Photoelectrochemical Investigation of an Azopyridinium Compound ((CH3)2NC6H4-N=N-C5H4NC18H37I) Modified SNO2 Electrodes

Article abstract of DOI:10.1021/jp9609390

The steady photoelectric response of an azopyridinium compound, (CH3)2NC6H4-N=N-C5H4NC18H37I (AI), monolayer modified SnO2 electrodes prepared by the Langmuir-Blodgett technique has been investigated using a conventional photoelectrochemical cell.Under ambient condition, a cathodic photocurrent is observed when the AI-SnO2 electrode is illuminated by white light (the IR was filtered).The action spectrum is coincident with the absorption of the electrode in the region around 450 nm, indicating the AI aggregates in LB film are responsible for the photocurrent.Dependencies of the photocurrent on some factors have been studied, which may change the megnitude of the observed photocurrent, namely, the light intensity of the irradiation, the concentration of the donor or acceptor in electrolyte solution, bias voltage, and the saturation degree of O2 or N2.Based on the changes of absorption spectra of AI in benzene solvent before and after irradiation, the electrochemical behaviors of AI in LB films are studied and the calculations of bond length, charge distribution, and total energy of different state for an analogue of AI, (CH3)2NC6H4-N=N-C5H4N(1+)CH3 (AM), were performed by using MINDO/3 method.The photocurrent generation is believed to result from photoinduced tautomerism upon irradiation.

Full text of DOI:10.1021/jp9609390

J. Phys. Chem. 1996, 100, 15525-15531
15525
Photoelectrochemical Investigation of an Azopyridinium Compound
((CH3)2NC6H4sNdNsC5H4NC18H37I) Modified SnO2 Electrodes
W. S. Xia, C. H. Huang,* C. P. Luo, L. B. Gan, and Z. D. Chen
State Key Laboratory of Rare Earth Material Chemistry and Applications, Peking UniVersity,
Beijing, 100871, China
X
ReceiVed: March 28, 1996; In Final Form: June 13, 1996
The steady photoelectric response of an azopyridinium compound, (CH
monolayer modified SnO
using a conventional photoelectrochemical cell. Under ambient condition, a cathodic photocurrent is observed
when the AI-SnO electrode is illuminated by white light (the IR was filtered). The action spectrum is
3 2 6 4 5 4
) NC H sNdNsC H NC18H37I (AI),
2
electrodes prepared by the Langmuir-Blodgett technique has been investigated
2
coincident with the absorption of the electrode in the region around 450 nm, indicating the AI aggregates in
LB film are responsible for the photocurrent. Dependencies of the photocurrent on some factors have been
studied, which may change the magnitude of the observed photocurrent, namely, the light intensity of the
irradiation, the concentration of the donor or acceptor in electrolyte solution, bias voltage, and the saturation
degree of O
2 2
or N . Based on the changes of absorption spectra of AI in benzene solvent before and after
irradiation, the electrochemical behaviors of AI in LB films are studied and the calculations of bond length,
+
charge distribution, and total energy of different state for an analogue of AI, (CH
AM), were performed by using MINDO/3 method. The photocurrent generation is believed to result from
photoinduced tautomerism upon irradiation.
3 2 6 4 5 4 3
) NC H sNdNsC H N CH
(
1
5
Introduction
used in the area of information storage, nonlinear optics,16
1
7
photochemical switch, etc. Basic studies on this kind of
compounds are very valuable in “molecular engineering”. Since
aromatic chromophores have π-electron which may facilitate
the electron transfer reaction between donors and acceptors
There is much interest focused on designing artificial systems
on various surfaces such as modified electrode, bilayer lipid
membrane, liquid crystal, and LB films, etc.1 to simulate the
processes occurring in biological systems in which molecules
organized themselves into functional entities, e.g. the photo-
synthetic unit. As one of these prospective techniques, Lang-
muir-Blodgett film may be used in constructing molecular
-4
18,19
linked by them as bridge,
it is anticipated that such molecular
structure with a strong polarity effect may produce considerable
photoelectric response signals. On the basis of this thought,
we first reported a highly polar azopyridinium compound,
electronic and photonic devices in future.5
,6
Photoinduced
(
CH3)2NC6H4sNdNsC5H4NC18H37I (AI), which has an elec-
electron transfer reactions in LB films are among the most
important process which may be controlled at the molecular
level to achieve such purposes. In order to explore more
effective photoelectric response systems, molecular designs have
been concentrating on developing both intramolecular electron
transfer reaction, in which an electron donor and an electron
acceptor moiety are linked into one molecule with an alkyl chain
or a rigid hydrocarbon spacer as with the folded type S-A-D or
linear type A-S-D, etc., and intermolecular electron transfer
reaction by depositing an electron donor or an electron acceptor
tron donor-dimethylamino group and an electron acceptor-
positive pyridinium ion linked by a rigid π-conjugated system
as an electron bridge. Since this highly polar molecule can form
intramolecular charge transfer complex easily, it is very interest-
ing to study its photochemical behaviors which may have
potential uses in the area of photoelectric conversion, photo-
chemical switch, and the electron transfer mechanism within a
molecule itself. Although photoelectrically active bilayer lipid
membrane (BLM) shows strong potential in the construction
4
with sensitizer in LB film through a sandwich depositing mode.
The mechanism of photoinduced electron transfer (PET) has
been vigorously investigated from theoretical and experimental
standpoints by many researchers on account of their potential
20
of biological electronic devices (BED), solving its stability is
an urgent problem facing scientists. In order to overcome the
extreme fragility of the BLM and avoid the limitation of the
solubility of the photosensitizer in the supporting material of
the BLM and also to increase the numbers of sensitizer
molecules per unit area, in the present study, we have used the
Langmuir-Blodgett technique to deposit AI on a SnO2 substrate.
A photocurrent is observed when the film is irradiated and the
quantum yield is about 10 , considerably larger than some of
the artificial photoelectric systems in LB films. The good
correspondence of the action spectrum of AI with the UV-vis
absorption of AI on SnO electrode suggests that the aggregates
in LB films are responsible for the photocurrent generation.
From various photochemical reactions of AI in solution or in
LB film and the quantum chemical calculations, a possible
mechanism for the photoconductive phenomenon is also pro-
posed.
7
8
9
uses in photoinduced charge storage and photosynthetic
10
prototype. Although the detailed molecular arrangements and
the film structures have not been well clarified, the character-
istics of photoelectric energy conversion have been examined.
As examples of these attributions, several artificial photoelec-
-2
11
trochemically active systems in LB films such as merocyanine,
triad molecules, squaraine, Ru(bpy) derivative, etc. have
been investigated. It is found that quantum yields of these
artificial systems were in the order of 10 -10 .
Azo compounds which have many good photoelectrochemical
properties are one kind of potential materials which may be
12
13
3+
14
-
4
-3
2
*
To whom correspondence should be addressed.
Abstract published in AdVance ACS Abstracts, August 15, 1996.
X
S0022-3654(96)00939-2 CCC: $12.00 © 1996 American Chemical Society

1
5526 J. Phys. Chem., Vol. 100, No. 38, 1996
Xia et al.
Experimental Section
Materials. The synthesis and charaterization of AI have been
reported elsewhere.16 Methyl viologen diiodide (MV ) was
synthesized by the reaction of 4,4′-dipyridyl with excess methyl
iodide in refluxing ethanol for 6 h. The product was filtered
and then washed by ethanol for four times and its purity was
2+
1
characterized by H NMR and elemental analysis. The pure
water was in-house deionized water purified by passing through
an EASYpure RF compact ultrapure water system (Barnstead
6
Co., USA) and the resistivity is over 18 × 10 Ω‚cm. All other
reagents were of analytical grade and were used without further
purification except hydroquinone (HQ) which was recrystallized
from water before use. AA represents ascorbic acid.
π-A Isotherms and Film Deposition. The monolayers of
AI were formed by dropping its CHCl3 solution (∼0.5 mg/mL)
on a pure water subphase (20 ( 1 °C, pH 5.6) or 0.1 mol/L
NaI aqueous solution subphase on a British NIMA Technology
Langmuir-Blodgett Model 622 trough. After vaporization of
CHCl3, the surface pressure-area (π-A) isotherms were
recorded. Transparent SnO2 glass substrates with a lateral
Figure 1. Absorption spectra of AI in SnO
CHCl solvent (---).
2
electrode (s) and that in
3
-1
-2
isotherms is about the same, ∼-3.0 mN‚m ‚Å , but their
collapse pressure is not identical, 58.7 mN/m for the former
and 65 mN/m for the latter. Furthermore, the solid phase region
can be extended from 10.7-58.7 mN/m in pure water subphase
to 6.7-65 mN/m in 0.1 mol/L NaI aqueous subphase. All these
facts indicate that the film formation properties are improved
-1
resistance of 10 ohms‚cm were hydrophilically pretreated. The
LB films were fabricated by dipping the substrate into the
aqueous subphase and raised at a rate of 5 mm/min with a
moderate surface pressure (35 mN/m). The multiple layers are
deposited in Y-type mode. The transfer ratios were about 1.
Electrochemical Measurements. The conventional three-
-
with the presence of I ions in the subphase. Comparing with
the area per molecule of AI in the two cases, a difference about
2
6
Å between them is observed. For the former the area per
2
electrode setup with a 0.8 cm effective contact area was used
2
molecule is 36 Å and for the latter the area per molecule is 42
for all irradiation experiments. The film-modified SnO2 glass
was used as a working electrode (WE), a Pt electrode as a
counter electrode (CE), and a Ag/AgCl electrode as a reference
electrode (RE). The electrolyte solution for the measurement
of photocurrent was 1 mol/L KCl aqueous solution. Another
electrolyte solution for electrochemical measurement was the
Brintton-Robinson buffer solution (pH 2-10) containing 1
mol/L KCl. The Britton-Robinson buffer solution consists of
acetic acid, phosphoric acid, boric acid, and hydroxide sodium.
Before every measurement of cyclic voltammetry, the electrolyte
solution was deaerated with high-purity N2 gas for 5 min. All
measurements were made at room temperature on a Model 600
voltammetric analyzer (CH Instruments, USA). The different
concentrations of donor or acceptor in electrolyte solution are
obtained by adding solid sample into the solution directly.
Photochemical Experiments. A 500 W xenon lamp (Ushio
Electric, Japan) was used as the excitation source for the
photochemical experiments. Different wavelengths were ob-
tained by using various filter with a certain band-pass, for
instance, the absorption centered at 450 nm (Toshiba KL-45,
Japan) with a half band-pass of 15 nm. The intensity of incident
beam was checked up by model LM-91 Photopower meter
2
-
Å , indicating that I ions occupy a certain space in the
-
formation of the film. Since I is relatively large (the radius is
2
about 2 Å and the section area is about 12 Å ), it may sustain
the molecules to stand perpendicularly to the electrode surface
resulting in better film formation. In addition, I ions existing
in the subphase may decrease the solubility of AI due to the
co-ions effect.
Figure 1 shows the absorption spectrum of AI on SnO2
electrode and that in CHCl3 solvent. Both spectra have an
absorption at 560 nm which may be attributed to the π f π*
transition of sNdNs group. However, the absorption of AI
on SnO2 electrode shows a broad absorption. The absorption
at 450 nm which has 110 nm blue shift compared with 560 nm
may result from the interaction of the packed azo groups in the
-
21
LB film, the formation of H-aggregates in LB film. The
absorption centered at 560 nm in the LB film may be contributed
by the absorption of monomeric AI. The broad absorption is a
mixture of these two kinds of existing state on the AI-SnO2
electrode. In order to prove the correctness of the above
assignment, we used a lower polar solvent, benzene, to substitute
chloroform as the spread solution. As a result, the absorption
spectrum of AI monolayer on SnO2 electrode prepared from
benzene solution also shows a broad absorption with a shoulder
absorption at 450 nm. Therefore, the blue shift of the absorption
may result from the formation of aggregates, rather than the
change of solvent polarity.
Photocurrent Generation from AI-SnO2 Electrodes. A
steady cathodic photocurrent is observed when the AI-SnO2
electrode is illuminated by white light or 450 nm as shown in
Figure 2. Among the more than 20 parallel data resulting from
AI-SnO2 electrodes, the photocurrent observed ranges from 380
to 580 nA in 1 mol/L KCl electrolyte solution without any bias
(National Institute of Metrology, Beijing, China).The IR light
was filtered for all the experiments with Toshiba IRA-25S filter
(Japan).
Spectroscopic Measurement. The UV-vis spectra were
recorded on a Shimadazu UV-3100 spectrometer (Japan).
Quantum Chemical Calculation. MINDO/3 calculation
2
method from Ceris software was adopted in HP9000/720(PA-
RISC) workstation. In order to make the calculation more easy,
+
we selected a model molecule [(CH3)2NC6H4sNdNsC5H4N -
CH3 (AM)] with a short tail methyl to replace the longer tail
octadecyl group in AI.
2
voltage under a light intensity of 240 mW/cm (white light)
corresponding with an open circuit voltage (Voc) of ∼160 mV.
The film shows a very stable photoelectric response for tens of
times of switching on and off the light. In order to identify the
contribution of SnO2 to photocurrent, three independent un-
coated SnO2 electrodes were tested. It was found that an anonic
Results and Discussions
Fabrication of AI-SnO2 Electrode. The title compound
AI can form stable monolayers both on pure water surface and
0
.1 mol/L NaI aqueous solution surface. The slope of their

Photoelectric Response of Modified SnO2 Electrodes
J. Phys. Chem., Vol. 100, No. 38, 1996 15527
Figure 2. Photocurrent generation from AI-SnO
2
electrode upon
2
irradiation of white light at 240 mW/cm .
Figure 4. Dependence of photocurrent generation from AI-SnO
2
electrode with vitamin A (AA) and hydroquinone (HQ) in 1 mol/L
2
KCl electrolyte solution upon irradiation of 240 mW/cm .
investigated. There is a linear relationship between the observed
photocurrent and the bias voltage with a slope of 2.25 nA/mV
in the region of -0.4-0.18 V (∼Ag/AgCl). An increase of
the cathodic photocurrent with increasing negative bias to the
electrode can be seen and vice versa. When the bias voltage is
more negative than -0.4 V, the photocurrent reaches a steady
value and does not increase with the negative bias. On the other
hand, when the bias voltage is more positive than 0.18 V, the
photocurrent converse, an anionic photocurrent appearing. The
reason may be that the applied negative voltage forms a strong
electric field in the LB film which may accelerate the photo-
induced electron-hole separation and the velocity of hole shift
to the surface of SnO2 and the electrons flowing to the
electrolyte solution. This decreases the probability of capture
of the photocarrier (electron) by recombination or “trapping”
and lengthens the lifetime of charge separation.
Figure 3. Action spectrum of AI-SnO
2
electrode. The intensities of
different wavelengths were all normalized.
photocurent about 20 nA could be observed when uncoated
SnO2 was used as electrode in the same electrolyte solution upon
2
irradiation of 240 mW/cm white light. While using 450 nm
For the AI-SnO electrodes, it can be seen that the photo-
2
irradiation, no photoresponse was observed from the uncoated
SnO2 electrode (the reason is that the absorption of SnO2 is
below 350 nm). A photocurrent of 120-150 nA was obtained
in 1 mol/L KCl electrolyte solution without bias at 450 nm
current increases linearly with the light intensity in the lower
2
intensity range (<100 mW/cm ), and it gradually reaches a
2
saturation state in the higher intensity range (>240 mW/cm ),
suggesting that all the active molecules may contribute to the
photocurrent at high light intensity.
16
2
irradiation (1.8 × 10 photons/(cm ‚s)), so the quantum yield,
which is cited per absorbed photon, is about 0.14% (the
absorbency ratio of the monolayer on the SnO2 electrode at 450
nm is about 4%). To test the stability of AI monolayer modified
SnO2 electrode, we selected three such electrodes submerged
in 1 mol/L KCl electrolyte solution. At the first 5-6 days, the
photocurrent remained almost unchanged and then decreased
gradually at a rate of 1 nA per day in the testing range of 1
month.
Effect of Electron Donor and Acceptor. As expected, when
2+
the well-known electron acceptor, methyl viologen (MV ) is
added to 1 mol/L KCl electrolyte solution, a linear increase of
2+
photocurrent with the concentration of MV in electrolyte
solution is abserved. The increasing magnitude of photocurrent
2+
is about 50 nA with the addition of 1 mmol/L MV into 1
mol/L KCl solution without any bias voltage under irradiation
2
of 240 mW/cm white light. Contrarily, when the electron
Figure 3 shows the action spectrum of AI on SnO2 electrodes.
The most effective wavelength centers at 450 nm. Comparing
the action spectrum with the absorption spectrum (see Figure
donors ascorbate acid (AA) and hydroquinone (HQ) were added
to electrolyte solution, a sharp decrease of photocurrent at lower
concentration and gradually to be saturated is seen in Figure 4.
HQ is a relatively stronger photocurrent quencher than AA. HQ
did not change the direction of the photocurrent here, unlike
1
), we can see that the action spectrum is coincident with the
absorption spectrum of AI on SnO2 electrode very well in the
region around 450 nm, suggesting that the aggregates of AI
deposited in the LB film are responsible for the photocurrent
generation. However, the monomeric AI molecules cannot
generate photocurrent, indicating that electron transfer within
aggregates is the key event subsequent to photoexcitation of
aggregates.
Effect of Bias Voltage and Light Intensity. The observation
of cathodic photocurrent indicates that electrons flow from the
electrode through the LB film to the electrolyte solution. To
further probe the electron-transfer process between the SnO2
electrode and the LB film, the effect of bias voltage was
1
3
that observed in the squaraine system.
Since O can accept an electron to form a superoxide anion
2
radical, it is a potential electron acceptor. The magnitude of
photocurrent is 2-fold raised after the electrolyte solution is
-
3
13
saturated by O ([O ] ∼ 1.2 × 10 mol/L) (about 5 min or
2
2
longer ventilating) compared with the ambient condition, e.g.,
-4
22
[O2] ∼ 2.7 × 10 mol/L. However, when N2 gas was used
to degas the O2 in electrolyte solution, the effect to the
photocurrent is not significant. The decrease ratio of the
photocurrent is smaller than 5% and it can be recovered after
the old electrolyte solution is updated by a new one.

1
5528 J. Phys. Chem., Vol. 100, No. 38, 1996
Xia et al.
TABLE 1: Photocurrent Value of Multiple Layers of
AI-SnO Deposited from 0.1 mol/L NaI Aqueous Solution
under 10 mW/cm Irradiation Using 1 mol/L NaI Electrolyte
2
2
2
+
Solution Containing 10 mmol/L MV
no. of layers
1
3
5
7
9
photocurrent (nA)
500
1000
1100
1155
970
TABLE 2: Electrochemical Data (vs Ag/AgCl) of AI in LB
Film in Different pH Electrolyte Solution
pH
E
c
(V)
E
a
(V)
∆E (V)
2
4
6
0.07
-0.10
-0.24
0.24
0.11
0.02
0.17
0.21
0.26
Figure 5. Photoelectric response of AI-SnO
2
, deposited from 0.1
mol/L NaI aqueous solution subphase, in 1 mol/L KCl electrolyte
photocurrent for all the multiple-layer-modified SnO2 electrodes
in 1 mol/L KCl electrolyte solution. For instance, only a saturate
photocurrent of 90-120 nA can be observed for the SnO2
electrode which was coated with nine layers of AI under 10
2
solution under an irradatiation of 10 mW/cm .
Comprehensive Effect of Favorable Factors. From the
above results, it can be seen that negative bias voltage, electron
acceptor, and the concentration of O2 in electrolyte solution are
all positive factors in enhancing the photocurrent. To test
whether these factors produce contradictory contributions to the
photocurrent when they are applied simultaneously, we selected
the above positive factors. It is interesting to find that for the
four independent testing samples, the photocurrent may be
enhanced 8 times when the conditions of -100 mV bias, 8 mM
2
mW/cm irradiation in 1 mol/L KCl electrolyte solution. When
2+
1
mol/L NaI aqueous solution together with 10 mmol/L MV
was used to substitute the 1mol/L KCl electrolyte solution into
the electrolyte solution, a stable photocurrent for SnO2 electrode
2
coated by one layer of about 500 nA under 10 mW/cm
irradiation can be obtained. The results are listed in Table 1.
The magnitude of photocurrent did not increase with increasing
layer numbers after three layers. It may reach a stable value
about 1100 nA. As a comparison, three independent testing
samples of two-layer modified SnO2 deposited from pure water
subphase in 1 mol/L KCl electrolyte solution give only about
2
+
MV in 1 mol/L KCl electrolyte solution, and O2-saturated
electrolyte solution were used than that resulting from 1 mol/L
KCl electrolyte solution only. As a result, the quantum yield
at 450 nm reached about 0.9-1.1%.
2
20-280 nA stable photocurrent under an irradiation of 240
2
mW/cm .
Since AI contains a long alkyl chain, the resistance will
Effect of I- and the Layer Numbers of AI. From the
discussion of the Langmuir-Blodgett deposition, we found that
increase with the number of layers. That is why the two layers
-
-
I ions occupy a certain area in the formation of the film. The
of AI on SnO deposited from pure water subphase without I
2
presence of I- ions in the LB film makes the photocurrent
generation much different from the photocurrent generated from
the AI-SnO2 electrode deposited from pure water subphase.
ions generate a smaller photocurrent than one layer film
deposited from 0.1 M NaI aqueous subphase. The presence of
-
I ions in LB films and in electrolyte solution depresses diffuse
2
-
Under a lower irradiation power 10 mW/cm , a peak current
of I ions from LB film to the electrolyte solution. The increase
(
∼900 nA) following a gradual decreasing to reach a steady
in the photocurrent may result from the decrease of film
resistance.
Photoelectrochemical Behavior of AI in LB Film. It is
value (∼240 nA) was observed as shown in Figure 5. After
that, only the steady value can be repetitive without the
instantaneous peak current. The addition of electron acceptor
MV cannot help to obtain a stable photocurrent for every new
LB film of AI deposited from 0.1 mol/L NaI aqueous subphase.
Since there are no I ions in electrolyte solution, I ions may
diffuse to the electrobilayer at the interface between the electrode
and the electrolyte solution to reach an equilibrium slowly and
consequently, the photocurrent may reach a steady state gradu-
ally. On the basis of this idea, 1 mol/L NaI was used as
electrolyte solution, the peak photocurrent still appeared for
well-known that the redox reaction of azo group must involve
2+
23,24
two protons.
Table 2 lists the different electrochemical data
of AI in LB film at different pH. From it, one can conclude
that the redox process becomes more irreversible with the
increase in pH value. When the pH value is larger than 7, the
reduction peak cannot be distinguished clearly. Furthermore,
with the pH value increasing, the reduction and oxidation
potentials shift to negative direction. In each pH value, many
cycles of scanning were carried out and the signals were very
stable. There is a certain variation of the reduction potential
-
-
-
every new AI LB films containing I ions, but the magnitude
of the steady photocurrent is larger than that in 1 mol/L KCl
for different film; for example, the E varies from -0.24 to
c
-
electrolyte solution. The reason may be that I ions in LB film
-0.32 V for the films deposited from pure water subphase at
can decrease the film resistance. When some MV2 is added
+
different times, while the oxidation potential shows little
to the 1 mol/L NaI electrolyte solution to make the concentration
variation, E ) 0.02 V. That may result from the different
a
2+
of MV about 10 mmol/L, only one stably steady photocurrent
aggregating states of AI in LB film.
-
appeared for every new AI LB film containing I ions. The
Therefore, the electrode reaction of AI in lower pH value
may occur as in Scheme 1.
To better understand the electrochemical behavior of AI in
LB film, the relationship between the peak current of reduction
value of the photocurrent reaches about 500 nA for the AI LB
2
film modified SnO2 electrode under 10 mW/cm irradiation. The
quantum yield at 450 nm is about 0.7%. That is a considerable
enhancement of the photocurrent compared with the AI-SnO2
or oxidation, I , and the square root of different scan rates has
been studied in pH 4.0 B-R buffer solution. At lower scan rates
p
-
without the presence of I ions.
Since I- can enhance the photocurrent from the AI-SnO2
electrode and improve the film formation properties, multiple
layers of LB films were deposited in 0.1 mol/L NaI aqueous
subphase. Just like the results mentioned above, there is a high
peak photocurrent following a slow decay to reach a saturate
(e100 mV/s), the peak current of reduction (Ip,red) is linear to
the square root of different scan rates, but it deviates from the
linear relationship at higher scan rates while the peak current
of oxidation (Ip,ox) on SnO2 electrode keeps a good linear
relationship to the scan rate in the studied range (20-800 mV/

Photoelectric Response of Modified SnO2 Electrodes
J. Phys. Chem., Vol. 100, No. 38, 1996 15529
2
Figure 6. Change of CV curves of AI-SnO electrode (1) before
2
irradiation, (2) under irradiation of 240 mW/cm , (3) keeping on
irradiation, and (4) switching off irradiation at pH 6 (R-B buffer). Scan
rate is 10 mV/s.
s). The linear relationship indicates that the redox reaction of
AI in LB film is controlled by electron transfer process. With
the increment of scan rate, the difference between the reduction
and the oxidation peak position increased in the CV curves from
AI-SnO2 electrode and the process becomes more irreversible.
Unlike other organic materials, AI has no detectable fluo-
rescent properties, so the excited states cannot be studied by
using fluorescent characterization. Fortunately, its electro-
chemical activity provides a new method to detect the excited
state when the AI-SnO2 electrode is illuminated. Figure 6
shows the CV curve of the AI-SnO2 electrode at pH 6.0 before
Figure 7. Calculated bond lengths (give in Å) and atomic charges
(
underline, give in +e) of AM molecule skeleton. All hydrogen atomic
are omitted for clarity.
simultaneously while there is only one strong absorption at 555
nm in a higher polarity solvent, for example, ethanol. The
similar results can be obtained in acetone (λmax ) 553 nm) and
in chloroform (λmax ) 560 nm) solvent. The process may be
demonstrated as Scheme 2. Considering the change in Scheme
2
, we can infer that AI may exist as a mixture of state A and
2
irradiation and under irradiation of white light of 240 mW/cm .
state B in benzene solvent, but in polar solvent, it may exist as
only one structure, state B.
From Figure 6, it can be seen that AI in SnO2 electrode
undergoes a redox process before irradiation due to the redox
reaction of azo group. After the light is switched on, the
reduction peak current increases and the potential position shows
little shift. However, the oxidation peak shows about 130 mV
negative shift, e.g. Ea shifts from 0.02 to -0.11 V, meaning
that the excited state is much easier to be oxidized. When the
light irradiation is kept on, the shape of the CV curve shows
little change, but it recovers to the original state rapidly after
the light is switched off. The same results were obtained after
repeating more than 10 times. It is interesting that the lower
the pH value is, the smaller the negative shift of the oxidation
potential.
By using the MINDO/3 method based on the assumption of
neglecting the interaction of the molecules, the two possible
+
states of simplified molecule, (CH3)2NC6H4sNdNsC5H4N -
CH3 (AM), have been calculated. The positive charge distribu-
tion and fitting bond length of AM corresponding to the lowest
total energy after optimization are shown in Figure 7.
From MINDO/3 calculations, the dipole moments of state A
and state B are 15.16 and 4.38 D, respectively. Data also show
that the bond lengths are quite different in the two states. In
state B, all the bond lengths are nearly uniform; for instance,
the bond lengths of ph-N-N-py for state A are 1.443, 1.273,
and 1.433 Å, respectively. Conversely, in state B, they are
1.302, 1.405, and 1.301 Å. Although both state A and state B
bear a net positive charge, the electrons of state B delocalized
much stronger than that of state A, suggesting that state B has
a larger π-conjugate system than state A. Consequently, a
Mechanisms of the Photocurrent Generation. Tautomer-
ism of AI in Solution. Like other solvachromic materials, the
difference of the solvent polarity can lead to the absorption
variation of AI. In lower polarity solvent, such as benzene, an
absorption at 400 nm and an absorption at 540 nm appear
SCHEME 1
SCHEME 2

1
5530 J. Phys. Chem., Vol. 100, No. 38, 1996
Xia et al.
Figure 9. Logarithmic plot of absorption of AI in benzene solvent
after irradiation under dark in room temperature via time (minutes).
Figure 8. Change of absorbency of AI in benzene solvent before (---)
and after (s) irradiation with white light.
totally has +0.84e for state A, while the corresponding total
charges of state B are +0.49e and +0.71e, respectively. Since
state B has a larger delocalized π-system than that of state A,
it may be easy to reallocate the distribution of energy after
excitation. As a result, the stronger dipole moment of state A
makes the intermolecule charge transfer happen easily to become
state B. From MINDO/3 calculations, we know that the total
energy of state A (trans-AI) is -2785.7 eV and that of state B
is -2784.8 eV (∆E ) 0.9 eV), while the total energy of cis-AI
is -2717.8 eV, much higher than that of trans-AI. Since 550
nm corresponding to 2.26 eV energy can excite trans-AI in
benzene effectively, this energy value locates in the region of
the energy difference (∼1 eV) of state A and state B and the
difference (∼5 eV) of HOMO and LUMO of state A (HOMO
) -10.14 eV, LUMO ) -4.28 eV). However, if cis-AI can
be formed from trans-AI, about 60 eV energy will be necessary.
Such an amount of energy demand is impossible by 550 nm
irradiation. In addition, this photoinduced tautomerization
cannot be observed in relatively higher polar solvent such as
ethanol, indicating the tautomerism may take place as shown
in Scheme 3. Otherwise, the absorption spectrum of AI in
ethanol or aceton may show significant change if AI undergoes
a trans-cis isomerism. All these facts support that photocurrent
longer wavelength absorbency was observed when AI was
dissolved in higher polarity solvent.
The similar absorption of variation of AI in benzene solvent
can be observed after irradiation by white light or at 550 nm
for a few minutes as shown in Figure 8. Before irradiation, AI
in benzene solvent shows two absorption peaks at 400 and 540
nm. After irradiation by white light or 550 nm for 3 min, there
is only one strong absorption peak at 530 nm, but the post-
irradiated state is not thermally stable, following a rapid recovery
process under dark abiding by a first-order kinetic process.
Figure 9 is the kinetic plot of this change where the kinetic
-
1
constant k is 0.25 min and the half-life is 2.8 min. However,
this phenomenon cannot be seen when AI-ethanol solution is
irradiated at the same condition. Comparing the absorption
change of AI before and after irradiation with the similar
spectrum changes of AI in a weaker or stronger polarity solvent,
it is reasonable to suggest that the tautomerization as shown in
Scheme 3 may take place when AI is dissolved in benzene
solvent.
This proposed photoinduced tautomerism is also supported
by our calculations. From Figure 7, it can be seen that the left
side of azo group totally has +0.25e charge and the right side
SCHEME 3
SCHEME 4

Photoelectric Response of Modified SnO2 Electrodes
J. Phys. Chem., Vol. 100, No. 38, 1996 15531
is generated from photoinduced tautomerism other than trans-
cis isomerism. The unification between experimental data and
calculation results shows that the above assumption of neglecting
the interaction of the molecules does not lead to much error
and is indicative of the proposed mechanism.
improve the quantum yield with the help of electron acceptor
in electrolyte solution with a quantum yield about 0.7%. All
these results indicate that AI is an ideal material to be used in
the field of photogenerated charge storage, photoelectric conver-
sion, and other photoelectronic devices. In addition, the
behavior of the novel dye AI also provides a set of basic data
for designing and finding other more effective photoelectronic
materials.
Electron Transfer Mechanisms on AI-SnO2 Electrode. From
the absorption spectrum change of AI in benzene solvent when
the solution is illuminated, a tautomerization takes place upon
irradiation and follows a first-order kinetic process after
illumination under dark. However, when the floating AI was
transferred from the air-water interface to the SnO2 substrate,
the tautomerization cannot be observed using the conventional
UV-vis absorption method. This phenomenon may result from
the strong interaction among the chromophores in the organized
packing of the LB film, therefore resulting in the rapid decay
process from excited state. However, it can be observed clearly
by using cyclic voltammetry under irradiation. The recovery
of the CV curve to the original state rapidly after irradiation
Acknowledgment. The authors thank the National Climbing
Plan A and B and the National Natural Science Foundation of
China for financial support of this project.
References and Notes
(
1) Kotov, N. A.; Kuzmin, M. G. J. Electroanal. Chem. 1992, 341,
2) Ivanov, I., Ed. Thin Liquid Films; Marcel Dekker: New York, 1988;
47.
(
pp 827-1057.
(
(
3) Tien, H. Ti. Prog. Surface Sci. 1992, 41, 337.
4) Fujihira, M. Nanostructrure Based Molecular Materials; Geopel,
(
see Figure 9) is good support of that conclusion.
The flatband potential (æfb) of SnO2 is -0.34V25 (vs SCE)
W., Ziegler, C., Eds.; VCH: Weinheim, Germany, 1992; pp 27-46.
(5) For example: (a) N a¨ bauer, A.; Berg, P.; Ruge, I. Nanastructrure
at pH 6. When Ag/AgCl electrode is used as reference, the
flatband potential is about -0.32 V. In the present study, the
reduction and oxidation potentials are -0.32 V and 0.02 V (vs
Ag/AgCl), respectively, before irradiation at pH 6. After
irradiation the redox process become more reversible: Ea )
Based Molecular Materials; Geopel, W., Ziegler, C., Eds.; VCH: Weinheim,
Germany, 1992; pp 72-78. (b) Kuhn, H. Thin Solid Films 1989, 178, 1.
(6) For example: (a) Ashwell, G. J., Ed. Molecular Electronics:
Research Studies; John Wiley & Sons: New York, 1992. (b) Carter, F. L.,
Ed. Molecular Electronic DeVices; Dekker: New York, 1983.
(7) For example: (a) Moser, C. C.; Keske, J. M.; Warncke, K.; Farid,
-
0.13 V and Ec ) -0.32 V (∼Ag/AgCl) as shown in Figure
R. S.; Dutton, P.L. Nature 1992, 335, 796. (b) Kakitani, T,; Mataga, N. J.
Phys. Chem. 1988, 92, 5059.
6
. The appearance of more negative oxidation potential of the
(
8) For example: (a) Trans-Thi, T. H.; Lipskier, J. F.; Houde, D.; Pepin,
C.; Keszei, E.; Jay-Gerin, J-P. J. Chem. Soc., Faraday Trans. 1992, 88,
129. (b) Willig, F.; Eichberger, R.; Bitterling, K.; Durfee, W. S.; Stock,
excited state makes it easier to give out electrons than that for
the ground state. The reversible redox signals of AI on SnO2
electrode under irradiation may result from the excited state,
probably state B. Based on the oxidation potential of AI
aggregates (∼0.02 V, vs Ag/AgCl), the oxidation potential of
excited AI aggregates (∼-0.13 V, ∼Ag/AgCl), and the energy
level of the SnO2 electrode, a probable mechanism for this
2
W. Ber. Bunsenges. Phys. Chem. 1987, 91, 867. (c) Sakomura, M.; Fujihira,
M. Thin Solid Films 1994, 243, 616. (d) Macdowell, L. M.; Kirmaier, C.
Holten, D. J. Phys. Chem. 1991, 95, 3379.
(9) (a) Naito, K.; Miure, A. J. Am. Chem. Soc. 1993, 115, 5185. (b)
Naito, K.; Miure, A. Thin Solid Films 1994, 242, 191.
(10) Fujihira, M. New Functionality Materials. Vol. C, Synthetic Process
and Control of Functionality Materials; Tusurta, T., Doyama, M., Seno,
M., Eds.; Elservier Science Publishers B. V.: Amsterdam, 1993; pp 473-
conductive process can be shown in Scheme 4.
The reduction potential of MV2+ is -0.10 V (vs Ag/AgCl)
4
80.
in the present study at pH 6, close to the oxidation potential of
the AI excited state. It is reasonable for MV2 to act as an
effective electron acceptor and enhance the cathodic photocur-
rent. Conversely, strong donors such as AA and HQ cannot
enhance the cathodic photocurrent probably because their energy
does not match the energy level of the excited AI on SnO2.
Because the oxidation potential of HQ is about -0.05 V (vs
(
11) For example: Saito, K.; Yokoyama, H. Thin Solid Films 1994, 243,
+
526.
(
(
12) Fujihira, M.; Sakomura, M. Thin Solid Films 1989, 179, 471.
13) Kim, K-s.; Liang, K.; Law, K-Y.; Whitten, D. G. J. Phys. Chem.
1
994, 98, 984.
14) For example: Fujihira, M.; Aoki, K.; Inowe, S.; Takemura, H.;
Muraki, A.; Aoyagui, S. Thin Solid Films 1985, 132, 221.
15) Liu, Z. F.; Hashimoto, K.; Fujishima, A. Nature 1990, 347, 658.
(
(
2
6
(16) Li, H.; Huang, C. H.; Zhou Y. F. Zhao, X. S.; Xie, X. H.; Li T. K.;
Bei J. J. Mater. Chem. 1995, 5 (11), 1871.
Ag/AgCl) at pH 6 and that of AA is about 0.10 V (∼Ag/
AgCl) which are higher than the reduction potential of AI, the
electron transfer from HQ to AI is impossible. Many
(17) Tachibana, H.; Nakamura, T.; Matsmoto, M.; Komizu, H.; Manda,
E.; Niino, H.; Yabe, A.; Kawabata, Y. J. Am. Chem. Soc. 1989, 111, 3080.
(18) Li, H.; Huang, C. H.; Zhao, X. S.; Xie, X. M.; Xu, L. G.; Li, T. K.
Langmuir 1994, 10, 3794.
1
3,27-29
reports
indicate that O2 may play a vital role in the
electron transfer process through accepting an electron to form
a superoxide radical. In the present AI-SnO2 system, O2 can
enhance the photocurrent as an electron acceptor probably
through the same mechanism. However, when the electrolyte
solution is degassed by N2, it does not decrease in a large scale
(19) Matsuo, T.; Itoh, K.; Takuma, K.; Hashimoto, K.; Nagamura, T.
Chem. Lett. 1980, 1009.
(20) Tien, H. Ti. Bilayer Lipid Membranes, Theory and Practice; Marcel
Dekker: New York, 1974.
(
21) Machae, E. G.; Kasha, M. Physical Processes in Radiation Biology;
Academic Press: New York 1964; p 23.
22) Morov, S. L. Handbook of Photochemistry; Marcel Dekker: New
York, 1973; p 89.
23) Liu, Z. F.; Loo, B. H.; Hashimoto, K.; Fujishima, A. J. Electroanal.
Chem. 1991, 297, 133.
24) Xia W. S.; Huang C. H.; Ye X. Z.; Luo C. P.; Gan L. B.; Liu Z.
F. J. Phys. Chem. 1996, 100, 2244.
25) Mollers, F.; Memming, R. Ber. Bunsenges. Phys. Chem. 1972, 76,
469.
(26) Meites, L., Zuma, P., Rupp, E., Eds. CRC Handbook Series in
Organic Electrochemistry; CRC Press: Boca Raton, FL, 1983; Vol. 4.
27) Haraguchi, A.; Yonezawa, Y.; Hanawa, R. Photochem. Photobiol.
(<5%). From this fact, it is easy to think that O2 is a helpful
(
factor to the photocurrent, but it is not a decisive one at ambient
condition.
(
(
Conclusion
(
The present data from both experiments and theoretical
calculations suggest that the cathodic photocurrent is generated
from the excited state AI. The mechanism is also supported
by studying the effect of bias voltage, the addition of the donors
and acceptors, and light intensity on the photocurrent generation.
(
1
990, 52, 307.
(28) Hada, H.; Yonezawa, Y. Synth. Met. 1987, 18, 791.
(29) Hada, H.; Yonezawa, Y.; Inaba, H. Ber. Bunsenges. Phys. Chem.
981, 85, 425.
-
The addition of I ions can enhance the photocurrent to a large
1
scale. The quantum yield is 0.14% at ambient condition and it
-
can be raised to 1.0% at optimal conditions. I ions also
JP9609390