of the former. In addition, the
intensities of the Q band of hexadecasubstituted species were
calculated to be about 30 50% stronger than those of the
octasubstituted species. Among the octasubstituted species,
the intensity of those substituted at a positions is stronger by
several tens of percent than those substituted at b positions.
It is empirically known that the B band of Pcs is about twice
as strong as the Q band.[27] This feature is reproduced for
unsubstituted and octaalkoxy b-substituted Pcs. However, Pcs
having the Q band at longer wavelength, such as 8a- and
Figure 7. Calculated transition energies and oscillator strengths f for
pyrrole-proton-deprotonated Pcs (Pc2À). Numbers indicate the values of
the Q band.
5128
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5123 5134

Phthalocyanines
5123 5134
Table 5. Calculated transition energies, oscillator strengths f, and configurations.[a]
The numbers of configura-
tions in Table 5 further suggest
that the purity of the Q band
increases for species having the
Q band at longer wavelengths.
Figure 8 shows some frontier
orbitals of representative spe-
cies. As previously reported,[27]
the HOMO and next HOMO of
general Pcs have a1u and a2u
symmetry, respectively, in the
D4h point group. However, one
of the most important features in
this figure is that the next HO-
MO of Pcs having the Q band at
longer wavelength, such as
8a(O- or S-alkyl)- and 16(O- or
S-alkyl)Pcs, has b1u symmetry,
which is generally the 4th or
lower HOMO in Pcs having the
Q band at shorter wavelength
(see also Figure 9). Data in Ta-
ble 5 and Figure 8 therefore sug-
gest that the second lowest tran-
sition for 8a(O- or S-alkyl)- and
16(O- or S-alkyl)Pcs is a transi-
tion from this b1u orbital to the
LUMO or LUMO ‡ 1. In accord
with this assignment, a Faraday
A term is seen for the 500 nm
band of 8a(SBu)ZnPc (Fig-
ure 3).
Energy [nm]
f
Configurations
Pc2À
662
662
332
332
302
302
0.84
0.84
2.19
2.19
0.25
0.25
21 ! 22(87%) 20 ! 23(12%)
21 ! 23(87%) 20 ! 22(12%)
20 ! 23(67%) 13 ! 23(12%) 21 ! 22(11%)
20 ! 22(67%) 13 ! 22(12%) 21 ! 23(11%)
16 ! 22(23%) 21 ! 27(20%) 13 ! 22(15%) 16 ! 23(14%) 21 ! 26(12%)
16 ! 23(23%) 21 ! 26(20%) 13 ! 23(15%) 16 ! 22(14%) 21 ! 27(12%)
8bOPc2À
685
685
366
366
342
342
318
318
0.86
0.86
0.16
0.16
2.65
2.65
0.34
0.34
29 ! 30(87%)
29 ! 31(87%)
25 ! 31(47%) 28 ! 31(20%) 24 ! 31(20%)
25 ! 30(47%) 28 ! 30(20%) 24 ! 30(20%)
28 ! 30(57%) 25 ! 30(14%)
28 ! 31(57%) 25 ! 31(14%)
24 ! 30(54%) 21 ! 31(19%) 25 ! 30(18%)
24 ! 31(54%) 21 ! 30(19%) 25 ! 31(18%)
8bSPc2À
694
694
402
402
371
371
327
327
0.89
0.89
0.38
0.38
1.35
1.35
1.32
1.32
0.21
0.21
29 ! 31(87%)
29 ! 30(87%)
25 ! 31(35%) 28 ! 31(31%)
25 ! 30(35%) 28 ! 30(31%)
28 ! 30(50%) 25 ! 30(41%)
28 ! 31(50%) 25 ! 31(41%)
24 ! 31(58%) 21 ! 30(29%)
24 ! 30(58%) 21 ! 31(29%)
21 ! 31(49%) 24 ! 30(20%)
21 ! 30(49%) 24 ! 31(20%)
308
308
16OPc2À
738
738
389
389
342
335
320
320
1.38
1.38
0.53
0.27
0.70
1.15
0.48
0.60
37 ! 38(94%)
37 ! 39(94%)
36 ! 39(88%)
36 ! 38(92%)
33 ! 38(45%) 32 ! 38(31%)
33 ! 39(69%) 32 ! 39(15%)
32 ! 39(78%) 33 ! 39(18%)
32 ! 58(59%) 33 ! 38(38%)
8aOPc2À
Figure 9 displays a partial mo-
lecular orbital energy diagram
for some Pcs substituted with O-
or S-alkyl groups. The following
points emerge from the inspec-
tion of this figure: 1) substitu-
tion at the b positions raises the
energy of all frontier orbitals;
2) this effect is slightly more
pronounced for the HOMO than
for the LUMO; 3) this effect is
more pronounced for S-alkyl-
than for O-alkyl-substituted
Pcs; 4) compared with substitu-
tion at b positions, that at a
positions results in larger desta-
bilization of the HOMO and
slight destabilization of the LU-
MO; 5) substitution at the a
positions lifts a low-lying b1u
orbital (the 4th HOMO in un-
substituted Pc) significantly, so
that it becomes the HOMO À 1;
and 6) the energy gap between
the HOMO and LUMO de-
creases in the following order:
unsubstituted Pc ꢀ 8b(O-alkyl)
744
744
399
399
345
345
301
301
16SPc2À
770
770
449
449
386
373
354
1.09
1.09
0.67
0.67
0.93
0.93
0.35
0.35
29 ! 30(76%) 29 ! 31(14%)
29 ! 31(76%) 29 ! 30(14%)
28 ! 30(70%) 28 ! 31(12%)
28 ! 31(70%) 28 ! 30(12%)
25 ! 30(53%) 24 ! 31(21%) 28 ! 30(10%)
25 ! 31(53%) 24 ! 30(21%) 28 ! 31(10%)
24 ! 30(40%) 24 ! 31(20%) 25 ! 31(10%)
24 ! 31(40%) 24 ! 30(20%) 25 ! 30(10%)
1.55
1.55
0.30
0.15
0.52
0.62
0.38
0.32
0.26
0.23
37 ! 38(95%)
37 ! 39(95%)
36 ! 39(92%)
36 ! 38(94%)
33 ! 38(47%) 32 ! 38(39%)
33 ! 39(59%) 32 ! 39(34%)
32 ! 39(63%) 33 ! 39(35%)
32 ! 38(55%) 33 ! 38(44%)
37 ! 42(93%)
351
315
315
37 ! 43(94%)
8aSPc2
799
799
460
460
369
369
319
319
301
301
1.24
1.24
0.26
0.26
0.19
0.19
0.35
0.35
1.18
1.18
29 ! 30(92%)
29 ! 31(92%)
28 ! 30(91%)
29 ! 31(91%)
25 ! 30(67%) 24 ! 31(24%)
25 ! 31(67%) 24 ! 30(24%)
29 ! 34(21%) 29 ! 35(20%) 24 ! 30(16%) 24 ! 31(16%)
29 ! 35(21%) 29 ! 34(20%) 24 ! 31(16%) 24 ! 30(16%)
29 ! 34(22%) 24 ! 31(17%) 23 ! 31(16%) 29 ! 35(12%)
29 ! 35(22%) 24 ! 30(17%) 23 ! 30(16%) 29 ! 34(12%)
[a] Excited states with energy less than about 3.9 eV and f greater than 0.15 are shown.
Chem. Eur. J. 2003, 9, 5123 5134
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5129

N. Kobayashi et al.
FULL PAPER
compounds differ from each
other in the number, position,
and electrostatic effect of the
substituents. As electron-re-
leasing groups, mostly alkoxyl
and alkyl(phenyl)thio groups
were used, while as electron-
withdrawing groups, nitro and
phenylsulfonyl groups were
employed. When electron-re-
leasing groups are bound to
four a-benzo positions of the
Pc skeleton, the B and Q bands
shift to longer wavelength. Rel-
ative to this shift, the effect of
introducing the same electron-
releasing groups at the other
four a positions amounts to
about 1.6 2.0. Although the
effect is not always clearly seen,
introduction of electron-releas-
ing groups in the b-benzo posi-
tions of the Pc skeleton gener-
ally shifts the Q band to shorter
wavelength. The effect of elec-
tron-withdrawing groups is ex-
actly the opposite, that is, four
such groups in a or b positions
shift the Q band to shorter and
longer wavelength, respective-
ly. These effects can be reason-
ably explained by taking into
account the size of the carbon
atomic orbital coefficients de-
rived from the molecular orbi-
tal calculations. Specifically,
since the coefficients of a car-
bon atoms are larger than those
of the b carbon atoms in the
HOMO, the extent of destabi-
lization of this orbital by intro-
ducing
electron-releasing
groups is larger when they are
linked to the a positions, which
makes the HOMO LUMO
gap smaller. As a result, the Q
band, which is composed main-
ly of the HOMO LUMO tran-
Figure 8. Some frontier orbitals of pyrrole-proton-deprotonated 8bSPc2À, Pc2À, 16SPc2À, and 8aSPc2À
.
> 8b(S-alkyl) > 16(O-alkyl)ꢀ 8a(O-alkyl) > 16(S-alkyl) ꢀ 8a-
(S-alkyl). Of these, points 3) and 4) are consistent with
experiments, since the Q band is nearly described as a single
transition from the HOMO to LUMO.
sition, shifts to longer wavelength, particularly when the
electron-releasing groups are introduced at the a positions.
The following intriguing phenomena were observed in the
experiments:
1) The splitting of the Q band of metal-free species decreases
with increasing wavelength of the Q band.
1
2) Judging from the pyrrole proton signals in the H NMR
Conclusion
spectra, the ring current of Pcs which have the Q band at
longer wavelength have smaller ring currents.
3) The absorption coefficients of the Q bands of Pcs with 16
electron-releasing substituents are larger than those of the
The effect of substituents on the position and intensity of the
electronic absorption and fluorescence spectra of phthalocya-
nines has been examined for 35 Pc compounds. These
5130
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5123 5134

Phthalocyanines
5123 5134
Figure 9. Partial energy diagram obtained by PPP calculations for the pyrrole-proton-deprotonated Pcs (Pc2À).
Chemicals: THF was distilled from sodium. DMF, n-pentanol, n-butanol,
and 1-hexanol were dried over activated 4 ä molecular shieves. Pyridine
was dried over KOH and distilled. All other commercially available
reagents and solvent were used without further treatment.
corresponding tetra- and octasubstituted Pcs by several
tens of percent.
4) Pcs showing S1 emission maxima at wavelengths longer
than about 740 nm generally have quantum yields of less
than 0.1.
Molecular orbital calculations within the frameworkof the
PPP approximations reproduced observations in points 1) and
3) and furthermore suggest that the absorption coefficient of
the Q band of Pcs substituted by more strongly electron
releasing substituents is larger. The second HOMO of Pcs
showing the Q band at longer wavelength has b1u symmetry, in
contrast to the a2u symmetry of normal Pcs.
The following phthalonitrile (1,2-dicyanobenzene) starting materials were
synthesized according to published procedures: 3,6-[33a] and 4,5-dibutyl-
oxy-,[33b] 4-butylthio-,[33c] 4,5-dibutylthio-,[33d] 3- and 4-phenylthio-,[33e] 3-tert-
butyl-5-phenylthio- and 3-tert-butyl-5-phenylsulfonyl-,[33f] and 3,4,5,6-tet-
rakis(2',2',2'-trifluoroethoxy)-1,2-dicyanobenzene.[33g] 3-n-Butyloxy-1,2-di-
cyanobenzene, 3-n-butylthio-1,2-dicyanobenzene, and 4-n-butylthio-1,2-
dicyanobenzene were prepared by a minor modification of the procedure
of Leznoff et al.[34] for the preparation of alkoxy- and alkylthio-substituted
phthalonitriles. Phthalocyanines not listed below were reported in earlier
g]
work.[33e
3-n-Butyloxy-1,2-dicyanobenzene: The crude product was purified by
column chromatography on silica gel using CH2Cl2/hexane (1:1) as eluant
to give a 47.3% yield of a white solid of 3-butyloxy-1,2-dicyanobenzene.
M.p. 131 1328C; elemental analysis (%) calcd for C12H12N2O (200.24): C
71.98, H 6.04, N 13.99; found: C 72.07, H 6.11, N 14.19; 1H NMR (400 MHz,
CDCl3, TMS): d ˆ 0.58 2.07 (m, 7H; CH2CH2CH3), 3.75 4.38 (m, 2H;
OCH2), 6.85 7.75 ppm (m, 3H; Ar-H).
Experimental Section
Measurements: Electronic spectra were measured with a Hitachi U-3410
spectrophotometer. Magnetic circular dichroism (MCD) measurements
were made with a JASCO J-725 spectrodichrometer equipped with a
JASCO electromagnet that produced magnetic fields of up to 1.09 T with
parallel and antiparallel fields. Its magnitude was expressed in terms of
molar ellipticity per tesla [q]M/104 degmolÀ1 dm3 cmÀ1 TÀ1. The 400 MHz
3-n-Butylthio-1,2-dicyanobenzene: The crude product was recrystallized
from ethanol to give a 95% yield of white needles. M.p. 70 718C;
elemental analysis (%) calcd for C12H12N2S (216.31): C 66.63, H 5.59, N
12.95, S, 14.82; found: C 66.49, H 5.62, N 12.93, S 14.52; 1H NMR
(400 MHz, CDCl3, TMS): d ˆ 0.61 (t, 3H; CH3), 1.49 (tq, 2H;
S(CH2)2CH2), 1.69 (tt, 2H; SCH2CH2), 3.05 (t, 2H; SCH2), 7.51 7.62 ppm
(m, 3H; Ar-H).
1H NMR spectra were recorded in CDCl3 on
spectrometer.
a JEOL GSX-400
Method of calculation: The Pc structures were constructed from standard
phthalocyanine X-ray structural data[29] and by making the ring perfectly
planar and adopting D4h symmetry. Molecular orbital (MO) calculations
were performed for the pyrrole proton-deprotonated dianionic species
4-n-Butylthio-1,2-dicyanobenzene: After recrystallization from ethanol,
this compound was obtained as off-white needles in 93% yield. M.p. 62
638C; elemental analysis (%) calcd for C12H12N2S (216.31): C 66.63, H 5.59,
N 12.95, S, 14.82; found: C 66.43, H 5.55, N 12.78, S 14.82; 1H NMR
(400 MHz, CDCl3, TMS): d ˆ 0.97 (t, 3H; CH3), 1.50 (tq, 2H;
S(CH2)2CH2), 1.70 (tt, 2H; SCH2CH2), 3.01 (t, 2H; SCH2), 7.48 (d, 1H;
Ar-H, C6), 7.54 (s, 1H; Ar-H, C2), 7.63 ppm (d, 1H; Ar-H, C5).
[30]
within the frameworkof the Pariser Parr Pople (PPP) approximation
with employment of recently published semiempirical parameters.[31] These
are atomic valence state ionization potentials of 11.16 (carbon), 20.21
(central nitrogen), and 14.21 eV (imino nitrogen), together with atomic
valence affinities of 0.03 (carbon), 5.32 (central nitrogen), and 1.78 eV
(imino nitrogen). The central nitrogen atoms were assumed to be
equivalent, supplying 1.5 electrons each to the p system. In addition, s
polarizability was taken into account according to Hammond.[32a] Reso-
nance integrals were taken to be À2.48 (bCN) and À2.42 (bCC).[31] Two-
center repulsion integrals were calculated by the method of Mataga and
Nishimoto.[32b] The choice of configuration was based on energetic
considerations, and all singly excited configurations up to 56458 cmÀ1 were
included.
1,2-Dicyano-3,6-bis(trifluorosulfonyl)benzene: Triethylamine (5.0 mL,
36 mmol), CH2Cl2 (25 mL), and dimethylaminopyridine (0.20 g, 1.6 mmol)
were added to 2,3-dicyanohydroquinone (2.4 g, 15 mmol) under nitrogen
while maintaining the temperature of the reaction mixture below À708C,
and the mixture was stirred for 24 h. Trifluoromethanesulfonic anhydride
(6.0 mL, 37 mmol) was added dropwise, and the mixture stirred for 24 h.
The reaction mixture was poured into CH2Cl2 (200 mL), 0.5n HCl (5 mL)
added, and the organic layer separated and evaporated under reduced
Chem. Eur. J. 2003, 9, 5123 5134
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5131

N. Kobayashi et al.
FULL PAPER
pressure. The residue was subjected to chromatography on a silica gel
column with CHCl3 as eluent to give in 85% yield 5.4 g of white needles.
M.p. 106 1078C; elemental analysis (%) calcd for C10F6H2N6O6S2 (424.26):
C 28.31, H 0.48, N 6.60, S 15.12; found: C 28.02, H 0.82, N 7.77, S 15.24;
1H NMR (400 MHz, CDCl3, TMS): d ˆ 7.87 ppm (s, 2H; Ar-H).
C64H80N8O8Zn (1154.77): C 66.57, H 6.98, N 9.70; found for 8a(OBu)ZnPc:
C 67.44, H 7.37, N 9.68; found for 8b(OBu)ZnPc: C 67.31, H 7.51, N 8.93.
8a(OBu)ZnPc is unstable in solution and in air.
16(OCH2CF3)H2Pc and 16(OCH2CF3)ZnPc: These were reported in our
earlier study.[33g]
1,2-Dicyano-3,6-dibutylthiobenzene:
1,2-Dicyano-3,6-bis(trifluorosulfo-
4a(SBu)H2Pc (mixture of regioisomers): 3-n-Butylthio-1,2-dicyanoben-
zene (1.0 g, 4.6 mmol) was added to lithium (20 mg, 2.88 mmol) dissolved in
dry n-pentanol (6 mL), and the solution was refluxed for 1 h. After cooling,
the solution was poured into methanol (200 mL) containing 3 drops of
concentrated HCl, and the resulting precipitate was collected by filtration
and purified on a silica gel column with pyridine as eluent. The front-
running green portion was recrystallized from CH2Cl2/methanol to give a
59% yield (0.59 g) of the desired compound as a green powder. Elemental
analysis (%) calcd for C48H50N8S4 (867.25): C 64.48, H 5.81, N 12.92, S 14.79;
found: C 64.17, H 5.98, N 12.27, S 13.77; 1H NMR (400 MHz, CDCl3, TMS):
d ˆ À1.14 (s, 2H; pyrrole), 1.1 1.3 (m, 12H; CH3), 1.6 1.9 (m, 8H;
S(CH2)2CH2), 1.90 2.20 (m, 4H; SCH2CH2), 3.41 (br, 8H; SCH2), 7.4 8.1
(m, 8H; Ar-H), 8.65 9.15 ppm (m, 4H; Ar-H).
nyl)benzene (3.0 g, 7.0 mmol) was added to a dry DMF (10 mL) solution
containing K2CO3 (1.2 g, 8.8 mmol) and butylmercaptan (1.4 g, 15 mmol),
and the solution was stirred at room temperature with slight bubbling of
nitrogen for 5 h, after which K2CO3 (1.2 g, 8.8 mmol) was added again, and
stirring continued for further 48 h. The reaction mixture was poured into
water (200 mL), and the solid collected by filtration and washed with water.
After recrystallization of the solid from ethanol, 0.40 g (19% yield) of the
desired dinitrile was obtained as pale yellowish needles. M.p. 85 868C;
elemental analysis (%) calcd for C16H20N2S2 (304.48): C 63.12, H 6.62, N
1
9.20, S 21.06; found: C 62.86, H 6.64, N 9.03, S 20.91; H NMR (400 MHz,
CDCl3, TMS): d ˆ 0.94 (t, 6H; CH3), 1.48 (tq, 4H; S(CH2)2CH2), 1.66 (tt,
4H; SCH2CH2), 3.02 (t, 4H; SCH2), 7.50 ppm (m, 2H; Ar-H).
1,2-Dicyano-4,5-diphenylthiobenzene: A mixture of 1,2-dicyano-4,5-di-
iodobenzene[35] (195 mg, 1 mmol), thiophenol (660 mg, 6 mmol), and
K2CO3 (2.36 g, 32 mmol) in dry DMSO (2 mL) was heated at 908C for
30 min.[36] After cooling, ice was added, and the precipitated yellow solid
was filtered off, and subjected to chromatography on silica with hexane/
CH2Cl2 (3:10) to give a 61% yield (211 mg) of the desired dinitrile as a
while solid. M.p. 197 1998C; elemental analysis (%) calcd for C20H12N2S2
(344.47): C 69.74, H 3.51, N 8.13, S 18.62; found: C 69.47, H 3.70, N 7.90, S
18.46. 1H NMR (400 MHz, CDCl3, TMS): d ˆ 6.99 (s, 2H; Ar-H), 7.26
7.57 ppm (m, 10H; Ar-H).
4a(SBu)ZnPc (mixture of regioisomers): A mixture of 1-hexanol (5 mL),
Zn(OAc)2, and 4a(SBu)H2Pc (0.2 g, 0.23 mmol) was refluxed for 1 h and
then poured into methanol (200 mL), and the resulting precipitate was
filtered off and subjected to chromatography on a silica gel column with
CHCl3 as eluent. The front-running blue-green portion was collected and
recrystallized from CHCl3/methanol to give a 47% yield (0.10 g) of a green
solid. Elemental analysis (%) calcd for C48H48N8S4Zn (930.64): C 61.95, H
1
5.20, N 12.04, S 13.78; found: C 60.49, H 5.03, N 11.04, S 13.64; H NMR
(400 MHz, CDCl3, TMS): d ˆ 0.90 1.22 (m, 3H; CH3), 1.64 2.10 (m, 4H;
S(CH2)2CH2 and SCH2CH2), 3.77 3.90 (br, 2H; SCH2), 7.04 7.75 (m, 3H;
Ar-H), 8.65 9.15 ppm (m, 4H; Ar-H).
4a(OBu)H2Pc (mixture of regioisomers): A mixture of lithium (30 mg,
4.32 mmol) and 3-n-butyloxy-1,2-dicyanobenzene (500 mg, 2.5 mmol) in
dry n-pentanol (3 mL) was refluxed for 3 h, the solvent was evaporated to
about half-volume under reduced pressure, and the solution was refluxed
for a further 40 min. After removing the solvent under vacuum, the residue
was subjected to chromatography on a silica gel column with CH2Cl2 as
eluent to give a 25.9% yield (130 mg) of a blue solid. Elemental analysis
(%) calcd for C48H50N8O4(802.98): C 71.80, H 6.28, N 13.95; found: C 70.78,
H 6.44, N 13.69; 1H NMR (400 MHz, CDCl3, TMS): d ˆ À0.73 (m, 2H;
pyrrole), 1.13 1.40 (m, 12H; CH3), 1.70 2.15 (m, 8H; CH2), 2.22 2.47 (m,
8H; CH2), 4.40 5.00 (m, 8H; OCH2), 7.20 8.10 (m, 8H; Ar-H), 8.30
9.00 ppm (m, 4H; Ar-H).
4b(SBu)H2Pc (mixture of regioisomers): As for the preparation of
4a(SBu)H2Pc, this compound was synthesized from 4-n-butylthio-1,2-
dicyanobenzene (1.0 g, 4.6 mmol) in 29% yield (0.29 g). Elemental analysis
(%) calcd for C48H50N8S4 (867.25): C 64.48, H 5.81, N 12.92, S 14.79; found:
1
C 66.02, H 5.58, N 12.79, S 14.70; H NMR (400 MHz, CDCl3, TMS): d ˆ
À5.00 (s, 2H; pyrrole), 1.18 (m, 12H; CH3), 1.74 (m, 8H; S(CH2)2CH2),
1.94 (m, 4H; SCH2CH2), 3.16 (br, 8H; SCH2), 7.00 7.75 ppm (m, 12H; Ar-
H).
4b(SBu)ZnPc (mixture of regioisomers): As for the preparation of
4a(SBu)ZnPc from 4a(SBu)H2Pc, zinc was inserted into 4b(SBu)H2Pc
(0.2 g, 0.23 mmol) to give 0.20 g (93%) of the desired compound.
Elemental analysis (%) calcd for C48H48N8S4Zn (930.64): C 61.95, H 5.20,
N 12.04, S 13.78; found: C 59.88, H 4.74, N 11.12, S 12.87; 1H NMR
(400 MHz, CDCl3, TMS): d ˆ 0.93 (m, 12H; CH3), 1.27 1.67 (m, 16H;
S(CH2)2CH2 and SCH2CH2), 2.60 (t, 8H; SCH2), 6.50 7.80 ppm (m, 12H;
Ar-H).
4b(OBu)H2Pc (mixture of regioisomers): This compound was prepared by
the procedure described for 4a(OBu)H2Pc, but with 4-n-butyloxy-1,2-
dicyanobenzene (500 mg, 2.5 mmol) in place of 3-n-butyloxy-1,2-dicyano-
benzene. Yield: 120 mg (23.9%); elemental analysis (%) calcd for
C48H50N8O4(802.98): C 71.80, H 6.28, N 13.95; found: C 69.70, H 6.21, N
13.12. 1H NMR (400 MHz, CDCl3, TMS): d ˆ À5.24 (s, 2H; pyrrole), 1.18
1.27 (t, 12H; CH3), 1.68 1.79 (m, 8H; CH2), 1.93 2.04 (m, 8H; CH2),
3.90 4.15 (t, 8H; OCH2), 6.60 8.00 ppm (m, 12H; Ar-H).
8b(SBu)H2Pc: 3,6-Dibutylthio-1,2-dicyanobenzene (0.4 g, 1.3 mmol) was
added to lithium (20 mg, 2.9 mmol) dissolved in dry n-pentanol (6 mL), and
the solution was refluxed for 1 h. After cooling, the solution was poured
into methanol (100 mL) containing 3 drops of concentrated HCl, and the
resulting precipitate was collected by filtration and purified on a silica gel
column with CHCl3 as eluent. The dull green portion was recrystallized
from CHCl3/methanol to give a 38% yield (0.15 g) of a green solid.
Elemental analysis (%) calcd for C64H82N8S8 (1219.95): C 63.01, H 6.78, N
4a(OBu)ZnPc and 4b(OBu)ZnPc (mixtures of regioisomers): 4a(O-
Bu)H2Pc or 4b(OBu)H2Pc (50 mg, 0.062 mmol) and an approximately
equal volume of Zn(OAc)2 were refluxed in a mixture of 1,2-dichloro-
ethane (3 mL) and ethanol (3 mL) for 18 h in the dark. After evaporation
of the solvent, the residue was subjected to chromatography on an alumina
column with CH2Cl2 as eluent. The green-blue fraction was recrystallized
from CH2Cl2/hexane to yield 16 mg (4a(OBu)ZnPc) or 18 mg (4b(OB-
u)ZnPc) of a blue-green powder (29.7 and 33.4%, respectively). Elemental
analysis (%) calcd for C48H48N8O4Zn (866.34): C 66.55, H 5.58, N 12.93;
found for 4a(OBu)ZnPc: C 66.94, H 6.11, N 12.00; found for 4b(OB-
u)ZnPc: C 67.40, H 6.01, N 12.30.
1
9.19, S 21.03; found: C 62.32, H 6.84, N 8.74, S 21.01; H NMR (400 MHz,
CDCl3, TMS): d ˆ 0.28 (s, 2H; pyrrole), 1.04 (t, 24H; CH3), 1.65 (tq, 16H;
S(CH2)2CH2), 1.91 (tt, 16H; SCH2CH2), 3.23 (br, 16H; SCH2), 7.69 ppm
(br, 8H; Ar-H).
8a(SBu)ZnPc:
A mixture of 1-hexanol (4.5 mL), pyridine (1.5 mL),
8a(SBu)H2Pc (40 mg, 0.033 mmol), and Zn(OAc)2 was refluxed for 1 h,
after which the reaction mixture was poured into methanol (80 mL), and
the precipitate formed was filtered off and subjected to chromatography on
a silica gel column with CHCl3 as eluent. The front-running colored portion
was collected and recrystallized from CH2Cl2/methanol to give a 93% yield
(39 mg) of a green solid. Elemental analysis (%) calcd for C64H80N8S8Zn
(1283.34): C 59.90, H 6.28, N 8.73, S 19.99; found: C 59.60, H 6.67, N 8.06, S
18.55; 1H NMR (400 MHz, CDCl3, TMS): d ˆ 0.97 (t, 24H; CH3), 1.63 (br,
16H; S(CH2)2CH2), 1.56 (br, 16H; SCH2CH2), 2.89 (br, 16H; SCH2),
7.38 ppm (br, 8H; Ar-H).
8a- and 8b(OBu)H2Pc: These compounds were obtained from 3,6- and 4,5-
dibutyloxy-1,2-dicyanobenzene, respectively, by a method reported in the
literature.[33a, b] Elemental analysis (%) calcd for C64H82N8O8 (1091.41): C
70.43, H 7.57, N 10.27; found for 8a(OBu)H2Pc: C 70.07, H 7.63, N 10.09;
found for 8b(OBu)H2Pc: C 70.14, H 7.60, N 10.34. In the case of
8a(OBu)H2Pc, prolonged heating in n-pentanol causes substitution of
some of butoxyl groups by pentyl groups so that reaction in n-butyl
alchohol appears safer.
8a- and 8b(OBu)ZnPc: These compounds were obtained as for 4a(O-
Bu)ZnPc and 4b(OBu)ZnPc by zinc insertion into 8a- and 8b(OBu)H2Pc in
15.5 and ca. 30% yield, respectively. Elemental analysis (%) calcd for
8b(SPh)ZnPc: 1,2-Dicyano-4,5-diphenylthiobenzene (172 mg, 0.5 mmol)
was added to lithium (2 mg, 0.29 mmol) dissolved in dry n-pentanol (1 mL),
5132
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5123 5134

Phthalocyanines
5123 5134
and the solution was refluxed for 2 h. Then dried Zn(OAc)2 (92 mg) was
added and refluxing was continued for 2 h in the dark. After cooling, the
[19] The term ™Soret band (region)∫ has long been used. This is now often
referred to as ™the B band (region)∫ (see for example ref. [27]).
[20] In addition to p p* transitions, an n p* transition of the ether
oxygen atom lies at the longer wavelength tail of the B band,[21] which
makes the analysis more difficult.
[21] N. Kobayashi, A. B. P. Lever, J. Am. Chem. Soc. 1987, 109, 7433 7441;
N. Kobayashi, M. Togashi, T. Osa, K. Ishii, S. Yamauchi, H. Hino, J.
Am. Chem. Soc. 1996, 118, 1073 1085.
solution was poured into methanol (20 mL) containing
1 drop of
concentrated HCl, and the resulting precipitate was collected by filtration
and purified on an alumina column with CH2Cl2/pyridine as eluent. The
first-eluted green portion was collected and recrystallized from pyridine/
methanol to give a 19.4% yield (35 mg) of green-blue powder. Elemental
analysis (%) calcd for C80H48N8S8Zn (1443.27): C 66.58, H 3.35, N 7.76, S
17.77; found: C 65.92, H 3.59, N 7.33, S 17.12; 1H NMR (400 MHz, CDCl3,
TMS): d ˆ 7.42 7.48 (m, 24H; Ar-H), 7.62 7.64 (d, 16H; Ar-H), 9.06 ppm
(s, 8H; Ar-H).
[22] a) A. Tajiri, J. Winkler, Z. Naturforsch. A. 1983, 38, 1263 1269; b) A.
Kaito, T. Nozawa, T. Yamamoto, M. Hatano, Y. Orii, Chem. Phys. Lett.
1977, 52, 154 160.
[23] Our aim was to collect as accurate absorption coefficients as possible
for a series of ZnPcs. However, some ZnPcs were unstable in solution.
For example, since 8a(OBu)ZnPc was unstable, we could not have
confidence in the comparison of the absorption coefficients among 4a-,
4b-, 8a-, and 8b-(OBu)ZnPc and 16(OCH2CF3)ZnPc. On the other
hand, although 4a-, 4b-, 8a-, and 8b(SBu)ZnPc were stable, we could
not obtain 16(SCH2CF3)ZnPc, depite many attempts.
[24] L. Guo, D. E. Ellis, B. M. Hoffman, Y. Ishikawa, Inorg. Chem. 1996,
35, 5304 5312.
[25] K. Ishii, N. Kobayashi in The Porphyrin Handbook, Vol. 16, (Eds.:
K. M. Kadish, K. M. Smith, R. Guilard), Academic Press, New York,
2002, Chapter 102, pp. 1 42.
Acknowledgement
This research was partially supported by the Ministry of Education,
Science, Sports and Culture, a Grant-in-Aid for the COE project, Giant
Molecules and Complex Systems, 2003. E.L. thanks Moscow City Govern-
ment and the Ministry of Science and Technology of Russia for financial
support.
[26] We also calculated MOs of tetranitro- and tetrasulfonyl-substituted,
pyrrole-proton-deprotonated species. However, calculations were not
definitive. For example, a single Q band was not obtained, although
the Q band was calculated at higher energy for tetrasubstitution at a
positions. Accordingly, discussion of Pcs substituted by electron-
withdrawing groups were not included in this section. The results of
the PPP calculations on planar porphyrin oligomers have been
reported recently (H. S. Cho, D. H. Jeong, S. Cho, D. Kim, Y.
Matsuzaki, K. Tanaka, A. Tsuda, A. Osuka, J. Am. Chem. Soc. 2002,
124, 14642 14654). Although MO calculations using the ZINDO/S
and DFT Hamiltonians have been used for simple, high-symmetry
systems such as porphyrins and Pcs with D4h symmetry, and the results
are often good, these newer methods do not necessarily produce
better results than the older PPP method. For example, in the case of
these porphyrin oligomers, the HOMO is an a1u-based molecular
orbital when calculated with the ZINDO/S and DFT Hamiltonians, so
that the transitions at the longest wavelength become symmetrically
forbidden and are not at all consistent with the experimental results
[1] a) A. B. P. Lever, Adv. Inorg. Chem. Radiochem. 1965, 7, 27 114;
b) Phthalocyanines-Properties and Applications, Vols. 1
4 (Eds.:
C. C. Leznoff, A. B. P. Lever), VCH, New York, 1989, 1993, 1993,
1996; c) F. H. Moser, A. H. Thomas, The Phthalocyanines, Vols. 1 and
2, CRC Press, Boca Raton, FL, 1983; d) H. Shirai, N. Kobayashi,
Phthalocyanines–Chemistry and Functions, IPC, Tokyo, 1997;
e) N. B. Mckeown, Phthalocyanine Materials, Cambridge University
Press, Cambridge, 1998.
[2] J. Simon, J. -J. Andre in Molecular Semiconductors, Springer, Berlin,
1985, Chapter 3.
[3] D. Wohrle, D. Meissner, Adv. Mater. 1991, 3, 129 138.
[4] R. O. Loutfy, A. M. Hor, C. K. Hsiano, G. Baranyi, P. Kazmaier, Pure
Appl. Chem. 1988, 60, 1047 1054.
[5] K. Abe, H. Sato, T. Kimura, Y. Ohkatsu, T. Kusano, Makromol. Chem.
1989, 190, 2693 2701.
[6] B. Simic-Glavaski in Phthalocyanines–Properties and Applications,
Vol. 3 (Eds.: C. C. Leznoff, A. B. P. Lever), VCH, New York, 1993,
Chapter 3, pp. 119 166.
(the experimental results are reproduced when the HOMO is an a2u
-
based orbital). Furthermore, to date MO calculations using the
ZINDO/S Hamiltonian on the same p system but with different type,
number, and position of substituents have not been carried out, so that
the fact that the ZINDO/S calculations do not reproduce sufficiently
the experimental order of even the longest-wavelength band among
these species has not been reported.
[7] a) M. J. Cook, A. Dunn, M. F. Daniel, R. C. O. Hart, R. M. Richard-
son, S. J. Roser, Thin Solid Films 1988, 159, 395 404; b) S. Palacin, P.
Lesieur, I. Stefanelli, A. Barraud, Thin Solid Films 1988, 159, 83 90.
[8] M. M. Nicholson, in Phthalocyanines–Properties and Applications,
Vol. 3, (Eds.: C. C. Leznoff, A. B. P. Lever), VCH, New York, 1993,
Chapter 2, pp. 71 117.
[27] a) J. Mack, M. J. Stillman in The Porphyrin Handbook, Vol. 16 (Eds.:
K. M. Kadish, K. M. Smith, R. Guilard), Academic Press, New York,
2002, Chapter 103, pp. 43 116; b) M. J. Stillman in Phthalocyanines–
Properties and Applications, Vol. 1 (Eds.: C. C. Leznoff, A. B. P.
Lever), VCH, New York, 1989, Chapter 3, pp. 133 289.
[9] T. J. Marks, Angew. Chem. 1990, 102, 886 908; Angew. Chem. Int. Ed.
Engl. 857 879.
[10] A. Snow, W. R. Barger in Phthalocyanines–Properties and Applica-
tions, Vol. 1 (Eds.: C. C. Leznoff, A. B. P. Lever), VCH, New York,
1989, Chapter 5, pp. 341 392.
[28] N. Kobayashi, J. Mack, K. Ishii, M. J. Stillman, Inorg. Chem. 2002, 41,
5350 5363.
[11] a) J. F. Van der Pol, E. Neeleman, J. W. Zwikker, R. J. M. Nolte, W.
Drenth, J. Aerts, R. Visser, S. J. Picken, Liq. Cryst. 1989, 6, 577 592;
b) J. Simon, C. Sirlin, Pure Appl. Chem. 1989, 61, 1625 1629; c) J. C.
Swarts, E. H. G. Langner, N. Krokeide-Hove, M. J. Cook, J. Mater.
Chem. 2001, 11, 434 443, and references therein.
[29] J. M. Robertson, I. Woodward, J. Chem. Soc. 1937, 219 230; P. A.
Barrett, C. E. Dent, R. P. Linstead, J. Chem. Soc. 1936, 1719 1736;
C. J. Brown, J. Chem. Soc. A 1968, 2488 2493, 2494 2498; J. F.
Kirner, W. Dow, W. R. Scheidt, Inorg. Chem. 1976, 15, 1685 1690.
[30] a) R. Pariser, R. G. Parr, J. Chem. Phys. 1953, 21, 466 471, 767 776;
b) J. A. Pople, Trans. Faraday Soc. 1953, 49, 1375 1385.
[31] S. Tokita, K. Matsuoka, Y. Kogo, K. Kihara, Molecular Design of
Functional Dyes–PPP Method and Its Application, Maruzen, Tokyo,
1990.
[12] G. de la Torre, T. Torres, F. Agullo-Lopez, Adv. Mater. 1997, 9, 265
269.
[13] H. Ali, J. E. Lier, Chem. Rev. 1999, 99, 2379 2450.
[14] A. B. P. Lever, M. R. Hempstead, C. C. Leznoff, W. Liu, M. Melnik,
W. A. Nevin, P. Seymour, Pure Appl. Chem. 1986, 58, 1467 1476.
[15] N. Kobayashi, H. Konami in Phthalocyanines–Properties and Appli-
cations, Vol. 4 (Eds.: C. C. Leznoff, A. B. P. Lever), VCH, New York,
1996, Chapter 9, pp. 343 404; N. Kobayashi, H. Konami, J. Porphyr-
ins Phthalocyanines 2001, 5, 233 255.
[16] E. A. Luk×yanets, Electronic Spectra of Phthalocyanines and Related
Compounds, NIOPIK, Moscow, 1989.
[17] H. Konami, H. Hatano, Chem. Lett. 1988, 1359 1362.
[18] N. Kobayashi, N. Sasaki, Y. Higashi, T. Osa, Inorg. Chem. 1995, 34,
1636 1637.
[32] a) H. A. Hammond, Theor. Chim. Acta 1970, 18, 239 249; b) N.
Mataga, K. Nishimoto, Z. Phys. Chem. 1957, 13, 140 157.
[33] a) M. J. Cook, A. J. Dunn, S. D. Howe, A. J. Thomson, K. I. Harrison,
J. Chem. Soc. Perkin Trans. 1 1988, 2453 2467; b) D. Wohrle, V.
Schmidt, J. Chem. Soc. Dalton Trans. 1988, 549 551; c) H. Matsuda, S.
Okada, A. Masaki, H. Nakanishi, Y. Suda, K. Shigehara, A. Yamada,
Proc. SPIE Int. Soc. Opt. Eng. 1990, 1337, 105 113; d) M. T. M. Choi,
P. P. S. Li, D. K. P. Ng, Tetrahedron 2000, 56, 3881 3887; e) V. M.
Chem. Eur. J. 2003, 9, 5123 5134
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5133

N. Kobayashi et al.
FULL PAPER
Derkacheva, E. A. Luk×yanets, Zh. Obshch. Khim. 1980, 50, 2313
2318; f) O. V. Dolotova, N. I. Bundina, V. M. Derkacheva, O. L.
Kaliya, E. A. Luk×yanets, Zh. Obshch. Khim. 1988, 58, 2173; g) H.
Weitman, S. Schatz, H. E. Gottlieb, N. Kobayashi, B. Ehrenberg,
Photochem. Photobiol. 2001, 73, 473 481.
[35] D. S. Telekov, K. J. M. Nolan, C. R. McArthur, C. C. Leznoff, J. Org.
Chem. 1996, 61, 3034 3040.
[36] B. del Rey, U. Keller, T. Torres, G. Rojo, F. Agullo-Lopez, S. Nonell, C.
Marti, S. Brasselet, I. Ledoux, J. Zyss, J. Am. Chem. Soc. 1998, 120,
12808 12817.
[34] C. C. Leznoff, S. M. Marcuccio, S. Greenberg, A. B. P. Lever, K. B.
Tomer, Can. J. Chem. 1985, 63, 623 631.
Received: February 6, 2003
Revised: June 24, 2003 [F4834]
5134
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5123 5134