Photoinduced charge separation of phenothiazine derivatives in layered zirconium phosphate at room temperature

Article abstract of DOI:10.1039/a901801h

The photoionization of alkylphenothiazines in layered α-zirconium phosphate (α-ZrP or ZrP) has been studied by electron spin resonance (ESR) and diffuse reflectance spectroscopies. Alkylphenothiazines (PC(n) where n = 1, 2, 3, 4, 6, 8, 10, 16) were synthesized and used to study the effects of the alkyl chain length. Phenothiazine and N-alkylphenothiazines were incorporated into ZrP by impregnation and ion-exchange methods. By both methods N-alkylphenothiazines are only incorporated onto the external surface of zirconium phosphate rather than into the interlayer space. The alkylphenothiazines photooxidize at room temperature to form stable alkylphenothiazine cation radicals (PC(n)⁺) which are measured by ESR and diffuse reflectance. The framework of ZrP is suggested to be the electron acceptor. Both the photoyield and the decay rate of the alkylphenothiazine cation radical depend on the alkyl chain length. As the alkyl chain length of the PC(n)⁺ cation radical increases from methyl to hexyl, the photooxidation yield increases; and as the alkyl chain length increases further from hexyl to hexadecyl, the photooxidation yield decreases. However, the decay rate of the PC(n)⁺ cation radical gradually increases from methyl to hexadecyl which is explained in terms of a greater inductive effect for longer alkyl chains on the PC(n)⁺ cation radicals. Also, the resolution of the ESR spectrum decreases with increasing alkyl chain length which is interpreted as being due to decreasing mobility of the radical which should also decrease the decay rate. Both impregnation and ion-exchange methods are reasonably effective for incorporating alkylphenothiazine molecules into ZrP for efficient photooxidation. The photoyields of N-alkylphenothiazines in ZrP are larger than in silica gel which suggests that ZrP assemblies can be utilized for solar energy conversion and storage.

Full text of DOI:10.1039/a901801h

View Article Online / Journal Homepage / Table of Contents for this issue
Photoinduced charge separation of phenothiazine derivatives in
layered zirconium phosphate at room temperature
R. M. Krishna, Vadim Kurshev and Larry Kevan
Department of Chemistry, University of Houston, Houston, T X-77204-5641, USA
Received 8th March 1999, Accepted 6th April 1999
The photoionization of alkylphenothiazines in layered a-zirconium phosphate (a-ZrP or ZrP) has been studied
by electron spin resonance (ESR) and di†use reÑectance spectroscopies. Alkylphenothiazines (PC where n \ 1,
n
2
, 3, 4, 6, 8, 10, 16) were synthesized and used to study the e†ects of the alkyl chain length. Phenothiazine and
N-alkylphenothiazines were incorporated into ZrP by impregnation and ion-exchange methods. By both
methods N-alkylphenothiazines are only incorporated onto the external surface of zirconium phosphate rather
than into the interlayer space. The alkylphenothiazines photooxidize at room temperature to form stable
alkylphenothiazine cation radicals (PC`) which are measured by ESR and di†use reÑectance. The framework
n
of ZrP is suggested to be the electron acceptor. Both the photoyield and the decay rate of the
alkylphenothiazine cation radical depend on the alkyl chain length. As the alkyl chain length of the PC`
n
cation radical increases from methyl to hexyl, the photooxidation yield increases; and as the alkyl chain length
increases further from hexyl to hexadecyl, the photooxidation yield decreases. However, the decay rate of the
PC` cation radical gradually increases from methyl to hexadecyl which is explained in terms of a greater
n
inductive e†ect for longer alkyl chains on the PC` cation radicals. Also, the resolution of the ESR spectrum
n
decreases with increasing alkyl chain length which is interpreted as being due to decreasing mobility of the
radical which should also decrease the decay rate. Both impregnation and ion-exchange methods are
reasonably e†ective for incorporating alkylphenothiazine molecules into ZrP for efficient photooxidation. The
photoyields of N-alkylphenothiazines in ZrP are larger than in silica gel which suggests that ZrP assemblies
can be utilized for solar energy conversion and storage.
Phenothiazine and N-alkylphenothiazines with longer alkyl
Introduction
chains are efficient electron-donor molecules. By chemical
methods, they can be easily oxidized to form stable alkyl-
phenothiazine cation radicals which can be monitored by ESR
and di†use reÑectance spectroscopic methods.12h15 These
alkylphenothiazine molecules can also be photooxidized by
near-ultraviolet irradiation. Photoinduced electron transfer of
phenothiazine and methylphenothiazine in micelles has been
studied previously16 using Ñash photolysis with transition
metal ions as electron acceptors at micelle surfaces. Those
studies showed the conversion of the phenothiazine and
methylphenothiazine molecules into phenothiazine and
methylphenothiazine cation radicals. Recent ESR and optical
absorption results of N-alkylphenothiazine derivatives with
longer alkyl chains in silica gels17 demonstrate that long-lived
charge separation can be achieved at room temperature. The
photoionization of alkylphenothiazine molecules in micelles
and vesicles at low temperature has been studied previously
by ESR and electron spin echo modulation (ESEM) methods
to explore the e†ects of alkyl chain length, which controls the
location of alkylphenothiazine molecules in micelles and vesi-
cles.18h21
Production of long-lived photoinduced charge separation is
an important goal in the development of artiÐcial photoredox
systems for solar energy conversion and storage.1h5 The
charge separation in these systems generally involves a redox
reaction between an excited photosensitizer donor (D*) and
an acceptor (A), resulting in the production of radical ion
pairs as follows:
D ] hl ] D*
D* ] A ] D` ] A~
D` ] A~ ] D ] A
(a)
(b)
(c)
The cation and anion produced in this way are better oxidants
and reductants, respectively, than either of the neutral ground
state molecules. To harvest the light energy put into this
system, the oxidizing and reducing power of the photoinduced
species must be used before the electrons are back transferred
as in reaction (c). A considerable challenge is to control this
thermally fast back electron transfer reaction, which is photo-
chemically unproductive. A number of artiÐcial systems have
been tried to improve the efficiency of the energy storage by
preventing the rapid back electron transfer reaction.6h10
Long-lived charge separation is generally achieved by care-
fully designing heterogeneous environments for carrying out
the reaction where the molecular mobility is restricted.1,2,11
The net yield of charge separation has been increased and
back electron transfer has been slowed down signiÐcantly in
heterogeneous systems compared with the same values in
homogeneous solutions. The lifetimes of the radical pairs in
these heterogeneous systems at room temperature range from
seconds to hours.
Inorganic layered phosphates such as a-zirconium phos-
phate, [Zr(HPO ) É H O], referred to as a-ZrP,22 provide
4
2
2
unique microenvironments for the construction of supramo-
lecular assemblies.23 ZrP consists of a layer of zirconium
atoms [Zr(IV)] lying in a plane bridged through tetrahedral
phosphate groups, located alternately above and below the
metal atom plane.24 The metal atoms are octahedrally coordi-
nated by oxygen atoms, with the three oxygens of each phos-
phate tetrahedra bound to three metal atoms. This
arrangement forces the organic groups to lie above and below
Phys. Chem. Chem. Phys., 1999, 1, 2833È2839
2833

View Article Online
the inorganic layer. Each phosphate group in ZrP carries one
ionizable hydroxy group, making these materials highly
acidic. They can be readily deprotonated and the resulting
ionic sites serve to bind cations. Yeates et al.25 used visible
photoacoustic spectroscopy and ESR at room temperature to
study the light-induced electron transfer of tris(2,2@-bipyridine)
ruthenium(II) incorporated into layered zirconium phosphate
lattice with triethylamine as an electron donor.
In the present study, we synthesized a-ZrP and used it as a
heterogeneous host for long-lived photoinduced electron
transfer from N-alkylphenothiazines. The photoionization
behavior of phenothiazine and N-alkylphenothiazine mol-
ecules with longer alkyl chains or di†erent charged groups on
the alkyl chains in layered zirconium phosphate was examined
using ESR and UV-visible di†use reÑectance spectroscopic
techniques. By impregnation or ion-exchange methods the
alkylphenothiazine molecules are incorporated into ZrP.
Attempts to incorporate alkylphenothiazine molecules into
ZrP intercalated with butylamine did not work by either
impregnation or ion-exchange methods. The photoionization
of N-alkylphenothiazine molecules in ZrP was compared for
molecules with di†erent alkyl chain lengths, the alkyl chain
charge and various methods of incorporating the molecules
into ZrP.
Experimental
Synthesis of zirconium phosphate
a-Zirconium phosphate was prepared by the method of Clear-
Ðeld and Stynes.26 One L of orthophosphoric acid (2 M) and
0.9 L of zirconyl chloride (1 M) were added dropwise in a 1: 2
ratio to 100 mL of cold water contained in a 3 L beaker. The
contents were vigorously stirred during the addition, agitation
was continued for about 2 h and the mixture was allowed to
stand overnight. The supernatant liquid was then decanted
and the solid was Ðltered o† and washed on a Ðlter with
approximately 5 L of distilled water. The solid was then dried
at 50È60 ¡C. An X-ray pattern showed only a few very broad
and indistinct peaks characteristic of an amorphous gel.
This gelatinous zirconium phosphate was reÑuxed in 50 g
portions for 48 h in 10 M orthophosphoric acid at 80È100 ¡C.
The mixture was then cooled to about 40 ¡C and Ðltered to
recover the solid. The solid was washed with water until the
effluent attained a pH of 3È4 and was then dried at room
temperature in vacuum or at 70È80 ¡C. An X-ray power dif-
fraction pattern showed very distinct peaks characteristic of
crystalline a-zirconium phosphate (Fig. 1).
Fig. 1 X-ray powder di†raction patterns for crystalline a-ZrP and
N-alkylphenothiazine molecules (PC with n \ 1, 6, 8, 10, 16) incorp-
n
orated into a-ZrP samples.
sumed, the reaction was quenched by pouring the solution
into a separatory funnel, followed by the addition of 100 mL
of water and 100 mL of ethyl acetate. The organic layer was
separated, and the aqueous layer was washed with ethyl
acetate. The combined organic extracts were washed with
water, and the remaining solvent was concentrated by evapo-
ration under reduced pressure. The resulting crude products
were puriÐed by liquid chromatography (Aldrich silica gel,
60È200 mesh, n-hexane eluant). The compounds were identi-
Ðed by NMR spectroscopy. The 1H NMR of PC in CDCl
2
3
showed absorption peaks at 1.12 (CH ), 3.91 (CH N), and
3
2
6
.90È7.45 ppm (phenothiazine ring protons). The 1H NMR of
Synthesis of the N-alkylphenothiazines
PC in CDCl showed absorption peaks at 1.05 (CH ), 1.92
(CH CN), 3.94 (CH N), and 6.94È7.45 ppm (phenothiazine
ring protons). The 1H NMR of PC in CDCl showed absorp-
tion peaks at 1.02 (CH ), 1.88 (CH CN), 1.10È1.42 (C(CH )C),
3
2
3
3
Commercial methylphenothiazine (Aldrich) was used as
2
received. Seven N-alkylphenothiazines (PC where n \ 2, 3, 4,
n
4
2
3
6
, 8, 10, 16) were synthesized27 together with PC tri-
6
3
2
methylammonium bromide and PC sodium sulfonate.
3.89 (CH N), and 6.80È7.40 ppm (phenothiazine ring protons).
6
During the puriÐcation of PC we used ethyl acetate instead
n
2
of methylisobutylketone as used earlier. The synthetic details
are as follows.
N-Hexylphenothiazine (PC ). Twelve g of phenothiazine
6
(
Aldrich, 98%) in 300 mL of anhydrous tetrahydrofuran were
N-Ethylphenothiazine (PC ), propylphenothiazine (PC ) and
reacted with 3 g of sodium hydride (Aldrich) and this solution
was stirred vigorously for 8 h at room temperature. The reac-
tion system was kept under nitrogen. After 8 h of stirring at
room temperature, 14 g of 1-bromohexane were added, and
the solution was stirred for 48 h. The progress of the reaction
was monitored by TLC. When all of the phenothiazine was
consumed, the reaction was quenched by pouring the solution
into a separatory funnel, followed by the addition of 100 mL
of water and 100 mL of ethyl acetate. The organic layer was
separated, and the aqueous layer was washed with ethyl
acetate. The remaining solvent was removed by evaporation
under reduced pressure. The resulting crude products were
2
3
butylphenothiazine (PC ). Ten g of phenothiazine (Aldrich,
4
9
8%) were dissolved in 250 mL of anhydrous tetrahydrofuran
and 2.5 g of sodium hydride (Aldrich) were added to the solu-
tion slowly while the solution was stirred. The reaction system
was kept under nitrogen. After the solution was stirred for 4 h
at room temperature, 8 g of ethyl bromide (Aldrich, 99]%),
1
0 g of 1-bromopropane (Aldrich, 99%), or 12 g of 1-
bromobutane (Aldrich, 99]%) were added, and the solution
was reÑuxed at room temperature for 24 h. The progress of
the reaction was monitored with thin-layer chromatography
(
TLC, n-hexane). When all of the phenothiazine was con-
2834 Phys. Chem. Chem. Phys., 1999, 1, 2833È2839

View Article Online
puriÐed by liquid chromatography (Aldrich silica gel, 60È200
mesh, n-hexane eluant). The compounds were identiÐed by
NMR spectroscopy. The 1H NMR of PC in CDCl showed
product was extracted with ethanol (95%). The Ðnal com-
pound was identiÐed with 1H NMR. The 1H NMR of the
product in D O showed absorption peaks at 1.08È1.32
6
3
2
absorption peaks at 1.03 (CH ), 1.10È1.60 (C(CH ) C), 1.90È
(C(CH ) C), 2.52 (CCH SO~), 3.91 (CCH N), and 6.87È7.53
3
2 3
2 4
2
3
2
1
.93 (CH CN), 3.91 (CH N), and 6.94È7.56 ppm
ppm (phenothiazine ring protons).
2
phenothiazine ring protons).
2
(
Preparation of samples and measurements
N-Octylphenothiazine (PC ), decylphenothiazine (PC ) and
10
hexadecylphenothiazine (PC ). Fifteen g of phenothiazine
8
Incorporation of the N-alkylphenothiazine molecules into
ZrP was performed by impregnation and ion-exchange. For
impregnation, 0.5 g of ZrP powder was immersed in 5 mL of
1
6
were dissolved in 350 mL of anhydrous tetrahydrofuran and
.5 g of sodium hydride (Aldrich) were added to this solution
3
1
0~2
M
PC 1C H , PC TMA`Br~1H O or PC S-
slowly while the solution was stirred under nitrogen at room
temperature. After 10 h of stirring at room temperature, 16 g
of 1-bromooctane (Aldrich, 99%), 18 g of 1-bromodecane
n n n
6 6
2
Na`1H O solution overnight, where n is the number of
carbon atoms in the alkyl chain. The samples prepared with
2
benzene were evaporated by blowing nitrogen gas onto the
surface of the solution to remove benzene, whereas the
samples prepared with water were Ðltered using Ðlter paper
under reduced pressure. All samples were then evacuated
below 1 Torr. For ion-exchange, 10 mL of 10~2 M PC 1C H
(
Aldrich, 98%), or 1-bromohexadecane (Aldrich, 97%) were
added, and the solution was reÑuxed for 48 h at room tem-
perature. The progress of the reaction was monitored by thin
layer chromatography. When the phenothiazine was con-
sumed completely, the reaction was quenched by pouring the
solution into a separatory funnel, followed by the addition of
n
6 6
was added to 1 g of ZrP and the mixture was stirred in the
dark for 5 h and evaporated by blowing nitrogen gas through
the solution to remove benzene. All samples were then evac-
uated below 1 Torr.
1
00 mL of water and 100 mL of ethyl acetate. The organic
layer was separated, and the aqueous layer was washed with
ethyl acetate. The remaining solvent was concentrated by
evaporation under reduced pressure. The resulting crude pro-
ducts were puriÐed by liquid chromatography (Aldrich silica
gel, 60È200 mesh, n-hexane eluant). The compounds were
For ESR measurements the same amount (D20 mm in
height) of dried ZrP1PC was Ðlled into a Supasil quartz tube
n
(
2 mm id ] 3 mm od) which was sealed at one end. The
sample tube was evacuated below 1 Torr for about 6 h and
was Ñame-sealed. All the samples were prepared in the dark to
avoid exposure to visible light. For di†use reÑectance mea-
surements the sample was Ðlled into a cylindrical quartz
sample cell (19 mm diameter by 1 mm path length) which was
evacuated below 1 Torr and Ñame-sealed.
identiÐed by NMR spectroscopy. The 1H NMR of PC in
8
CDCl showed absorption peaks at 0.89 (CH ), 1.27È1.42
3
3
(
C(CH ) C), 1.82È1.87 (CH CN), 3.42 (CH N), and 6.83È7.25
2
5
2
2
ppm (phenothiazine ring protons). The 1H NMR of PC in
1
0
CDCl showed absorption peaks at 1.03 (CH ), 1.38È1.42
3
2
3
(
(
C(CH ) C), 1.95 (CH CN), 3.55 (CH N), and 6.97È7.29 ppm
7
2
2
X-ray powder di†raction (XRD) patterns of all ZrP1PC
phenothiazine ring protons). The 1H NMR of PC in CDCl
n
1
6
3
(
where n \ 0, 1, 2, 3, 4, 6, 8, 10, 16) samples were obtained on
showed absorption peaks at 0.91 (CH ), 1.12È1.64
3
a Philips 1840 powder di†ractometer using Cu Ka radiation
of wavelength 1.541 A, a 0.025¡ step size, and 1 s step time
(
(
C(CH ) C), 1.86 (CH CN), 3.41 (CH N), and 6.85È7.25 ppm
2
13
2
2
Ž
phenothiazine ring protons).
over the range 1.5¡ \ 2h \ 40¡. Samples were mounted on
aluminium slides by sprinkling the powder onto the slides in
order to avoid preferential ordering.
6
-(10º-Phenothiazinyl)hexane-1-trimethylammonium
A Cermax 150 W xenon lamp (ILC-LX150F) was used for
photoirradiation. The light was passed through a 10 cm water
Ðlter and a Corning 7-54 glass Ðlter to give 320 nm irradiation
of 1.3 ] 106 erg cm~2 s~1 in the range from 240 to 400 nm.
This was of sufficient intensity to carry out monophotonic28
photoionization of N-alkylphenothiazine. Each sample was
irradiated in a quartz Dewar that was rotated at 4 rpm to
ensure even irradiation of the samples.
bromide (PC TMA‘Br—). Fifteen g of phenothiazine in 350
6
mL of anhydrous tetrahydrofuran were reacted with 5 g of
sodium hydride and stirred for 4 h at room temperature to
form the phenothiazine N-sodium salt. Then this reaction
mixture was slowly added dropwise into a 1,6-dibromohexane
solution (20 mL of 1,6-dibromohexane (Aldrich, 96%) in 15
mL of dry tetrahydrofuran) to form phenothiazine N-
alkylbromide. The resulting product mixture was Ðltered to
remove solids and the solvent was evaporated under reduced
pressure. The product, 6-(10@-phenothiazinyl)hexyl bromide,
was puriÐed by liquid chromatography (Aldrich silica gel,
ESR was used to detect the photoinduced N-
alkylphenothiazine cation radicals (PC`). ESR spectra were
n
recorded on a Bruker ESP 300 ESR spectrometer with 4 G
magnetic Ðeld modulation and a microwave power of 2 mW
60È200 mesh, n-hexane eluant). Then 5 g of 6-(10@-pheno-
at ambient temperature. The photoinduced PC` cation
thiazinyl)hexyl bromide were dissolved in 30 mL of acetonitri-
le and trimethylamine gas was Ñowed slowly into the solution
at room temperature until there was no more turbid product.
The crude product was then Ðltered under reduced pressure.
The Ðnal compound was identiÐed with 1H NMR of the
product in D O showed peaks at 3.31 [(CH ) N`C)], 3.17
n
radical yields were determined by double integration of the
ESR spectra using the spectrometer software. Each photoyield
is an average of three determinations.
A Perkin-Elmer UV-vis (Model 300) spectrophotometer
with integrating sphere accessory was used to record the
di†use reÑectance spectra. ZrP without N-alkylphenothiazines
was used as a reference sample.
2
3 3
(
7
CCH N`), 3.93 (CCH N), 1.95È2.08 (C(CH ) C), and 6.92È
2 2
2 4
.50 ppm (phenothiazine ring protons).
Results
6
-(10º-Phenothiazinyl)hexane-1-sodium
sulfonate
(
PC S—Na‘). The 6-(10@-phenothiazinyl)hexyl bromide was
The synthesis of crystalline ZrP, its structure and its ion-
exchange properties have been extensively studied.22,24 ZrP is
a layered two-dimensional material with an interlayer distance
6
prepared as described above. Then, 5 g of the resulting
product were added to a saturated aqueous solution of
Na SO and reÑuxed for a few hours for sulfonation. The
of 7.56 A
Ž
which has negatively charged phosphate groups on
apart. The XRD patterns of
2
3
solvent was evaporated, and the solid was washed with a
mixure of dry tetrahydrofuran and hexane to remove the
unreacted 6-(10@phenothiazinyl)hexyl bromide. The solid
the surface spaced about 5.3 A
Ž
ZrP and ZrP1PC with n \ 1, 6, 8, 10, 16 are shown in Fig. 1.
The well resolved, narrow peaks in the XRD patterns are
n
Phys. Chem. Chem. Phys., 1999, 1, 2833È2839
2835

View Article Online
indicative of the excellent crystallinity of the samples. The Ðrst
peak in the pattern has been interpreted24 as the d-spacing
between the layers of the crystal. The exchange of suitable
cations into crystalline ZrP results in enhancement of the
interlayer distance which is dependent on the cation size.
When crystalline ZrP is impregnated or ion-exchanged with
turn dark pink after irradiation, which is characteristic of N-
alkylphenothiazine cation radicals.29 The ESR signal inten-
sities
with
irradiation
time
of
impregnated
methylphenothiazine samples in gel ZrP1PC and crystalline
1
ZrP1PC forms are shown in Fig. 3 for comparison. The ESR
1
signal intensities slowly increase in gel ZrP1PC during the
Ðrst 30 min and reach a plateau, whereas in crystalline
1
PC molecules, no e†ect on the XRD patterns of the ZrP indi-
n
cates that the molecules of PC are adsorbed only on the
ZrP1PC the signal intensities increase continuously even
n
1
external surface of the ZrP (Fig. 1).
after 180 min irradiation time. Figs. 4 and 5 show the increase
Samples of ZrP impregnated or ion-exchanged with N-
alkylphenothiazine molecules show very weak ESR signals
before irradiation (Fig. 2). This indicates that some alkyl-
in the ESR signal intensities of impregnated N-
alkylphenothiazine-containing samples ZrP1PC with n \ 2,
n
3, 4, 6, 8, 10, 16 with irradiation time. The variations in ESR
phenothiazine cation radicals (PC`) are produced during the
sample preparation. The samples showed strong ESR signals
signal intensities are di†erent from those observed in
n
ZrP1PC samples for the same irradiation times. The net
1
(
Fig. 2) typical of PC` cation radicals at room temperature
photoionization yields of ZrP1PC are plotted as a function of
n
n
after being irradiated by 320 nm light at room temperature.17
This indicates that a substantial number of stable cation rad-
icals are generated by UV radiation. The observed ESR
spectra are broad, asymmetric and only partially resolved.
the pendent alkyl chain length in Fig. 6. An irradiation time of
60 min was selected for comparative yields.
The di†use reÑectance spectra of ZrP1PC
and ZrP1PC3
1
recorded before irradiation and after being irradiated at room
temperature are shown in Figs. 7 and 8. The di†use reÑectance
spectrum shows a strong absorption in the UV region before
irradiation, which is attributed to N-alkylphenothiazine mol-
ecules (PC and PC ) trapped in the ZrP. The absorption
The ESR spectra of samples containing PC are similar to
1
those of the methylphenothiazine cation radical (PC`) in
1
homogeneous solution and in silica gels at room tem-
perature.13,17 The weak ESR signals arising before irradiation
have the same line shape as those observed after irradiation.
The background signals before irradiation were subtracted
from the observed signals after irradiation to estimate the net
photoyield. The samples prepared both by impregnation or
ion-exchange are light pink in color before irradiation and
1
3
spectrum in the visible range recorded after di†erent UV irra-
diation times is the same as the absorption spectrum of alkyl-
phenothiazine cation radicals, PC` in liquids, micelles and
n
silica gels.13,16,17 This indicates that some stable PC` cation
n
Fig. 3 Increase of ESR intensities of gel ZrP1PC and crystalline
1
ZrP1PC at room temperature versus photoirradiation time. The
1
insert shows that the photoyield rapidly plateaus for the gel ZrP1PC
1
system.
Fig. 2 ESR spectra of (a) ZrP1PC , (b) ZrP1PC , (c) ZrP1PC
10 ,
Fig. 4 Increase of ESR intensities of ZrP1PC (where n \ 2, 3, 4, 6)
1
6
n
and (d) ZrP1PC versus photoirradiation time.
at room temperature versus photoirradiation time.
16
2836
Phys. Chem. Chem. Phys., 1999, 1, 2833È2839

View Article Online
Fig. 8 UV-vis di†use reÑectance spectra of ZrP1PC samples at
3
room temperature before irradiation and after 1, 5 and 30 min room
temperature irradiations at 320 nm.
Fig. 5 Room temperature ESR spectral intensities of ZrP1PC
n
(
where n \ 8, 10, 16) versus 320 nm irradiation time.
radicals are generated by UV irradiation and conÐrms the
conversion of PC molecules into PC` cation radicals in ZrP.
after 320 nm irradiation are shown in Figs. 9 to 11. The stabil-
n
n
The variations in the intensities of the photoinduced PC`
ity of the PC` cation radicals increases from methyl-
n
n
ESR signals measured at room temperature versus decay time
phenothiazine to hexadecylphenothiazine in ZrP. Also, the
stabilities of the PC` cation radicals in impregnated samples
n
Fig. 9 Decay of relative ESR intensities at room temperature for
ZrP1PC and ZrP1PC with time after 90 min room temperature
0 1
irradiation at 320 nm.
Fig. 6 Room temperature photoyield values of ZrP1PC with n \ 0,
n
1
, 2, 3, 4, 6, 8, 10, 16 measured after 60 min photoirradiation at room
temperature versus the pendent alkyl chain length of the pheno-
thiazines.
Fig. 7 UV-vis di†use reÑectance spectra of ZrP1PC at room tem-
Fig. 10 Decay of relative ESR intensities at room temperature for
ZrP1PC , ZrP1PC and ZrP1PC with time after 90 min room tem-
perature irradiation at 320 nm.
1
perature before irradiation and after 1, 10 and 80 min room tem-
2
3
6
perature irradiations at 320 nm.
Phys. Chem. Chem. Phys., 1999, 1, 2833È2839
2837

View Article Online
Discussion
The methods of impregnation and ion-exchange were chosen
to incorporate the phenothiazine and N-alkylphenothiazine
molecules into ZrP but the interlayer distance of the ZrP was
not a†ected. This indicates that the alkylphenothiazine mol-
ecules do not enter between the ZrP layers and are adsorbed
only on the external surface. This conclusion is based on the
observation of similar di†raction patterns for ZrP and
ZrP1PC (Fig. 1).
n
Phenothiazine and N-alkylphenothiazine molecules are
known to undergo slow oxidation by single-electron steps in
solution into their respective cation radicals.31 The hydrogen
atom attached to nitrogen on phenothiazine can adopt two
distinct conÐgurations, called quasi-equatorial and quasi-
axial.32 The ionization potential of the quasi-equatorial con-
Ðguration is lower than that of the quasi-axial
conÐguration.33 Introduction of an alkyl group at the nitro-
gen of phenothiazine renders the conÐguration quasi-axial due
to steric hindrance.34 Thus, the ionization potential increases
with alkyl substitution. In the present work initially prepared
samples are light pink indicating oxidation and this light pink
Fig. 11 Decay of relative ESR intensities at room temperature for
ZrP1PC , ZrP1PC and ZrP1PC with time after 90 min room
temperature irradiation at 320 nm.
8
10
16
are similar to those in ion-exchanged samples and the stability
increases as the alkyl chain length increases. Thus, the photo-
yield and stability of the alkylphenothiazine cation radicals
are a†ected by changing the alkyl chain length from 1 to 16
carbons.
color decreases in intensity for PC with longer alkyl chains.
Thus the amount of dark reaction decreases with increasing
alkyl chain length, although the reason for this trend is not
understood.
n
The di†use reÑectance and ESR results clearly conÐrm the
photooxidation of alkylphenothiazine molecules into alkyl-
phenothiazine cation radicals in ZrP by UV irradiation at
room temperature. The ZrP framework is the likely electron
acceptor and may involve surface hydroxy groups. In other
work, the hydroxy groups at a silica gel surface have been
found to act as electron acceptors to form hydrogen atoms
which are stably trapped at 77 K in silica gel and detected by
ESR.17,35 The hydrogen atom signal intensity decreases with
smaller pore size.17,35 Since ZrP has a very small pore size
The photoyield of PC` in ZrP1PC` is about 25-fold higher
6
6
than the photoyields of PC TMA2` and PC S~` radicals in
6
6
ZrP1PC TMA` and ZrP1PC S~ for the same irradiation
6
6
conditions as shown in Fig. 12. The stabilities of the photoin-
duced alkylphenothiazine cation radicals are not a†ected by
changing the charge of the alkyl chain. The stabilities of the
PC` cation radicals are similar to those of PC TMA2` and
6
6
6
PC S~` radicals at room temperature which is consistent
with the relative photoyields.
For the photooxidation of N-alkylphenothiazine molecules
there should be an electron acceptor in ZrP. By evacuation,
solvent C H , oxygen, and physisorbed water are removed.
(D3 AŽ ), the lack of trapped hydrogen atoms is consistent with
these earlier studies on silica gel.17
A remarkable change in both the photoyield and decay rate
of the alkylphenothiazine cation radicals with the alkyl chain
length is observed (Figs. 3È6). The alkylphenothiazine cation
radical photoyield increases monotonically from methyl to
hexyl, and as the alkyl chain length further increases from
hexyl to hexadecyl the photoyield slowly decreases. However,
6
6
The hydroxy groups on the surface of ZrP are possible elec-
tron acceptors. They may accept an electron and form hydro-
gen atoms. In some systems hydrogen atoms are stably
trapped at 77 K and are detectable by ESR.17,30 The result
that no trapped hydrogen atoms are observed by ESR sug-
gests that the hydrogen atoms are not trapped in the small
the decay rate of the PC` cation radicals monotonically
n
pores of ZrP or that the photoionization of PC in small pores
decreases from methyl to hexadecyl. The decrease in decay
n
does not generate hydrogen atoms.17
rate can be explained in terms of an increased inductive e†ect
of the PC` cation radicals as a function of longer alkyl chain
n
length.36h39
A plot of relative ESR intensities versus alkyl chain length
of PC` cation radicals gives a bell-shaped curve with
n
maximum photoyield centered at ZrP1PC` (Fig. 6). A pos-
6
sible model for this trend is as follows. When the PC alkyl
n
chain length is short, some of the PC molecules react with the
n
atmospheric oxygen during sample preparation, oxidize and
turn light pink. Increasing the alkyl chain length of the PC
molecules reduces the extent of this dark reaction with atmo-
n
spheric oxygen. If the PC molecules are distributed uniformly
n
on the surface of the ZrP independent of alkyl chain length,
then the photooxidation yields are expected to increase mono-
tonically. Thus, the decreased photooxidation yields for
ZrP1PC` to ZrP1PC` may suggest that PC to PC are
7
16
7
16
bound nonuniformly on ZrP in clusters which makes oxida-
tion more difficult.35,38 In other work, N-alkylphenothiazines
with longer alkyl chain length form clusters and change their
reactivity toward oxidizing agents.40 Clusters may also self-
quench their excitation by intermolecular transfer to a surface
trap on ZrP.41
Fig. 12 Relative PC`, PC S~`, and PC TMA2` photoyields of
6
6
6
ZrP1PC`, ZrP1PC S~` and ZrP1PC TMA2`. The ESR spectra
6
6
6
The e†ect of alkyl chain charge on the radical photoyield
were recorded at room temperature after 60 min room temperature
irradiation at 320 nm.
(
Fig. 12) indicates that only the neutral PC increases the
n
2838
Phys. Chem. Chem. Phys., 1999, 1, 2833È2839

View Article Online
5
6
7
J. S. Connolly, Photochemical Conversion and Storage of Solar
Energy, Academic Press, New York, 1981.
A. Slama-Schwok, D. Avnir and M. Ottolenghi, Photochem. Pho-
tobiol., 1991, 54, 525.
D. Gust, T. A. Moore, P. A. Liddell, G. A. Nemeth, L. R.
Makings, A. L. Moore, D. Barrett, P. J. Pessiki, R. V. Bensasson,
M. Rougee, C. Chachuty, F. C. De Schryver, M. van der
Auweraer, A. R. Holzwarth and J. S. Connolly, J. Am. Chem.
Soc., 1987, 109, 846.
P. D. Dutta and J. A. Incavo, J. Phys. Chem., 1987, 91, 4443.
T. J. Meyer, Acc. Chem. Res., 1989, 22, 163.
T. Nakato, K. Kazuyuki and C. Koto, Chem. Mater., 1992, 4,
128.
photoyield. The photoyield and stability of PC` cation rad-
n
icals in ZrP1PC prepared by either impregnation or ion-
n
exchange are similar. Therefore, both impregnation and
ion-exchange seem equally good for incorporating PC mol-
n
ecules onto ZrP for efficient photooxidation.
The relative mobilities of the PC` cation radicals are
n
revealed by the partially resolved ESR spectra in Fig. 2. The
results suggest a greater mobility for PC` with smaller alkyl
n
8
9
0
chain lengths which correlates with greater probability for
decay by back electron transfer from ZrP or by some other
1
process (Figs. 9È11). The lifetimes of the PC` cation radicals
n
in ZrP range from a few days to several weeks at room tem-
11 A. Slama-Schwok, M. Ottolenghi and D. Avnir, Nature, 1992,
55, 40.
W. F. Forbes and P. D. Sullivan, J. Am. Chem. Soc., 1966, 88,
862.
3
perature, depending on the alkyl chain length of the PC`
n
1
2
3
cation radicals (Figs. 9È11).
2
The results obtained here for the photoionization of N-
alkylphenothiazines in ZrP with longer alkyl chain lengths are
di†erent from previous work on the photooxidation of N-
alkylphenothiazines with longer alkyl chain lengths in silica
gels.17 As a function of alkyl chain length in ZrP, the photo-
yield increases nearly threefold when compared to silica gel.17
1
H. J. Shine, D. R. Thompson and C. Veneziani, J. Heterocycl.
Chem., 1967, 4, 517.
14 E. R. Biehl, H. S. Chiou, J. Keepers, S. Kennard and P. C. Reeves,
J. Heterocycl. Chem., 1975, 12, 397.
1
1
1
5
6
7
D. Clarke, B. C. Gilbert and P. Hanson, J. Chem. Soc., Perkin
T rans., 1978, 2, 1103.
S. A. Alkaitis, G. Beck and M. Gratzel, J. Am. Chem. Soc., 1975,
97, 5723.
Also, the stability of PC` cation radicals is higher (from days
n
to several weeks) in ZrP at room temperature as compared to
B. Xiang and L. Kevan, L angmuir, 1994, 10, 2688.
silica gel where the PC` cation radicals are stable only for a
18 M. Sakaguchi, M. Hu and L. Kevan, J. Phys. Chem., 1990, 94,
n
few hours at room temperature. This suggests that the photo-
870.
1
2
9
0
M. Hu and L. Kevan, J. Phys. Chem., 1990, 94, 5348.
Y. S. Kang, P. Baglioni, H. J. D. McManus and L. Kevan, J.
Phys. Chem., 1990, 94, 870.
oxidation mechanism of PC` cation radicals in ZrP is more
n
efficient than in silica gel for similar molecules.
21
22
23
Y. S. Kang, H. J. D. McManus and L. Kevan, J. Phys. Chem.,
1
993, 97, 2027.
Conclusions
Inorganic Ion Exchange Materials, ed. A. ClearÐeld, CRC, Boca
Raton, FL, 1982.
The e†ect of the alkyl chain length on the photooxidation
yield of a series of N-alkylphenothiazine molecules in ZrP was
determined by ESR and di†use reÑectance spectroscopy. The
S. B. Ungashe, W. L. Wilson, H. E. Katz, G. R. Scheller and T.
M. Putrvinski, J. Am. Chem. Soc., 1992, 114, 8717.
A. ClearÐeld and G. D. Smith, Inorg. Chem., 1969, 8, 431.
R. C. Yeates, S. T. Kuznicki, L. B. Lloyd and E. M. Eyring, J.
Inorg. Nucl. Chem., 1981, 43, 2355.
2
2
4
5
photoyield and stability of the PC` cation radicals depends
n
strongly on the pendent alkyl chain length. The highest photo-
yield is obtained for ZrP1PC` whereas the maximum radical
26 A. ClearÐeld and J. A. Stynes, J. Inorg. Nucl. Chem., 1964, 26,
17.
27 I. Gozlan, D. Ladkani, M. Halpern, M. Rabinovitz and D. J.
6
1
1
stability is found for ZrP1PC` . The photoyield and stability
6
also depend upon the mobility of the alkylphenothiazine
cation radicals. Impregnation and ion-exchange methods are
equally good for incorporating alkylphenothiazine molecules
onto ZrP for efficient photooxidation. By both impregnation
or ion-exchange, the N-alkylphenothiazine molecules are
adsorbed only on the external surface of ZrP. These results
clearly highlight the utility of a-ZrP in controlling the photo-
Anoir, J. Heterocycl. Chem., 1984, 21, 613.
2
8 S. A. Alkaits, G. Beck and M. Gratzel, J. Am. Chem. Soc., 1975,
97, 5723.
29 M. C. Hovey, J. Am. Chem. Soc., 1982, 104, 4196.
30 H. Hase and L. Kevan, J. Phys. Chem., 1970, 94, 3355.
3
1
A. M. Braun, M.-A. Gilson, M. Krieg, M.-T. Maurett, P. Mura-
secco and E. Oliveros, in Organic Phototransformations in Non-
homogeneous Media, ACS Symp. Ser. 278, ed. M. A. Fox.
American Chemical Society, Washington, DC, 1985, p. 79.
N. I. Wakayama, Bull. Chem. Soc. Jpn., 1971, 44, 2847.
P. Kabasakalian and J. MacGlotten, Anal. Chem., 1959, 31, 431.
C. Bodea and I. Silberg, Advn. Heterocycl. Chem., 1968, 9, 321.
B. Xiang and L. Kevan, L angmuir, 1995, 11, 860.
L. S. Levitt and H. F. Widing, in Progress in Physical Organic
Chemistry, ed. R. W. Taft, Wiley, New York, 1976, vol. 12, ch. 5.
process of PC molecules.
n
32
33
34
35
36
Acknowledgements
This research work was supported by the Division of Chemi-
cal Sciences, Office of Basic Energy Sciences, Office of Energy
Research, US Department of Energy and by the Texas
Advanced Research Program.
37 R. R. Gupta and K. Mahendra, in Phenothiazines and 1,4-Benzo-
thiazines: Chemical and Biomedical Aspects, ed. R. R. Gupta,
Elsevier, Amsterdam, 1988.
3
8
L. Louis, T. N. Tozer, L. D. Tuck and D. B. Loveland, J. Med.
References
Chem., 1972, 15, 898.
3
9
A. Ortiz and I. Poyato and J. I. Fernandez-Alonso, J. Pharm. Sci.,
1
Photoinduced Electron T ransfer, ed. M. A. Fox and M. Chanon,
Elsevier, Amsterdam, 1988.
1
983, 72, 50.
4
4
0
1
E. Pelizzetti and E. Mentasi, Inorg. Chem., 1979, 18, 583.
R. van Grondelle, Biochim. Biophys. Acta, 1985, 811, 147.
2
M. Gratzel, Heterogeneous Photochemical Electron T ransfer,
CRC, Boca Raton, FL, 1989.
3
4
J. H. Fendler, J. Phys. Chem., 1985, 89, 2730.
G. J. Kavarnos and N. J. Turro, Chem. Rev., 1986, 86, 401.
Paper 9/01801H
Phys. Chem. Chem. Phys., 1999, 1, 2833È2839
2839