Search for squaraine derivatives that can be sublimed without thermal decomposition

Article abstract of DOI:10.1021/jp013698r

To search for sublimable squaraine (SQ) dyes for fabricating thin films under vacuum, we have synthesized a series of 2,4-bis[4-(N,N-dialkylamino)-2,6-dihydroxyphenyl]squaraines [SQ(OH)₄] and studied their properties. The investigation of the b

Full text of DOI:10.1021/jp013698r

4370
J. Phys. Chem. B 2002, 106, 4370-4376
Search for Squaraine Derivatives That Can Be Sublimed without Thermal Decomposition
Minquan Tian,*,† Makoto Furuki,† Izumi Iwasa,† Yasuhiro Sato,† Lyong Sun Pu,† and
Satoshi Tatsuura‡
Corporate Research Center, Fuji Xerox Co., Ltd., 430 Sakai, Nakai-machi, Ashigarakami-gun,
Kanagawa 259-0157, Japan, and The Femtosecond Technology Research Association,
5-5 Tokodai, Tsukuba, Ibaraki 300-2635, Japan
ReceiVed: October 3, 2001; In Final Form: February 18, 2002
To search for sublimable squaraine (SQ) dyes for fabricating thin films under vacuum, we have synthesized
a series of 2,4-bis[4-(N,N-dialkylamino)-2,6-dihydroxyphenyl]squaraines [SQ(OH)4] and studied their properties.
The investigation of the behavior of their Langmuir films at the air-water interface revealed that SQ(OH)4
molecules with branched N-alkyl groups such as sec-butyls and isobutyls have larger limiting areas and tend
to form J-aggregates in the monolayers, whereas molecules with straight N-alkyl chains have smaller limiting
areas and are apt to form H-aggregates. This behavior is attributable to the much larger steric hindrance of
the branched N-butyl groups than that of the straight ones. The thermal stability of these dyes was investigated
by differential thermal analysis (DTA) and thermogravimetry (TG), and their sublimation ability was evaluated
through heating under vacuum. As a result, we verified that the four hydroxyls at the 2′,6′-positions of the
two phenyl rings significantly enhanced the thermal stability and the sublimation ability of an anilino SQ dye
molecule, and the introduction of branched N-butyls further promoted the sublimation ability of the target
SQ(OH)4 molecules. These phenomena may be attributed to the intramolecular hydrogen bonding between
the hydroxyls and the CO groups of the four-membered ring, and the much larger intermolecular repulsive
force between branched N-alkyls, respectively. In particular, the SQ(OH)4 dye with four N-isobutyls could be
sublimed without any decomposition. These results suggest that SQ(OH)4 molecule with branched N-butyls
is the most effective structure for realizing a high sublimation ability. Furthermore, pure SQ dye thin films
have been successfully fabricated by molecular beam deposition of the SQ(OH)4 dye with four N-isobutyls.
The vacuum-deposited thin films of such SQ dyes have potential applications in various fields such as
electrophotography, solar energy conversion, optical recording, and nonlinear optics.
1. Introduction
the SQ materials employed for fabricating vacuum-deposited
films were mainly 2,4-bis[4-(N,N-dialkylamino)-2-hydroxyphen-
yl]squaraines [SQ(OH)2],4-6,8,10 2,4-bis[4-(N,N-dimethylamino)-
2-methylphenyl]squaraine,5,8,10 and 2,4-bis[4-(N,N-dimethylami-
no)phenyl]squaraine (SQ11).8,10 To fabricate SQ thin films with
high uniformity for nonlinear optics, we repeated the preparation
of vacuum-deposited films from the above-mentioned SQ
materials. As a result, we found that pure vacuum-deposited
films could not be obtained using these known SQ materials,
because their thermal decomposition occurred during the process
of sublimation. Moreover, we also noted that SQ(OH)2 showed
better thermal stability and higher sublimation ability than the
corresponding analogues without any hydroxyl. For example,
the thermal decomposition temperatures for SQ11 and 2,4-bis-
[4-(N,N-dimethylamino)-2-hydroxyphenyl]squaraine [SQ11-
(OH)2], which were obtained by TG, were 277 and 326 °C,
respectively. However, we observed that SQ11 was initially
sublimed at 200 °C and decomposed at 220 °C in a vacuum of
(0.4-1) × 10-3 Pa, whereas SQ11(OH)2 was initially sublimed
at 230 °C and decomposed at 260 °C in the same vacuum
system. Moreover, SQ11 was found to be sublimed less easily
and to decompose more easily than SQ11(OH)2 in the sublima-
tion experiments. Intramolecular hydrogen bonding between the
hydroxyls and the CO groups of the four-membered ring may
be the reason for this behavior. Thus, we think an anilino SQ
dye molecule with four hydroxyls at the 2′,6′-positions of the
two phenyl rings will show even better thermal stability and
Squaraine (SQ) dyes are expected to have significant tech-
nological applications in many areas such as electrophotogra-
phy,1-3 solar energy conversion,4-7 optical recording,8-10
nonlinear optics,11-17 and electroluminescence.18,19 For practical
application, the fabrication of thin films with high uniformity
is strongly required. Although there have been a large number
of reports on SQ thin films fabricated by solution-coating tech-
niques such as Langmuir and Langmuir-Blodgett (LB) meth-
ods,15,16,20-27 dip and bar coating,28-30 or casting and spin
coating,5,6,8,26,27,31,32 several fatal limitations of the solution-
coating method cannot be easily overcome. First, SQ thin films
prepared by solution coating usually do not have a smooth
surface, which is very important for the optical application of
SQ dyes. Second, the thickness homogeneity of SQ thin films
prepared by solution coating is poor. Moreover, the desired
thickness of SQ thin films is difficult to control by solution
coating. These limitations may be overcome by the vacuum-
deposition method, because dye molecules may be evaporated
very slowly and deposited onto the substrate one by one in
ultrahigh vacuum. Nevertheless, only a few reports on vacuum-
deposited films of SQ dyes can be found in the literature, and
* Telephone: +81-465-80-2276. Fax: +81-465-81-8997. E-mail: Tian.
Minquan@fujixerox.co.jp.
† Fuji Xerox Co., Ltd.
‡ The Femtosecond Technology Research Association.
10.1021/jp013698r CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/03/2002

Squaraine Derivatives That Can Be Sublimed
J. Phys. Chem. B, Vol. 106, No. 17, 2002 4371
2. Experimental Section
2.1. General Techniques. IR spectra were recorded on a
JASCO FTIR-23 Fourier transform infrared spectrometer, using
1
potassium bromide pellets. H NMR spectra with tetramethyl-
silane as the internal standard were recorded on a JEOL JNM-
AL400 Fourier transform nuclear magnetic resonance spec-
trometer (400 MHz). Electron ionization (EI) mass spectra were
recorded on a Shimadzu QP-5000 mass spectrometer. UV-
visible absorption spectra were measured on a Hitachi U-4100
spectrophotometer in a quartz cell of path length 10 mm.
Reagent chemicals and solvents were used as purchased without
further purification.
2.2. Materials. SQ dyes 1-7 were synthesized by a modi-
fication of the reference method.14 5-N,N-Dialkylamino-1,3-
dihydroxybenzene was prepared by the nucleophilic substitution
of phloroglucinol with a secondary amine in a toluene-1-butanol
solvent system under reflux with azeotropic distillation of water.
The condensation of 5-N,N-Dialkylamino-1,3-dihydroxybenzene
and squaric acid in the toluene-1-butanol solvent system under
azeotropic reflux produced the desired SQ dye.47,48 For the
preparation of dyes 1 and 5-7, a one-pot reaction procedure
similar to the reference method was used; i.e., the corresponding
intermediate anilines were condensed with squaric acid without
any separation. However, for the preparation of dyes 2-4, the
corresponding intermediate anilines were separated and purified
before condensation with squaric acid. The latter modified
method gave a better yield of the desired SQ dye and is more
valid for the preparation of SQ dyes with branched N-alkyls.
New compounds 1-4 were synthesized for the first time. A
representative procedure for the synthesis of this class of SQ
dyes is described as follows.
2,4-Bis[4-(N-sec-butyl-N-n-propylamino)-2,6-dihydroxy-
phenyl]squaraine (3). Under dry nitrogen atmosphere, a mix-
ture of 1,3,5-trihydroxybenzene (7.92 g, 62.8 mmol), N-sec-
butyl-n-propylamine (14.48 g, 125.7 mmol), 1-butanol (50 mL),
and toluene (150 mL) was refluxed for 6 h with azeotropic
distillation of water. The pale-brown solution was then cooled,
and the solvents were evaporated under reduced pressure to give
a brown liquid. The resulting crude product was purified by
column chromatography on silica gel with n-hexane/acetone
(volume ratio 4:1 ∼ 3:2) as the eluents to give a pale brown-
yellow viscous liquid of 5-N-sec-butyl-N-n-propylamino-1,3-
dihydroxybenzene, 4.61 g (32.9%). Then, a mixture of this
aniline intermediate (4.47 g, 20 mmol), squaric acid (1.14 g,10
mmol), 1-butanol (40 mL), and toluene (120 mL) was stirred
under dry nitrogen atmosphere and refluxed for 5 h with
azeotropic distillation of water. After toluene was distilled,
1-butanol was removed by azeotropic distillation with cyclo-
hexane (70 mL). The final cyclohexane solution of ap-
proximately 15 mL was cooled, and the black-green pre-
cipitate was filtered and washed with n-hexane and meth-
anol. Yellow-green crystals were recrystallized from dichlo-
romethane-methanol to give shining yellow-green needle mi-
crocrystals, 4.16 g (79.3%). Total yield: 26.1%. IR: νmax
3430, 3087, 2965, 2925, 2870, 2695, 1625 (CdO), 1552, 1516,
1431, 1397, 1328 (C-N), 1285, 1251 (OH), 1160, 1119, 1057,
989, 965, 902, 867, 819, 796, 735 cm-1. 1H NMR (CDCl3): δ
10.941 (s, 4H, OH), 5.861 (s, 4H, Harom), 3.946 (m, 2H,
NCH), 3.172 (m, 4H, NCH2), 1.728∼1.606 (m, 8H, CH2),
1.234 (d, J ) 6.59 Hz, 6H, NCCH3), 0.952 (t, J ) 7.32 Hz,
6H, CH3), 0.893 (t, J ) 7.32 Hz, 6H, CH3). MS (EI): m/z 524
(M+, 50%), 495 (M+-C2H5, 100%). Anal. Calcd for C30H40O6N2
(524.65): C, 68.7; H, 7.68; N, 5.34. Found: C, 67.2; H, 7.55;
N, 5.18.
Figure 1. Molecular structure of the SQ derivatives synthesized.
higher sublimation ability and may be well sublimed without
any thermal decomposition.
On the other hand, some derivatives of SQ(OH)4, such as
the N,N-dimethyl,14,33,34 N,N-di-n-butyl,14,35-40 N,N-di-n-pent-
yl,35-37,41-44 N,N-di-n-hexyl,14,35-37,41-45 N-methyl-N-alkyl,27,46
and N,N-bis(2-n-hexyloctyl) analogues,47 have been prepared,
and their xerographic and nonlinear optical properties in
solution-coated films or in solution have also been investigated.
Moreover, Dirk et al. reported that the melting point of the N,N-
di-n-butyl derivative of SQ(OH)4 was higher by 45 °C than that
of the corresponding SQ(OH)2 analogue.14 This result suggests
that SQ(OH)4 dyes should have better thermal stability and
higher sublimation ability than the corresponding SQ(OH)2
analogues. However, no information about the sublimation
ability of SQ(OH)4 dyes has been reported thus far.
In order for an anilino SQ dye to be sublimable, we think
that the single SQ molecule should be thermally stable.
Moreover, the interactions between molecules of such an SQ
dye should be sufficiently small for the molecules to be
evaporated in a vacuum. Therefore, the substituents that can
improve the thermal stability and reduce the intermolecular
interactions of an anilino SQ dye are favorable for its sublima-
tion ability.
With such knowledge in mind, we designed and synthesized
a series of SQ(OH)4 derivatives (Figure 1), to find the most
effective molecular structure for thin-film fabrication under
vacuum. There are several reasons for such a molecular design.
First, the introduction of four hydroxyls at the 2′,6′-positions
of the two phenyl rings may enhance the thermal stability of
the target SQ dyes to such an extent that they could be sublimed
without any thermal decomposition. Second, N-alkyl chains were
varied to control intermolecular interactions so that the most
effective molecular structure for high sublimation ability could
be picked out. In particular, branched N-alkyl chains such as
isobutyls and sec-butyls, which exhibit larger steric hindrance
effects, may suppress the intermolecular hydrogen bonding
between the hydroxyls and further promote the sublimation
ability of the target SQ materials. In addition, the synthetic
feasibility was also considered in the design of such SQ dye
molecules.
In this paper, we report the synthesis, thermal stability, and
sublimation ability of these SQ derivatives. The behavior of
their Langmuir films at the air-water interface was also
investigated in order to understand their intermolecular interac-
tions in the solid phase. Consequently, we found the most
effective structure of SQ molecules for realizing a high
sublimation ability. Furthermore, pure SQ thin films were
fabricated by the molecular beam deposition of those sublimable
SQ dyes.

4372 J. Phys. Chem. B, Vol. 106, No. 17, 2002
Tian et al.
2,4-Bis[4-(N,N-diethylamino)-2,6-dihydroxyphen-
yl]squaraine (1). Shining green microcrystals, total yield:
20.0%. IR: νmax 3418, 3087, 2980, 2929, 2702, 1627 (CdO),
1565, 1528, 1432, 1402, 1329 (C-N), 1266 (OH), 1165, 1134,
1118, 1077, 981, 845, 793, 780, 733 cm-1. 1H NMR (CDCl3):
δ 10.986 (s, 4H, OH), 5.807 (s, 4H, Harom), 3.431 (q, J ) 7.08
Hz, 8H, NCH2), 1.243 (t, J ) 7.08 Hz, 12H, CH3). MS (EI):
m/z 440 (M+, 100%), 425(M+-CH3, 45.5%). Anal. Calcd for
C24H28O6N2 (440.49): C, 65.4; H, 6.41; N, 6.36. Found: C,
64.1; H, 6.19; N, 6.22.
2,4-Bis[4-(N,N-dipropylamino)-2,6-dihydroxyphen-
yl]squaraine (2). Shining blue-green needle microcrystals, total
yield: 25.5%. IR: νmax 3422, 3087, 2964, 2929, 2877, 2695,
1626 (CdO), 1559, 1433, 1410, 1329 (C-N), 1298, 1251 (OH),
Figure 2. UV-visible absorption spectrum of dye 4 in 1,2-dichloro-
ethane.
1
1164, 1133, 1093, 919, 867, 819, 795, 733 cm-1. H NMR
TABLE 1: Absorption Properties of Dyes 1-7 in
1,2-Dichloroethane
(CDCl3): δ 10.974 (s, 4H, OH), 5.781 (s, 4H, Harom), 3.307 (t,
J ) 7.81 Hz, 8H, NCH2), 1.670 (m, 8H, CH2), 0.953 (t, J )
7.57 Hz, 12H, CH3). MS (EI): m/z 496 (M+, 95.6%), 467 (M+-
C2H5, 100%). Anal. Calcd for C28H36O6N2 (496.60): C, 67.7;
H, 7.31; N, 5.64. Found: C, 66.4; H, 7.18; N, 5.48.
2,4-Bis[4-(N,N-diisobutylamino)-2,6-dihydroxyphen-
yl]squaraine (4). Shining green-brown needle microcrystals,
total yield: 23.9%. IR: νmax 3444, 3087, 2959, 2926, 2871,
2708, 1630 (CdO), 1550, 1498, 1430, 1402, 1349, 1322
(C-N), 1257 (OH), 1163, 1119, 1095, 961, 948, 926, 890, 870,
808, 797, 743 cm-1. 1H NMR (CDCl3): δ 10.963 (s, 4H, OH),
5.805 (s, 4H, Harom), 3.237 (d, J ) 7.32 Hz, 8H, NCH2), 2.124
(m, 4H, NCH2CH), 0.925 (d, J ) 6.59 Hz, 24H, CH3). MS
(EI): m/z 552 (M+, 43.3%), 509 (M+-C3H7, 100%). Anal.
Calcd for C32H44O6N2 (552.70): C, 69.5; H, 8.02; N, 5.07.
Found: C, 67.8; H, 7.97; N, 4.94.
SQ dye
λ
max (nm)
ꢀ (dm3 mol-1 cm-1
)
1
2
3
4
5
6
7
644.0
648.0
651.0
652.2
649.2
650.2
650.6
3.01 × 105
3.45 × 105
3.32 × 105
3.20 × 105
3.36 × 105
3.39 × 105
3.47 × 105
an ultrahigh vacuum of (0.2-2) × 10-6 Pa. The dye film was
evaporated onto a 3×3-cm2 glass substrate of 22 °C, and the
deposition rate was monitored using a quartz crystal oscillator
thickness meter. Heating temperature was controlled to give a
deposition rate of 0.1-0.9 Å/s.
3. Results and Discussion
2.3. Langmuir Films. A moving-wall LB-film deposition
apparatus (Nippon Laser & Electronics Lab., NL-LB240-MWA)
with a Wilhelmy balance was employed for the preparation of
Langmuir films. Formation of Langmuir films of the SQ dye
was achieved by spreading a 1,2-dichloroethane solution at a
concentration of 8.33 × 10-4 mol/dm3 (i.e., 20 µL ) 1 × 1016
molecules) onto the pure water subphase of the trough with the
water temperature maintained at 15 °C. In situ absorption spectra
of the floating monolayers were obtained as a function of surface
pressure by means of a fiber-optic multichannel photodiode array
spectrometer (Otsuka Denshi, MCPD-100).
2.4. Thermal Analysis. The thermal stability of dyes 1-7
was investigated using a Shimadzu DTG-50 simultaneous DTA-
TG measuring device. The measurements were carried out from
room temperature to 500 °C in an atmosphere of argon gas with
a flow rate of 30 mL/min, at a heating rate of 2 °C/min. Melting
points and thermal decomposition temperatures were determined
from the resulting DTA and TG curves.
2.5. Evaluation of Sublimation Ability. The sublimation
ability of dyes 1-7 was evaluated through heating under
vacuum. The SQ sample was put in a glass crucible, and the
crucible was located at one end of a 60-cm-long, 3-cm-diameter
transparent glass test tube. The tube was connected to an
ULVAC VPC-250FA vacuum pump unit through a liquid
nitrogen trap and evacuated to a vacuum lower than 1 × 10-3
Pa. Then, the sample was gradually heated under high vacuum
using a tubular electric furnace. The thermal behavior of the
sample was monitored during the heating process.
2.6. Fabrication of SQ Thin Films by Molecular Beam
Deposition. Thin films of those sublimable SQ(OH)4 dyes were
prepared using an ULVAC BC4558 molecular beam epitaxy
apparatus. The SQ sample was dispersed across the entire
bottom of a quartz crucible and slowly heated in a K-cell under
3.1. Spectroscopic Properties in Solution. Except for 1 and
2, SQ dyes 3-7 exhibit good solubility in halogenated solvents
such as dichloromethane, 1,2-dichloroethane, and chloroform.
The solution spectra of this class of SQ dyes are similar and,
for example, in 1,2-dichloroethane have a sharp absorption band
at approximately 650 nm with a well-defined shoulder at about
600 nm (Figure 2). This asymmetry is a familiar property of
SQ dyes and may be attributed to aggregation, even at low
concentrations.27 The maximum absorption wavelengths (λmax
)
and the molar extinction coefficients (ꢀ) of dyes 1-7 in 1,2-
dichloroethane are summarized in Table 1. It is noted that the
order of λmax variation for these compounds is 1 < 2 < 5 < 6
< 7 < 3 < 4. The reason for this phenomenon may be that the
steric hindrance of N-alkyl causes the amino groups to rotate
slightly or to vary in its ability to sp2 hybridize to participate in
delocalization. We think that the larger steric hindrance effect
of two neighboring N-alkyl chains conduces to the better
participation of the amino groups in π-electron delocalization,
resulting in the redshift of peak absorption. Since the steric
hindrance of straight N-alkyl chain increases as the alkyl chain
length grows, and branched N-alkyl chains such as sec-butyls
and isobutyls have even larger steric hindrance than the straight
ones, the λmax of the corresponding dyes will vary in the order
as observed.
3.2. Langmuir Films at the Air-Water Interface. There
is a report in the literature on the Langmuir films of dye 5 at
the air-water interface.38 Our investigation confirmed the earlier
study of Ashwell et al.38 As a representative result, the surface
pressure-area (π-A) isotherms of dyes 2-5 and 7 are shown
in Figure 3. Similar to the behavior of the N-methyl-N-alkyl
analogues at the air-water interface,27 the π-A isotherms of
this class of SQ dyes are generally featureless. The limiting
areas for dyes 2-5 and 7, obtained by extrapolating the high-

Squaraine Derivatives That Can Be Sublimed
J. Phys. Chem. B, Vol. 106, No. 17, 2002 4373
Figure 5. Typical absorption spectra of dyes 3 (solid line) and 4
(broken line) at the air-water interface for surface pressures of 10 e
π e 20 mN m-1
.
Figure 3. Surface pressure versus area isotherms of dyes 2-5 and 7
at ca. 15 °C.
Figure 4. Typical absorption spectra of dyes 2 (solid line) and 7
(broken line) at the air-water interface for surface pressures of 3 e π
e 10 mN m-1
.
Figure 6. DTA-TG analysis curves for dye 3 (R1 ) n-propyl, R2 )
sec-butyl).
pressure region of the isotherm to π ) 0, are 28, 46.5, 45, 38,
and 40 Å2 molecule-1, respectively. This result indicates that
the longer straight N-alkyl chain has a larger steric hindrance
than the shorter one, and the branched N-alkyl chains such as
sec-butyls and isobutyls possess even larger steric hindrance
than the straight ones, as we stated above. Since the van der
Waals dimensions of the SQ chromophore, including the first
methylene attached to each of the nitrogen atoms but without
any hydroxyl group, are estimated to be ca. 19 Å (length) × 8
Å (width) × 3.4 Å (thickness),27 the limiting areas for dyes
1-7 are smaller than 19 Å (length) × 3.4 Å (thickness). This
fact suggests that the SQ(OH)4 molecules not only have very
good coplanarity between the phenyl rings and the four-
membered ring, but also exhibit a large intermolecular aggrega-
tion effect at the air-water interface. The good molecular
planarity may be attributed to the intramolecular hydrogen
bonding between the hydroxyls and the centric CO groups, and
the large aggregation effect between SQ(OH)4 molecules is
believed to come from the intermolecular hydrogen bonding
between the hydroxyls.
The absorption spectra of the Langmuir films at the air-
water interface are dependent not only on the alkyl chain length
but also on whether the alkyl chains are branched or not. The
straight-chain analogues 1, 2, and 5-7 have a similar absorption
type, i.e., main H-aggregate bands with maxima at ca. 540 nm
and monomeric bands with maxima at 650-700 nm (Figure
4). The N-alkyl chain length has an influence on the maximum
absorption position of the monomeric band. For example, dyes
5-7 have monomeric absorption bands with maxima at ca. 650
nm, whereas dyes 1 and 2 exhibit monomeric absorption bands
with maxima at ca. 700 nm. On the other hand, the branched-
butyl analogues 3 and 4 show a different absorption type, i.e.,
sharp J-aggregate bands with maxima at ca. 700-730 nm and
a weak shoulder in the range of 550-660 nm (Figure 5). These
phenomena can be explained from the intermolecular interac-
tions in the Langmuir films. In the case of the straight-chain
analogues, we think that their monolayer structure is dominated
by chromophore-induced interactions that consist mainly of the
intermolecular hydrogen bonding between the aromatic hy-
droxyls. Since such intermolecular hydrogen bonding favors a
parallel face-to-face arrangement of the chromophores, the
formation of H-aggregates at the air-water interface will be
markedly promoted.49 Conversely, it is assumed that the van
der Waals interactions between branched alkyl chains dictate
the monolayer structure of the branched-alkyl analogues, and
as a result, the strong steric hindrance effects stemming from
branched butyls can suppress the intermolecular hydrogen
bonding between the hydroxyls, and favor not the parallel face-
to-face arrangement but the parallel face-to-breast alignment of
the chromophores; therefore, the formation of J-aggregate in
the Langmuir films will be significantly facilitated.49 The
behavior of the branched-butyl analogues 3 and 4 at the air-
water interface demonstrated that the introduction of branched
N-alkyls could provide sufficiently large intermolecular steric
hindrance for the target SQ(OH)4 molecules to overcome the
intermolecular hydrogen bonding. In particular, because dye 3
has a slightly larger molecular limiting area than 4, and the
Langmuir films of 3 show J-aggregate absorption bands at the
longer wavelength side than of 4, we believe that the intermo-
lecular steric hindrance of dye 3 is slightly larger than of 4.
3.3. Thermal Stability. Representative DTA-TG analysis
curves for dyes 3 and 4 are shown in Figures 6 and 7,
respectively. Dye 3 exhibited a crystal phase transition at 183.6
°C and a clear melting point at 219.2 °C. From the TG curve,
its thermal decomposition temperature was found to be 237.0
°C. This result is consistent with the DTA analysis. Meanwhile,

4374 J. Phys. Chem. B, Vol. 106, No. 17, 2002
Tian et al.
TABLE 3: Evaluation Results of the Sublimation Ability of
Dyes 1-7a
initial sublim.
initial decomp.
sublim. compl.
SQ dye
temp. (°C)
temp. (°C)
temp. (°C)
1
2
3
4
5
6
7
180
180
150
170
170
180
190
230
260
220
260
220
250
230
220
220
a The vacuum for sublimation experiments is (10∼4) × 10-4 Pa.
chain to decompose on heating will be higher than those for
short straight alkyl chains such as n-propyls and n-butyls. For
these reasons, 3 shows a much lower melting point and a lower
thermal decomposition temperature than 2, 5, and 6. On the
other hand, 4 has four branching points that are relatively far
from the phenyl ring and symmetrically distributed at the two
ends of the dye molecule. Such substitution reduces the
possibility for the isobutyl chains to decompose on heating, and
compared with 3, 4 exhibits more-compact spatial stacking of
molecules. Because 2 and 4-7 have the same molecular
symmetry and the same backbone structure, in this case, it is
assumed that their melting behavior is dominated by the
flexibility of the four N-alkyl chains. Obviously, bulky isobutyl
groups will have much lower flexibility than n-propyl, n-butyl,
n-pentyl, and n-hexyl groups. Therefore, 4 will have much lower
molecular flexibility than 2 and 5-7. These may be the reasons
why 4 shows a much higher melting point and a much higher
thermal decomposition temperature than 2, 3, and 5-7.
To further verify that an SQ(OH)4 dye will show even better
thermal stability than the corresponding dihydroxy analogue,
we compare the thermal stability of 2,4-bis[4-(N,N-dipropyl-
amino)-2-hydroxyphenyl]squaraine [SQ33(OH)2]50 with that of
2. The thermal decomposition temperatures of SQ33(OH)2 and
dye 2, which were measured by DTA-TG analysis, are 226.0
and 277.5 °C, respectively. Apparently, dye 2 shows much
higher thermal stability than SQ33(OH)2. Such a high thermal
stability helps to improve the sublimation ability of the same
dye molecules and hence is preferable for sublimable dyes.
3.4. Sublimation Ability. The evaluation results of the
sublimation ability of SQ dyes 1-7 are summarized in Table
3. Dyes 1, 6, and 7 could be slightly sublimed under high
vacuum when heated to 180, 180, and 190 °C, respectively,
and on further heating, their sublimation rate increased very
slowly, with most of their molecules decomposing before
sublimation at higher temperatures. Dye 5 could be sublimed
without decomposition at 170-230 °C under high vacuum, and
its sublimation rate increased rapidly when it began to melt at
230 °C, owing to the significant improvement in the thermal
conduction of the melting sample as compared with that of the
unmelted sample powders. However, slight decomposition of
its molecules also occurred at 230 °C. Similarly, dye 2 could
be sublimed without decomposition at 180-260 °C under high
vacuum, and its sublimation rate was sufficiently high at 250
°C to allow most of its molecules to evaporate. However, slight
decomposition of its molecules also occurred at 260 °C.
Regarding dye 3, its sublimation started even at 150 °C under
high vacuum, and the color of the sample at 190 °C changed
from yellow-green to brown-green, and finally to golden brown,
indicating the occurrence of a crystal phase transition. This
compound could be well sublimed without decomposition at
temperatures below 220 °C, and its sublimation rate increased
rapidly when heated at a temperature near its melting point.
Figure 7. DTA-TG analysis curves for dye 4 (R1 ) R2 ) isobutyl).
TABLE 2: Melting Points and Thermal Decomposition
Temperatures of Dyes 1-7
SQ dye
melting point (°C)
thermal decomp. temp. (°C)
1
2
3
4
5
6
7
>290
277.1
219.2
294.0
235.1
224.0
204.6
288.6
277.5
237.0
299.0
242.0
244.0
239.1
dye 4 melted at 294.0 °C and decomposed at 299.0 °C. The
melting points and thermal decomposition temperatures of dyes
1-7 are given in Table 2. For the straight-chain analogues, both
the melting points and the thermal decomposition temperatures
decreased as the N-alkyl chain length increased. This phenom-
enon may be due to the influence of flexible N-alkyl chains.
The longer straight N-alkyl chain has larger steric hindrance
and hence would lead to less-compact spatial stacking of the
corresponding SQ(OH)4 molecules. Moreover, the longer N-
alkyl chain will exhibit higher possibility to decompose on
heating. These effects result in the decrease of the melting points
and the thermal stability of the corresponding SQ(OH)4 dyes
with increasing chain length. However, for the branched-butyl
analogues, their thermal behavior is slightly peculiar. Compound
3, which has two N-n-propyls and two N-sec-butyls, shows a
lower melting point and a lower thermal decomposition tem-
perature than 2, 5, and 6, which are substituted with four
n-propyls, n-butyls, and n-pentyls, respectively. However, 4,
which has four N-isobutyls, exhibits the highest melting point
and the highest thermal decomposition temperature among 2-7.
These phenomena may be due to the influence of the branching
of N-alkyl chains. The branching position and the number of
branching points determine the total influence of branched
N-alkyl chains on the thermal properties of the corresponding
SQ dye. The branching position closer to the phenyl ring would
result in a larger steric hindrance and a higher possibility for
the branched N-alkyl chain to decompose on heating. On the
other hand, the number of branching points will influence the
symmetry and spatial stacking of SQ molecules. Compound 3
has two sec-butyls, and the two branching points are directly
connected to the nitrogen atoms. Such a molecular structure
will result in three effects. The first one is that 3 will have lower
molecular symmetry because of the possible free rotation of
C-N bonds; thus, molecules of 3 have looser spatial stacking
in the solid state. The second effect is that 3 will have the largest
molecular free volume in the solid-state owing to the largest
intermolecular steric hindrance, as revealed by its Langmuir
films. The third effect is that the possibility for the sec-butyl

Squaraine Derivatives That Can Be Sublimed
J. Phys. Chem. B, Vol. 106, No. 17, 2002 4375
Most of this sample could be sublimed at 220 °C, although very
slight decomposition of its molecules also occurred at 220 °C.
In contrast, dye 4 could be well sublimed without decomposition
from 170 °C under high vacuum, and its sublimation rate
increased with heating temperature. We found that it was
completely sublimed without any decomposition when heated
to 250 °C, which was far lower than its melting point and
thermal decomposition temperature.
It is necessary for us to explain what high sublimation ability
means to a dye. Generally speaking, heating is required for a
dye to be sublimed under vacuum. Thus, a dye with high
sublimation ability should, first of all, be sufficiently thermally
stable during the process of sublimation; i.e., its initial decom-
position temperature should be higher than its initial sublimation
temperature, and a larger difference between these two tem-
peratures is preferable. Second, for such a dye, its sublimation
rate should increase with heating temperature and be control-
lable. Moreover, it should be possible for such a dye to be
sublimed completely far below its initial decomposition tem-
perature. In fact, it is difficult to judge whether a dye has high
sublimation ability without conducting practical sublimation
experiments.
Figure 8. UV-visible absorption spectrum of a 57.5-nm-thick, mo-
lecular-beam-deposited film of dye 4.
three and four. After that, the attractive force between the alkyl
chains dominates again as the alkyl chain length increases
further. This is why dyes 2 and 5 exhibit the maximum
sublimation ability among the straight-chain analogues. More-
over, when branched butyls replace the straight alkyl chains,
the additional steric hindrance caused by the branching of butyls
enhances the intermolecular repulsive force and suppresses the
intermolecular hydrogen bonding between the hydroxyls, as
demonstrated in the Langmuir films of these SQ(OH)4 dyes.
Thus, the sublimation ability of the corresponding SQ dye
increases. This is why dyes 3 and 4 show lower temperatures
of initial sublimation than the other SQ(OH)4 dyes. In addition,
although dye 4 has a much higher melting point, its sublimation
ability depends mainly on the intermolecular repulsive interac-
tions. Since the presence of four bulky isobutyls would result
in the largest intermolecular repulsive force, dye 4 exhibits the
highest sublimation ability among all the seven compounds.
These results suggest that an SQ(OH)4 molecule with branched
N-butyls is the most effective structure for realizing a high
sublimation ability. Although we have not succeeded in the
preparation of SQ(OH)4 molecules with four N-sec-butyls or
N-tert-butyls, or with a combination of isobutyl, sec-butyl, and
tert-butyl groups at the nitrogen atoms, these molecules can also
be expected to show even higher sublimation ability.
To verify the effectivity of our molecular design, the
sublimation ability of SQ33(OH)2 was also investigated for
comparison. SQ33(OH)2 could only be slightly sublimed when
heated at 150 °C under high vacuum. On further heating, its
sublimation rate increased very slowly, with most of its
molecules decomposing before sublimation at temperatures
higher than 180 °C. Undoubtedly, SQ33(OH)2 exhibited a much
worse sublimation ability than dye 2. These results demonstrate
that the four hydroxyls at the 2′,6′-positions of the two phenyl
rings enhance the sublimation ability of an anilino SQ dye
molecule much more significantly than the two hydroxyls at
the same 2′-positions. Furthermore, with the introduction of
branched N-butyls, the target SQ(OH)4 molecules can have
sufficiently high sublimation ability for thin-film preparation
under vacuum.
Some trends regarding the sublimation ability of these SQ
dyes can be seen from Table 3. For the straight-chain analogues,
the sublimation ability increases with the carbon number of the
N-alkyl chain first, reaches a maximum at carbon numbers three
and four, and then decreases again as the N-alkyl chain length
increases further. Moreover, the two branched-butyl analogues
show higher sublimation ability than the straight-chain ones,
and dye 4 is the most sublimable material among these dyes.
These phenomena can be explained from the inter- and
intramolecular interactions of SQ molecules. The intramolecular
hydrogen bonding between the hydroxyls and the centric CO
groups greatly enhances both the thermal stability and the
sublimation ability of the corresponding SQ(OH)4 molecules
to such an extent that the initial sublimation temperature is much
lower than the thermal decomposition temperature. Thus, SQ
dyes 1-7 were rendered sublimable. On the other hand, the
N-alkyl chains have two influences on the intermolecular
interactions. One is the attractive force between the alkyl chains,
which grows with the chain length, and the other one is the
repulsive force coming from the steric hindrance of the alkyl
chains. The balance of these two forces determines the total
influence of the N-alkyl chains on the sublimation ability of
the SQ analogues. In general, the sublimation ability benefits
from the intermolecular repulsive force. For the straight-chain
analogues, the attractive force dominates when the alkyl chain
has a carbon number less than three, while the repulsive force
dominates as the carbon number of the alkyl chain increases to
3.5. Molecular-beam-deposited Films. Since only dye 4
could be completely sublimed without any decomposition, this
dye was employed for thin-film fabrication under vacuum. Pure
thin films of dye 4 were successfully fabricated by molecular
beam deposition at a rate of 0.1-0.8 Å/s when the sample was
heated at 245 °C under an ultrahigh vacuum of (0.3-2) × 10-6
Pa. The UV-visible absorption spectrum of such a film with a
thickness of 57.5 nm is shown in Figure 8. The molecular-beam-
deposited (MBD) film of dye 4 exhibited a broad absorption
band with the peak located at 680 nm. Compared to the
absorption of its solution, the absorption of the MBD film is
redshifted by 28 nm. This indicates that J-like aggregates were
formed in the MBD films. Such aggregates are very important
for application in nonlinear optics.17,31,32 The behavior of dye
4 in the MBD films is similar to that in its monolayers at the
air-water interface. Because MBD films have higher surface
flatness and better thickness homogeneity than solution-coated
films, and their thicknesses are precisely controllable, such thin
films of sublimable SQ dyes such as 4 are indispensable for
fabricating many optical and electrooptic devices containing SQ
films.
4. Conclusions
As an effort to obtain sublimable SQ materials superior to
known SQ dyes, we have synthesized a series of tetrahydroxy
anilino SQ derivatives, SQ(OH)4 dyes 1-7 and studied their

4376 J. Phys. Chem. B, Vol. 106, No. 17, 2002
Tian et al.
(12) Zu, Y. Z.; Shi, R. F.; Garito, A. F.; Grossman, C. H. Opt. Lett.
1994, 19, 786.
(13) Andrews, J. H.; Khaydarov, J. D. V.; Singer, K. D. Opt. Lett. 1994,
properties. The behaviors of their Langmuir films at the air-
water interface demonstrate that the longer straight N-alkyl chain
has a larger steric hindrance than the shorter one, and the
branched N-alkyl chains such as sec-butyls and isobutyls possess
even larger steric hindrance than the straight ones. It is also
manifested that the intermolecular hydrogen bonding between
the aromatic hydroxyls favors the formation of SQ(OH)4
H-aggregates at the air-water interface. However, the branching
of N-alkyl chains greatly suppresses the formation of H-
aggregates and facilitates the formation of J-aggregates in the
Langmuir films, owing to the strong steric hindrance of the
bulky branched N-alkyls. The evaluation of the monolayers at
the air-water interface proved effective in assessing intermo-
lecular interactions, especially intermolecular steric hindrance.
Moreover, from investigations of dyes 1-7 by DTA and TG,
as well as by heating under vacuum, we demonstrated that the
four hydroxyls at the 2′,6′-positions of the two phenyl rings
enhanced the thermal stability and the sublimation ability of an
anilino SQ dye molecule much more significantly than the two
hydroxyls at the same 2′-positions, and the introduction of
branched N-butyls further lowered the sublimation temperatures
of the target SQ(OH)4 molecules. It was also clarified that the
intramolecular hydrogen bonding between the hydroxyls and
the centric CO groups, and the branching of N-alkyl chains can
greatly enhance the sublimation ability of the corresponding SQ-
(OH)4 molecules, whereas the intermolecular hydrogen bonding
between the hydroxyls should be suppressed to improve their
sublimation ability. In particular, dye 4, which has four N-iso-
butyls, exhibited excellent sublimation ability and could be
sublimed without any decomposition. Furthermore, pure SQ dye
thin films that have been obtained only by solution coating so
far were successfully fabricated by molecular beam deposition
of dye 4. Similar to the behavior in the Langmuir films at the
air-water interface, not H-aggregates but J-like aggregates were
formed in the MBD thin films of dye 4. Such thin films of SQ
dyes may find technological applications in many areas such
as electrophotography, solar energy conversion, optical record-
ing, and nonlinear optics.
19, 984.
(14) Dirk, C. W.; Herndon, W. C.; Cervantes-Lee, F.; Selnau, H.;
Martinez, S.; Kalamegham, P.; Tan, A.; Campos, G.; Velez, M.; Zyss, J.;
Ledoux, I.; Cheng, L.-T. J. Am. Chem. Soc. 1995, 117, 2214.
(15) Ashwell, G. J.; Jefferies, G.; Hamilton, D. G.; Lynch, D. E.; Roberts,
M. P. S.; Bahra, G. S.; Brown, C. R. Nature 1995, 375, 385.
(16) Ashwell, G. J. AdV. Mater. 1996, 8, 248.
(17) Furuki, M.; Pu, L. S.; Sasaki, F.; Kobayashi, S.; Tani, T. Appl.
Phys. Lett. 1998, 21, 2648.
(18) Mori, T.; Miyachi, K.; Kichimi, T.; Mizutani, T. Jpn. J. Appl. Phys.
1994, 33, 6594.
(19) Zhang, B.; Zhao, W.; Cao, Y.; Wang, X.; Zhang, Z.; Jiang, X.;
Xu, S. Synth. Met. 1997, 91, 237.
(20) Swalen, J. D.; Tacke, M.; Santo, R.; Rieckhoff, K. E.; Fisher, J.
HelV. Chem. Acta 1978, 61, 960.
(21) Kawabata, Y.; Sekiguti, T.; Tanaka, T.; Nakamura, T.; Koizumi,
H.; Honda, K.; Manda, E.; Saito, M.; Sugi, M.; Iizima, S. J. Am. Chem.
Soc. 1985, 107, 5270.
(22) Kim, S.; Furuki, M.; Pu, L. S.; Nakahara, H.; Fukuda, K. J. Chem.
Soc., Chem. Commun. 1987, 1201.
(23) Kim, S.; Furuki, M.; Pu, L. S.; Nakahara, H.; Fukuda, K. Thin
Solid Films 1988, 159, 1201.
(24) Law, K. Y.; Chen, C. C. J. Phys. Chem. 1989, 93, 2533.
(25) Liang, K.; Law, K. Y.; Whitten, D. G. J. Phys. Chem. 1994, 98,
13379.
(26) Ashwell, G. J.; Wong, G. M. S.; Bucknall, D. G.; Bahra, G. S.;
Brown, C. R. Langmuir 1997, 13, 1629.
(27) Ashwell, G. J.; Jefferies, G.; Rees, N. D.; Williamson, P. C.; Bahra,
G. S.; Brown, C. R. Langmuir 1998, 14, 2850.
(28) Loutfy, R. O.; Hsiao, C. K.; Kazmaier, P. M. Photogr. Sci. Eng.
1983, 27, 5.
(29) DiPaola-Baranyi, G.; Hsiao, C. K.; Kazmaier, P. M.; Burt, R.;
Loutfy, R. O.; Martin, T. I. J. Imaging Sci. 1988, 32, 60.
(30) Law, K. Y. J. Imaging Sci. 1990, 34, 38.
(31) Tatsuura, S.; Furuki, M.; Tian, M.; Sato, Y.; Pu, L. S. Mater. Res.
Soc. Symp. Proc. 1999, 561, 105.
(32) Tatsuura, S.; Tian, M.; Furuki, M.; Sato, Y.; Pu, L. S.; Wada, O.
Jpn. J. Appl. Phys. 2000, 39, 4782.
(33) Baranyi, G.; Burt, R. A.; Hsiao, C. K.; Kazmaier, P. M.; Carmichael,
K. M.; Horgan, A. M. U. S. Patent 4,471,041, 1984.
(34) Kazmaier, P. M.; Burt, R.; DiPaola-Baranyi, G.; Hsiao, C. K.;
Loutfy, R. O.; Martin, T. I.; Hamer, G. K.; Bluhm, T. L.; Taylor, M. G. J.
Imaging Sci. 1988, 32, 1.
(35) Garvey, D. W.; Zimmerman, K.; Young, P.; Tostenrude, J.;
Townsend, J. S.; Zhou, Z.; Lobel, M.; Dayton, M.; Wittorf, R.; Kuzyk, M.
G.; Sounick, J.; Dirk, C. W. J. Opt. Soc. Am. B 1996, 13, 2017.
(36) Mathis, K. S.; Kuzyk, M. G.; Dirk, C. W.; Tan, A.; Martinez, S.;
Gampos, G. J. Opt. Soc. Am. B 1998, 15, 871.
Acknowledgment. The New Energy and Industrial Technol-
ogy Development Organization (NEDO) supported this work
within the framework of the Femtosecond Technology Research
Project.
(37) Kuzyk, M. G.; Garvey, D. W.; Vigil, S. R.; Welker, D. J. Chem.
Phys. 1999, 245, 533.
(38) Ashwell, G. J.; Williamson, P. C.; Green, A.; Bahra, G. S.; Brown,
C. R. Aust. J. Chem. 1998, 51, 599.
(39) Zhong, T. X.; Workman, R. K.; Yao, X.; Jabbour, G. E.; Peterson,
C. A.; Sarid, D.; Dirk, C. W.; de la Cruz, D.; Nagarur, A. Thin Solid Films
1998, 315, 294.
(40) Yang, M.; Jian, Y. Phys. Chem. Chem. Phys. 2001, 3, 4213.
(41) Kruhlak, R. J.; Young, J.; Kuzyk, M. G. Proc. SPIE-Int. Soc. Opt.
Eng. 1997, 3147, 118.
References and Notes
(1) Meiz, R. J.; Champ, R. B.; Chang, L. S.; Chiou, C.; Keller, G. S.;
Liclian, L. C.; Neiman, R. B.; Shattuck, M. D.; Weiche, W. J. Photogr.
Sci. Eng. 1977, 21, 73.
(2) Tam, A. C. Appl. Phys. Lett. 1980, 37, 979.
(3) Law, K. Y. Chem. ReV. 1993, 93, 449.
(4) Merritt, V. Y.; Hovel, H. J. Appl. Phys. Lett. 1976, 29, 414.
(5) Merritt, V. Y. IBM J. Res. DeV. 1978, 22, 353.
(6) Musser, M. E.; Dahlberg, S. C. Appl. Surface Sci. 1980, 5, 28.
(7) Yamin, P.; Piechowski, A. P.; Bird, G. R.; Morel, D. L. J. Phys.
Chem. 1982, 86, 3796.
(8) Jipson, V. B.; Jones, C. R. IBM Technol. Disclosure Bull. 1981,
24, 298.
(9) Gravesteijn, D. J.; Steenbergen, C.; van der Ween, J. Proc. SPIE-
Int. Soc. Opt. Eng. 1983, 420, 327.
(42) Kruhlak, R. J.; Kuzyk, M. G. Proc. SPIE-Int. Soc. Opt. Eng. 1998,
3473, 57.
(43) Kruhlak, R. J.; Kuzyk, M. G. J. Opt. Soc. Am. B 1999, 16, 1749.
(44) Kruhlak, R. J.; Kuzyk, M. G. J. Opt. Soc. Am. B 1999, 16, 1756.
(45) Garvey, D. W.; Kuzyk, M. G. Proc. SPIE-Int. Soc. Opt. Eng. 1999,
3796, 13.
(46) Ashwell, G. J. J. Mater. Chem. 1998, 8, 373.
(47) Meier, H.; Petermann, R.; Gerold, J. Chem. Commun. 1999, 977.
(48) Sprenger, H. E.; Ziegenbein, W. Angew. Chem., Int. Ed. Engl. 1966,
5, 894.
(49) McRae, E. G.; Kasha, M. J. Chem. Phys. 1958, 28, 721.
(50) Law, K. Y. J. Phys. Chem. 1987, 91, 5184.
(10) Sporer, A. H. Appl. Optics 1984, 23, 2738.
(11) Dirk, C. W.; Kuzyk, M. G. Chem. Mater. 1990, 2, 4.