Kinetic analysis of a photosensitive chelator and its complex with Zn(II)

Article abstract of DOI:10.1562/0031-8655(2002)075<0211:KAOAPC>2.0.CO;2

The reversible sequestration and release of metal ions is an important objective in biological and environmental research. Unfortunately, although there have been dramatic examples of metal ion activity control, there are very few quantitative investigations of stoichiometry, equilibria and kinetics. A significant contributor to this lack of quantitative work is the complexity of many photochromic systems. Therefore, we have attempted to create a simple, reversible photochromic metal-ion chelator that can be analyzed quantitatively. The chelator should have certain other attributes as well, namely, that it binds to divalent metal ions (because of their extreme biological importance) and that it binds metal ions in the dark so that light is used to release metal ions rather than sequester them. The photochromic chelator (1) binds to divalent metal ions [Zn(II), Cu(II), Pb(II), Hg(II), Fe(II), Co(II) and Cd(II); other metal ions have not yet been tested] in the dark with a significant binding strength. In both methanol (by spectrophotometry) and methanol-water (by voltammetry), the stoichiometry of the 1-Zn(II) complex is 2:1. The binding constant (K₁K₂) is on the order of 10¹²-10 ¹⁴ M^-2 in methanol and 5.0 × 10⁸ M ^-2 in 50% aqueous methanol. The chelator 1 is photolabile, yielding 2 with a quantum efficiency of 0.91. In a solution containing excess Zn(II), so that over 99% of the ligand exists as the monodentate complex, photolysis produces 2 with a quantum efficiency of 0.15. A kinetic analysis leads to the conclusion that the complex itself is photolabile.

Full text of DOI:10.1562/0031-8655(2002)075<0211:KAOAPC>2.0.CO;2

Kinetic Analysis of a Photosensitive Chelator and its Complex with Zn(II)
Author(s): Melinda R. Stephens, Caroline D. Geary, Stephen G. Weber
Source: Photochemistry and Photobiology, 75(3):211-220.
Published By: American Society for Photobiology
https://doi.org/10.1562/0031-8655(2002)075<0211:KAOAPC>2.0.CO;2
URL: http://www.bioone.org/doi/full/10.1562/0031-8655%282002%29075%3C0211%3AKAOAPC
%3E2.0.CO%3B2
BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and
environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published
by nonprofit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of
BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries
or rights and permissions requests should be directed to the individual publisher as copyright holder.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research
libraries, and research funders in the common goal of maximizing access to critical research.

Photochemistry and Photobiology, 2002, 75(3): 211–220
Kinetic Analysis of a Photosensitive Chelator and its Complex with Zn(II)¶
Melinda R. Stephens, Caroline D. Geary and Stephen G. Weber*
Department of Chemistry, University of Pittsburgh, PA
Received 27 July 2001; accepted 20 December 2001
ABSTRACT
ture. Such reversible, switchable metal-ion binding activity will
lead to systems that can sequester metal ions while in the avid
form and to systems that can exchange metal ions rapidly while
in the other. These systems have applications in biological re-
search, sensors and extraction. A variety of metal-binding pho-
tochromics have been investigated, including thioindigos
(5,32,33), azobenzenes (13,15,34–36), spirooxazines (30,37–
40), spiropyrans (20,21,41–50), chromenes (23,24), styryl dyes
(31) and triphenylmethanes (14,16,18,51). Despite the many
systems analyzed, few are intended to bind rapidly with metal
ions in the dark (14,16–18,23), a characteristic that is a practical
necessity if these compounds are to be used in biological re-
search, sensors or extraction processes.
An important question that has not been answered for any
divalent metal–binding systems and has only recently been
asked for crowned systems binding to alkali metals involves
the photosensitivity of the complexes formed between metal
ion and photosensitive ligand. If the complexes are photol-
abile, then light can be used to force the dissociation of the
complex. However, if only the ligand is photosensitive, then
light can only influence the metal–ligand binding indirectly.
The latter case is not technically attractive. By a kinetic anal-
ysis, we indirectly show that, for a newly designed and syn-
thesized photoactive chelator, the complex itself can be pho-
tolyzed.
Often, the measurement of binding constants and stoichi-
ometry of these systems is complicated. Thus, a quantitative
analysis of these photochromic complexes has been limited
in number and in the types of systems analyzed. The ma-
jority of quantitative reports have involved azobenzene, styr-
yl or thioindigo structures, which undergo a cis–trans isom-
erization when photolyzed. (13,25,31,34,52–54). These sys-
tems are relatively easy to understand because only two pho-
toproducts exist, cis and trans. In all cases thus far,
metal-binding occurs only in the cis form and is controlled
only by changes in geometric conformation. Although sev-
eral metal-binding spirooxazines, spiropyrans and chro-
menes have been reported (4,17,21,24,49,55–60), only a few
have been analyzed quantitatively (24,46), a result of the
complicated photochemistry of these compounds. In con-
trast, triphenylmethanes exhibit very simple photochromism.
For example, when malachite green (MG) is irradiated with
UV light of 225–300 nm, heterolytic cleavage of the C–R
bond occurs, and a charged photoproduct is formed with
minimal changes in geometric conformation (Scheme 1).
As our goal is to photodissociate tightly bound metal ions,
we thought that the photochromic chelator 1 would be a
The reversible sequestration and release of metal ions is
an important objective in biological and environmental
research. Unfortunately, although there have been dra-
matic examples of metal ion activity control, there are
very few quantitative investigations of stoichiometry,
equilibria and kinetics. A significant contributor to this
lack of quantitative work is the complexity of many pho-
tochromic systems. Therefore, we have attempted to cre-
ate a simple, reversible photochromic metal-ion chelator
that can be analyzed quantitatively. The chelator should
have certain other attributes as well, namely, that it
binds to divalent metal ions (because of their extreme
biological importance) and that it binds metal ions in the
dark so that light is used to release metal ions rather than
sequester them. The photochromic chelator (1) binds to
divalent metal ions [Zn(II), Cu(II), Pb(II), Hg(II), Fe(II),
Co(II) and Cd(II); other metal ions have not yet been
tested] in the dark with a significant binding strength. In
both methanol (by spectrophotometry) and methanol–
water (by voltammetry), the stoichiometry of the 1–
Zn(II) complex is 2:1. The binding constant (K1K2) is on
the order of 1012–1014
M ؋
2 in methanol and 5.0 108 M
Ϫ
Ϫ2
in 50% aqueous methanol. The chelator 1 is photolabile,
yielding 2 with a quantum efficiency of 0.91. In a solution
containing excess Zn(II), so that over 99% of the ligand
exists as the monodentate complex, photolysis produces
2 with a quantum efficiency of 0.15. A kinetic analysis
leads to the conclusion that the complex itself is photol-
abile.
INTRODUCTION
Photochromic compounds have been used to control protein
and enzyme activity (1–12) and metal-ion binding (13–31). In
metal-ion binding systems, light can be used to drive photo-
chromic molecules from one stable or metastable avid binding
structure to another stable or metastable loosely binding struc-
¶Posted on the website on December 31, 2001.
*To whom correspondence should be addressed at: Department of
Chemistry, University of Pittsburgh, Chevron Science Center, Pitts-
burgh, PA 15260, USA. Fax: 412-624-8611; e-mail: sweber@
pitt.edu
Abbreviations: DI, deionized; MG, malachite green; NMR, nuclear
magnetic resonance; THF, tetrahydrofuran; TMP, tetramethylpyr-
idine; TETA, triethylenetetraamine; UV, ultraviolet.
᭧
2002 American Society for Photobiology 0031-8655/02
$5.00ϩ0.00
211

212 Melinda R. Stephens et al.
Scheme 2.
1H-NMR (300 MHz, C3D6O): ␦ ϭ 7.2–7.33 (m, 10H), 7.11 (d, 2H),
6.54 (d, 2H), 4.31 (s, 4H). 13C-NMR (70 MHz, C3D6O): ␦ ϭ 176,
130, 129, 128, 127, 111, 70, 58.
4-[Bis(carboxymethyl-amino)-phenyl] diphenyl methanol (H21)
via 4-[bis(ethoxycarbonylmethyl-amino)-phenyl] diphenyl methanol
(5). Ethyl ester 5 (0.42 g, 0.94 mmol) was dissolved in 20 mL
tetrahydrofuran (THF) and 6 mL DI H2O. The solution was cooled
on ice, and lithium hydroxide monohydrate (88.2 mg, 2.1 mmol)
was added. The reaction mixture was allowed to warm to room
temperature and react for 2 h. The THF was removed in vacuo, and
the cloudy solution that remained was diluted with CH2Cl2 and
poured into a separatory funnel. Slowly, 0.5 N HCl was added until
a dark-red precipitate formed. The product was extracted several
times with CH2Cl2. The organic layers were combined, dried over
Na2SO4, filtered and concentrated in vacuo to yield a shiny orange–
red solid. This solid was dissolved in a small amount of CH2Cl2 and
precipitated with hexanes. After removal of the solvent in vacuo,
H21 remained as a fluffy bright-orange solid (0.2318 g, 63% yield),
which dissolved in methanol and water to yield a colorless solution.
1H-NMR (300 MHz C3D6O): ␦ ϭ 7.2–7.32 (m, 10H), 7.11 (d, 2H),
6.56 (d, 2H), 4.31 (s, 4H). MS–fast atom bombardment (negative)
meta-nitrobenzoic acid 391.
4-[Bis(trimethylsilyl-amino)-phenyl] diphenyl methanol (3). 4-
Bromo-N,N-bis(trimethylsilyl)aniline (Aldrich, St. Louis, MO; 3.8
mL, 13.4 mmol) was added to 40.0 mL anhydrous ether in a 250
mL flask equipped with a reflux condenser, a stirbar and a nitrogen
inlet. A solution of 1.6 M BuLi in hexanes (Aldrich; 8.4 mL, 13.4
mmol) was added, and the mixture was stirred at room temperature
for 1 h. Benzophenone (EM Science, Gibbstown, NJ; 2.27 g, 12.5
mmol) was dissolved in anhydrous ether and added slowly. The
solution was stirred with gentle reflux for 45 min and was then
quenched with 30 mL DI water. The product was extracted with
ether, dried with Na2SO4 and evaporated to yield a yellow oil (crude
Scheme 1.
good target. With only one heteroatom participating in res-
onance stabilization, the effect of the photolysis of the tri-
arylmethanol on the binding constant will be maximal
(Scheme 2).
We will discuss the synthesis, photochromism and metal-
binding ability of 1. Because of 1’s simple photochromism
and unique design, we were able to observe binding of metal
ions to 1 rapidly in the dark, analyze 1 quantitatively and
determine the photosensitivity of a 1–Zn(II) complex. Each
of these analyses are rare in themselves, but, to our knowl-
edge, have never been accomplished in one system.
MATERIALS AND METHODS
General methods. UV–Vis absorption spectra were recorded on an
HP 8452 Diode Array Spectrometer. To allow simultaneous photol-
ysis and UV–Vis observations, a 1 mm cuvette was placed at an
angle of 45Њ. Photolysis was accomplished using a 150 W Xenon
lamp filtered by water for IR and a CVI UG-5 bandpass filter.
(Transmission in the UV is maximal at 370 nm [17.6% T] and falls
to 8% T at 240 and 404 nm and to 2% T at 229 and 428 nm.)
Electrochemical measurements were performed on a Cypress Sys-
tems Model CS-1087 potentiostat controlled by a DTK Keen-2530
weight, 5.44 g; crude yield,
ϳ100%). Crude 3 could be used in the
next step without purification or could be purified by column chro-
matography. During purification, some cleavage of the silyl groups
occurs, but the deprotected amine 3 can be collected. 1H-NMR (300
MHz, CDCl3): ␦ ϭ 7.20 (m, 10H), 7.00 (d, 2H), 6.77 (d, 2H), 2.80
(s, 1H), 0.00 (s, 18H).
computer with Model CS-1087 software. H- and 13C–nuclear mag-
1
netic resonance (NMR) were measured using a Bruker 300 NMR
spectrometer.
4-Aminophenyldiphenyl methanol (4). Crude 3 (ϳ27 mmol) was
Chemicals. Anhydrous ether was prepared with stirring over so-
dium and distillation over Na–benzophenone. Zinc perchlorate (GFS
Chemicals, Powell, OH) and sodium perchlorate (GFS Chemicals)
were recrystallized from water before use. Water for use in all ex-
periments was doubly deionized (DI) with a Millipore Milli-Q sys-
tem.
placed in a flask equipped with a stirbar, nitrogen inlet and reflux
condenser. Deprotection of the amine was accomplished by refluxing
in 35 mL methanol for 2 h. The reaction was allowed to cool, and
the methanol was evaporated to yield an orange fluffy solid (6.9 g,
100% yield). 1H-NMR (300 MHz, CDCl3): ␦ ϭ 7.28 (m, 10H), 7.01
(d, 2H), 6.60 (d, 2H), 3.6 (broad, 2H), 2.8 (broad, 1H).
4-[Bis(carboxymethyl-amino)-phenyl] diphenyl methanol (H21)
via 4-aminophenyldiphenyl methanol. 4-Aminophenyldiphenyl
methanol (3.11 g, 11.3 mmol) was placed in a 250 mL flask
equipped with a reflux condenser and a stirbar. DI H2O (100 mL),
potassium iodide (0.59 g, 3.5 mmol), sodium hydroxide (1.83 g, 45.8
mmol) and bromoacetic acid (3.45 g, 24.8 mmol) were added. After
refluxing for 1.5 h, additional bromoacetic acid (3.85 g, 27.7 mmol)
was added to ensure complete alkylation of the amine. As the so-
lution was refluxed, NaOH was added to maintain the pH above 7.
This can be monitored by maintaining a yellow color because the
solution turns red as it becomes acidic. Once the pH remained con-
stant for 30 min, the basic solution was decanted off, and several
milliliters of concentrated HCl were added to the hot solution. A
dark-red product was extracted with dichloromethane. During the
extraction, a white precipitate formed. This was recovered by filtra-
tion, analyzed by NMR and found to be pure H21 (394.7 mg, 9%
yield). The combined organic layers were dried with sodium sulfate
and evaporated to yield crude H21 as a flaky red solid (700 mg),
which dissolved in methanol and water to yield a colorless solution.
4-[Bis(ethoxycarbonylmethyl-amino)-phenyl] diphenyl methanol
(5). Compound 4 (0.3869 g, 1.4 mmol) was placed in a 25 mL flask
equipped with a reflux condenser, stirbar and nitrogen inlet. Sodium
iodide (115.5 mg), 1,8-bis(dimethyl-amino) naphthalene (0.67 g, 3.1
mmol) and 5 mL dry acetonitrile were added. Ethyl bromoacetate
(0.35 mL; d
reaction was heated to reflux for 3 days. The reaction was diluted
with toluene and washed with phosphate buffer (pH 2) until the
ϭ
1.506 g mLϪ1, 1.6 mmol) was also added, and the
ϭ
pH remained 2. The organic layer was dried over Na2SO4, filtered
and concentrated in vacuo to yield 0.6339 g of a tan oil. The crude
product was purified on silica gel with a mobile phase of 20% ethyl
acetate in hexanes to give pure 5 (0.4207 g, 67%) as a light-tan oil.
1H-NMR (300 MHz, CDCl3): ␦ ϭ 7.26 (s, 10H), 7.07 (d, 2H), 6.52
(d, 2H), 4.22 (q, 4H), 4.1 (s, 4H), 2.7 (s, 1H), 1.27 (t, 6H). 1H-NMR
(300 MHz, C2D6O): ␦ ϭ 7.26 (m, 10H), 7.04 (d, 2H), 6.53 (d, 2H),
4.2 (s, 4H), 4.14 (q, 4H), 2.8 (s, 1H), 1.2 (t, 6H). 13C-NMR (70
MHz, CDCl3): ␦ ϭ 171, 147.4, 147.2, 136.9, 129.3, 128.1, 128.0,
127.2, 112.0, 81.9, 61.4, 53.6, 14.4. MS–EI m/z (Mϩ) 447. Found
447.2053, Calculated 447.2046 for C27H29N1O5.

Photochemistry and Photobiology, 2002, 75(3) 213
1
LϪ sϪ1. Because the volume of the cell was 3 mL, the total moles
9
of MGϩ produced per unit time is 2.1
ϫ
10Ϫ mol MGϩ sϪ1. The
total moles of MGϩ produced can be used to calculate the total
number of photons by dividing the previous value by the quantum
9
1
yield (␾ ϭ 0.91) (62) to give 2.3
ϫ .
10Ϫ einstein sϪ
RESULTS AND DISCUSSION
Synthesis
Compound H21 was synthesized by the methods outlined in
Scheme 3. The amine 4 was prepared by organometallic ad-
dition to benzophenone, followed by reaction with methanol
as described by Walton (63) and Hellwinkel and Fristch
(64). Alkylation of the amine with bromoacetic acid via the
method of Cox and Smith (65) gave H21 in 9% yield in three
steps. Alternatively, H21 could be prepared via alkylation of
4 to form the ethyl ester 5, followed by hydrolysis with
lithium hydroxide. By this path, H21 is synthesized in 42%
yield in four steps.
Scheme 3.
Electrochemistry. The reduction of Znϩ at a hanging mercury
drop electrode was used to determine the formation constant and
stoichiometry of 1 with Zn(II). A BAS RE-5 Ag–AgCl reference
electrode and Pt grid auxiliary electrode were used in all experi-
ments. The solution analyzed was 0.1 M NaNO3 in 50% aqueous
2
methanol, 20
␮M Zn(NO3)2 and 5 mM N-2-hydroxythylpiperazine-
N
Ј
-2-ethane-sulfonic acid (HEPES) (SigmaUltra, St. Louis, MO)
buffer (pH 7.0). Before voltammograms were collected, the elec-
trolyte solution was degassed with N2 for 20 min. Between scans,
the solutions were degassed with N2 for 5 min. The square wave
Photochromism
A typical photochromic response for 1 in methanol is shown
in Fig. 1. Initially, the UV–Vis spectrum of 1 has a large
absorbance band from 230 to 280 nm (A). During irradiation
with UV light, the C–O bond breaks heterolytically to form
a colored species 2. The UV spectrum of the colored pho-
toproduct alone is often difficult to obtain because both tri-
phenylmethanol and its photoproduct exist at the same time
in solution. We can, however, conclude that the UV–Vis
spectrum for 2 is characterized by a visible absorption band
at 464 nm and a decrease in molar absorptivity in the 260
nm band (B). When the UV light is removed, the hydroxide
ion and 2 recombine slowly. We observe that 3.5 min after
the light is turned off, the band at 464 nm has disappeared,
whereas that at 260 nm has regained 94% of its initial ab-
sorbance (C). The small difference in absorbance between
initial and final conditions is most likely a result of photo-
degradation. A lack of photostability is a common disadvan-
tage for the triarylmethanols (66).
parameters were as follows: step height ( 2 mV, pulse am-
⌬
E)
ϭ
plitude (Esw) 30 mV and square wave period (␶) ϭ 100 ms. For
ϭ
the determination of the rate of dissociation, data were acquired
with the square wave period set to 30 ms, with varying concentra-
tions of 1.
Job Plot experiment. Stock solutions of 133
␮M tetramethylpyr-
idine (TMP), 133 M H21 and 133 M Zn(ClO4)2·6H2O were pre-
␮
␮
pared in 0.1 M NaClO4–methanol. A series of solutions containing
H21 and Zn(II) were prepared with varying quantities of each stock
solution but a total volume of 500 ␮L. An additional set of solutions
was prepared using the same amount of 1 but no Zn(II). Two equiv-
alents of TMP (based on the triphenylmethanol concentration) were
added to every solution, and the solutions were diluted to 5 mL with
0.1 M NaClO4 in methanol. The difference in absorbance between
the Zn(II)-containing and the Zn(II)-free solutions was plotted vs the
mole fraction of the ligand.
Preparation of MG cyanide. MG cyanide or MG leuconitrile was
prepared following the method of Holmes (61). Almost white crys-
tals (very light green) were collected as pure MG cyanide. Melting
1
point: 180–182
Њ
C. H-NMR (300 MHz, CDCl3): ␦ ϭ 7.3 (m, 5H),
7.05 (d, 4H), 6.7 (br.s 4H), 2.95 (s, 12H). MS–EI m/z (Mϩ) 355.
Found 355.2050, Calculated 355.2048 for C24H25N3.
For biological and sensor applications, measurements will
likely be performed in aqueous solutions; so it is necessary
to design chelators that will photoreversibly bind metal ions
in water. Although 1 is soluble and photochromic in aqueous
solutions, it degrades to 53% of the initial absorbance after
only five cycles, probably because of the hydrolytic insta-
bility of the aryliminium functional group (2). This is a dis-
advantage accompanying our design that places a significant
positive charge on the nitrogen in the product. Such photo-
degradation is not as evident in methanol, in which only a
9% reduction in photochromism is seen after nine cycles (i.e.
about 1% per cycle). In methanol–water (50:50), we see
about a 5% decrease in absorbance over 30 s in the absence
of the Zn(II) and only 0.1% decrease over 30 s of photoylsis
in the presence of excess Zn(II). Thus, methanol and meth-
anol–water (50:50) were selected as suitable alternatives to
water. Because of their solubility in methanol, the perchlo-
rate salts of metal ions were used in all metal-binding stud-
ies.
Chemical actinometry. From the point of view of actinometry, it
is best to have the source wavelength range smaller than the ab-
sorption wavelength range of the dye. Then, an accurate measure-
ment of all of the light output by the source can be made. On the
other hand, from the point of view of speed, a broad source that is
capable of providing a large flux of absorbable photons is required.
In an effort to speed up the photochemical reactions, we have chosen
to use a source wavelength range that is wider than the absorption
band of H21 or Zn1. Therefore, we have used MG, the spectrum of
which is almost identical to that of H21 and Zn1, for actinometry.
MG stock solutions (1 mM in 1 mM HCl in ethanol) were kept
frozen and in the dark when not in use and were used within 1 day
of preparation. A 2 mm
ϫ 5 mm mask was placed in front of a 1
cm cuvette in an HP 8452 diode array spectrophotometer. A refer-
ence solution of 1 mM HCl in ethanol was placed in the cuvette,
and spectra were acquired for 90 s with UG-5 filtered light illumi-
nation and stirring. The MG stock solution was diluted with 1 mM
HCl in ethanol until the absorbance at the desired wavelength(s) was
1.5. Absorption spectra of the illuminated stirred solutions were ac-
quired for 90 s. The initial linear region of absorbance (at 622 nm)
increase was used to calculate Io. A plot of absorbance (A) versus
time (t) produced a regression line of A
ϭ 0.0731t ϩ 0.1516. Thus,
1
the rate of production of MGϩ is 0.0731 sϪ in absorbance units.
The intercept corresponds to the presence of a small amount of pho-
Metal-binding ability
toproduct at the start of the experiment. Dividing by the molar ab-
1
105 cmϪ MϪ1) (62) and the cell length (1
The formation of a 1–Zn(II) complex was confirmed in both
methanol and water by a hypsochromic shift from 260 to
sorptivity (G622
ϭ
1.06
ϫ
cm), converts the rate of formation of MGϩ to 6.9
ϫ
10Ϫ7 mol MGϩ

214 Melinda R. Stephens et al.
Figure 2. Job Plots of 1–Zn(II). The difference in absorbance be-
tween solutions (in methanol 0.1 M NaClO4) containing varying
ratios of 1 and Zn(II) and similar solutions without Zn(II). (
Experimental and (A) theoretical results for 1 concentrations ranging
from 1.33 to 13.3 M. ( ) Experimental and (B) theoretical results
based on 1 concentrations from 0.53 to 5.32
ࡗ)
Figure 1. Photochromism of 1 in methanol. 100
mM NaClO4 in methanol. (A) in the dark, (B) after 30 s of UV-light
and (C) 3.5 min after the light is removed.
␮M H21 and 10
␮
Ⅲ
2
␮
M. K1K2 ϭ ;
1013 MϪ
1
K1
ϭ ϭ .
107 MϪ1; K2 106 MϪ
244 nm. This blue shift is a common result of metal-ion
binding and is caused by stabilization of the nitrogen lone
lower in the more polar solvent. However, binding in both
media is avid.
pair, which increases the energy that is required for the n
→
␲* transition. Addition of a competitive chelator such as
There are two pathways for the formation of 2 from pho-
tolysis of the 1–Zn(II) complex (Scheme 4): Path ‘‘A’’ is
indirect and is possible if the rate of dissociation of the ML
complex is rapid. Path ‘‘B’’ leads to direct dissociation as
the complex itself is photolyzed. The second pathway is the
preferred one and requires that the complex be photosensi-
tive.
triethylenetetraamine (TETA) should disrupt the metal-ion
complex and cause a bathochromic shift of the 244 nm band.
Indeed, addition of a 10-fold excess of TETA to a 1:1 so-
lution of 1 and Zn(II) resulted in a shift of the 244 nm band
to longer wavelengths (260 nm). Zn(II) is used to minimize
spectroscopic competition; 1 also binds to Cu(II), Pb(II),
Hg(II), Fe(II), Co(II) and Cd(II). Other metal ions have not
yet been tested.
The photosensitivity of the complex can be determined by
examining its photochemical response at large excess con-
centrations of metal ion, which diminishes the concentration
of free L and discourages path A. The photochemical yield
of 2 (represented by the absorbance at 464 nm), resulting
from 10 s of photolysis of solutions containing 1, decreases
as the concentration of metal ion increases (Fig. 4). Data
were collected for solutions containing 0–0.997 mole frac-
tion zinc, which corresponds to M/L ratios from 0 to 332.
Even when the ligand exists as greater than 99% complex,
there is a significant amount of 2 resulting from photolysis.
Because we know the stoichiometry and have an estimate
The stoichiometry of the 1–Zn(II) complex was deter-
mined with a Job Plot to be 2:1 in methanol (Fig. 2). To
eliminate complications that may result from the presence of
the carboxylic acid protons, two equivalents of the base
TMP were added to the Job Plot solutions. For this ligand,
the binding constant is too large to obtain an accurate value
from the Job Plot data. Curve fitting using Eqs. 1–5 indicates
Ϫ
2
that K1K2 is on the order of 1012–1014 M , and K1 is equal
to or slightly greater than K2, where K2
K1 ML/M·L.
ϭ
ML2/ML·L and
ϭ
Ϫ
2
of the binding constant (1013 M ) for this system in meth-
anol, we can estimate the contribution from paths A and B
in Scheme 4. If only free 1 is photolyzed (path A), the pro-
duction of 2 will be proportional to the amount of free 1,
assuming that the rate of dissociation is slow (curve A in
Fig. 4). Thus, the absorbance will be determined by Eq. 6,
where k is an adjustable constant and [L] is the concentration
of free 1. If the 1–Zn(II) complex is also photolyzed, even
when zinc is in excess, there will be significant production
of 2 (curve B). In this case, the absorbance will be described
T
MT
ϭ
ϭ
ϭ
ϭ
ϭ
MT
ϩ
LT
(1)
(2)
(3)
(4)
(5)
M
L
ϩ
ML
ML
ϩ ML2
LT
ϩ
ϩ
2
ϫ ML2
ML
ML2
K1
K2
ϫ
ϫ
M
ϫ
L
ML
ϫ L
A more quantitative analysis of the formation constant of
the 1–Zn(II) complex was accomplished with square wave
voltammetry. With an increase in concentration of 1 from 0
to 30 equivalents, the reduction potential of zinc shifts from
by Eq. 7, where k, kЈ and kЉ are adjustable constants, [ML]
is the concentration of the 1:1 complex and [ML2] is the
concentration of the 2:1 complex. Initially, it appears that
the complex is photolyzed with about one-tenth of the effi-
ciency of the free ligand.
Ϫ
963 to
mation (Fig. 3). A plot of the change in potential (
log[1] confirmed the stoichiometry of the 1–Zn(II) complex
Ϫ
1021 mV, which is consistent with complex for-
⌬E) versus
to be 2:1. With these data, a binding constant, K, of 5.0
ϫ
A
A
ϭ
ϭ
k[L]
k[L]
(6)
(7)
Ϫ
2
108 M (log K
ϭ
8.70) in 50% aqueous methanol was cal-
ϩ kЈ[ML] ϩ kЉ[ML2]
culated. As expected, the formation constant is significantly

Photochemistry and Photobiology, 2002, 75(3) 215
Scheme 4.
kh
␯
2
*
ML2
→
kh2
→
knr2
ML2
(17)
(18)
(19)
ϩ
Ϫ
*
ML2
ML ϩ L ϩ OH
*
ML2
→
ML2
Figure 3. Electrochemical evidence of zinc binding to 1. The re-
duction of 20 M Zn(II) in 50% aqueous methanol with addition of
(A) 0, (B) 1, (C) 5, (D) 10 or (E) 30 equivalents of 1. Details are
Calculation of kh by chemical actinometry
␯
␮
Equations 8, 13 and 17 describe the formation of an excited
in Materials and Methods.
state species with rates kh , kh and kh 2. These rates depend
␯
␯
1
␯
on the total number of photons incident on the sample (Io),
which was determined by actinometry (see Materials and
Methods).
Kinetic analysis
Only a fraction of the total Io is absorbed by 1 because
the average absorbance (A) for the photochromic band (224
to 284 nm) was less than 2 when 1 was photolyzed. Iabs can
In the preceding analysis, we assumed that the rate of dis-
sociation of the metal complex is slow. We have performed
a detailed kinetic analysis of this system to examine the va-
lidity of this assumption and determine the photosensitivity
of 2. Ultimately, we will simulate the photolysis of 1 with
a 100-fold excess of Zn(II) in 50% aqueous methanol. This
solvent is chosen because we need a rate that is determined
by electrochemistry, and 100% methanolic solutions are not
compatible with the hanging mercury drop working elec-
trode used in the determination.
The set of Eqs. 8–19 describe our system. All the rate
constants can be calculated directly from experimental data,
except those for the sets of Eqs. 14 and 15 and Eqs. 18 and
19. These pairs of equations describe the unknown quantum
efficiency for formation of the colored photoproduct from
ML and ML2. The individual rate constants are so fast that
only the relative values of the constants matter in the current
analysis. In the following paragraphs, we will describe the
determination of the rate constants and the kinetic analysis.
¯
¯
be calculated from Eq. 20, where b
ϭ 1 cm and A is the
¯
average absorbance from 224 to 284 nm. The A for both 1
and the 1–Zn(II) complex is approximately 0.35. This results
Ϫ
7
Ϫ
1
Ϫ1
in a value of 4.2
ϫ 10 mol photons L s .
¯
Iabs
ϭ
[Io(1
Ϫ
exp(2.303A))]/b
(20)
Finally, we can calculate a rate of formation of excited 1,
Ϫ
7
Ϫ1
kh , from Eq. 21 where Iabs
and [1] is 40
same as that for 1, we have kh
ϭ
4.2
ϫ 10 mol photons L
␯
1
Ϫ
s
␮M. Because the Iabs for the complex is the
Ϫ
1
ϭ
kh
ϭ
kh ϭ
␯2
0.011 s
␯
␯
1
kh
ϭ Iabs/[1]
(21)
␯
Calculation of kr
We first need to determine the molar absorptivity of the
product 2. This is a transient species; so a direct measure-
kh
␯
L
L
→
L*
(8)
(9)
kh
→
ϩ
Ϫ
L
ϩ
OH
knr
L*
→
L
(10)
(11)
kr
→
ϩ
Ϫ
L
ϩ
OH
L
k1
M
ϩ
L
s
ML
(12)
kϪ
1
kh
␯
1
ML
→
ML*
(13)
(14)
(15)
kh1
→
ϩ
Ϫ
ML*
ML*
M
ϩ
L
ϩ
OH
Figure 4. Photosensitivity of the 1–Zn(II) complex. (
mental data for photolysis of 44 M 1 in 0.1 M NaClO4 in methanol
with increasing Zn(II) (0–17 mM). Solid lines are calculated based
ࡗ) Experi-
knr1
→
␮
ML
2
upon K
k[L]
and kЉ ϭ 100 MϪ
ϭ
ϩ
1013 MϪ for (A) Path A: A
ϭ
ϭ
k[L] and (B) Path B: A
k2
1
ϭ
kЈ
[ML] [ML2], where k
ϩ
kЉ
900 MϪ1, kЈ ϭ 100 MϪ
ML
ϩ
L
s
ML2
(16)
1
kϪ
2
.

216 Melinda R. Stephens et al.
1
ment of the molar absorptivity
␧
of 2 is precluded. This
258
value can be obtained in two ways. First, 2 can be generated
independently of photolysis by titration with acid. In this
case, complete ionization of 1 occurs; so the concentration
of 2 can be calculated based on the initial number of moles
Ϫ
Ϫ
of 1. A molar absorptivity of 22 000 cm 1 M 1 is determined
with this method. This number agrees well with the molar
absorptivity of an analog, the salt of 4-N,N-dimethylamino-
phenyldiphenyl methanol, which is solvent sensitive but in
Ϫ
1
Ϫ1
the range of about 1.5–3.5
ϫ
104 cm
M
(67).
Alternatively, can be calculated from mass balance
␧
480
during photolysis. At short photolysis times, the decrease in
concentration of 1 will correspond to the increase in con-
centration of 2. After 10 s of UV photolysis, we see a de-
Figure 5. Steady state absorbance and recombination. 40
two equivalents of TMP and 0.1 M NaClO4 in 50% aqueous meth-
anol after 60 s photolysis while stirring (b 1 cm).
␮M 1 with
1
crease of 0.0245 in absorbance at 258 nm (
␧
ϭ
15 300
ϭ
258
Ϫ
1
Ϫ1
cm M ) and an increase of 0.0354 in absorbance at 480
Ϫ
1
Ϫ1
nm. Thus, a molar absorptivity of 22 100 cm
culated. The two values agree well.
M
is cal-
Using Eq. 23, we can assign values for the last two rate
constants. Equations 9 and 10 describe the fraction of ex-
cited molecules that undergo heterolysis versus nonradiative
decay. With a quantum yield of 0.91, we can determine that
the ratio of kh to knr must be 10. Because we know that these
processes are rapid, we will assign arbitrary, but large, rates
Equation 11 describes the recombination of 2 and hy-
droxide to regenerate 1. The rate of change of absorbance,
which is proportional to the rate of recombination (kr), is
calculated from measurements of absorbance vs time follow-
ing the cessation of photolysis. Because this is a bimolecular
process, we anticipate a second-order reaction. A plot of 1/
Ϫ
1
Ϫ1
of 106 s to kh and 105 s to knr.
A versus t gives a linear response (r2
ϭ 0.999) with a slope
kh
Ϫ
1
Ϫ1
␾ ϭ
(23)
equal to 6.4823 s (absorbance units) . The recombination
kh
ϩ knr
Ϫ
1
Ϫ1
rate is thus 143 000 M s .
Simulation results
Calculation of K1 and K2
Simulations were completed using the chemical kinetics sim-
ulator (IBM Inc., White Plains, NY). This simulator does
not integrate sets of coupled differential equations, but uses
stochastic algorithms, to predict product concentrations (69).
Figure 6 shows the results of the simulations of Eqs. 8–19.
In the top figure, the experimental data for photolysis of 1
in 50% aqueous methanol are plotted against the simulated
As discussed earlier, the overall binding constant (K1K2) in
50% aqueous methanol is 5.0
Ϫ
2
ϫ
108 M . Because K1 is
equal to or slightly greater than K2 in methanol, the same
relationship is likely to exist in 50% methanol. Fitting to
Ϫ
1
experimental data suggests that K1 is 5
is 1
ϫ
104 M and K2
Ϫ
1
ϫ
104 M (Eqs. 12 and 16).
The rate of dissociation of the complex was determined
with square wave voltammetry according to the method of
Correia dos Santos et al. (68). A fit of experimental data to
published model curves (68) results in a value of 2.24 for
Ϫ
5
results, where L
ϭ 4 ϫ 10 M and M ϭ 0. In this case,
only Eqs. 8–11 are actually necessary. Thus, all the rate con-
stants are known. Although there is some noise in the data
because of the stirring of the solution while collecting the
spectra, the match is quite good.
Ϫ
1
kϪ
M
Ϯ
0.14
ϫ
103 s . This requires that k1
ϭ 1.36 ϫ
106
1
Ϫ
1
Ϫ1
s . With such a large excess of metal ion (100:1), ML2
The pairs of equations, Eqs. 14 and 15 and Eqs. 18 and
19, describe quantum yields for ML and ML2. With a 100-
fold excess of Zn(II), the concentration of ML2 will be neg-
ligible; so we need to concern ourselves with the quantum
yield from ML only. Simulation of experimental results for
photolysis of 1 with 100-fold excess Zn(II) in 50% aqueous
methanol, with adjustment of the rate constants in Eqs. 14
and 15, indicates that the quantum yield for the complex is
0.15 (Fig. 6B). The constants are estimated as follows: kh1
and its related equations and rate constants are of no signif-
icant importance; however, they were included in the sim-
ulations. The ‘‘on’’ rate for ML2 is assigned a nearly dif-
Ϫ
1
Ϫ1
fusion controlled limit of 109 M s , with kϪ then equal
2
Ϫ
1
to 105 s .
Calculation of kh and knr
The quantum yield can be calculated from Eq. 22 where kr
Ϫ
1
Ϫ1
ϭ
1.8
There are two side reactions that we are justified in ig-
ϫ ϭ ϭ 0.18 ϫ knr1).
105 s and knr1 106 s (i.e. kh1
is the rate of recombination, kh is the rate of formation of
␯
ϩ
ϩ
excited L, [L ]ss is the concentration of L at steady state
noring. Photodegradation in the presence of Zn(II) is negli-
gible on the timescale of these experiments; so it is not in-
cluded. The recombination of 2 with methoxide is possible.
However, we have not seen differences in the UV spectra
resulting from photolysis and relaxation in the dark. Some
rough calculations show that the effect of the potential for-
mation of the methoxy compound is small. The excited state
conditions and [LT] is the total concentration of ligand (1).
Ϫ
1
Ϫ
1
Ϫ1
In this case, kr
ϭ
143 000 M s , kh
ϭ
0.011 s , LT
ϭ
40
M. The value of L is obtained from
experimental data for 40 M 1 in 50% aqueous methanol
␯
ϩ
ϩ
␮M and L ϭ 1.65 ␮
␮
after 60 s photolysis (Fig. 5). Thus, we calculate a quantum
yield of 0.91.
kr[L ]s2s
ϩ
Ϫ
1
production rate is 0.011 s , and the quantum efficiency in
␾ ϭ
(22)
ϩ
ϩ
the presence of Zn2 [Zn(II)] is 0.15. Thus, the overall first-
kh([LT]
Ϫ [L ]ss)

Photochemistry and Photobiology, 2002, 75(3) 217
Figure 7. Simulation of photoionization of the Zn(II)-complex of
1. (
ࡗ) Experimental data for photolysis of 40 ␮M 1 with two equiv-
alents of TMP and 0.1 M NaClO4. The solid lines are simulation
5
results with M
stants: kh
0.011 sϪ1, kh
106 sϪ1, knr2
ϭ ϭ 4 ϫ
10Ϫ3, L 10Ϫ and the following rate con-
ϭ
ϭ
ϭ
106 sϪ1, knr
ϭ
105 sϪ1, kr
ϭ
ϫ
143 000
␯
MϪ sϪ1, knr1
ϭ
106 sϪ and (A) k1
ϭ
1.34
106 MϪ
1
1
1
1
sϪ1, kϪ
ϭ
2236.2 sϪ1, kh1
ϭ
180 000 sϪ1, k2
ϭ
109 MϪ sϪ1, kϪ
ϭ
ϭ
ϭ
ϭ
1
2
1
1
105 sϪ1, kh2
ϭ
170 000 sϪ (B) k1
ϭ
1.34
ϫ
106 MϪ sϪ1, kϪ
1
2236.2 sϪ1, kh1
ϭ
10Ϫ sϪ1, k2
ϭ
109 MϪ sϪ1, kϪ
ϭ
105 sϪ1, kh2
7
1
2
10Ϫ sϪ (C) k1
ϭ
109 MϪ sϪ1, kϪ
ϭ
105 sϪ1, kh1
ϭ
10Ϫ sϪ1, k2
7
1
1
6
1
1
6
1
109 MϪ sϪ1, kϪ
ϭ ϭ .
105 sϪ1, kh2 10Ϫ sϪ
2
lax nonradiatively. Thus, the quantum yield for nonradiative
decay is 1 and knr1 and knr2 need only be set to large values.
Therefore, there are no adjustable parameters for this system.
By maintaining the calculated rate constants and eliminating
ϩ
photolysis of the complexes, we see that barely any L is
produced (curve B).
Our previous conclusion that the metal complex is pho-
tosensitive relied on the assumption that the rate of disso-
ciation is slow. Even though the measured rate of dissocia-
Ϫ
1
Ϯ 0.14 ϫ
103 s , we have simulated the results
Figure 6. Simulation of photoionization of 1. The solid lines are
the simulation results, and the diamonds are the experimental data
of (A) photolysis in 50% aqueous methanol of 40 ␮M 1 with two
tion is 2.24
with rates of dissociation that are large (kϪ
Ϫ
1
ϭ
kϪ2 ϭ 105 s )
1
and rates of association that are nearly diffusion controlled
equivalents of TMP and 0.1 M NaClO4 (B) with 100-fold excess
Ϫ
1
Ϫ1
109 M s ). The results are shown in curve C.
Zn(ClO4)2. Simulation rate constants are: kh
ϭ
ϭ
0.011 sϪ1, kh
106 sϪ1, knr2
ϭ
ϭ
106
(k1
ϭ
k2
ϭ
␯
1
sϪ1, knr
ϭ
ϭ
105 s1, kr
ϭ
143 000 MϪ sϪ1, knr1
106
ϩ
Even in this case, insufficient L is produced.
1
1
sϪ1, k1
1.34
ϫ
106 MϪ sϪ1, kϪ
ϭ
2236.2 sϪ1, kh1
ϭ
180 000 sϪ
,
1
From these simulations, we must conclude that photolysis
of the complex is the only method to achieve sufficient pro-
duction of 2. Thus, the complex is photosensitive.
1
1
k2
ϭ
109 MϪ sϪ1, kϪ
ϭ
105 sϪ1, kh2
ϭ
170 000 sϪ
.
2
ϩ
order rate constant for production of the product L is
There is one lingering question that we have not answered
satisfactorily with these experiments. That is the question of
whether the Zn(II) is truly liberated from the complex or
whether it remains associated, e.g. as an ion pair with a
carboxylate. We have tried to minimize this possibility by
using a high concentration of salt, NaClO4, in the experi-
ments. The presence of the visible band in this class of mol-
ecules is associated with a significant positive charge on ni-
trogens para to the methanol carbon. Therefore, in com-
pound 2, there is no driving force for metal-ion coordination
by the nitrogen. In fact, an electrostatic repulsion between
the nitrogen and zinc would be expected. In recent studies
in a similar system, Allgeier et al. (70) placed ligands near
Ϫ
1
0.0017 s . From this, we can estimate that about 10% of
the original complex has been photolyzed in 60 s. The pre-
ponderant form of L at the end of the 60 s measurement,
ϩ
even if all L reacted to form the methoxy compound, which
is unlikely, is the hydroxy form.
We have hypothesized that the production of L results
ϩ
from photolysis of the ML complex (path B). If this is so,
then using only path A (by eliminating photolysis of ML
and ML2), it will not be possible to simulate the experimen-
tal results with reasonable parameters. We can turn off com-
plex photolysis in the simulation by making log kh1 and log
kh2
K 0. With these changes, we can test the productivity of
path A by adjusting the values for the rates of dissociation
and the binding constants, K1 and K2. The results of these
simulations are shown in Fig. 7. The same experimental data
and simulated results for L from Fig. 6B are shown for
reference (curve A). When we remove kh1 and kh2 from the
the iron atom of a ferrocene. The ϩ1 change in charge
caused by oxidation of the iron center led to a 106 to 107
decrease in binding of Rh(I) and nearly a 1010-fold decrease
in binding of Pd(II) to the adjacent ligands. Kimura et al.
(16) showed a similar decrease in a cation’s affinity for a
positively charged ligand in comparing a crowned leuconi-
ϩ
simulation, we are specifying that all excited complexes re-

218 Melinda R. Stephens et al.
and photoenzymatically cleavable DNA. J. Am. Chem. Soc. 114,
9266–9269.
11. Willner, I., M. Lion-Dagan, S. Rubin, J. Wonner, F. Effenberger
and P. Bauerle (1994) Photoregulation of alpha-chymotrypsin
activity in organic media: effects of bioimprinting. Photochem.
Photobiol. 59, 491–496.
12. Inouye, M., K. Kim and T. Kitao (1992) Selective coloration of
spiro Pyridopyrans for Guanosine derivatives. J. Am. Chem. Soc.
114, 778–780.
13. Blank, M., L. M. Soo, N. H. Wassermann and B. F. Erlanger
(1981) Photoregulated ion binding. Science 214, 70–72.
14. Kimura, K., R. Mizutani, M. Yokoyama, R. Arakawa and Y.
Sakurai (2000) Metal-ion complexation and photochromism of
triphenylmethane dye derivatives incorporating monoaza-15-
crown-5 moieties. J. Am. Chem. Soc. 122, 5448–5454.
15. Kimura, K. (1996) Photocontrol of ionic conduction by photo-
chromic crown ethers. Coord. Chem. Rev. 148, 41–61.
16. Kimura, K., R. Mizutani, M. Yokoyama, R. Arakawa, G. Mat-
subayashi, M. Okamoto and H. Doe (1997) All-or-none type
photochemical switching of cation binding with malachite green
carrying a bis(monoazacrawn ether) moiety. J. Am. Chem. Soc.
119, 2062–2063.
17. Kimura, K., M. Kaneshige and M. Yokoyama (1995) Cation
complexation, photochromism, and photoresponzive ion con-
ducting behavior of crowned malachite green leuconitrile.
Chem. Mater. 7, 945–950.
18. Kimura, K., M. Kaneshige and M. Yokoyama (1994) Anoma-
lous photochromism and photochemical cation binding control
of crowned malachite green leuconitrile based on intramolecular
electrostatic repulsion. J. Chem. Soc., Chem. Commun. 1103–
1104.
trile with an azacrowned leuconitrile. As in 1 and 2, the
nitrogen in Kimura’s azacrown, by virtue of using its lone
pair to relax the positive charge formed at the central carbon
atom after photolysis of the leuconitrile, did not bind metal
ions, whereas the crowned analog (no nitrogen) did. Their
compounds were stable in the colored form, allowing, for
example, NMR experiments to reveal the metal-free photo-
product. Thus, photolysis of 1–Zn(II) should result in a con-
siderable decrease of the affinity of the iminodiacetic acid
moiety for the metal ion. This is hard to confirm, however.
Product 2 is not stable, reverting to 1 rapidly. It’s steady
state concentration is a mere few percent of the starting con-
centration of 1. The visible absorption spectrum should not
be particularly sensitive to changes in Zn(II) concentration.
That is because the carboxylate functional groups are not
part of the chromophore, and the nitrogen, which is part of
the chromophore, is no longer basic. We do see a 12 nm
shift (468 to 456 nm) in the visible band’s wavelength of
maximum absorbance (
centration from 0 to 14.7 mM ([1]
hand, protonation of 2 gives a
␭max) from a change in Zn(II) con-
ϭ
44.0
␮M). On the other
␭
max
of 484 nm. Thus, the
spectroscopic shift is small and bathochromic for protonation
of the carboxylates and small and hypsochromic with in-
creasing Zn(II) ion concentration. Although the evidence is
not conclusive, it does not support a strong interaction of
Zn(II) with the carboxylates in 2.
Acknowledgement The authors would like to thank the Office of
Naval Research for support of this work.
19. Inouye, M., K. Akamatsu and H. Nakazumi (1997) New crown
spirobenzopyrans as light- and ion-responsive dual-mode signal
transducers. J. Am. Chem. Soc. 119, 9160–9165.
20. Atabekyan, L. S., A. I. Lilikin, G. V. Zakharova and A. K.
Chibisov (1996) Transient species in the spironapthoxazine pho-
tocoloring reaction. High Energy Chem. 30, 409–412.
21. Zhou, J.-W., Y.-T. Li and X.-Q. Song (1995) Investigation of
the chelation of a photochromic spiropyran with Cu(II). J. Pho-
tochem. Photobiol. A: Chem. 87, 37–42.
REFERENCES
1. Lion-Dagan, M., S. Marxtibbon, E. Katz and I. Willner (1995)
Photoswitchable electrical communication of glucose oxidase
and glutathione reductase with electrode surfaces through pho-
toisomerizable redox mediators. Angew. Chem. Int. Ed. Engl.
34, 1604–1606.
2. Willner, I., V. Helegshabtai, R. Blonder, E. Katz and G. L. Tao
(1996) Electrical wiring of glucose oxidase by reconstitution of
FAD modifies monolayers assembled onto AU electrodes. J.
Am. Chem. Soc. 118, 10321–10322.
3. Willner, I., R. Blonder, E. Katz, A. Stocker and A. F. Buckmann
(1996) Reconstitution of apo glucose oxidase with a nitrospi-
ropyran modified FAD cofactor yields a photoswitchable bio-
catalyst for amperometric transduction of recorded optical sig-
nals. J. Am. Chem. Soc. 118, 5310–5311.
4. Willner, I., M. Lion-Dagan and E. Katz (1996) Photostimulation
of dinitrospiropyran modified glucose oxidase in the presence
of DNP anitibody: a biphase switch for the amperometric trans-
duction of recorded optical signals. J. Chem. Soc., Chem. Com-
mun. 623–624.
22. Stauffer, M. T., J. A. Grosko, K. Z. Ismail and S. G. Weber
(1995) Optical switching of the redox activity of a hydroxy-
chromene. J. Chem. Soc., Chem. Commun. 1695–1696.
23. Stauffer, M., D. Knowles, C. Brennan, L. Funderburk, F.-T. Lin
ϩ
and S. G. Weber (1997) Optical control over Pb2 binding to a
crown ether-containing chromene. J. Chem. Soc., Chem. Com-
mun. 287–288.
24. Stauffer, M. T. and S. G. Weber (1999) Optical control of di-
valent metal ion binding to a photochromic catechol: photo-
reversal of tightly bound Zn2ϩ. Anal. Chem. 71, 1146–1151.
25. Shinkai, S., T. Minami, Y. Kusano and O. Manabe (1983) Photo-
responsive crown ethers. 8. Azobenzenophane-type ‘‘switched-on’’
crown ethers which exhibit an all-or-nothing change in ion-binding
ability. J. Am. Chem. Soc. 105, 1851–1856.
26. Shinkai, S., T. Ogawa, T. Nakaji, Y. Kusano and O. Nanbe
(1979) Photocontrolled extraction ability of azobenzene-bridged
azacrown ether. Tetrahedron Lett. 20, 4569–4572.
27. Shinkai, S., T. Nakaji, T. Ogawa, K. Shigematsu and O. Manabe
(1981) Photoresponsive crown ether. 2. Photocontrol of ion ex-
traction and ion transport by a bis(crown ether) with a butterfly-
like motion. J. Am. Chem. Soc. 103, 111–115.
28. Shinkai, S., T. Nakaji, Y. Nishida, T. Ogawa and O. Manabe
(1980) Photoresponsive crown ethers. 1. Cis–trans isomerism of
azobenzene as a tool to enforce conformational changes of
crown ethers and polymers. J. Am. Chem. Soc. 102, 5860–5865.
29. Shinkai, S., K. Shigematsu, Y. Honda and O. Manabe (1984)
Photoresponsive crown ethers. 13. Synthesis of photoresponsive
NS2O crown ethers and application of the copper(I) complexes
to oxygen binding. Bull. Chem. Soc. Jpn. 57, 2879–2884.
30. Alfimov, M. V., A. V. Balakin, S. P. Gromov, Y. V. Zaushitsyn,
5. Rosengaus, J. and I. Willner (1995) Control of the thermal cis
isomerizations to trans isomerizations of azobenzene and supra-
molecular H-bonded assemblies. J. Phys. Org. Chem. 8, 54–62.
6. Zahavy, E. and I. Willner (1995) Application of eosin-modified
reconstituted Co(II)–myoglobin as semisynthetic photoenzyme
for cyclic photosynthesis. J. Am. Chem. Soc. 117, 10581–10582.
7. Willner, I. and S. Rubin (1996) Control of the structure and
functions of biomaterials by light. Angew. Chem. Int. Ed. Engl.
35, 367–385.
8. Valeur, B. and E. Bardez (1995) Application of eosin-modified
reconstituted Co(II)–myoglobin as semisynthetic photoenzyme
for cyclic photosynthesis. Chem. Br. 31, 216–220.
9. Pfeifer-Fukumura, U. (1997) Control of the thermal reaction of
a photochromic spirobenzopyran by the enzyme-like activity of
albumins. J. Photochem. Photobiol. A: Chem. 111, 145–146.
10. Nadjii, S., C. Wang and J. S. Taylor (1992) Photochemically

Photochemistry and Photobiology, 2002, 75(3) 219
O. A. Fedorova, N. I. Koroteev, A. V. Pakulev, A. Y. Resny-
anskii and A. P. Shkurinov (1999) Femtosecond spectrochron-
ography of the reverse photochromic transition in derivatives of
spiro compounds. Russ. J. Phys. Chem. 73, 1685–1694.
31. Alfimov, M. V., S. P. Gromov, Y. V. Fedorov, O. A. Fedorova,
A. I. Vedernikov, A. V. Churakov, L. G. Kuz’mina, J. A. K.
Howard, S. Bossmann, A. Braun, M. Woerner, D. F. Sears and
J. Saltiel (1999) Synthesis, structure, and ion selective com-
plexation of trans and cis isomers of photochromic dithia-18-
crown-6 ethers. J. Am. Chem. Soc. 121, 4992–5000.
32. Irie, M. and M. Kato (1985) Photoresponsive molecular twee-
zers. Photoregulated ion capture and release using thioindigo
derivatives having ethylenedioxy side groups. J. Am. Chem. Soc.
107, 1024–1028.
33. Rahman, S. M. F.-U. and K. Fukunishi (1994) Metal-ion bind-
ing by photoresponsive thioindigo crown ethers. J. Chem. Soc.,
Chem. Commun. 917–918.
34. Lednev, I. K., R. E. Hester and J. N. Moore (1997) A cation-
specific, light-controlled transient chromoionophore based on a
benzothiazolium styryl azacrown ether dye. J. Am. Chem. Soc.
119, 3456–3461.
35. Anzai, J. I., T. Osa (1994) Photosensitive artificial membranes
based on azobenzene and spirobenzopyran derivatives. Tetra-
hedron 50, 4039–4070.
36. Wu, L. P., Y. Suenaga, T. Kuroda-Sowa, M. Maekawa, K. Fu-
ruichi and M. Munakata (1996) Syntheses, structures and prop-
erties of palladium(II) complexes with photochromic 4-meth-
oxyazobenzene. Inorg. Chim. Acta 248, 147–152.
37. Zhou, J., F. Zhao, Y. Li, F. Zhang and X. Song (1995) Detailed
investigation on a negative photochromic spiropyran. J. Photo-
chem. Photobiol. A: Chem. 92, 193–199.
50. Kimura, K., T. Yamashita and M. Yokoyama (1992) Syntheses,
cation complexation, isomerization, and photochemical cation-
binding control of spirobenzopyrans carrying a monoazacrown
moiety at the 8-position. J. Chem. Soc., Perkin Trans. 2 613–
619.
51. Srivastava, H. P. and D. Tiwari (1996) Triphenylmethane group
of dyes as a biamperometric reagent for the determination of
metal ions at semi trace level. Indian J. Chem., Sect. A 35, 253–
254.
52. Lednev, I. K., S. P. Gromov, E. N. Ushakov, M. V. Al’fimov,
J. N. Moore and R. E. Hester (1992) Photochromic crown ether
complexes: a Raman spectroscopic study. Spectrochim. Acta A
48A, 799–809.
53. Shinkai, S., Y. Honda, K. Ueda and O. Manabe (1984) Photo-
responsive crown ethers. 13. Synthesis of photoresponsive
NS2O crown ethers and application of the copper(I) complexes
to oxygen binding. Bull. Chem. Soc. Jpn. 57, 2144–2149.
54. Shinkai, S., K. Miyazaki and O. Manabe (1987) Photorespon-
sive crown ethers. Part 18. Photochemically switched-on crown
ethers containing an intra-annular azo substituent and their ap-
plication to membrane transport. J. Chem. Soc., Perkin Trans.
1 449–456.
55. Pimienta, V., G. Levy, D. Lavabre, A. Samat, R. Guglielmetti
and J. C. Micheau (1994) Computer analysis of the thermo-
reversible photochromism of spiropyran compounds: evaluation
of absorption spectrum and quantum yield. Mol. Cryst. Liq.
Cryst. Sci. Technol., Sect. A 246, 283–286.
56. Preigh, M. J., F.-T. Lin, K. Z. Ismail and S. G. Weber (1995)
Bivalent metal ion-dependent photochromism and photofluoro-
chromism from a spiroquinoxazine. J. Chem. Soc., Chem. Com-
mun. 2091–2092.
38. Kimura, K., M. Kaneshige, T. Yamashita and M. Yokoyama
(1994) Cation complexation, photochromism, and reversible
ion-conducting control of crowned spironaphthoxazine. J. Org.
Chem. 59, 1251–1256.
39. Elmaloulibibout, M., P. Larginie, L. Noussi, A. Samat and R.
Guglielmetti (1994) Spironaphthoxazine linked crown ethers:
synthesis, photochromic and ionophoric properties. Mol. Cryst.
Liq. Cryst. Sci. Technol., Sect. A 246, 177–181.
57. Miyasaka, H., S. Araki, A. Tabata, T. Nobuto, N. Mataga and
M. Irie (1994) Picosecond laser photolysis studies on photo-
chromic reactions of 1,2 bis (2,4,5 trimethyl 3 thienyl)maleic
anhydride in solutions. Chem. Phys. Lett. 230, 249–254.
58. Irie, M., A. Menju and K. Hayashi (1979) Photoresponsive poly-
mers. Reversible solution viscosity change of poly(methyl meth-
acrylate) having spirobenzopyran side groups. Macromolecules
12, 1176–1180.
40. Tamaki, T. and K. Ichimura (1989) Photochromic chelating spi-
ronaphthoxazines. J. Chem. Soc., Chem. Commun. 1477–1478.
41. Sasaki, H., A. Ueno, J. Anzai and T. Osa (1986) Benzo-15-
crown-5 linked spirobenzopyran. I. Photocontrol of cation bind-
ing ability and photoinduced membrane potential changes.
Chem. Soc. Jpn. 59, 1953–1956.
42. Oda, H. (1993) Photostabilisation of photochromic materials:
contribution of amphoteric counter-ions on the photostability of
spiropyrans and related compounds. Dyes Pigm. 23, 1–12.
43. Winkler, J. D., K. Deshayes and B. Shao (1989) Photodynamic
transport of metal ions. J. Am. Chem. Soc. 111, 769–770.
44. LeBaccon, M. and R. Guglielmetti (1988) Coordination of aza-
heterocyclic spiropyrans by first row transition metals and zinc
halides and thiocyanates. Physico-chemical characteristics. N. J.
Chem. 12, 825–833.
45. Duerr, H. and C. Kranz (1994) A new type of photochromic
spirodihydroindolizines and their ability of cation binding. Mol.
Cryst. Liq. Cryst. Sci. Technol., Sect. A 246, 135–138.
46. Inouye, M. and Y. Noguchi (1994) Selective coloration of crypt-
and type spirobenzopyran for alkaline earth metal cations. Mol.
Cryst. Liq. Cryst. Sci. Technol., Sect. A 246, 187–189.
47. Cabera, I., A. Dittrich and H. Ringsdorf (1991) Thermally ir-
reversible photochromic liquid crystal polymers. Angew. Chem.
Int. Ed. Engl. 30, 76–78.
59. Kimura, K., H. Tokuhisa, M. Kaneshige, T. Yamashita and M.
Yokoyama (1994) Design of photochromic crown compounds
and their applications to photoswitchable ion conducting mate-
rials. Mol. Cryst. Liq. Cryst. Sci. Technol., Sect. A 246, 173–
176.
60. Tsivgoulis, G. M. and J. M. Lehn (1995) Photonic molecular
devices: reversibly photoswitchable fluorophores for nonde-
structive readout for optical memory. Angew. Chem. Int. Ed.
Engl. 34, 1119–1122.
61. Holmes Jr., E. (1966) The effect of the properties of solvents
of various dielectric constants and structures on the photoioni-
zation of the leucocarbinols and leucocyanides of malachite
green, crystal violet, and sunset orange and related phenomena.
J. Phys. Chem. 70, 1037–1046.
62. Fisher, G. J., J. C. LeBlanc and H. E. Johns (1967) A calori-
metric determination of the quantum yield for the ionization of
malachite green cyanide by ultraviolet radiation. Photochem.
Photobiol. 6, 757–767.
63. Walton, D. R. M. (1966) The protection of aminophenyl groups
in organometallic syntheses. J. Chem. Soc., Chem. Commun.
1706–1707.
64. Hellwinkel, D. and H. Fritsch (1990) Phenylazo-substituierte tri-
phenylmethylium-systeme Chem. Berichte. 123, 2207–2226.
65. Cox, J. R. and B. D. Smith (1964) The preparation of N,N,N
ЈNЈ-
48. Walther, D. and E. G. Jaeger (1975) Effect of chelate complex
formation on the photochromism of 1,3,3-trimethylspi-
tetramethyl-p-phenylenediamine. J. Org. Chem. 29, 488–489.
66. Duerr, H. (1990) Photochromism: molecules and systems. In
Photochromism (Edited by H. Duerr and H. Bouas-Laurent), pp.
727–737. Elsevier, Amsterdam.
ro(indoline-2,2Ј-benzopyran)-8Ј-carboxylic acid. Z. Chem. 15,
236–237.
49. Kobayashi, N., S. Sato, K. Takazawa, K. Ikeda and R. Hirohashi
(1995) A new polymer electrolyte for reversible photorespon-
sive ionic conduction. Electrochim. Acta 40, 2309–2311.
67. Walba, H. and G. E. R. Branch (1951) The absorption spectra
of some N-substituted p-aminotriphenylmethyl ions. J. Am.
Chem. Soc. 73, 3341–3348.

220 Melinda R. Stephens et al.
68. Correia dos Santos, M. M., M. L. Simo˜es Goncalves and J. C.
Roma˜o (1996) Study of CE mechanisms by square wave vol-
70. Allgeier, A. M., C. S. Slone, C. A. Mirkin, L. M. Liable-Sands,
G. P. A. Yap and A. L. Rheingold (1997) Electrochemically
controlling ligand binding affinity for transition metals via
RHLs: the importance of electrostatic effects. J. Am. Chem. Soc.
119, 550–559.
tammetry: Cd(II)
ϩ nitrilotriacetic acid and Cd(II) ϩ aspartic
acid systems. J. Electroanal. Chem. 413, 97–103.
69. IBM Corporation(1998) Chemical Kinetics Simulator Software
http://www.almaden.ibm.com/st/msim/cksfacts.html, p. 1.