Solution and thin-film aggregation studies of octasubstituted dendritic phthalocyanines

Article abstract of DOI:10.1560/IJC.49.1.9

The synthesis and solution and thin-film characterization of eight octasubstituted dendritic phthalocyanines (Pcs) and their zinc complexes are reported. The Pc chromophore was substituted in the 2,3,9,10,16,17,23,24- positions with three generations of b

Full text of DOI:10.1560/IJC.49.1.9

Israel Journal of Chemistry Vol. 49
DOI: 10.1560/IJC.49.1.9
2009
pp. 9–21
Solution and Thin-Film Aggregation Studies of Octasubstituted Dendritic
Phthalocyanines
Casey a. Kernag and dominiC V. mCgrath*
Department of Chemistry, University of Arizona, P.O. Box 210041, Tucson, Arizona 85721, USA
(Received 5 February 2008 and in revised form 16 December 2008)
Abstract. The synthesis and solution and thin-film characterization of eight
octasubstituted dendritic phthalocyanines (Pcs) and their zinc complexes are re-
ported. The Pc chromophore was substituted in the 2,3,9,10,16,17,23,24-positions
with three generations of benzylaryl ether dendrons with either a benzyl (3a–3c) or
3,5-di-t-butylbenzyl periphery (3d–3f). Visible spectra in solution (CH2Cl2–EtOH
mixtures, toluene, THF, dioxane, acetone, and EtOAc) indicated a varying degree
of chromophore aggregation that depended on solvent, dendrimer generation, and
whether the Pc was metallated. Variable-concentration visible spectroscopic studies
were analyzed using a nonlinear least-squares fitting procedure giving Kd values.
These values further quantitated the observations that the t-butyl-substituted den-
drimers 3d–3f were all less prone to aggregation in solution than the unsubstituted
dendrimers 3a–3c, with a monotonic decrease in Kd across the series 3a → 3b → 3c
→ 3d → 3e → 3f. Second-generation t-butyl-substituted dendrimer 3f showed little
to no aggregation in all solvents studied. Thin-film studies indicated that the largest
members of the two dendrimer groups, third-generation 3c and second-generation
3f, were largely monomeric as evidenced by split Q-bands, similar to that seen in
dilute CH2Cl2 solution when deposited via spin-coating onto glass slides. The metal-
lated zinc Pcs 4a–4f all exhibited significantly less tendency toward aggregation in
both solution and thin-films than their unmetallated analogues.
IntroductIon
ranging degrees of success. McKeown and cowork-
ers,20–24 Ng and coworkers,25–28 Kobayashi and cowork-
Phthalocyanines (Pcs) are widely used in a large number
of materials applications.1–3 These exceedingly robust
chromophores are known to aggregate, forming stable,
non-covalent structures, especially in polar solvents and
thin-films.4 Aggregation of Pcs leads to excited state
quenching and decreased molar absorptivities. Current
research has focused on the design of non-aggregating
Pcs for specific applications in the fields of optical lim-
iting5–7 and photodynamic therapy (PDT).8–10 Control of
the aggregation state of Pcs in thin-films is also critical
to their performance in photovoltaic devices.11–13
31,32
ers,29,30 and Kasuga and coworkers
have studied the
placement of bulky groups along the periphery of Pc
chromophores or as axial substituents. Four peripheral
dendritic substituents, unless coupled with electrostatic
repulsion,25,26,29,30 are ineffective at preventing Pc self-
association, even up to third-generation dendrimers.20–24
Axial dendritic substituents attached to the central atom
of a SiPc effectively prevented aggregation.23,24 However,
this approach limits the range of metals in the Pc. Tuning
the properties of Pcs by changing the internal metal is
critical to the application of Pcs as sensors, photosensitiz-
ers, catalysts, and information storage systems.2,3,13
*Author to whom correspondence should be addressed.
E-mail: mcgrath@u.arizona.edu
Several approaches have been attempted to reduce
Pc self-association14 including using dendritic substitu-
ents,15–17 which are known to be particularly effective
at site-isolation.18,19 These attempts have been met with

10
partitioned between a mixture of CH2Cl2 and water (1:1,
50 mL), separated, and the aqueous layer was further extracted
with CH2Cl2 (3 ´ 25 mL). The combined organic extracts were
dried (MgSO4), filtered, and evaporated to dryness. The crude
product was purified by flash column chromatography (SiO2,
CH2Cl2) to give a colorless foam.
We have reported that placing eight dendritic sub-
stituents on the periphery of phthalocyanine chromo-
phores can successfully prevent condensed phase ag-
gregation.33 This contribution details the full scope of
materials prepared in this new class of octasubstituted
dendritic phthalocyanines, their structural characteriza-
tion, and investigation of their self-association behavior
in solution and the condensed phase.
4,5-Bis(4-[G1]oxyphenoxy)phthalonitrile (2a). Accord-
ing to the general procedure, [G1]-Br (0.12 g, 0.3 mmol), 1
(50 mg, 0.15 mmol), potassium carbonate (0.21 g, 1.5 mmol),
and 18-crown-6 (4 mg, 0.02 mmol) in acetone (10 mL) at re-
1
flux gave 2a (121 mg; 87%): H NMR (250 MHz, CDCl3) δ
ExpErImEntAl
7.47–7.25 (m, 20 H), 7.05 (s, 2 H), 7.01 (s, 8 H), 6.69–6.68 (d,
J = 2.3 Hz, 4 H), 6.59-6.57 (t, J = 2.2 Hz, 2 H), 5.03 (s, 8 H),
5.00 (s, 4 H); 13C NMR (62.9 MHz, CDCl3) δ 160.25, 156.65,
152.40, 147.33, 138.96, 136.70, 128.61, 128.06, 127.55,
121.45, 120.64, 116.60, 115.56, 106.42, 104.32, 101.63,
70.49, 70.16; MS (FAB) m/z 949.35 (M+).
4,5-Bis(4-[G2]oxyphenoxy)phthalonitrile (2b). Accord-
ing to the general procedure, [G2]-Br (0.12 g, 0.15 mmol), 1
(25 mg, 0.07 mmol), potassium carbonate (0.1 g, 0.75 mmol),
and 18-crown-6 (2 mg, 0.01 mmol) in acetone (10 mL) at
reflux gave 2b (97 mg; 74%): 1H NMR (250 MHz, CDCl3) δ
7.48-7.19 (m, 40 H), 7.03 (s, 2 H), 7.00 (s, 8 H), 6.68-6.63 (m,
12 H), 6.58-6.54 (m, 6 H), 5.01 (s, 16 H), 4.99 (s, 4 H), 4.97 (s,
8 H); 13C NMR (125.7 MHz, CDCl3) δ 160.17, 160.11, 156.63,
152.36, 147.30, 139.14, 138.94, 136.72, 128.60, 128.02,
127.55, 121.44, 120.60, 116.59, 115.12, 109.74, 106.45,
106.41, 101.67, 101.56, 70.47, 70.12, 70.02; MS (FAB) m/z
1798.69 (M+).
General
NMR spectra were recorded on either a Bruker 250 MHz
or 500 MHz NMR spectrometer. Chemical shifts are reported
in ppm (δ) referenced to internal residual solvent protons
(1H). 13C NMR spectra were recorded as proton-decoupled
spectra. Electrospray ionization (ESI), fast atom bombard-
ment (FAB), and matrix assisted laser disorption ionization
(MALDI) mass spectra were performed at the University of
Arizona Mass Spectrometry Facilities. All chemicals were
purchased from commercial suppliers and used as received.
THF was distilled from benzophenone and potassium metal.
Toluene was distilled from calcium hydride. Flash chroma-
tography using silica (Natland International Corp., silica gel
200–400 Mesh) was performed by the method of Still et al.34
Gel permeation chromatography (GPC) was performed using
a Shimadzu HPLC system (LS-10AT pump, RID-6A refrac-
tive index detector) with CH2Cl2 as the eluent at ambient tem-
perature (21 ºC). The flow rate was 1 mL/min through three
250 ´ 10 mm Jordi DVB columns of 104, 103, and 500 Å pore
size. Preparatory GPC was done on a 500 ´ 30 mm Jordi gel
DVB mixed-bed column at a flow rate of 10 mL/min. Ana-
lytical thin-layer chromatography (TLC) was performed on
precoated TLC plates (Merck precoated 0.25 mm silica gel
60 F254 plates). The synthesis of the [Gn]-Br dendrons,35
4,5-bis-(4-hydroxyphenoxy)phthalonitrile (1),36 and 4-ni-
trophthalonitrile (7),37 was performed according to published
procedures. 3,5-Di-tert-butyl [Gn]-Br dendrons were pre-
pared from 3,5-di-t-butylbenzyl bromide38 by the procedure
of Zimmerman and coworkers.39 Synthesis of 3,5-di-t-butyl-
benzyl bromide38 was performed from 3,5-di(t-butyl)toluene
(Aldrich) by oxidation40 to the 3,5-di-tert-butylbenzoic acid,41
subsequent reduction to 3,5-di-tert-butylbenzyl alcohol,41
and then bromination with N-bromosuccinimide (NBS) and
triphenylphosphine. All spectroscopic measurements were
performed with spectrophotometric grade solvents (Aldrich).
UV-Visible spectra were recorded on a Shimadzu UV-2401Pc
spectrophotometer and temperature measurements were done
with the attached Shimadzu CPS-Controller. Corrected fluo-
rescence spectra were collected on a Spex Fluorolog-2 spec-
trophotometer.
4,5-Bis(4-[G3]oxyphenoxy)phthalonitrile (2c). Accord-
ing to the general procedure, [G3]- Br (0.25 g, 0.15 mmol), 1
(25 mg, 0.07 mmol), potassium carbonate (0.1 g, 0.75 mmol),
and 18-crown-6 (2 mg, 0.01 mmol) in acetone (10 mL) at re-
flux gave 2c (171 mg; 67%): 1H NMR (250 MHz, CDCl3) δ
7.52-7.13 (m, 80 H), 7.01 (s, 2 H), 6.97 (s, 8 H), 6.68-6.61 (m,
28 H), 6.59-6.51 (m, 14 H), 4.99 (s, 32 H), 4.95 (s, 12 H), 4.93
(s, 16 H); 13C NMR (125.7 MHz, CDCl3) δ 160.15, 160.13,
160.06, 156.61, 152.32, 147.25, 139.17, 139.13, 138.82,
136.74, 128.56, 127.98, 127.53, 121.44, 120.34, 116.54,
115.22, 109.71, 106.44, 106.38, 106.36, 102.12, 101.59,
101.57, 70.42, 70.08, 70.02, 69.98; MS (MALDI) m/z 3534.32
(M + K+)
.
4,5-Bis(4-((t-Bu)2[G0])phenoxy)phthalonitrile (2d).
According to the general procedure, (t-Bu)2-[G0]-Br (0.5 g,
1.8 mmol), 1 (0.3 g, 0.85 mmol), potassium carbonate
(0.52 g, 3.8 mmol), and 18-crown-6 (45 mg, 0.17 mmol) in
1
acetone (25 mL) at reflux gave 2d (551 mg; 87%): H NMR
(250 MHz, CDCl3) δ 7.42–7.41 (m, 2 H), 7.28–7.27 (m, 4 H),
7.06(s, 8 H), 7.04(s, 2H), 5.04(s, 4H), 1.33(s, 36H); 13C NMR
(62.9 MHz, CDCl3) δ 157.05, 152.49, 151.19, 147.11, 135.34,
122.42, 122.23, 121.52, 120.41, 116.53, 115.16, 109.67, 71.45,
34.88, 31.44; MS (MALDI) m/z 748.40 (M+).
4,5-Bis(4-((t-Bu)4[G1])phenoxy)phthalonitrile (2e).
According to the general procedure, (t-Bu)4-[G1]-Br (1.1 g,
1.8 mmol), 1 (0.3 g, 0.85 mmol), potassium carbonate
(0.69 g, 5.0 mmol), and 18-crown-6 (45 mg, 0.17 mmol) in
acetone (25 mL) at reflux gave 2e (721 mg; 59%): H NMR
(250 MHz, CDCl3) δ 7.41–7.40 (m, 4 H), 7.28–7.27 (m, 8 H),
General Procedure for the Preparation of Phthalonitriles
Potassium carbonate (10 equiv), 1 (1 equiv), 18-crown-6
(0.2 equiv), and the dendritic bromide (2 equiv) were stirred
at an elevated temperature for 2 d. The reaction mixture was
cooled, and the solvent removed in vacuo. The residue was
1
Israel Journal of Chemistry
49
2009

11
7.07 (s, 2 H), 7.04 (s, 8 H), 6.74–6.73 (m, 4 H), 6.67–6.65 (m,
2 H), 5.03 (s, 4 H), 5.01 (s, 8 H), 1.33 (s, 72 H); 13C NMR
(62.9 MHz, CDCl3) δ 160.50, 156.75, 152.42, 151.08, 147.28,
138.80, 135.59, 122.32, 122.27, 121.48, 120.55, 116.60,
115.41, 109.74, 106.45, 101.57, 71.06, 70.62, 34.86, 31.45;
MS (ESI) m/z 1419.52 (M + Na+).
4,5-Bis(4-((t-Bu)8[G2]oxy)phenoxy)phthalonitrile (2f).
According to the general procedure, (t-Bu)8-[G2]-Br (2.3 g,
1.8 mmol), 1 (0.31 g, 0.87 mmol), potassium carbonate
(0.69 g, 5.0 mmol), and 18-crown-6 (45 mg, 0.17 mmol) in
acetone (25 mL) at reflux gave 2f (1.73 g; 74%): 1H NMR
(500 MHz, CDCl3) δ 7.40 (bs, 8 H), 7.28 (bs, 16 H), 7.05 (s,
2 H), 7.03 (s, 8 H), 6.74 (bs, 8 H), 6.72 (bs, 4 H), 6.65 (bs,
4 H), 6.62 (bs, 2 H), 5.02 (bs, 12 H), 5.01 (s, 16 H), 1.33 (s, 144
H); 13C NMR (62.9 MHz, CDCl3) δ 160.46, 160.27, 156.75,
152.43, 151.06, 147.22, 138.97, 138.92, 135.67, 122.31,
122.30, 121.59, 120.35, 116.59, 115.16, 109.70, 106.49,
106.48, 101.69, 101.55, 71.06, 70.52, 70.19, 34.87, 31.48; MS
(ESI) m/z 2716.44 (M + Na+).
General Procedure for the Cyclization of
Phthalonitriles
A mixture of phthalorilnitrile (4 equiv) and 1-pentanol was
heated to reflux under Ar. Lithium bromide (1 equiv), and 1,8-
diazabicyclo[5.4.0]-7-undecene (DBU) (4 equiv) for 3a and
3b, or lithium metal (20 equiv) for 3c–3f and 9 was added and
the reaction was allowed to proceed for 24 h. The mixture was
cooled to 0 ºC and the solvent decanted. The crude product
was precipitated from CH2Cl2 by addition of hexanes and pu-
rified by flash column chromatography (SiO2, CH2Cl2) giving
a dark green solid.
2,3,9,10,16,17,23,24-octakis(4-[G1]oxyphenoxy)phthal
ocyanine (3a). By using the general procedure, phthalonitrile
2a (0.5 g, 0.53 mmol), lithium bromide (11.3 mg, 0.13 mmol),
and DBU (0.20 mg, 0.53 mmol) in 1-pentanol (25 mL) gave
1
3a (140 mg; 27%): H NMR (500 MHz, CDCl3) δ 8.91 (s,
2 H), 7.42–7.21 (br m, 104 H), 6.98–6.96 (d, J = 8.5 Hz, 16 H),
6.66 (s, 16 H), 6.52 (s, 8 H), 4.93 (s, 32 H), 4.90 (s, 16 H); 13C
NMR (62.9 MHz, CDCl3) δ 160.25, 156.65, 147.30, 139.14,
138.94, 136.71, 128.61, 128.06, 127.55, 127.52, 121.45,
120.64, 118.24, 116.60, 108.21, 106.42, 101.63, 70.49, 70.16,
70.02; MS (MALDI) m/z 3798.39 (M+).
2,3,9,10,16,17,23,24-octakis(4-[G2]oxyphenoxy)phtha
locyanine (3b). According to the general procedure, phtha-
lonitrile 2b (0.35 g, 0.19 mmol), lithium bromide (4.3 mg,
0.05 mmol), and DBU (29.0 mg, 0.19 mmol) in 1-pentanol
(10 mL) gave 3b (77 mg; 22%): 1H NMR (500 MHz, CDCl3)
δ 8.89 (s, 2 H), 7.31–7.18 (m, 168 H), 7.17–7.15 (d, J =
9 Hz, 16 H), 7.00–6.99 (d, J = 9 Hz, 16 H), 6.68–6.65 (m, 16
H), 6.60–6.57 (m, 32 H), 6.52–6.48 (m, 8 H), 6.47–6.44 (m,
16 H), 4.95 (s, 16 H), 4.88 (s, 64 H), 4.86 (s, 32 H); 13C NMR
(125 MHz, CDCl3) δ 160.26, 160.15, 156.65, 152.17, 147.32,
139.14, 138.94, 136.72, 128.61, 128.06, 127.55, 127.52,
121.45, 120.64, 118.21, 116.60, 109.74, 108.21, 106.42,
101.63, 101.59, 70.49, 70.17; MS (MALDI) m/z 7194.74
(M+).
2,3,9,10,16,17,23,24-octakis(4-[G3]oxyphenoxy)phth
alocyanine (3c). By using the general procedure, phthaloni-
trile 2c (0.2 g, 0.05 mmol) and lithium metal (7 mg, 1 mmol)
in 1- pentanol (5 mL) gave 3c, which was further purified by
preparatory gel permeation chromatography using CH2Cl2 as
the eluent (32 mg; 16%): 1H NMR (500 MHz, CDCl3) δ 8.85(
s, 2 H), 7.40-7.12 (m, 344 H), 6.95-6.93 (d, J = 9 Hz, 16 H),
6.68-6.62 (m, 24 H), 6.59–6.52 (m, 96 H), 6.44–6.39 (m, 48
H), 4.89 (s, 16 H), 4.82 (s, 128 H), 4.79 (s, 32 H), 4.75 (s, 64
H); 13C NMR (62.9 MHz, CDCl3) δ 160.26, 160.14, 160.06,
156.61, 152.21, 147.25, 139.17, 139.13, 138.82, 136.74,
128.56, 127.57, 127.53, 121.44, 120.34, 118.24, 116.57,
109.71, 108.25, 106.38, 106.36, 101.65, 101.59, 101.57,
70.42, 70.08, 70.02, 69.98; MS (MALDI) m/z 13983.21
(M+).
4-[G2]oxyphenol (6). Hydroquinone (5) (0.66 g, 6 mmol),
[G2]-Br (1.0 g, 1.2 mmol), potassium carbonate (2.1 g,
15 mmol), and 18-crown-6 (63 mg, 0.24 mmol) were stirred
in acetone (40 mL) at reflux for 2 d. The reaction mixture
was cooled, and the solvent removed in vacuo. The residue
was partitioned between a mixture of CH2Cl2 and water (1:1,
50 mL) and the aqueous layer was extracted with CH2Cl2 (3 ´
25 mL). The combined organic extracts were dried (MgSO4),
filtered, and evaporated to dryness. The crude product was
purified by flash column chromatography (SiO2, CH2Cl2) to
give 6 (500 mg; 48%) as a colorless foam: 1H NMR (500 MHz,
CDCl3) δ 7.41–7.30 (m, 20 H), 6.81–6.79 (d, J = 9.0 Hz, 2 H),
6.70–6.68 (d, J = 8.9 Hz, 2 H), 6.66–6.65 (d, J = 2.2 Hz, 4 H),
6.63–6.62 (d, J = 2.1 Hz, 2 H), 6.56–6.55 (t, J = 2.2 Hz, 2 H),
6.52–6.51 (t, J = 2.1 Hz, 1 H), 5.01 (s, 8 H), 4.95 (s, 4 H),
4.91 (s, 2 H); 13C NMR (62.9 MHz, CDCl3) δ 160.15, 160.00,
149.62, 146.51, 139.77, 139.24, 136.75, 128.58, 127.99,
127.55, 116.05, 116.02, 106.36, 106.33, 101.58, 101.53, 70.63,
70.10, 69.96; MS (FAB) m/z 837.16 (M+)
.
4-(4-[G2]oxyphenoxy)phthalonitrile (8). 4-Nitrophtha-
lonitrile (7) (0.12 g, 0.7 mmol), 6 (0.45 g, 0.54 mmol), and po-
tassium carbonate (0.1 g, 0.7 mmol) were stirred in dry DMF
(10 mL) under Ar and at 60 ºC for 2 d. The reaction mixture
was cooled and poured into 100 mL cold water. The resulting
cloudy solution was extracted with CH2Cl2 (3 ´ 30 mL). The
combined organic extracts were dried (MgSO4), filtered, and
evaporated to dryness. The crude product was purified by flash
column chromatography (SiO2, CH2Cl2) to give 8 (319 mg;
62%) as an off-white foam: 1H NMR (500 MHz, CDCl3) δ
7.66–7.64 (d, J = 8.7 Hz, 1 H), 7.39–7.25 (m, 20 H), 7.21–7.20
(d, J = 2.5 Hz, 1 H), 7.16-7.14 (dd, J = 2.5 Hz, 8.7 Hz, 1 H),
6.99–6.93 (m, 4 H), 6.66–6.65 (d, J = 2.2 Hz, 2 H), 6.65–6.64
(d, J = 2.1 Hz, 2 H), 6.56–6.55 (t, J = 2.2 Hz, 2 H), 6.54–6.53 (t,
2,3,9,10,16,17,23,24-octakis(4-((t-Bu)2-[G0])phenoxy)p
hthalocyanine (3d). By using the general procedure, phtha-
J = 2.1 Hz, 1 H), 5.01 (s, 8 H), 4.99 (s, 2 H), 4.97 (s, 4 H); 13
C
NMR (62.9 MHz, CDCl3) δ 160.18, 160.12, 156.72, 147.12, lonitrile 2d (0.25 g, 0.33 mmol) and lithium metal (46 mg,
139.12, 139.04, 136.69, 135.30, 133.61, 128.58, 128.02, 6.6 mmol) in 1-pentanol (10 mL) gave 3d, which was further
127.52, 121.77, 121.05, 120.85, 116.63, 116.05, 106.38, purified by preparatory gel permeation chromatography us-
1
106.36, 101.51, 101.50, 70.41, 70.11, 70.00; MS (ESI) m/z ing CH2Cl2 as the eluent (64 mg; 26%): H NMR (500 MHz,
985.43 (M + Na+).
CDCl3) δ 8.79 (s, 2 H), 7.42–7.41 (m, 8 H), 7.28–7.27 (m,
Kernag and McGrath / Aggregation Studies of Dendritic Phthalocyanines

12
24 H), 6.62 (s, 32 H), 4.92 (s, 16 H), 1.28 (s, 144 H); 13C NMR 116.66, 108.21, 106.42, 101.63, 70.49, 70.16, 70.02; MS (ESI)
(62.9 MHz, CDCl3) δ 154.03, 151.49, 149.19, 147.11, 137.34, m/z 3861.76 (M+).
121.52, 120.41,118.42, 116.53, 115.16, 114.23, 109.67, 70.43,
34.76, 31.41; MS (MALDI) m/z 2998.52 (M+).
2,3,9,10,16,17,23,24-octakis(4-[G2]oxyphenoxy) zinc
1
phthalocyanine (4b). Dark green solid (2.4 mg; 72%): H
2,3,9,10,16,17,23,24-octakis(4-((t-Bu)4-[G1])phenoxy)p NMR (500 MHz, CDCl3) δ 8.54 (s, 8 H), 7.36–7.24 (m,
hthalocyanine (3e). By using the general procedure, phtha- 160 H), 7.17–7.15 (d, J = 9 Hz, 16 H), 7.00–6.98 (d, J = 9 Hz,
lonitrile 2e (0.25 g, 0.18 mmol) and lithium metal (25 mg, 16 H), 6.69–6.65 (m, 16 H), 6.60–6.56 (m, 32 H), 6.52– 6.47
3.6 mmol) in 1-pentanol (10 mL) gave 3e, which was further (m, 8 H), 6.45–6.43 (m, 16 H), 4.93 (s, 16 H), 4.87 (s, 64 H),
purified by preparatory gel permeation chromatography us- 4.85 (s, 32 H); 13C NMR (62.9 MHz, CDCl3) δ 160.26, 160.15,
1
ing CH2Cl2 as the eluent (58 mg; 23%): H NMR (500 MHz, 156.65, 152.17, 149.32, 142.33, 138.94, 136.72, 128.61,
CDCl3) δ 8.85 (s, 2 H), 7.41–7.40 (m, 16 H), 7.28–7.27 (m, 128.06, 127.56, 127.51, 121.45, 120.64, 118.21, 116.64,
32 H), 6.74–6.73 (m, 16 H), 6.67–6.60 (m, 48 H), 4.93 (s, 109.74, 108.21, 106.52, 101.63, 101.57, 70.49, 70.17; MS
16 H), 4.91 (s, 32 H), 1.28 (s, 288 H); 13C NMR (62.9 MHz, (MALDI) m/z 7258.90 (M+).
CDCl3) δ 160.49, 154.05, 151.42, 149.18, 147.11, 137.34,
135.59, 121.52, 120.41, 118.42, 116.58, 115.16, 114.27, phthalocyanine (4c). Dark green solid (2.7 mg; 83%): H
109.74, 106.45, 101.57, 71.06, 70.52, 34.76, 31.45; MS NMR (500 MHz, CDCl3) δ 8.52 ( s, 8 H), 7.42–7.14 (m,
2,3,9,10,16,17,23,24-octakis(4-[G3]oxyphenoxy) zinc
1
(MALDI) m/z 5594.92 (M+).
336 H), 6.97–6.94 (d, J = 9 Hz, 16 H), 6.65–6.59 (m, 24 H),
2,3,9,10,16,17,23,24-octakis(4-((t-Bu)8-[G2])phenoxy) 6.57–6.51 (m, 96 H), 6.42–6.36 (m, 48 H), 4.91 (s, 16 H), 4.85
phthalocyanine (3f). By using the general procedure, phtha- (s, 128 H), 4.81 (s, 32 H), 4.79 (s, 64 H); 13C NMR (125 MHz,
lonitrile 2f (0.5 g, 0.19 mmol) and lithium metal (26 mg, 3.8 CDCl3) δ 160.26, 160.14, 160.06, 156.61, 152.21, 149.25,
mmol) in 1-pentanol (10 mL) gave 3f, which was further 142.17, 139.18, 138.82, 136.79, 128.66, 127.59, 127.53,
purified by preparatory gel permeation chromatography us- 121.44, 120.34, 118.24, 116.57, 109.71, 108.45, 106.38,
1
ing CH2Cl2 as the eluent (90 mg; 18%): H NMR (500 MHz, 106.36, 101.63, 101.59, 101.57, 70.42, 70.28, 70.02, 69.98;
CDCl3) δ 8.87 (s, 2 H), 7.40 (bs, 32 H), 7.28 (bs, 64 H), MS (MALDI) m/z 14050.32 (M+).
6.74–6.62 (br m, 112 H), 4.92 (s, 48 H), 4.91 (s, 64 H), 1.28
2,3,9,10,16,17,23,24-octakis(4-((t-Bu)2-[G0])oxyphe-
(s, 576 H); 13C NMR (62.9 MHz, CDCl3) δ 160.46, 160.27, noxy) zinc phthalocyanine (4d). Dark green solid (3.1 mg;
1
154.05, 151.43, 149.16, 147.12, 138.97, 137.32, 135.57, 94%): H NMR (500 MHz, CDCl3) δ 8.51 (s, 8 H), 7.41–
121.59, 120.35, 118.45, 116.59, 115.16, 114.23, 109.70, 7.40 (m, 8 H), 7.25–7.23 (m, 16 H), 6.59 (s, 32 H), 4.97 (s,
106.49, 106.48, 101.69, 101.55, 71.06, 70.52, 70.19, 34.77, 16 H), 1.29 (s, 144 H); 13C NMR (62.9 MHz, CDCl3) δ 154.03,
31.48; MS (MALDI) m/z 10785.31 (M+).
151.49, 149.19, 149.11, 142.34, 121.56, 120.37, 118.22,
2,9(10),16(17),23(24)-tetrakis(4-[G2]oxyphenoxy) 116.53, 115.16, 114.29, 109.61, 70.43, 34.76, 31.41; MS
phthalocyanine (9). By using the general procedure, phtha- (MALDI) m/z 3060.39 (M+).
lonitrile 8 (0.17 g, 0.17 mmol) and lithium metal (24 mg,
2,3,9,10,16,17,23,24-octakis(4-((t-Bu)4-[G1])oxyphe-
3.4 mmol) in 1-pentanol (5 mL) gave 9, which was further noxy) zinc phthalocyanine (4e). Dark green solid (2.9 mg;
purified by preparatory gel permeation chromatography us- 87%): 1H NMR (500 MHz, CDCl3) δ 8.55 (s, 8 H), 7.41–7.38
1
ing CH2Cl2 as the eluent (43 mg; 25%): H NMR (500 MHz, (m, 16 H), 7.28–7.25 (m, 24 H), 6.77–6.73 (m, 16 H), 6.67–
CDCl3) δ 8.91 (s, 2 H), 7.39–7.14 (br m, 92 H), 6.66–6.53 6.58 (m, 48 H), 4.95 (s, 16 H), 4.93 (s, 32 H), 1.29 (s, 288
(br m, 36 H), 4.91 (s, 32 H), 4.89 (s, 8 H), 4.87 (s, 8 H); 13
C
H); 13C NMR (62.9 MHz, CDCl3) δ 160.49, 154.05, 151.42,
NMR (62.9 MHz, CDCl3) δ 160.18, 160.12, 154.02, 147.12, 149.18, 149.12, 142.37, 135.51, 121.52, 120.41, 118.41,
139.12, 139.04, 136.69, 135.30, 133.61, 128.58, 128.02, 116.38, 115.11, 114.27, 109.74, 106.42, 101.58, 71.01, 70.32,
127.52, 121.05, 120.85, 118.37, 116.63, 116.05, 106.38, 34.71, 31.55; MS (MALDI) m/z 5656.10 (M+).
106.36, 101.51, 101.50, 70.41, 70.11, 70.00; MS (MALDI)
m/z 3854.24 (M+).
2,3,9,10,16,17,23,24-octakis(4-((t-Bu)8-[G2])oxyphe-
noxy) zinc phthalocyanine (4f). Dark green solid (2.7 mg;
81%): H NMR (500 MHz, CDCl3) δ 8.57 (s, 2 H), 7.42 (bs,
1
General procedure for the synthesis of zinc phthalocyanines
32 H), 7.21 (bs, 56 H), 6.71–6.60 (br m, 112 H), 4.97 (s, 48 H),
Zinc acetate (1.1 equiv) and the dendritic phthalocyanine 4.92 (s, 64 H), 1.29 (s, 576 H); 13C NMR (62.9 MHz, CDCl3)
(1 equiv) were added to dry DMF and heated at 60 ºC over- δ 160.46, 160.27, 154.07, 151.43, 149.16, 149.10, 142.31,
night under argon. The reaction was cooled and poured into 138.97, 135.57, 121.59, 120.30, 118.45, 116.57, 115.26,
cold water and the desired compound was obtained by filtra- 114.27, 109.70, 106.44, 106.41, 101.69, 101.57, 71.16, 70.72,
tion, taken up in CH2Cl2, dried (MgSO4), concentrated, and 70.49, 34.78, 31.58; MS (MALDI) m/z 10848.67 (M+).
evaporated to dryness.
2,9(10),16(17),23(24)-tetrakis(4-[G2]oxyphenoxy) zinc
2,3,9,10,16,17,23,24-octakis(4-[G1]oxyphenoxy) zinc phthalocyanine (10). By following the generalized procedure,
1
phthalocyanine (4a). Dark green solid (5.5 mg; 78%): H zinc acetate (3.3 mg, 0.02 mmol) and 9 (0.11 g, 0.02 mmol)
NMR (500 MHz, CDCl3) δ 8.56 (s, 8 H), 7.45–7.27 (br m, in dry DMF (10 mL) gave 10, as a dark green solid (2.9 mg;
96 H), 6.98–6.96 (d, J = 8.5 Hz, 16 H), 6.64 (s, 16 H), 6.57 89%): 1H NMR (500 MHz, CDCl3) δ 8.51 (s, 8 H), 7.38–7.17
(s, 8 H), 4.91 (s, 32 H), 4.89 (s, 16 H); 13C NMR (62.9 MHz, (br m, 84 H), 6.67-6.56 (br m, 36 H), 4.95 (s, 32 H), 4.93 (s,
CDCl3) δ 160.25, 156.65, 149.37, 142.34, 138.94, 136.74, 8 H), 4.90 (s, 8 H); 13C NMR (62.9 MHz, CDCl3) δ 160.18,
128.51, 128.26, 127.55, 127.52, 121.45, 120.44, 118.24, 160.14, 154.22, 149.12, 142.12, 139.14, 136.71, 135.38,
Israel Journal of Chemistry
49
2009

13
Scheme 1. Synthesis of benzyl aryl ether dendritic phthalonitriles and phthalocyanines with cyclization conditions (a) LiBr/
DBU/1-pentanol/∆ or (b) Li0/1-pentanol/∆.
133.61, 128.51, 128.02, 127.56, 121.15, 120.83, 118.37,
116.61, 116.25, 106.38, 106.36, 101.57, 101.50, 70.42, 70.21,
70.09; MS (MALDI) m/z 3917.40 (M+).
To compare the effect of four versus eight dendritic
substituents on Pc aggregation,22,23 we prepared tet-
rasubstituted dendritic phthalocyanine 9 (Scheme 2).
Unsubstituted second-generation dendron bromide was
allowed to react with an excess of hydroquinone (5,
4 equiv) along with potassium carbonate (10 equiv) and
18-crown-6 (0.2 equiv) in refluxing acetone. Dendritic
product 6 was isolated in 52% yield and coupled to 4-ni-
trophthalonitrile (7) to give mono-substituted dendritic
phthalonitrile 8 in a 62% yield. Cyclization in refluxing
pentanol with lithium metal gave the desired meso-tetra-
substituted phthalocyanine 9 in 25% yield (Scheme 2).
rEsults And dIscussIon
Synthesis of Dendritic Phthalocyanines
Our strategy for the preparation of the octasubsti-
tuted dendritic Pcs was the cyclotetramerization of
dendritic dialkoxyphthalonitriles. Coupling dihydroxy
phthalonitrile 136 with unsubstituted Fréchet-type den-
drons proceeded in the presence of 18-crown-6, po-
tassium carbonate, and acetone at reflux temperatures
to provide dendritic phthalonitriles 2a–c as colorless
foams in yields ranging from 67-87% (Scheme 1). Cou-
pling of the 3,5-di-t-butyl-substituted dendrons39 to 1
was also carried out under similar conditions to provide
2d–f in good yields.
Initial cyclization of the phthalonitriles was per-
formed using lithium bromide as the template salt and
1,8-diazabicyclo[5.4.0]-7-undecene (DBU) as the base
in refluxing pentanol. The removal of the lithium ion
in the acidic workup provided an unmetallated phthalo-
cyanine. These conditions worked well for the unsubsti-
tuted first- and second-generation phthalonitriles 2a and
2b providing phthalocyanines 3a and 3b in yields of
27% and 22%, respectively (Scheme 1). Third-genera-
tion phthalonitrile 2c and all the di-t-butyl-substituted
phthalonitriles 2d–2f did not undergo successful cycli-
zations under these conditions. However, lithium metal
in refluxing pentanol proved effective in producing
third-generation phthalocyanine 3c as well as the t-butyl
dendritic phthalonitriles 3d–3f (Scheme 1).42
Scheme 2. Synthesis of a tetrasubstituted dendritic Pc 10.
Kernag and McGrath / Aggregation Studies of Dendritic Phthalocyanines

14
Reacting the unmetallated phthalocyanines 3a–3f and
9 with zinc acetate in DMF at 40 ºC gave the desired
metallated phthalocyanines 4a–4f and 10 without exten-
sive purification and in yields ranging from 72–94%.
2
1
0
1
Structural Characterization of Dendritic
Phthalocyanines
1
All dendritic Pcs were characterized by H and
13C NMR spectroscopies and MALDI mass spectrom-
etry, confirming both identity and purity of the desired
materials. In addition, gel permeation chromatography
(GPC) was used as an indicator of relative hydrody-
namic size of the dendritic Pcs. As expected, the chro-
matograms of both unsubstituted dendrimers 3a–3c
and t-butyl-substituted dendrimers 3d–3f exhibited a
decrease in elution volume with increasing generation,
corresponding to an increase in hydrodynamic volume
(Fig. 1). Comparison of the GPC molecular weights,
calculated using linear polystyrene standards, indicated
that the smaller generations of both series exceeded the
actual MW. The GPC MW for the largest structures
(third-generation 3c and second-generation 3f) was
significantly smaller than the actual value, consistent
with a more compact globular structure for the dendritic
structures relative to the extended chain of the polysty-
rene standards. Interestingly, the presence of the periph-
eral t-butyl groups on first- and second-generation den-
drimers 3e and 3f results in a much larger hydrodynamic
volume than corresponding first- and second-generation
dendrimers 3a and 3b. Indeed, second-generation 3f is
comparable in size to third-generation 3c, merely by the
presence of t-butyl groups on the periphery.
0
500
600
700
800
Wavelength (nm)
300
400
500
600
700
Wavelength (nm)
Fig. 2. UV-Vis spectra of 3a (—), 3b (- -), and 3c (— -) in
CH2Cl2 (ca. 13 μΜ); Inset: Non-aggregated 3a (20 μΜ) in
CH2Cl2 and fully aggregated 3a (20 μΜ) in 1:1 CH2Cl2–EtOH
at 25 ºC.
Solution Aggregation Studies
Non-aggregated unmetallated Pcs exhibit a split Q-
band in the 680–720 nm region with vibrational side
bands appearing between 640–600 nm in the UV-vis-
ible absorption spectrum. UV-Vis absorption spectra
obtained of dilute CH2Cl2 solutions (13 µM) were con-
sistent with reported spectra of non-aggregated Pcs for
3a–3f, with a split Q-band at 700 and 665 nm, vibra-
tional bands at 640 and 610 nm, and B-band at 340 nm
(Fig. 2).43 The band arising from absorption of the ben-
zylaryl ether dendrons at 280 nm increases appropriate-
ly with increasing generation. In 1:1 EtOH-CH2Cl2 the
aggregated species becomes the predominant form of
the chromophore, and the Q-band undergoes significant
hypsochromic and hypochromic shifts (Fig. 2, inset).
Controlled aggregation studies were performed by
adding increasing amounts of ethanol to CH2Cl2 solu-
tions of 3a–3f and monitoring the Q-band intensity at
700 nm (Fig. 3). The aggregation of the Pcs was in-
dicated by the hypsochromic shift of the Q-band at
700 nm and gradual broadening of the Q-band region
in a hypochromic fashion. As the size of the dendron
along the periphery increased (i.e., 3a→3b→3c), the
onset of aggregation of the Pc was shifted to higher
ethanol concentrations. There was a noticeable drop in
the absorbance of the first-generation 3a when the solu-
tion contained 10% EtOH, a drop that continued with
increasing EtOH content until a broad peak indicative
(a)
(b)
GPC
MW
GPC
MW
MW
MW
3a 3795 (4500)
3b 7189 (7420)
3c 13986 (8750)
3d 3030 (3780)
3e 5594 (6770)
3f 10785 (8760)
25
30
35
40
25
30
35
40
Elution Vol. (mL)
Elution Vol. (mL)
Fig. 1. GPC comparisons of (a) 3a (- - -), 3b (- –), and 3c (—);
(b) 3d (- - -), 3e (- –), and 3f (—).
Israel Journal of Chemistry
49
2009

15
monomeric in solution at comparatively higher con-
centrations of EtOH in CH2Cl2 (Fig. 3, right). Zeroth
generation 3d exhibited a decrease in the Q-band ab-
sorptivity at 10% EtOH in CH2Cl2, although the con-
tinued hypochromic shift with increasing EtOH content
was not as precipitous as with first-generation 3a. Both
first-generation 3e and second-generation 3f began to
aggregate only at 40% EtOH in CH2Cl2, and the hy-
pochromic shift was much more pronounced for the
smaller 3e compared to 3f. Studies beyond 50% EtOH
content were not possible because of low solubility of
the Pcs. It is interesting to note that the t-butyl-substi-
tuted dendrimers tend to resist aggregation in solution
more than the unsubstituted dendrimers of larger size.
It is possible that the increased van der Waals size of
the t-butyl groups on the periphery is contributing to a
greater excluded volume around the Pc chromophores,
although we surmise that there is likely a contributing
entropic (solvophobic) effect as well.
2
1
0
1
0
0
10 20 30 40 50
% EtOH
0
10 20 30 40 50
% EtOH
Fig. 3. (Left) Comparison of the Q-band absorbance in CH2Cl2
at 700 nm of first-generation 3a (●), second-generation 3b
(■), second-generation 9 (£), and third-generation 3c (▲) (20
μM). (Right) Comparison of the Q-band molar absorptivities
at 700 nm between zeroth generation 3d (●), first-generation
3e (■), and second-generation 3f (▲) (20 μM).
Beer’s Law plots constructed for compounds 3a–3f
and 9 illustrate the aggregation behavior of these com-
pounds at varying concentrations. As expected, all den-
of full aggregation was seen at 1:1 CH2Cl2-EtOH. By drimers remained monomeric in 100% CH2Cl2 through-
this measure second-generation 3b did not begin to sig- out a 10–7–10–5 M concentration range as evidenced
nificantly aggregate until the solution contained at least by linearity in the Beer’s Law plot with a slope corre-
20% EtOH, while third-generation 3c exhibited no no- sponding to the observed absorptivity (e.g., Fig. 4a). As
ticeable aggregation until the concentration of ethanol observed above, 3a–3c appeared to be fully aggregated
exceeded 30% EtOH in CH2Cl2. The tetrasubstituted in 1:1 EtOH–CH2Cl2 solution, and the Beer’s Law plot
second-generation analogue 9 exhibited the same trend confirmed this for the entire 10–7–10–5 M concentration
as 3a with an immediate indication of aggregation after range. A much lower slope reflected the hypochromic
the introduction of EtOH. Precipitation of 3a–3c from shift of the Q-band relative to unaggregated samples.
solution occurred when the concentration of ethanol In the case of solutions that fell between both of these
exceeded 50%. solvent extremes, there was a noticeable deviation in
In contrast, t-butyl-substituted Pcs 3d–3f remained the Beer’s Law plot. In a 7:3 CH2Cl2–EtOH solvent
3
(a)
(b)
(c)
4
3
2
1
0
4
3
2
1
0
2
2
1
1
0
0
1
2
3
0
1
2
3
0
1
2
[3b] (• 10–5
)
[3a] (• 10–5
)
[3d] (• 10–5)
Fig. 4. (a) Beer’s law plots of the Q-band absorbance at 700 nm for 3b in CH2Cl2 (●), 7:3 CH2Cl2–EtOH (■), and 1:1 CH2Cl2–
EtOH (▲) at 25 ºC. (b) Beer’s law plots of the Q-band absorbance at 700 nm for 3a in dioxane (▲), THF (▼), toluene (◆),
EtOAc (£), and acetone (+) at 25 ºC. (c) Beer’s Law plots of the Q-band absorbance at 700 nm for 3d in dioxane (▲), THF (▼),
toluene (◆), EtOAc (£), and acetone (+) at 25 ºC.
Kernag and McGrath / Aggregation Studies of Dendritic Phthalocyanines

16
Fig. 5. (a) UV-Vis spectra of 4a in CH2Cl2 (–––), 3:2 CH2Cl2–EtOH (---), and 2:3 CH2Cl2–EtOH (– –) (2 ´ 10–5 M, 25 °C). (b)
Comparisons of the Q-band absorbance at 680 nm between first-generation 4a (●), second-generation 4b (■), third-generation
4c (▲), and second-generation 10 (£) as a function of % ethanol present in CH2Cl2 (20 μM, 25 ºC). (c) Comparisons of the Q-
band molar absorptivities at 680 nm between zeroth-generation 4d (●), first-generation 4e (■), and second-generation 4f (▲)
t-butyl-substituted dendritic zinc Pcs as a function of % ethanol present in CH2Cl2 (20 μM, 25 ºC).
mixture, dendrimer 3b underwent a shift from largely 9 to give 4a–4f and 10, respectively, had consequences
unaggregated to aggregated chromophore as the con- on the solution behavior of the macrocycle. The most
centration increased from 10–7 to 10–5 M. The decrease noticeable change between the absorption spectra of
in slope reflected the lower absorptivity of the aggre- the unmetallated and metallated Pcs is in the collapse
gated species at the wavelength monitored.
of the Q-band due to the change in symmetry (Fig. 5).
We further studied the aggregation properties of the The Q-band also undergoes a dramatic hypochromic
dendritic Pcs in several other solvents: acetone, ethyl shift as aggregation becomes more evident (e.g., in 2:3
acetate (EtOAc), dioxane, tetrahydrofuran (THF), and CH2Cl2–EtOH), although the wavelength of absorption
toluene (Fig. 4b). For unsubstituted dendritic Pcs 3a–3f remains largely unaffected. The spectra for each of the
and 9 the solvents could be characterized by their ability Zn Pcs were measured in CH2Cl2–EtOH mixtures, and
to induce aggregation. Dioxane and THF were similar to the aggregation properties were monitored as a func-
CH2Cl2 in that the dendrimer remained non-aggregated tion of EtOH content. In order to obtain comparable
at all concentrations studied. Acetone and EtOAc acted data between the pure CH2Cl2 solutions and those that
very much like 1:1 CH2Cl2–EtOH in that a large degree contained EtOH, it was necessary to add 0.1% pyridine
of aggregation was seen in all cases. Toluene induced to all solutions to prevent EtOH coordination that can
moderate amounts of aggregation, varying over the affect the intensity of the Q-band.32
10–7–10–5 M concentration range in each of the cases
studied.
In sharp contrast to the unmetallated Pcs, substitution
of the zinc into the core of the macrocycle resulted in
As expected based on the results of their aggregation disfavoring aggregation relative to a demetallated core.
CH2Cl2-EtOH mixtures, the t-butyl-substituted den- This was most evident in the spectra obtained of unsub-
drimers 3d–3f exhibited less aggregation than 3a–3c stituted Pcs 4a–4c, where first-generation 4a did not
in the same range of solvents (Fig. 4c). Aggregation show any signs of strong aggregation until 40% EtOH
was absent in THF, dioxane, and toluene. Even EtOAc, in CH2Cl2. Tetrasubstituted second-generation 10 did
shown to strongly induce aggregation in the unsubsti- not exhibit sign of aggregation until 1:1 EtOH–CH2Cl2
tuted dendrimers, only had a moderate effect on the (Fig. 5b). And second- and third-generation Pcs 4b and
zeroth-generation 3d that became negligible with sec- 4c showed signs of aggregation at 40% and 50% EtOH,
ond-generation 3f (not shown).
respectively. This major difference between the behavior
of the unmetallated and metallated Pcs is potentially due
to the apical coordination of pyridine to the Zn core, and
Solution Properties of Dendritic Zinc Phthalocyanines
Incorporation of zinc into the center of Pcs 3a–3f and the steric barrier this presents to aggregation. When the
Israel Journal of Chemistry 49 2009

17
Table 1. Kd valuesa for the unmetallated dendritic Pcs 3a–3f and 9
solvent
3a
c
3b
3c
9
3d
3e
3f
CH2Cl2
9:1b
–
–
–
–
–
–
–
9.34
12.2
44.1
–
–
–
–
–
–
40.2
–
–
–
–
–
–
–
3.30
–
–
2.40
3.65
30.4
128
562
–
0.95
1.11
9.64
89.8
156
–
0.134
0.397
0.518
15.7
106
–
1.12
2.83
17.7
97.5
236
–
8:2b
7:3b
6:4b
5:5b
Dioxane
THF
–
–
–
–
–
Toluene
Acetone
EtOAc
18.0
652
239
23.3
366
142
3.55
35.4
12.8
17.4
567
92.5
–
8.32
15.2
–
8.32
3.02
–
8.02
–
a×104 M–1. bCH2Cl2–EtOH. cDesignates a negligible value.
Table 2. Kd valuesa for the unmetallated dendritic Pcs 4a–4f and 10
solvent
4a
c
4b
4c
10
4d
4e
4f
CH2Cl2
9:1b
–
–
–
–
–
2.27
87.1
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
8:2b
7:3b
–
–
6:4b
0.954
3.43
0.867
43.2
0.852
3.04
–
0.343
–
0.458
–
0.173
5:5b
a×104 M–1. bCH2Cl2–EtOH. cDesignates a negligible value.
polarity increased sufficiently, disfavorable intermolecu-
Overall, the dimerization constants of the Pcs sup-
lar interactions due to these sterics, as well as possibly ported the general trends seen above: (a) increasing
charge repulsion of the metal core, were overcome by the amounts of EtOH in CH2Cl2 increased the observed
energetics of Pc association accompanied by exclusion solution aggregation for a given Pc; (b) increasing
of a polar solvent from a predominantly nonpolar den- dendrimer generation resulted in decreased solution
dritic environment. For the t-butyl-substituted dendritic aggregation; and (c) the presence of t-butyl groups on
Pcs, we observed a similar decrease in aggregation ten- a dendritic Pc of a given generation greatly decreased
dency on going from unmetallated 3d–3f to metallated the tendency for self-association. These trends are par-
4d–4f cores (Fig. 5c). No significant indication of aggre- ticularly evident when the Kd values are plotted as a
gation was present up to 1:1 EtOH–CH2Cl2.
function of dendrimer and solvent (Fig. 6). For a given
compound, 3a for example, a monotonic increase in Kd
value was clearly seen with increasing EtOH concen-
tration in CH2Cl2, as expected from the studies above
Determination of Thermodynamic Properties of
Aggregating Phthalocyanines
Dimerization constants, Kd, were determined using (i.e., Fig. 3). Also clearly evident was the monotonic
the nonlinear least-square fitting method of Ng44 that as- decrease in Kd value with increasing generation, e.g.,
sumes the absence of higher order aggregate formation 3a → 3b → 3c, seen most readily in 1:1 CH2Cl2–EtOH,
in dilute solution.45–49 Under this assumption, Ng used but also evident in other solvents. Interestingly, tetra-
a nonlinear least-square fitting method to obtain values substituted second-generation analogue 9 fell between
for Kd in which the observed absorbance for a 1 cm cell first- and second-generation octasubstituted dendrimers
can be expressed as:
Am = [εm (1 + 8 Kd[Ct])1/2 – 1] / 4 Kd
where Am is the absorbance of the split Q-band, εm is
3a and 3b in its ability to resist aggregation. The trend
in Kd values, in all solvents except EtOAc, followed the
sequence 3a > 9 > 3b > 3a.
(1)
The t-butyl-substituted dendrimers 3d–3f exhibited
the molar absorptivity of a nonaggregated solution, and markedly lower Kd values relative to their unsubstituted
[Ct] is the total concentration of Pc. We carried out all analogues. Compounds 3b and 3e have the same ap-
Kd determinations using this method in reference to the proximate size (Fig. 1), but exhibit a 4-fold difference
band at 700 nm (Tables 1 and 2).
in Kd. In fact, the monotonic decrease in Kd seenover the
Kernag and McGrath / Aggregation Studies of Dendritic Phthalocyanines

18
Fig. 6. (Left) Plot of Kd values for unmetallated Pcs 3a–3f and 9 in various solvents. (Right) Plot of Kd values for metallated Pcs
4a–4f and 10 in mixtures of CH2Cl2 and EtOH. Vertical axis values for both plots are 104 M–1.
range of unsubstituted dendrimers continues through the temperature range (278–340 K). First-generation t-bu-
t-butyl-substituted dendrimers, i.e., 3a → 3b → 3c → tyl-substituted Pc 3e and second-generation unsubstitut-
3d → 3e → 3f, seen most readily in 1:1 CH2Cl2–EtOH. ed Pc 3b both exhibit similar sizes in solution according
This trend also supported the observation of increasing to GPC (Fig. 1), yet have markedly different aggrega-
dendrimer generation resulting in decreased solution tion behavior in solution. The solvent toluene was used
aggregation for the t-butyl series. The behavior of the because it causes a moderate amount of aggregation in
different substitution types in various solvents followed the unsubstituted Pcs, and was thus thought to be the
the same trend. In toluene, 3e possessed a negligible Kd best solvent to illustrate a range in the extent of aggre-
while 3b was found to have a significant dimerization gation over the tempatures studied. The enthalpic data
constant (Kd = 2.3 ´ 105 M–1).
gained from the experiments show that 3e has a slightly
Determination of the dimerization constants for higher enthalpy of aggregation (–19.08 kJ mol–1) than
the metallated zinc Pcs 4a–4f and 10 was performed 3b (–15.63 kJ mol–1). The difference in the values for
following the same procedure as the unmetallated the unsubstituted Pc 3b (77.8 J mol–1 K–1) and the t-bu-
compounds with CH2Cl2-EtOH mixtures as solvent. tyl-substituted Pc 3e (31.9 J mol–1 K–1) indicate that the
Pyridine (0.1%) was added to the solutions of zinc aggregation of the dendritic Pcs is entropically driven.
Pcs to prevent the coordination of ethanol to the zinc
atom.32 For all compounds, negligible Kd values were Thin-Film Behavior of Unmetallated Dendritic
determined in solutions containing up to 30% ethanol Phthalocyanines
(Table 2, Fig. 6, right), consistent with the results in
Thin-films of 3a–f and 9 were prepared to investi-
Fig. 5. The presence of the t-butyl groups on the periph- gate the nature of the Pcs in the condensed state. Spin-
ery in 4d–4f prevented self-association in up to 40% coated films on glass substrates were fabricated from
ethanol. The anomalously low Kd for zeroth-genera- chloroform solutions of octasubstituted dendritic Pcs
tion 4a, relative to larger compounds 4b and 4c, in 1:1 3a–f and tetrasubstituted dendritic Pc 9. All compounds
CH2Cl2–EtOH may be due to the strong association of formed crack-free solid films exhibiting a peak in the
pyridine to the zinc core.
Q-band region indicative of varying states of aggrega-
The enthalpy and entropy of aggregation were deter- tion (Fig. 7). While the spectrum of first-generation 3a
mined for dendrimers 3b and 3e in toluene (20 µM) by appeared to be fully aggregated, the spectrum of second-
measuring their respective Kd values over a ca. 60 °C generation 3b suggests less than full aggregation. The
Israel Journal of Chemistry
49
2009

19
3a
4a
9
3b
10
4b
3c
4c
4d
3d
3e
3f
4e
4f
500
600
700
800
500
600
Wavelength (nm)
700
800
Wavelength (nm)
Fig. 7. (Top) UV-Vis spectra of thin-films of 3a, 3b, 9, and Fig. 8. (Top) UV-Vis spectra of the thin-films for 4a, 4b, 4c,
3c (—) Pcs. (Bottom) UV-Vis spectra of thin-films of 3d–3f and 10. (Bottom) UV-Vis spectra of the thin-films for 4d–4f
t-butyl-substituted Pcs.
t-butyl-substituted Pcs.
increase of peripheral dendron size decreases the abil- this striking similarity between the aggregation states of
ity of the Pc to aggregate. When compared to previous the two series of compounds confirms that compounds
reports, octasubstituted second-generation Pc 3b shows of similar molecular weight behave similarly in the
less thin-film aggregation than a tetrasubstituted third- condensed phase, it is nevertheless surprising when
generation Pc prepared by McKeown and coworkers.22 compared to the contrasting behavior that these two
Tetrasubstituted second-generation dendrimer 9 exhib- series of compounds exhibited in solution (Fig. 3). This
ited a similarly split Q-band, indicating an intermediate further supports the proposal that the aggregation of the
level of aggregation in this sample. Interestingly, a com- t-butyl-substituted Pcs 3d–3f relative to unsubstituted
pletely analogous dendritic Pc lacking the hydroquinone Pcs 3a–3c in solution is not merely a result of repulsive
linker was reported to be aggregated in the condensed van der Waals interaction, but is largely an entropic (sol-
phase,22 suggesting that the presence of the linker seems vophobic) effect.
to prevent full aggregation of the Pc. In contrast, third-
generation 3c exhibited a complete lack of aggregation Thin-Film Behavior of Dendritic Zinc Phthalocyanines
in the solid state as shown by the presence of a Q-band,
Thin-films of metalated Pcs 4a–4f and 10 were fab-
similar to that seen in dilute solution absorption spectra ricated by spin-casting from chloroform solutions onto
(Fig. 2), a characteristic of non-aggregating Pcs.43 This glass slides. The absorbance spectra of these films be-
complete lack of aggregation in thin-film Pcs utilizing tween 500 and 800 nm (Fig. 8) made it apparent that the
unsubstituted poly(aryl ether) dendrons has previously zinc Pcs exhibited much less aggregation when com-
been reported only when the substitution is in the center pared to the unmetallated Pcs in the solid state. For the
of the Pc, attached to a silicon atom.21
unsubstituted Pcs, there was a rapid progression from
Thin-film absorption spectra of the t-butyl-substitut- aggregated to unaggregated chromophore between first-
ed dendritic Pcs 3d–3f exhibited a simlar trend (Fig. 7). generation 4a and second-generation 4b. Interestingly,
The smallest compound, zeroth-generation 3d, showed tetrasubstituted second-generation Pc 10 appeared to
the absorption characteristics of a completely aggre- exist in an intermediate level of aggregation between
gated chromophore, while first-generation 3e provided these two octasubstituted compounds. Third-genera-
an absorption spectrum suggesting less than full aggre- tion 4c showed no sign of aggregation in the condensed
gation. Second-generation 3f exhibited a complete lack phase. The t-butyl-substituted Pc dendrimers followed
of aggregation, as indicated by its split Q-band. While a similar trend that was seen with the unsubstituted Pc
Kernag and McGrath / Aggregation Studies of Dendritic Phthalocyanines

20
(5) Dini, D.; Barthel, M.; Hanack, M. Eur. J. Org. Chem.
2001, 3759–3769.
(6) O'Flaherty, S.M.; Hold, S.V.; Cook, M.J.; Torres, T.;
Chen, Y.; Hanack, M.; Blau, W.J. Adv. Mater. 2003, 15,
19–32.
dendrimers, with a dramatic decrease in aggregation
for first-generation 4e relative to zeroth-generation 4d
(Fig. 8). Third-generation 4f appeared to contain only
monomeric chromophore.
(7) Chen, Y.; Hanack, M.;Araki, Y.; Ito, O. Chem. Soc. Rev.
2005, 34, 517.
conclusIons
(8) Bonnett, R. Chem. Soc. Rev. 1995, 24, 19–33.
(9) Pandey, R.K.; Zheng, G. Porphyrin Handbook 2000, 6,
157–230.
(10) Moreira, L.M.; dos Santos, F.V.; Lyon, J.P.; Maftoum-
Costa, M.; Pacheco-Soares, C.; da Silva, N.S. Austra-
lian J. Chem. 2008, 61, 741–754.
(11) Shaheen, S.E.; Ginley, D.S.; Jabbour, G.E. MRS Bul-
letin 2005, 30, 10–19.
(12) Yoo, S.-Y.; Domercq, B.; Donley, C.L.; Carter, C.; Xia,
W.; Minch, B.A.; O'Brien, D.F.; Armstrong, N.R.; Kip-
pelen, B. SPIE: Organic Photovoltaics IV 2004, 5215,
71–78.
(13) Brumbach, M.; Placencia, D.; Armstrong, N.R.J. Phys.
Chem. C 2008, 112, 3142.
(14) Brewis, M.; Hassan, B.M.; Li, H.; Makhseed, S.; McK-
eown, N.B.; Thompson, N.J. Porphyrin Phthalocya-
nines 2000, 4, 460–464.
(15) Ng, D.K.P. C. Rendus C. 2003, 6, 903-910.
(16) McKeown, N.B.J. Mater. Chem. 2000, 10, 1979–1995.
(17) McKeown, N.B. Adv. Mater. 1999, 11, 67–69.
(18) Gorman, C.B.; Smith, J.C. Acc. Chem. Res. 2001, 34,
60–71.
(19) Hecht, S.; Fréchet, J.M. J. Angew. Chem., Int. Ed. 2001,
40, 74–91.
(20) Brewis, M.; Clarkson, G.J.; Holder, A.M.; McKeown,
N.B. Chem. Commun. 1998, 969–970.
(21) Brewis, M.; Clarkson, G.J.; Goddard, V.; Helliwell, M.;
Holder, A.M.; McKeown, N.B. Angew. Chem. Int., Ed.
1998, 37, 1092–1094.
(22) Brewis, M.; Clarkson, G.J.; Helliwell, M.; Holder,
A.M.; McKeown, N.B. Chem. Eur. J. 2000, 6, 4630.
(23) Brewis, M.; Helliwell, M.; McKeown, N.B.; Reynolds,
S.; Shawcross, A. Tetrahedron Lett. 2001, 42, 813.
(24) Brewis, M.; Helliwell, M.; McKeown, N.B. Tetrahe-
dron 2003, 59, 3863–3872.
A series of benzylaryl ether dendrimers with an octa-
substituted Pc core was synthesized, as well as their
metallated zinc derivatives. Within this series we pre-
pared three generations of dendrimer with a benzyl
group periphery (3a–3c) and three generations with a
3,5-di-t-butylbenzyl periphery (3d–3f). GPC data veri-
fied the increase in hydrodynamic size with increasing
generation, and indicated the increased bulk associ-
ated with the t-butyl periphery on the latter group of
dendrimers. Investigation of the solution aggregation
behavior of these compounds, based on the intensity
of the Q-band in the absorbance spectrum, revealed
the inhibiting effect of increasing generation size, and
t-butyl periphery, on self-association. The dimerization
constants (Kd) measured using the protocol of Ng, sup-
ported these conclusions, and verified that the t-butyl-
substituted dendrimers 3d–3f were all less prone to ag-
gregation in solution than the unsubstituted dendrimers
3a–3c, with a monotonic decrease in Kd across the series
3a → 3b → 3c → 3d → 3e → 3f. A strong entropic con-
tribution to the solution aggregation behavior is sug-
gested by a temperature-dependent Kd study. Finally,
thin-film studies indicated that the octasubstitution on
the Pc chromophores was particularly effective at pre-
venting condensed state aggregation. The largest mem-
bers of the two dendrimer groups, third-generation 3c
and second-generation 3f exhibited split Q-bands, simi-
lar to that seen in dilute CH2Cl2 solution, when depos-
ited via spin-coating onto glass slides. The metallated
zinc Pcs 4a–4f all exhibited even less tendency toward
aggregation in both solution and thin-films than their
unmetallated analogues.
(25) Li, X.-Y.; He, X.; Ng, A.C.H.; Wu, C.; Ng, D.K.P. Mac-
romolecules 2000, 33, 2119–2123.
(26) Ng, A.C.H.; Li, X.-Y.; Ng, D.K.P. Macromolecules
1999, 32, 5292–5298.
(27) Ngai, T.; Zhang, G.; Li, X.-Y.; Ng, D.K.P.; Wu, C.
Langmuir 2001, 17, 1381.
Acknowledgments. This work was generously supported by
the Homeland Security Grant program of the University of
Arizona. CAK was the recipient of a University of Arizona
Imaging Fellowship.
(28) Sheng, Z.; Ye, X.; Zheng, Z.; Yu, S.; Ng, D.K.P.; Ngai,
T.; Wu, C. Macromolecules 2002, 35, 3681.
(29) Kimura, M.; Nakada, K.; Yamaguchi, Y.; Hanabusa,
K.; Shirai, H.; Kobayashi, N. Chem. Commun. 1997,
1215–1216.
(30) Kimura, M.; Sugihara, Y.; Muto, T.; Hanabusa, K.;
Shirai, H.; Kobayashi, N. Chem. Eur. J. 1999, 5, 3495–
3500.
rEfErEncEs And notEs
(1) Armstrong, N.R.J. Porphyrins Phthalocyanines 2000, 4,
414-417.
(2) de la Torre, G.; Claessens, C.G.; Torres, T. Chem. Com-
mun. 2007, 2000–2015.
(3) Phthalocyanines. Properties and Applications; Lever,
A.P.B.; Leznoff, C.C., Eds.; VCH: New York, 1989.
(4) Hassan, B.M.; Li, H.; McKeown, N.B.J. Mater. Chem.
2000, 10, 39–45.
(31) Kasuga, K.; Akita, T.; Matsuura, N.; Handa, M.; Sugi-
mori, T. Chem. Letters 2002, 31, 966.
Israel Journal of Chemistry
49
2009

21
(32) Kasuga, K.; Matsuura, N.; Inoue, K.; Handa, M.;
Sugimori, T.; Isa, K.; Nakata, M. Chem. Lett. 2002, 31,
352.
(33) Kernag, C.A.; McGrath, D.V. Chem. Commun. 2003,
1048–1049.
(41) Carpenter, M.S.; Easter, W.M. J. Org. Chem. 1954, 19,
87–102.
(42) Brewis, M.; Clarkson, G.J.; Humberstone, P.; Makh-
seed, S.; McKeown, N.B. Chem. Eur. J. 1998, 4, 1633–
1640.
(34) Still, W.C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43,
2923–2925.
(35) Hawker, C.J.; Fréchet, J.M.J. J. Am. Chem. Soc. 1990,
112, 7638–7647.
(43) Dodsworth, E.S.; Lever, A.B.P.; Seymour, P.; Leznoff,
C.C. J. Phys. Chem. 1985, 89, 5698–5705.
(44) Choi, M.T.M.; Li, P.P.S.; Ng, D.K.P. Tetrahedron 2000,
56, 3881.
(36) Wöhrle, D.; Eskes, M.; Shigehara, K.; Yamada, A. Syn-
thesis 1993, 194–196.
(37) Young, J.G.; Onyebuagu, W. J. Org. Chem. 1990, 55,
2155–2159.
(38) Meade, E.A.; Sznaidman, M.; Pollard, G.T.; Beau-
champ, L.M.; Howard, J.L. Eur. J. Med. Chem. 1988,
33, 363.
(45) Osburn, E.J.; Chau, L.-K.; Chen, S.-Y.; Collins, N.;
O’Brien, D.F.; Armstrong, N.R. Langmuir 1996, 12,
4784–4796.
(46) Nevin, W.A.; Liu, W.; Lever, A.B.P. Can. J. Chem.
1987, 65, 855–858.
(47) Schelly, Z.A.; Harward, D.J.; Hemmes, P.; Eyring, E.M.
J. Phys. Chem. 1970, 74, 3040–3042.
(39) Zeng, F.; Zimmerman, S.C.; Kolotuchin, S.V.; Reichert,
D.E.C.; Ma, Y. Tetrahedron 2002, 58, 825–843.
(40) Kyba, E.P.; Liu, S.T.; Chockalingam, K.; Reddy, B.R. J.
Org. Chem 1988, 53, 3513–3521.
(48) Yang, Y.C.; Ward, J.R.; Seiders, R.P. Inorg. Chem. 1985,
24, 1765.
(49) Monahan, A.R.; Brado, J.A.; DeLuca, A.F. J. Phys.
Chem. 1972, 76, 1994.
Kernag and McGrath / Aggregation Studies of Dendritic Phthalocyanines