Photoluminescent properties of cadmium selenide in contact with solutions and films of metalloporphyrins: Nitric oxide sensing and evidence for the aversion of an analyte to a buried semiconductor-film interface

Article abstract of DOI:10.1021/ja993855o

The band-edge photoluminescence (PL) intensity of etched n-CdSe single crystals is quenched reversibly by adsorption of the trivalent metalloporphyrins, MTPPCl (TPP = tetraphenylporphyrin; M = Mn, Fe, Co) in nitrogen-saturated methylene chloride solution. The PL responses are concentration dependent and can be fit to the Langmuir adsorption isotherm model to yield binding constants of ～10³-10⁴ M^-1. The MTPPCl compounds react irreversibly with NO in solution to form nitrosyl adducts, and these compounds reversibly enhance the CdSe PL intensity when adsorbed onto the semiconductor surface, also with binding constants of ～10³-10⁴ M^-1. Films of MTPPCl were prepared on CdSe substrates by solvent evaporation. These coatings serve as transducers for NO detection: while the bare CdSe surface shows no response to NO gas relative to N₂, the coated surfaces reversibly enhance the PL intensity (CoTPPCl) or quench it (MnTPPCl and FeTPPCl), with binding constants on the order of ～1 atm^-1. In contrast to the PL results, which are particularly sensitive to the semiconductor-film interface, electronic and IR spectral changes of the bulk film induced by NO binding were irreversible. The UV-vis and IR spectra could be spectroscopically mimicked by preformed nitrosyl adduct films that were prepared by solvent evaporation of MTPPCl (M = Co, Fe) and MTPP (M = Co) solutions that had been exposed to NO. These films, however, lack transduction capability, as the PL intensity is the same in NO and N₂ ambients. For the films prepared from FeTPPCl and CoTPPCl, the saturation of IR and UV-vis spectral changes occurs at NO pressures at least 10-fold lower than observed for PL changes. These results indicate that NO has a strong aversion to binding at the semiconductor-film interface as opposed to the bulk film environment. Steric and electronic contributions to these observed effects are discussed.

Full text of DOI:10.1021/ja993855o

J. Am. Chem. Soc. 2000, 122, 3731-3738
3731
Photoluminescent Properties of Cadmium Selenide in Contact with
Solutions and Films of Metalloporphyrins: Nitric Oxide Sensing and
Evidence for the Aversion of an Analyte to a Buried
Semiconductor-Film Interface
Albena Ivanisevic, Mark F. Reynolds, Judith N. Burstyn,* and Arthur B. Ellis*
Contribution from the Department of Chemistry, UniVersity of WisconsinsMadison,
1101 UniVersity AVenue, Madison, Wisconsin 53706
ReceiVed October 29, 1999
Abstract: The band-edge photoluminescence (PL) intensity of etched n-CdSe single crystals is quenched
reversibly by adsorption of the trivalent metalloporphyrins, MTPPCl (TPP ) tetraphenylporphyrin; M ) Mn,
Fe, Co) in nitrogen-saturated methylene chloride solution. The PL responses are concentration dependent and
can be fit to the Langmuir adsorption isotherm model to yield binding constants of ∼103-104 M-1. The
MTPPCl compounds react irreversibly with NO in solution to form nitrosyl adducts, and these compounds
reversibly enhance the CdSe PL intensity when adsorbed onto the semiconductor surface, also with binding
constants of ∼103-104 M-1. Films of MTPPCl were prepared on CdSe substrates by solvent evaporation.
These coatings serve as transducers for NO detection: while the bare CdSe surface shows no response to NO
gas relative to N2, the coated surfaces reversibly enhance the PL intensity (CoTPPCl) or quench it (MnTPPCl
and FeTPPCl), with binding constants on the order of ∼1 atm-1. In contrast to the PL results, which are
particularly sensitive to the semiconductor-film interface, electronic and IR spectral changes of the bulk film
induced by NO binding were irreversible. The UV-vis and IR spectra could be spectroscopically mimicked
by preformed nitrosyl adduct films that were prepared by solvent evaporation of MTPPCl (M ) Co, Fe) and
MTPP (M ) Co) solutions that had been exposed to NO. These films, however, lack transduction capability,
as the PL intensity is the same in NO and N2 ambients. For the films prepared from FeTPPCl and CoTPPCl,
the saturation of IR and UV-vis spectral changes occurs at NO pressures at least 10-fold lower than observed
for PL changes. These results indicate that NO has a strong aversion to binding at the semiconductor-film
interface as opposed to the bulk film environment. Steric and electronic contributions to these observed effects
are discussed.
Introduction
Because NO is known to react with a variety of metallopor-
phyrins, often causing substantial changes in metal coordination
environment and electronic structure, these compounds poten-
tially serve as a rich source of sensor transduction strategies.9-11
We and others have used the photoluminescence (PL) of
semiconductor substrates such as CdS and CdSe to demonstrate
that divalent metalloporphyrins (MPs) bind strongly to the
surfaces of these substrates, and that adsorption can be
substantially influenced by the presence of ligands such as
oxygen that bind to the MPs.12-16 Many adsorbates have the
ability to act as Lewis acids or bases toward the solid, shifting
electron density into or out of surface electronic states and
thereby expanding or contracting the depletion width of the
semiconductor, respectively. If a region on the order of the
depletion width is regarded as nonemissive (a “dead layer”)
insofar as PL is concerned, the electric field therein separates
Recent research developments have highlighted the critical
role that nitric oxide, NO, plays in a variety of physiological
processes.1-3 For example, NO has been found to serve as a
biological messenger in neurotransmission and to assist the
immune system in destroying tumor cells and intracellular
parasites. The importance of NO has prompted efforts to develop
sensors that can detect this species under a variety of experi-
mental conditions.4-8 Methods used to date have largely been
based on changes in electronic absorption spectra and electro-
chemical properties. For example, Malinski and co-workers have
developed an amperometric sensor based on a porphyrin-
catalyzed oxidation reaction.6 Sailor et al. have reported
reversible photoluminescence (PL) changes on binding of NO
to porous silicon.4
* Authors to whom correspondence should be addressed. E-mail:
Ellis@chem.wisc.edu; Burstyn@chem.wisc.edu.
(1) Traylor, T. G.; Sharma, V. S. Biochemistry 1992, 31, 2847.
(2) Butler, A. R.; Williams, D. L. H. Chem. ReV. Soc. 1993, 22, 233.
(3) Feldman, P. L.; Griffith, O. W.; Stuehr, D. J. Chem. Eng. News 1993,
Dec 20, 26.
(4) Harper, J.; Sailor, M. J. Anal. Chem. 1996, 68, 3713.
(5) Zhou, X.; Arnold, M. A. Anal. Chem. 1996, 68, 1748.
(6) Malinski, T.; Taha, Z. Nature 1992, 358, 676.
(7) Tschudi, M. R.; Mesaros, S.; Luscher, T. F.; Malinski, T. Hyperten-
sion 1996, 27, 32.
(9) Scheidt, W. R. Acc. Chem. Res. 1977, 10, 339.
(10) Scheidt, W. R.; Ellison, M. K. Acc. Chem. Res. 1999, 32, 350.
(11) Hoshino, M.; Laverman, L.; Ford, P. C. Coord. Chem. ReV. 1999,
187, 75.
(12) Ivanisevic, A.; Ellis, A. B. J. Phys. Chem. B 1999, 103, 1914.
(13) Chrysochoos, J. J. Phys. Chem. 1992, 96, 2868.
(14) Bhamro; Chrysochoos, J. J. Lumin. 1994, 60-61, 359.
(15) Isarov, A. V.; Chrysochoos, J. Langmuir 1997, 13, 3142.
(16) Isarov, A. V.; Chrysochoos, J. Proc. Indian Acad. Sci., Chem. Sci.
1998, 110, 277.
(8) Bennett, J. E.; Malinski, T. Chem. Mater. 1991, 3, 490.
10.1021/ja993855o CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/29/2000

3732 J. Am. Chem. Soc., Vol. 122, No. 15, 2000
IVaniseVic et al.
photogenerated electron-hole pairs rather than permits their
recombination. The PL response defines a “luminescent litmus
test”, with Lewis acids quenching and Lewis bases enhancing
the semiconductor’s PL intensity.17-22
Chart 1
Porphyrins and metalloporphyrins are known to organize
themselves into crystalline, supramolecular arrays, and the
geometry of such arrays has been shown to be highly dependent
upon the solvent and local chemical environment.23-33 Our
recent studies demonstrated that the interaction of many divalent
metalloporphyrins with single-crystal CdSe surfaces is affected
by oxygen, both in solution and in the gas phase.12 In the latter
case, we showed that films of divalent metalloporphyrins
deposited onto single-crystal CdSe substrates can serve as
transducers toward oxygen, coupling ligation chemistry with
changes in the PL of the underlying semiconductor. Because
these changes are readily reversible, such structures have the
potential to serve as on-line detectors for this analyte.
to yield 1‚A‚σ. Our evidence to date indicates that analyte-
induced semiconductor PL changes are sensitive specifically
to the film monolayer(s) nearest the semiconductor surface and
that the equilibrium constant Kinterface derived from PL changes
is associated with binding at the buried interface. The ratio of
Kinterface/Kfilm thus yields the equilibrium constant for eq 2, which
An important issue associated with such film-coated semi-
conductor structures is the nature of the buried semiconductor-
film interface. Our data for several coated semiconductor
structures suggest that partitioning of an analyte A between the
bulk film and the buried interface can be detected.21,34,35 The
relevant equilibria are given in eqs 1-3, where, for the present
study, 1 is the MP and A (g) is NO. Equation 1 describes the
is the distribution coefficient that expresses the relative prefer-
ence of the analyte for the two environments.
We report in this study the successful implementation of a
NO detection strategy using metalloporphyrin-mediated effects
on the PL of CdSe substrates. These interactions provide a
means to characterize a variety of semiconductor-metallopor-
phyrin interfaces. Specifically, we show that the CdSe PL
intensity is reversibly quenched by the adsorption of MTPPCl
(M ) Mn, Fe, Co) compounds (Chart 1A) in nitrogen-saturated
CH2Cl2 solution. When the solvent is instead saturated with NO,
the nitrosyl adduct of each MP is irreversibly formed in solution
(Chart 1B). When these nitrosyl adducts are brought into contact
with the semiconductor surface, they cause reversible enhance-
ments in the CdSe PL intensity. Deposition of films of MTPPCl
onto CdSe results in coatings that respond reversibly to NO in
the solid state, permitting their use as transducers for on-line
detection of this analyte. Furthermore, comparisons of binding
affinities estimated from PL and from electronic and IR
spectroscopic measurements reveal that NO has a strong
aversion to the buried semiconductor-film interface, character-
ized by a partition coefficient on the order of 10-1 or less. Steric
and electronic effects that might account for this remarkable
binding preference are discussed.
A(g) + 1(f) S 1‚A(f)
1‚A(f) + 1‚σ S 1‚A‚σ + 1(f)
A(g) + 1‚σ S 1‚A‚σ
(1)
(2)
(3)
equilibrium associated with binding of the analyte in the bulk
film, 1(f), to produce 1‚A(f), for which the equilibrium constant
Kfilm can be estimated, for example, through electronic absorp-
tion or IR spectral changes occurring in the bulk film. In
contrast, eq 3 represents binding of the analyte to a molecule
of 1 that is localized at a semiconductor surface site 1‚σ, i.e.,
at the generally different environment of the buried interface,
(17) Murphy, C. J.; Lisensky, G. C.; Leung, L. K.; Kowach, G. R.; Ellis,
A. B. J. Am. Chem. Soc. 1990, 112, 8344.
(18) Lorenz, J. K.; Kuech, T. F.; Ellis, A. B. Langmuir 1998, 74, 680.
(19) Ellis, A. B.; Brainard, R. J.; Kepler, K. D.; Moore, D. E.; Winder,
E. J.; Kuech, T. F.; Lisensky, G. C. J. Chem. Educ. 1997, 74, 680.
(20) Brainard, R. J.; Paulson, C. A.; Saulys, D.; Gaines, D. F.; Kuech,
T. F.; Ellis, A. B. J. Phys. Chem. B 1997, 101, 11180.
(21) Brainard, R. J.; Ellis, A. B. J. Phys. Chem. B 1997, 101, 2533.
(22) Kepler, K. D.; Lisensky, G. C.; Patel, M.; Sigworth, L. A.; Ellis,
A. B. J. Phys. Chem. 1995, 99, 16011.
(23) Sharma, C. V. K.; Broker, G. A.; Huddleston, J. G.; Baldwin, J.
W.; Metzger, R. M.; Rogers, R. D. J. Am. Chem. Soc. 1999, 121, 1137.
(24) Bhyrappa, P.; Wilson, S. R.; Suslick, K. J. Am. Chem. Soc. 1997,
119, 8492.
(25) Burrell, A. K.; Officer, D. L.; Reid, D. C. W.; Wild, K. Y. Angew.
Chem., Int. Ed. 1998, 37, 114.
(26) Abrahams, B. F.; Hoskins, B. F.; Robson, R. J. Am. Chem. Soc.
1991, 113, 3606.
(27) Abrahams, B. F.; Hoskins, B. F.; Michall, D. M.; Robson, R. Nature
1994, 369, 727.
(28) Drain, C. M.; Nifiatis, F.; Vasenko, A.; Batteas, J. D. Angew. Chem.,
Int. Ed. 1998, 37, 2344.
(29) Linke, M.; Chambron, J.; Heitz, V.; Sauvage, J. J. Am. Chem. Soc.
1997, 119, 11329.
(30) Turner, P.; Gunter, M. J.; Skelton, B. W.; White, A. H. Aust. J.
Chem. 1998, 51, 853.
(31) Jentzen, W.; Unger, E.; Song, X. Z.; Jia, S. L.; Tyrk, I. T.; Stenner,
R. S.; Dreybrodt, W.; Scheidt, W. R.; Shenutt, J. A. J. Phys. Chem. A 1997,
101, 5789.
(32) Kumar, R. K.; Balasubramanian, S.; Goldberg, I. Inorg. Chem. 1998,
37, 541.
(33) Lin, V. S. Y.; DiMagno, S. G.; Therien, M. J. Science 1994, 264,
1105.
(34) Meeker, K.; Ellis, A. B. J. Phys. Chem. 1999, 103, 995.
(35) Moore, D. E.; Lisensky, G. C.; Ellis, A. B. J. Am. Chem. Soc. 1994,
116, 9687.
Experimental Section
Materials and Sample Preparation. Samples of MnIIITPPCl,
FeIIITPPCl, CoIIITPPCl, and CoIITPP were purchased from Porphyrin
Products, Logan, UT, and used without further purification. Nitric oxide
gas was obtained from Aldrich (98.9%) and purified through a KOH
column immediately before each experiment to remove higher nitrogen
oxides.36 Chlorine gas (Cl2) was purchased from Aldrich (99.5%) and
used as received. Methylene chloride (Aldrich, 99+%) was distilled
from calcium hydride under nitrogen. 15NO was generated in situ from
the reaction of ascorbic acid (Aldrich, 99%) with sodium nitrite (15N,
98+%, Cambridge Isotope Laboratories, Inc, Andover, MA) in an
anaerobic aqueous medium. Single-crystal, vapor-grown c-plates of
n-CdSe, having a resistivity of ∼2 Ω‚cm, were obtained from Cleveland
Crystals, Inc. Prior to a PL experiment the crystals were etched in Br2/
MeOH (1:15 v/v), allowing the shiny Cd-rich (0001) face to be revealed
and later illuminated in the PL experiment. Prior to film deposition
the CdSe sample was given a second etch in a solution of concentrated
HCl, rinsed with methanol, and then dried in flowing nitrogen.
The nitrosyl adduct of each metalloporphyrin was prepared by
bubbling an excess of NO through a methylene chloride solution of
each MTPPCl compound, while maintaining rigorously anaerobic
conditions. The excess NO was removed by subsequently bubbling dry
nitrogen through the solution after nitrosyl formation.
(36) Richter-Addo, G. B.; Hodge, S. J.; Yi, G.; Khan, M. A.; Ma, T.;
Van Caemelbecke, E.; Guo, N.; Kadish, K. M. Inorg. Chem. 1996, 35, 6530.

Photoluminescent Properties of Cadmium Selenide
J. Am. Chem. Soc., Vol. 122, No. 15, 2000 3733
Nitrosyl chloride (NOCl) was prepared in situ in a solution of CH2-
Cl2.37 Initially CH2Cl2 was saturated with NO, by bubbling the gas
through the solution for 3 min. The presence of NO in the solution
compounds in this study exhibited negligible absorbance at the 633
nm excitation and 720 nm CdSe emission wavelengths, and the solution
spectra matched those previously reported.36,38-40 Solid-state UV-vis
spectroscopy was used to monitor the binding of NO to the CoTPPCl
and FeTPPCl complexes when deposited as films.
EPR Measurements. The X-band (9.2 GHz) EPR spectrum of the
NO adduct of FeIIITPPCl in methylene chloride was recorded at 77 K.
The spectrum was recorded on a Bruker ESP 300E X-band spectrometer
equipped with an EIP model 625 A CW microwave counter. A triplet
signal was observed, characteristic of five-coordinate FeTPP(NO).41-43
Proton NMR Measurements. The 1H NMR spectra of the NO
adducts of CoIIITPPCl and MnIIITPPCl in N2-saturated CDCl3 were
recorded with a Bruker AC-300 spectrometer. The chemical shifts
obtained were in good agreement with literature values for the known
MTPPNO complexes.36,38-40
was confirmed by its characteristic IR band, νNO, at 1846 cm-1
.
Subsequently Cl2 was bubbled through the solution, and the solution
turned dark orange, as previously reported.37 The observed shift in νNO
to 1852 cm-1 confirmed the formation of NOCl in solution.
Apparatus. The sample of CdSe was mounted on a glass rod
between two Teflon rings within a glass cell.17 In the solution
experiment a gas flow apparatus was assembled that allowed dry
nitrogen to flow through the glass cell where the solid semiconductor
crystal was illuminated.17 A flow meter was used to adjust the flow
rate, which varied from 80 to 120 mL/min, and the total gas pressure
over the sample was 1 atm.21 In the film experiments the glass cell
was directly connected to the same setup, and two flow meters were
used to measure the independently variable concentrations of N2 and
NO, as described previously.21,34
Results and Discussion
Optical Measurements. A Spectra-Physics He-Ne laser (632.8 nm)
provided excitation of the semiconductor. Incident intensities ranged
from 5 to 20 mW/cm2. Emission spectra were monitored using an Oriel
Instaspec II silicon diode array spectrophotometer. A cutoff filter was
used in the spectrophotometer to permit collection of the band-edge
PL at λmax ) 720 nm (Eg ) 1.7 eV). The entire PL spectrum was
monitored, while the PL intensity was tracked at a particular wavelength,
typically the band maximum, as a function of time; the band maximum
did not shift under the low-resolution (0.5 nm) spectral conditions
employed. The signal collected was analyzed by Oriel Instaspec
software for Windows.
Film Preparation. Two methods were used to prepare the MP films
and their nitrosyl adducts. In the direct method, the MPs were dissolved
in methylene chloride solution (∼400 µM). Prior to the PL experiments
two drops of the solutions were deposited on n-CdSe crystals and dried
under a stream of N2, forming films with an estimated thickness of 0.1
µm if uniform deposition is assumed. PL experiments were also
conducted with preformed films of the nitrosyl adducts of the MPs.
The MP nitrosyls were preformed in solution by bubbling NO through
the solution for at least 3 min. The nitrosyl adduct was then deposited
in the same fashion as the parent MP on the single crystal. Both methods
were used to prepare films for electronic spectroscopy experiments. In
these cases the films were evaporated under strictly anaerobic conditions
on the sides of a quartz cuvette.
IR Studies. Surface IR measurements of films of the MPs and their
nitrosyl adducts were conducted using the synchrotron source at the
University of Wisconsin Synchrotron Radiation Center. Samples of the
CdSe crystals were polished with 0.3 µm alumina prior to the deposition
of the film in order to slightly roughen the surface. The rough surface
produced better IR data than the etched semiconductor surface; the latter
resulted in oriented deposition of the metalloporphyrin parallel to the
surface and led to a double pass interference effect. To obtain good
quality spectra each polished crystal (ca. 1.0 × 1.0 × 0.1 cm) was
partially coated with a solution of the appropriate MP (∼400 µM) in
methylene chloride, followed by slow evaporation of the solvent in a
closed container under a stream of nitrogen. The samples were analyzed
in transmission mode using a NicPlan microscope enclosed in a dry
nitrogen purged chamber. The microscope was interfaced to a Nicolet
Magna 55011 IR spectrometer, equipped with a liquid nitrogen cooled
1/4 mm MCT detector. The mapping stage installed on the microscope
was used to survey the sample surface and showed that the film
thickness was highly nonuniform. The part of the CdSe surface that
was not coated with the film was used to collect the reference spectrum,
and patches of film areas were used to collect the sample IR spectra.
The spectra were collected using 200 scans and 1 cm-1 resolution at a
magnification of 32×. The mirror velocity was set at 0.6329 cm/s in
order to avoid interference from the 60 Hz component of the
synchrotron beam and provide an acceptable signal-to-noise ratio in
the spectra.
Samples of CdSe exhibit red band-edge PL (Eg ) 1.7 eV;
λmax ∼ 720 nm) when excited with ultra-band-gap excitation.
Because the metalloporphyrins of Chart 1 employed in this study
are highly colored, excitation was restricted to 633 nm, where
there is negligible absorption by the MTPPCl (M ) Mn, Fe,
Co) compounds. The CdSe samples show negligible PL intensity
changes referenced to N2 when exposed to NO, either in the
gas phase or in CH2Cl2 solution. However, in the presence of
a film or CH2Cl2 solution of MTPPCl compounds, addition of
NO causes substantial changes in PL properties. Sections below
describe the reactions of NO with solutions and films of
MTPPCl compounds, using PL changes to estimate adsorption
binding constants to the CdSe surface, and electronic spectral
changes to estimate film binding affinity for NO. In the final
section, we examine aspects of the coordination chemistry of
the MTPPCl compounds that could contribute to the spectro-
scopic effects observed.
I. Nitrogen-Saturated Solution Studies. When each of the
MTPPCl compounds is added to a nitrogen-saturated CH2Cl2
solution, quenching of the CdSe PL intensity is observed. Figure
1A presents typical data for titration with MnTPPCl. Quenching
typically onsets at a concentration of 20-50 µM, and the effect
saturates in the range 350-500 µM. The effect is readily
reversible, with a solvent rinse serving to restore the original
PL intensity, as illustrated in Figure 1A. PL quenching is
consistent with the analyte acting as a Lewis acid toward the
surface, possibly through withdrawal of electron density from
the semiconductor into the porphyrin’s aromatic system.19 Table
1 reveals that all three MTPPCl compounds elicit similar PL
responses, although our data do not allow us to know whether
the same kind and number of sites are used for binding in each
case. In our previous study with divalent metalloporphyrins, we
found that the PL response depended on the identity of both
the metal and the porphyrin.12
The concentration dependence of the PL changes observed
permits a rough estimate of the adsorption binding constant
through use of the Langmuir adsorption isotherm model.44 The
model is quantitatively represented by eq 4,
θ ) KC/(1 + KC)
or
1/θ ) (1/KC) +1 (4)
(38) Hoshimo, M.; Kogure, M. J. Phys. Chem. B 1989, 93, 5478.
(39) Scheidt, W. R.; Frisse, M. E. J. Am. Chem. Soc. 1975, 97, 17.
(40) Scheidt, W. R.; Hatano, K.; Rupprecht, G. A.; Piciulo, P. L. Inorg.
Chem. 1979, 18, 292.
Electronic Spectroscopic Measurements. The solution UV-vis
spectrum of each sample was obtained on a Hewlett-Packard HP89530A
spectrophotometer at room temperature between 200 and 800 nm. All
(41) Stone, J. R.; Sands, R. H.; Dunham, R.; Marletta, M. A. Biochem.
Biophys. Res. Commun. 1995, 207, 572.
(42) Wayland, B. B.; Olson, L. W. J. Am. Chem. Soc. 1974, 96, 6037.
(43) Bohle, D. S.; Hung, G. H. J. Am. Chem. Soc. 1995, 117, 9584.
(44) Atkins, P. W. Physical Chemistry; W. H. Freeman and Co: New
York, 1998.
(37) Vorobyova, O. I.; Dunaeva, K. M.; Ipolitova, E. A.; Tamm, N. S.
Practical Inorganic Chemistry; Mir Publishers: Moscow, 1987.

3734 J. Am. Chem. Soc., Vol. 122, No. 15, 2000
IVaniseVic et al.
Figure 2. Plot of θ, based on fractional PL changes of the data in
Figure 1A (eq 5), as a function of concentration for an etched CdSe
sample exposed to a N2-saturated CH2Cl2 solution of MnTPPCl. The
inset presents the same data as a double-reciprocal plot, yielding an
equilibrium constant of (8.0 ( 0.4) × 103 M-1. The PL was excited
with 633 nm light and monitored at 720 nm.
Figure 1. Changes in the PL intensity of an etched n-CdSe sample
resulting from exposure to (A) MnTPPCl dissolved in N2-saturated CH2-
Cl2 and (B) MnTPPCl dissolved in NO-saturated CH2Cl2, forming the
nitrosyl adduct. The PL was excited in each case with 633 nm light
and monitored at 720 nm. The downward spikes result from draining
the sample cell when changing solutions. The bottom of the plots in A
and B corresponds to zero PL intensity.
MTPPCl compounds yield similar binding constants of 103-
104 M-1, suggesting that variation in the metal exerts only a
modest influence on binding affinity.
II. Nitric Oxide Saturated Solution Studies. A control
experiment for our studies is the effect on CdSe PL intensity
of saturating the CH2Cl2 solvent with NO. We estimate that
the concentration of NO in this solvent at atmospheric pressure
is in the range ∼1-10 mM.46 We find no effect on the
semiconductor’s PL intensity on switching between N2- and NO-
saturated solvent.
It is well-known that NO reacts irreversibly with some
MTPPCl compounds in solution.36,38-40 The first equivalent of
NO reduces the metalloporphyin to MIITPP and forms
NOCl.42,47,48 The excess NO bubbled through the solution results
in the formation of the nitrosyl adduct (Chart 1B). In principle,
NOCl, too, could affect the semiconductor’s PL intensity. We
prepared a solution of NOCl in situ in CH2Cl2 by reacting NO
with Cl2.37 Although a dark orange solution was produced,
characteristic of NOCl, we saw no effect on the CdSe PL
intensity in this medium relative to the PL intensity in N2- or
NO-saturated CH2Cl2.
We have confirmed that the reaction of MTPPCl with NO
yields the nitrosyl adducts in CH2Cl2 solution by a variety of
spectroscopic techniques. Exposure to NO causes significant
and irreversible color changes, and the corresponding spectral
changes for all three metal complexes match literature re-
ports.9,39,40 Agreement with the literature was also found by EPR
measurements of the Fe complex and by 1H NMR measurements
of the Mn and Co complexes.36,38-43
Table 1. PL Results in Solutiona
c
e
compound
PL ratiosb
K (M -1
)
PL ratiosd
K (M -1
)
MnIIITPPCl
FeIIITPPCl
CoIIITPPCl
0.75
0.29
0.55
8 × 103
33 × 103
12 × 103
1.31
1.40
1.95
14 × 103
30 × 103
8 × 103
a PL ratios and binding constants based on adsorption of the indicated
MIIITPPCl compounds onto CdSe in N2- and NO-saturated CH2Cl2
solutions. All data were obtained using the same CdSe sample. b The
ratio of CdSe PL intensity upon addition of the indicated compound to
N2-saturated CH2Cl2 solution (when PL changes were saturated with
adsorbate concentration) relative to the PL intensity in N2-saturated
CH2Cl2 solvent. The excitation wavelength was 633 nm, and PL
intensity was measured at 720 nm. The error in a single measurement
was less than 5%. c Equilibrium binding constants for adsorption of
the indicated compound in nitrogen-saturated CH2Cl2 solution onto
CdSe, obtained from fits to the Langmuir adsorption isotherm model,
eq 4. The error in the Langmuir fit was generally less than 10%. d PL
ratio as defined in footnote b, except that all data were obtained with
NO-saturated CH2Cl2 solutions, leading to the formation of nitrosyl
metalloporphyrin adducts in the sample cell. e Equilibrium binding
constants as described in footnote c, except that all data were obtained
with NO-saturated CH2Cl2, leading to the formation of nitrosyl
metalloporphyrin adducts in the sample cell.
where K is the equilibrium constant for adsorption and C is the
molar concentration. The fractional surface coverage, θ, is
estimated from the fractional PL changes: When the PL changes
have saturated, PLsat, θ is taken to be 1; the PL intensity in the
reference ambient, PLref, corresponds to θ ) 0; at intermediate
covarages, the PL intensity is designated as PLx and defined
by eq 5.45
When MTPPCl complexes were dissolved in CH2Cl2 solu-
tions and NO was used to saturate the mixture, completely
different PL behavior is observed relative to the N2-saturated
CH2Cl2 solution experiments discussed above. As depicted in
Figure 1B, an enhancement of PL intensity is seen when the
nitrosyl adduct of the Mn complex is present. Table 1 indicates
that similar results are found for all three metal complexes
investigated. The PL response saturated above a ratio of ∼20%
NO in N2.
θ ) |(PLx - PLref)|/|(PLsat - PLref)|
(5)
A representative plot of θ-1 vs C-1 is shown in Figure 2 for
MnTPPCl and demonstrates a reasonably good fit, leading to a
value for K of 8 × 103 M-1. Table 1 shows that all three
(46) Fogg, P. G. T.; Gerrard, W. Solubility of Gases in Liquids; John
Wiley & Sons: New York, 1991.
(45) Winder, E. J.; Kuech, T. F.; Ellis, A. B. J. Electrochem. Soc. 1998,
145, 2475.
(47) Scheidt, W. R.; Hoard, J. L. J. Am. Chem. Soc. 1973, 95, 8281.
(48) Wayland, B. B.; Olson, L. W. Inorg. Chim. Acta 1974, 11, L23.

Photoluminescent Properties of Cadmium Selenide
J. Am. Chem. Soc., Vol. 122, No. 15, 2000 3735
Figure 4. PL intensity changes of an etched n-CdSe crystal coated
with a film of MnTPPCl and then exposed to N2, NO, and N2/NO
mixtures, as described in the Experimental Section; the partial pressure
of NO is indicated in the figure. The total pressure in all cases was 1.0
atm. The PL was excited with 633 nm light and monitored at 720 nm.
The bottom of the plot corresponds to zero PL intensity.
Figure 3. Plot of θ, based on fractional PL changes, as a function of
concentration for an etched CdSe sample exposed to a NO-saturated
CH2Cl2 solution of FeTPPCl. The inset represents the same data as a
double-reciprocal plot, yielding an equilibrium constant of (3.0 ( 0.1)
× 104 M-1. The PL was excited with 633 nm light and monitored at
720 nm.
substrate by placing a few drops of a CH2Cl2 solution on the
surface and allowing the solvent to evaporate in an anaerobic
ambient. If uniform coverage is assumed, the deposited material
would correspond to an average film thickness of approximately
0.1 µm.
As an independent test of the species causing the PL change,
we investigated the divalent precursor, CoTPP. When this
species was dissolved in CH2Cl2 and NO was bubbled through
the solution, the same product resulted, CoTPP(NO), as was
formed using CoTPPCl as a starting material, based on UV-
vis spectra.36,38-40 Likewise, we recorded identical PL enhance-
ments due to exposure of the crystal to a solution of CoTPP(NO)
prepared from the two different Co porphyrins. These data
confirm that the Co porphyrin nitrosyl species is responsible
for the observed PL response.
The direction of the PL change for the nitrosyl adduct
adsorbates is consistent with donation of electron density from
the adsorbate to the semiconductor, assuming electric field
effects are driving the PL changes. Whether the differences
shown in Table 1 for the three metal complexes are due to
differences in intrinsic basicity or to number and types of
binding sites cannot be addressed by our data. Table 1 also
includes binding constants for the nitrosyl solution adduct of
each metalloporphyrin with the CdSe surface based on PL
response, and a typical Langmuir plot is shown in Figure 3.
These values are almost the same as those for the corresponding
MTPPCl complexes, within experimental error, suggesting that
formation of a nitrosyl adduct has little effect on binding affinity
to the semiconductor surface.
Films of all three metalloporphyrins were found to mediate
a PL response to NO. A representative PL trace is presented in
Figure 4, which shows NO-induced PL quenching obtained with
a MnIIITPPCl-coated substrate. The quenching of the PL
intensity in response to NO is indicative of the conversion of
the film to a more electron-withdrawing structure upon addition
of NO. Quenching of the PL intensity was observed for both
the Mn and Fe derivatives; however, the Co complex showed
the opposite effect, an enhancement, indicating that the intro-
duction of NO causes donation of electron density to the
semiconductor relative to the unreacted film. Most of the
changes observed were modest, typically less than ∼20%
between ∼0.1 and 1.0 atm, but no attempt was made to optimize
the deposition procedure to enhance the response. In contrast,
another NO sensor based on porous silicon was reported to yield
larger PL responses that onset at ppm concentration levels.4 Also
noteworthy from Figure 4 is the fact that the NO-induced PL
changes are readily reversible once the ambient is restored to
pure N2, although some baseline drift occurs over the course of
the experiment.
One additional solution experiment that was conducted
involved trying to reverse the effects of FeTPP(NO) adsorption
on CdSe PL intensity by returning to a N2-saturated solution.
Once NO has been introduced, subsequent introduction of N2
with displacement of NO from the CH2Cl2 solution causes no
change in the CdSe PL intensity. This observation is consistent
with the strong affinity of NO for the metal center; N2 does not
displace NO from the metalloporphyrin. If the metal complex
is completely removed from the system, however, by repeated
solvent rinses, the original PL intensity in the N2-saturated
solvent can be recovered. The solution formation of a nitrosyl
adduct with the other two porphyrins, CoTPPCl and MnTPPCl,
was “tracked” by PL in a similar fashion with analogous results.
III. Film Studies. PL Results. When gaseous NO is allowed
to come into contact with the surface of etched CdSe, the PL
intensity of the semiconductor remains constant within experi-
mental error of its intensity in a N2 ambient. To test the ability
of MTPPCl films to serve as transducers for NO detection, a
film of each MTPPCl compound was prepared on a CdSe
The type of data gathered in Figure 4 permits an estimate of
film binding constants for NO with the three metalloporphyrin
films using the Langmuir adsorption isotherm model (eq 4,
expressed in terms of pressure). A typical result is shown in
Figure 5. Table 2 summarizes the film experiments and reveals
that binding constants for all of these samples are on the order
of ∼1 atm-1. As noted in the Introduction, we believe that these
values are rough measures of the affinity of NO to bind to the
buried MTPPCl-CdSe interface, Kinterface
.
We explored the question of whether deposition of a nitrosyl
adduct directly from CH2Cl2 solution could be used to prepare
the transducer film instead of deposition of the trivalent MTPPCl
film. We investigated this possibility by preparing nitrosyl
adducts by bubbling NO through a CH2Cl2 solution of FeTPPCl
or CoTPPCl and allowing drops of this solution placed on a
CdSe surface to evaporate in a container under a stream of N2.
In contrast to the properties of the MTPPCl films, neither of
these films could be used as a transducer: switching between
NO and N2 ambients caused no PL change in the CdSe substrate.

3736 J. Am. Chem. Soc., Vol. 122, No. 15, 2000
IVaniseVic et al.
Figure 6. IR spectrum of a film of the nitrosyl adduct of CoIITPP,
preformed in solution by bubbling excess 14NO(g) through a methylene
chloride solution of the metalloporphyrin. Drops of this nitrosyl adduct
solution were placed onto the semiconductor surface to yield the film
of CoIITPP(NO)preformed. The spectrum was collected in transmission
mode using 200 scans and 1 cm-1 resolution. See the Experimental
Section for complete conditions for these IR studies. Peaks at 1003,
1260, 1350, 1450, and 1600 cm-1 were seen in all IR film spectra of
all CoIITPP and CoIIITPPCl complexes and have been assigned to νC-H
rock (pyrrole), νdN-H in-plane, νdC-N stretch, νC-H bend (pyrrole), and
ν-CdC- (phenyl), respectively. The stretching frequency assignments
are based on refs 35, 46, and 47.
Figure 5. Plot of θ, based on fractional PL changes (eq 5) of the data
in Figure 4, vs partial pressure of nitric oxide. The inset shows the
double-reciprocal plot of the same data, yielding an equilibrium constant
of 4.5 ( 0.4 atm-1
.
Table 2. PL Results for Films
b
compound
PL ratiosa
K (atm-1
)
MnIIITPPCl
FeIIITPPCl
CoIIITPPCl
0.82
0.77
1.36
4.5
1.0
0.8
a The ratio of PL intensity for CdSe coated with a film of the
indicated compound (see Experimental Section) upon exposure to NO
(when PL changes were saturated) relative to the PL intensity in N2.
The excitation wavelength was 633 nm, and the PL intensity was
measured at 720 nm. b Equilibrium binding constants determined from
fits to the Langmuir adsorption isotherm model, eq 4. The error in the
Langmuir fit was generally less than 10%.
the substrate. Figure 6 presents the IR spectrum of a film
prepared from CoIITPP in this manner; the same product was
formed from both the CoIITPP and CoIIITPPCl starting materials,
and the spectra obtained matched that of CoIITPP(NO) previ-
ously reported.36
Direct formation of the same CoIITPP(NO) adduct was
observed in the solid state, when films of either CoIITPP or
CoIIITPPCl were exposed to gaseous NO (eq 6). Films of CoII-
TPP and CoIIITPPCl were also deposited directly from solution
and subsequently exposed to either 14NO or 15NO. The peak
positions of the bulk films of CoIITPP and CoIIITPPCl were
the same as those previously reported for these complexes;36,49,50
thus, the metalloporphyrin in the bulk film appears unaffected
by the semiconductor substrate. When the films of CoIITPP and
CoIIITPPCl were exposed to 14NO, a new peak appeared at
approximately 1700 cm-1. The assignment of this peak to νNO
was verified by isotopic substitution; when 15NO was substituted
for 14NO, the film exhibited a peak at approximately 1670 cm-1
(Figure 7 and Table 3). The spectrum obtained when a film of
either CoIITPP or CoIIITPPCl was exposed to 14NO (eq 6)
matched that of a CoIITPP(NO)preformed film (eq 7), suggesting
that the same chemistry occurs in the bulk film as in solution.
Since films of CoIITPP(NO) do not mediate a PL response to
NO, and the formation of CoIITPP(NO) in the bulk film appears
irreversible, the bulk chemistry cannot account for the observed
reversible PL response of CoIIITPPCl films to NO.
For comparison with the Kinterface values, we attempted to
estimate the equilibrium constant for uptake of nitric oxide by
the bulk metalloporphyrin film, Kfilm, using IR spectral changes,
as described previously.21 Coatings of FeIIITPPCl or CoIIITPPCl
were placed onto a large CdSe crystal, and the spectral changes
in the νNO spectral region were monitored as a function of NO
partial pressure. The conversion was monitored at νNO ) 1700
cm-1, and several different films were used to monitor the NO
uptake by the MP-coated surface. Due to the decaying current
from the synchrotron source, we were unable to obtain accurate
The IR spectra of films of MTPP(NO), deposited from solutions
of NO-exposed MTPPCl (M ) Fe or Co), also remained
unchanged after repeated exposure to excess NO or N2 gas. The
Mn case was not examined due to the extreme air sensitivity of
the nitrosyl film.
IR and Electronic Spectroscopy Results. Our PL studies
of films prompted us to examine other techniques to further
characterize the binding of NO to a solid metalloporphyrin film
deposited on a semiconductor substrate. In the previous section
we described two distinct methods that resulted in the formation
of a nitrosyl adduct on the CdSe surface, summarized by eqs 6
and 7. Equation 6 describes the equilibrium associated with
MP(f) + NO(g) S MP(NO)(f)direct
(6)
MP(solv) + NO(g) w MP(NO)(solv) w MP(NO)(f)preformed
(7)
binding NO gas to a bulk film of a metalloporphyrin (MP)
coated onto the semiconductor surface, to produce MP(NO)-
(f)direct, which is a nitrosyl adduct formed directly in the solid
state. Equation 7 represents the reaction of NO gas with a MP
in solution, resulting in the formation of MP(NO)(solv), which
is the nitrosyl adduct in solution. When drops of the solution
are placed onto the CdSe surface under a stream of nitrogen, a
film of MP(NO)(f)preformed is deposited. The strikingly different
PL sensing capabilities of the nitrosyl films prepared by these
two methods led us to examine further their properties by solid-
state IR and electronic absorption spectroscopy.
A variety of MP films were prepared on CdSe substrates,
and their NO binding behavior was studied by IR spectroscopy.
NO was bubbled through a CH2Cl2 solution of either CoIITPP
or CoIIITPPCl and the preformed nitrosyl product deposited on
(49) Thomas, D. W.; Martell, A. E. J. Am. Chem. Soc. 1959, 81, 5111.
(50) Shimomura, E. T.; Phillippi, M. A.; Goff, H. M.; Scholz, W. F.;
Reed, C. A. J. Am. Chem. Soc. 1981, 103, 6778.

Photoluminescent Properties of Cadmium Selenide
J. Am. Chem. Soc., Vol. 122, No. 15, 2000 3737
Figure 8. Solid-state UV-vis spectra on a quartz slide illustrating
the spectral changes of a film of CoIIITPPCl exposed to increasing
pressures of 14NO(g). The film preparation procedure and gas handling
apparatus are described in the Experimental Section.
Figure 7. IR spectra comparing the νNO region of films of CoIITPP
that were exposed to 14NO(g) and 15NO(g) to form directly the nitrosyl
adduct in the solid state. The two spectra were collected under identical
conditions in transmission mode using 200 scans and 1 cm-1 resolution,
using two different samples.
Table 3. IR Results for Filmsa
film
CoIITPP(f)b {direct}
ν
NO (cm-1
)
CoIITPP(f) + 14NO(g)c {direct}
CoIITPP(f) + 15NO(g)d {direct}
CoIITPP(solv) + 14NO(g)e {preformed}
CoIIITPPCl(f)b {direct}
1700
1670
1700
Figure 9. Normalized UV-vis responses (circles) of a film of CoIII-
TPPCl on a quartz cuvette and of PL responses (squares) of a film of
CoIIITPPCl on CdSe to changes in NO partial pressure. Responses were
normalized so that the UV-vis absorption (monitored at 431 nm) and
PL intensity changes at the saturation level were both set to 100%;
intermediate values are expressed as percentages of the saturation value.
CoIIITPPCl(f) + 14NO(g)c {direct}
CoIIITPPCl(f) + 15NO(g)d {direct}
CoIIITPPCl(solv) + (14NO)e {preformed}
1698
1668
1700
a The spectra were collected in transmission mode using 200 scans
and 1 cm-1 resolution. b The film was evaporated onto the surface of
the single-crystal CdSe, as described in the Experimental Section, from
a methylene chloride solution of the indicated metalloporphyrin under
strictly anaerobic conditions. c The metalloporphyrin film was prepared
as described in footnote b and subsequently exposed to a flow of
14NO(g) for at least 3 min. The container was subsequently purged
with dry nitrogen to eliminate the excess 14NO(g). d The film was
prepared as described in footnote c, except that 15NO(g) generated in
situ was used (see the Experimental Section for 15NO(g) generation
procedure). e The nitrosyl adduct of either CoIIITPPCl or CoIITPP was
preformed in solution by bubbling excess 14NO(g) through a methylene
chloride solution of the metalloporphyrin; the film was then prepared
by solvent evaporation (see Experimental Section and the text).
environments may be involved. The onset of the spectral
changes was observed at very low pressures of NO, similar to
those used for the IR studies, indicating that the binding profiles
observed by solid-state IR and solid-state electronic spectroscopy
are similar. Due to the complexity of the spectral changes, we
are unable to estimate K values directly from these data.
However, a substantial difference between interface and film
binding affinities is highlighted in Figure 9, which directly
compares the concentration profiles of the UV-vis and PL
measurements for films of CoIIITPPCl. Especially noteworthy
is the saturation of response seen by PL at roughly a 10-fold
higher concentration than is observed by UV-vis or IR
spectroscopy. On the basis of this comparison we can estimate
that there is on the order of at least a 10-fold preference for
NO to bind in the bulk film rather than at the buried
semiconductor-film interface.
IV. Binding Characteristics. Metalloporphyrins have docu-
mented ability to coordinate several different ligands to their
metal centers.51 In a nitrogen-saturated solution, the M(III) center
in the MP acts like a Lewis acid, causing a quench in the PL
intensity relative to nitrogen-saturated methylene chloride
solution. When nitric oxide saturated solutions were used, a
nitrosyl adduct of each MP was generated, and a Lewis basic
PL response was observed. We believe that the difference in
the PL response in solution between the two ambients is most
estimates of Kfilm by this method. However, we can infer that
values for Kfilm are considerably higher than corresponding
Kinterface values due to the complete conversion to a nitrosyl
adduct observed by IR at much lower partial pressures (<0.02
atm) than observed during the PL measurements (typically ∼1
atm). For the FeTPPCl case our experimental setup did not allow
us to create sufficiently low NO concentrations to determine at
what pressures of NO the adduct begins to form and saturate.
Solid-state electronic absorption spectroscopy was also used
to track the binding of NO to the MP films. As described in
the Experimental Section, a solution of FeTPPCl or CoTPPCl
was evaporated onto a quartz cuvette and subsequently repeat-
edly exposed to various pressures of NO in a sealed container.
As represented in Figure 8, a substantial red shift of the most
intense absorption peak was observed for all complexes,
indicating the formation of a nitrosyl adduct; the lack of an
isosbestic point suggests that multiple species and/or binding
(51) Bertini, I.; Gray, H. B.; Lippard, S. J.; Valentine, J. S. Bioinorganic
Chemistry; University Science Books: Sausalito, CA, 1994.

3738 J. Am. Chem. Soc., Vol. 122, No. 15, 2000
IVaniseVic et al.
film interface is different from that preformed in solution or in
the bulk film. Our reversible PL changes in the solid state lead
us to the conclusion that nitric oxide participates in some kind
of adduct interaction with the metalloporphyrin that is uniquely
mediated by the presence of the semiconductor surface. The
IR film data in Table 3 indicate that there is no difference
between the nitrosyls prepared by the two different methods
(direct and preformed).
The porphyrins present at the semiconductor-film interface
may be sterically restricted, and the unique environment of the
interfacial region may alter the chemistry observed when NO
is present. Although the bulk porphyrin may undergo reduction
by NO, with concomitant release of NOCl, similar reaction of
the porphyrin at the interface may be inhibited. Under such
conditions, binding of NO to the CoIIITPPCl provides a plausible
mechanism for the reversible PL response observed, as il-
lustrated schematically in Figure 10B, with structure as shown
in Chart 1C. Alternatively, CoIITPP(NO) may indeed form, but
the nitrosyl adduct may be subject to destabilizing steric or
electronic effects at the film-semiconductor boundary. The data
presented herein reveal that the interaction between NO and
the porphyrin is mediated by the presence of the surface atoms
on the semiconductor crystal, weakening the binding of NO at
the interface. The lowered affinity of the porphyrin at the film-
semiconductor interface is presumably responsible for the
reversible PL response. Further experimentation using other
surface-sensitive techniques may provide additional details as
to the nature of the adduct formed at the interface.
Figure 10. Idealized representation of the possible binding modes of
MTPPCl in the solution and solid state with nitric oxide.
likely due to a change in the coordination environment around
the metal center due to the formation of a nitrosyl adduct, as
postulated in Figure 10A. This irreversible change once NO is
introduced causes the analyte to donate additional electron
density to the bulk of the semiconductor, thereby shrinking the
depletion region and enhancing the PL intensity.19
The PL effects observed in the solid state are quite different
from the solution effects. The introduction of nitric oxide into
the ambient around the coated semiconductor resulted in
reversible PL changes. This observation is suggestive of the
formation of a weak nitrosyl adduct, that is readily dissociated
once NO is removed from the system, as depicted in Figure
10B. The proposed chemistry in the bulk film is given by eq 8.
Acknowledgment. We thank Dr. Dovas Saulys for help in
the preparation of NOCl, Sergei Ivanov for assistance with NMR
measurements, Ryan Parks for help with 15NO experiments, and
Michael Kavana for experimental help. We are grateful to Dr.
Robert Julian, Tim May, and Roger Hansen for assistance with
the IR measurements. We acknowledge the National Science
Foundation (A.B.E.) and the National Institutes of Health, Grant
HL54762 (J.N.B.), for support of this work. The Synchrotron
Radiation Center at UWsMadison is a facility funded by the
National Science Foundation through Grant DMR-95-31009.
MTPPCl(f) + NO(g) S MTPPCl(NO)(f)
(8)
Assembled MP arrays have been shown under certain
conditions to accommodate a variety of hosts,28 and therefore
it is not surprising that solid MTPPCl on CdSe can bind NO
from the gas phase. However, all of the PL data collected
suggests that the nitrosyl adduct formed at the semiconductor-
JA993855O