v

- 2...: ,

. 3:01
~ .Tffv.

: ....
. (RN-.2 s.
. x. . ..
2..., . he; ........ 5%
2.53.. s
5...... I.

f
.: ~:rn
r 1!
t... I... .4?
::§1t,r::.r.
: 1i
2:

.13.. t.
(.1:
. 2:71
«a.

fluid:
.2 nu...
:-
. .. 4
Sid:
2.

I .r In... .
2.1....

.::.s :r x:
xe::.::‘.r
:‘SIJI. n V. ‘l f.

7

.35.... .

31...:

; x ... . .r .x‘ z .
9......» wmmuhfamm . .w. ”manna”.

1.2;; .

.t. .12 :1...

;

1.. .9355...»
Z a

K.

t

w

5

Im

.1.
.r.
xv

 

 

 

 

 

 

GAN STATE UNIVE BRARIE

M8,. IIIIIHHII\Hlllllll‘lIIIlII I I III

3 1293 01420 3180

 

 

 

 

 

 

This is to certify that the
dissertation entitled
CHARACTERIZATION OF METALLOPORPHYRIN SINGLET EXCITED STATE
VIBRATIONS THROUGH PICOSECOND RAMAN SPECTROSCOPY AND
NATURAL ORBITAL CALCULATIONS

presented by
Douglas H. Kreszowski
has been accepted towards fulfillment

of the requirements for

Ph. D. degree in Chemistry

48%

Major professor

 

 

Date 12 May I995

 

MS U is an Afflrmatiw Action/Equal Opportunity Institution 0-1277!

 

LIBRARY
Michigan State
Universlty

 

 

 

PLACE IN RETURN BOX to remove thh chockout from your "cord.
TO AVOID FINES Mum on or bdon dd. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU!:.A.'E A.- .L A . — .

 

CHARACTERIZATION OF METALLOPORPHYRIN SINGLET EXCITED STATE
VIBRATIONS THROUGH PICOSECOND RAMAN SPECTROSCOPY AND
NATURAL ORBITAL CALCULATIONS

By

Douglas H. Kreszowski

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1995

 

ABSTRACT
CHARACTERIZATION OF METALLOPORPHYRIN SINGLET EXCITED STATE
VIBRATIONS THROUGH PICOSECOND RAMAN SPECTROSCOPY AND
NATURAL ORBITAL CALCULATIONS
By

Douglas H. Kreszowski

Photosynthesis in bacteria proceeds through photoexcitation of the special pair
bacteriochlorophyll dimer (BChlz) that is lodged within the reaction center. This dimer
resides in the excited singlet electronic state (8;) during the initial electron transfer to
bacteriopheophytin A (BPhA), and its conformation at the time of electron transfer is
unknown. Thus, a picosecond laser system was developed that can utilize time-resolved
Raman techniques to characterize structures of singlet state intermediates, as well as to
investigate vibrationally excited species. Due both to the complexity of the reaction
centers, and the limited knowledge about singlet excited states, it became necessary to do
initial experiments on model porphyrins. The S. state of zinc(II) octaethylporphyrin
(ZnOEP) displayed Raman spectra that are dominated by totally symmetric modes, that
generally shifi down in frequency, and have increased linewidths within the core modes .
This dominance of totally symmetric modes suggest that, contrary to triplet excited states,
the 81 state lacks significant Jahn-Teller distortion. The experimental results are in
agreement with theoretical predictions that configuration interaction (CI) in singlet states
will act to diminish Jahn-Teller distortion. The downshift in frequencies of core vibrations

('04, 02, 03) suggest an overall expansion of the porphyrin ring in the 81 state. This issue

 

had not been addressed in either theory or experimentation, so further experiments were
done involving ZnOEP, ZnOEP-d4, ZnOEP-15N4, zinc(II) etioporphyrin (ZnEtiO), and
magnesium octaethylporphyrin (MgOEP). These additional experiments led us to
conclude that mode composition changes do not significantly contribute to the S. state of
these porphyrins. To assist in the interpretation, composite orbital diagrams were drawn
using natural orbital calculations that take into account configuration interaction betweeen
electronic transitions. The diagrams predict the experimentally observed ring expansion as
well as a small b1, distortion in lower singlet excited states. We conclude that this
predicted structural inhomogeneity is reflected by the observed increased linewidths, and
that both the increased linewidth and the ring expansion are consequences of extensive

configuration interaction between the alues and azues transitions.

 

Copyright by
999M311 WSZQWSKI

1995

   

V ”mm! “Manic '— d “If“ TAP; “1‘... . I ‘3 4. I l I‘d-IL. :., ’H‘)‘ [

   
    
   
  
 

‘2‘.
.1639» Mummrtnwz I run: lawns“ .- a .- ‘ v - :.. Sm;
. , I _
‘ "w. as l.’ -:" *I- ' "
“WWW v.3. II; we v .. Rut
Raw-r .. ~ .
' “Meant ww-
’ mun“! slfai‘V2:‘ ”Q! II 7‘ O «at... .‘83

$0..“ continuum.“ . many. an. m ”3-way mum
. Prduum (km... ‘l’ique' Immut- '3‘ . .w e- m .m.

myearly gnu-m yrm Ethan; Rm.- Rnarsa ms; K's.- h-r- um
7. V}: i‘c.

1%, A 2- .gyfllinfmseu'ch. lasdy,lthulaltmmx,ug
. .‘ I“: ‘pfiembemn vision WM.

    
  

 

rtla

‘.

va

an

ACKNOWLEDGMENTS

My diverse list of acknowledgements include: I thank PrOfessor Jerry Babcock for
his guidance and financial support (federal research grant NIH GM25480). I also
appreciate Professor George Leroi’s contributions as second reader of my guidance
committee. I thank Professor’s Jeff Ledford and Eugene Legoff for serving on my
guidance committee. Dr. Geurt Deinum deserves special thanks for his collaborative
contributions within portions of this thesis, for he was an asset to the project. Dr. Tom
Carter was vital in both technical aid with experimental apparatus and help with computer
related operations. I thank Scott Sanderson of the electronics shop for help in designing
and implementing the shutter driver. Likewise, I thank Russ Geyer, Dick Menke, and Sam
Jackson for their skilled support involving equipment fabrications for detector mounts,
spectrometer and optics mounts. and many other items. The glass shop members, Keki ,
Manfred. and Scott, deserve credit for their support in glassware related projects.
Professor Jim Harrison was of immeasurable help with the natural orbital calculations.
Professor Chris Chang's contributions of isotopes and chlorins to our porphyrin studies
are greatly appreciated. Professor Costas Varotsis’s technical discussions was also
valuable input during my early graduate years. I thank Rob Richards and Kitz for room
and boarding during my last year of research. Lastly, I thank all other labmates and

schoolmates for teaching me the balance between vision and practicality.

 

TABLE OF CONTENTS

 

LIST OF TABLES ix

LIST OF FIGURES x

 

CHAPTER 1 INTRODUCTION TO THVIE-RESOLVED RAMAN

 

 

 

 

 

 

 

 

 

SPECTROSCOPY
Abstract 1
I. ' ‘ J 4 2
11. Reaction Centers 4
III. Ultrafast Raman Spectroscopy A H" " ---~ 14
Instrumental T ‘ '1 15
Vibrational Relaxation P 16
Photochemical Systems 17
Photobiological S, ‘ 19
IV. Metalloporphyrin Structure and Dynamics ........................... 22
List of References 35

' CHAPTER 2 INSTRUMENTATION

 

 

 

 

Abstract ......... 43
1. Experimental Apparatus 44
II. Flowing Liquid Jet Assembly... 60
III. Quasi-Simultaneous .‘ ' " 70
List of References 78

 

CHAPTER 3 RAMAN SPECTRA 0F METALLOPORPHYRIN

 

 

 

 

 

EXCITED STATES
Abstract 8 l
I. ' ‘ J " 82
II. Materials and Methods 83
III. Results 87
IV. Discussion , 107
List of References 1 10

 

vii

 

CHAPTER 4 VIBRATIONAL ANALYSIS OF ercm)
OCTAE'I'HYLPORPHYRIN IN THE LOWEST EXCITED

 

 

 

 

 

SINGLET STATE
Abstract 113
I. ' ‘ J " .......... 114
II. Materials and Methods 120
III. Results 122
IV. Discussion 137
List of References 148

 

CHAPTER 5 STRUCTURAL PREDICTIONS FROM NATURAL

 

 

 

 

 

 

ORBITAL CALCULATIONS
Abstract 153
I. ' ‘ " " 154
II. (‘ ' ' " 160
III. Results and Discussion 172
IV. Future Work 199
List of References 203

APPENDIX A 206

viii

 

la?

la

"‘4

LIST OF TABLES
Table 2.1 Flow Rate Requirements for Sample Replenishment. 66
Table 2.2 Flowing Jet Restrictions on Temporal Delay. 67
Table 3.] Raman shifis, polarizations and assignments of the observed 96

Raman modes of ZnOEP in its ground and S, and T. excited
states in tetrahydrofuran and pyridine.

Table 4.1 Vibrational Frequencies (cm'l), isotope shifis (cm'l), and 128
polarizations of ZnOEP, ZnEtio and MgOEP in the ground state,
(So), and the lowest excited singlet state, (8,).

Table 4.2 Predictions based on shifts of vibrational frequencies of ZnTPP 145
and ZnOEP anions, and ZnOEP and CuTPP cations compared
with the measured values of shifls in the S, state.

Table 5.1 Molecular orbital energies of metalloporphin.a 161

Table 5.2 Screened potential wavefunctions of low-lying excited 163
state metalloporphin.I

 

Figure 1.1

Figure 1.2

Figure 1.3

Figure 1.4

Figure 1.5

Figure 1.6

LIST OF FIGURES

Representation of coupled electron and proton transfer in
bacterial photosynthesis. Photoexcitation of the special pair
dimer (BChlz) results in electron transfer through
bacteriopheophytin A (BPhA) and quinone A (QA) to reduce

quinone (QB), which triggers proton transfer across the membrane.

Diagram taken from Okamura, M. Y.; Feher, G.
Annu. Rev. Biochem. 1992, 61, 861-896.

Reaction center structure of Rps. viridis. Structure taken from
Deisenhofer, J .; Epp, 0.; Miki, R.; Huber, R.; Michel, H.
Nature(London) 1985, 318, 618-624.

Electron transfer reactions in reaction centers of photosynthetic
bacteria. Transfer times for forward and recombination rates
are taken from Feher, G.; Allen, J. P.; Okamura, M. Y.; Rees,
D. C. Nature 1989, 339, 111-116.

3) Structural Diagram of a D41. metalloporphyrin. X and Y
indicates positions for peripheral substitution. Pyrrole core
carbons are assigned to a, b, and m positions to permit
interpretation of Raman peaks. b) Compositions of porphyrin
in-plane “skeletal” modes deemed important in 8, state studies
within this thesis.

Metalloporphyrin absorption characteristics for ground and
excited states. The spectra are simulated to represent those of
zinc(II) octaethylporphyrin in methylene chloride. The S1 and
T1 spectra are taken with reference to results of Rodriquez, J.;
Kirmaier, C.; Holten, D. W. J. Am. Chem. Soc. 1989, 111,
6500-6501.

Molecular orbital coefficients for the HOMO and LUMO orbitals
involved in lower electronic state transitions (as calculated

ab initio by Sekino, H.; Kobayashi, H. J. Chem. Phys. 1987,
86(9), 5045-5052.

24

27

29

 

Figure 1.7

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 2.6

Figure 2.7

Raman spectrum of zinc(II) octaethylporphyrin in tetrahydrofilran.
Taken with glass nmr tube in a backscattering apparatus at room
temperature (1877 triple spectrometer, 1200 grooves/mm,

A. = 413.1 nm, 10 mW, exposure time 60 seconds). Solvent peaks
are marked by an asterisk.

Instrumentation used for time—resolved picosecond Raman
experiment SHG: second harmonic generator using KTP crystal;
THG: third harmonic generator using BBO crystal; Shutter driver:
homebuilt device to control mechanical shutters; PI/CCD:
princeton instruments charge coupled device detector; OPTICS:

a collimating lens (focus length = 2.5 inch, diameter = 2 inch),
and a focusing lens ( focus length = 12 inch, diameter = 2 inch).

Laser dye efliciency curves for a) stilbene 420 in ethylene glycol
b) pyrromethene 567 in propyl phenyl ether (PPH), and

c) rhodamine 6G in ethylene glycol. Data were collected by
delivering 1 W 355 nm and 2 W 532 nm pulses from a Coherent
76-s Antares laser to a pair of Coherent 702 dye lasers with the
cavity dumper drivers operated at 3.825 MHz .

Pulse energies from Coherent 702 dye lasers as a function of
repetition rate for a) pyrromethene 567 in propyl phenyl ether
b) rhodamine 6G in ethylene glycol and c) stilbene 420 in
ethylene glycol. The pulse energy dependence was done at the
efliciency curve maxima of each dye, with 2 W 532 nm and 1 W
355 nm pump pulses. The actual repetition rate equals

38.25 MHz/Repetition divisor.

Spectral response curves for optical filters 520DFD22-SOS and
505FD64-508. The responsivity of the 520FD22-50S is shown
at incident beam angles of 0° and 450 to display the angle
dependence of transmission.

Experimental setup initially used in determining zero-time
overlap through a modified transient absorption experiment

DODC iodide transient absorption. a) Simulated absorption
spectra representing optical behavior of DODC iodide.

b) Time-averaged voltage pulse of “blue” laser showing intensity
decrease when in temporal overlap with “red” pulses

(as seen on Tektronix DSA 602 oscilloscope).

Flowing jet cell assembly used in high repetition rate
picosecond Raman experiments.

xi

33

46

51

53

56

58

63

 

Figure 2.8

Figure 2.9

Figure 2.10

Figure 2.11

Figure 2.12

Figure 3.1

Figure 3.2

Figure 3.3

Knife-edge technique used for determining beam waist size.

a) Experimental setup used with 5X microscope objective(569-199).

b) Plots of beam intensity versus changes in razor edge translation
taken at four vertical positions. The vertical positions are listed
in reference to the position that gave the minimal beam waist
value of 13 um.

Molecular excitations expected under rapid flowing jet setup.

Simultaneous method implementing mechanical shutters
(SI and 811) within laser beam pathways.

Schematic of shutter driver. a) voltage transformer.
b) analog switching permitting shutter operation.

Partitions of time intervals generated by shutter driver.

Instrumentation used for time-resolved nanosecond Raman
experiments. HG: hamonic generators giving 2'“! and 3"1 harmonic
wavelengths; Dye Laser: optionally used to generate wavelengths
not accessible by Raman shifter; HzRaman shifter: 1st anti-stokes
of 532 nm used (435.7 nm); PBP: Pellin-Broca prism used to
separate the stokes lines coming from the Raman shifler; 1459
Illuminator: mirrors used to direct Raman scatter into Spex 1877
Triplemate.

Time resolved resonance Raman spectra of ZnOEP in THF in

(a) low wavenumber region and (b) in the high wavenumber
region. The pump and probe beams were at 572 nm and 450 nm,
respectively. The time evolution of the Raman scattering is
represented by spectra obtained at time delays between the

pump and probe pulse of 100 ps, 800 ps, 2 ns, and 27 ns. The
concentration of ZnOEP was 50104 M. The spectra of the
ground state (gs) and the solvent (dotted line) were obtained

in the absence of the pump pulse.

Time resolved resonance Raman difference spectra of ZnOEP

in THF in (a) low-wavenumber and (b) high-wavenumber range.
Experimental spectra from Figure 3.2 were acquired “quasi-
simultaneously”. The ground state features observed in Figure 3.2

at various delays for high and low frequencies have been subtracted

out such that no negative peaks appear. The peak positions of the
vibrational modes are collected in Table 3.1. The low intensities
observed at 669 cm‘1 and 753 cm'1 suggest that the intensities of
ground state features in the excited state spectra are minimal.

xii

65

69

72

74

75

86

89

91

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Time-resolved resonance Raman difference spectra taken at 93
100 pS in (a) low and (b) high wavenumber regions. Spectra

are shown for all polarizations, parallel polarization, and

perpendicular polarization. Qualitative assignments to

symmetries of vibrational modes are labeled as either totally

symmetric (P) or nontotally symmetric (DP). Solvent spectra

(dotted lines) are obtained in the absence of a pump pulse and

peaks are marked by an asterisk.

Time-resolved resonance Raman difference spectra taken at 27 95
nanoseconds in (a) low and (b) high wavenumber regions.

Spectra are shown for all polarizations, parallel polarization, and
perpendicular polarization. Qualitative assignments to symmetries

of vibrational modes are labeled as either totally symmetric (P)

or nontotally symmetric (DP). Solvent spectra (dotted lines)

are obtained in absence of a pump pulse and peaks are marked by

an asterisk.

Time-resolved resonance Raman difference spectra taken using 99
nanosecond Q-switched Nd:YAG laser pulses as pump and probe.
Spectra are shown for (a) ground state (g.s.), (b) time delay

between pump and probe of 13 ns, (c) scaled difference spectrum,

(d) parallel polarization, and (e) perpendicular polarization. The
wavelength of the pump pulse was 532 nm, and the probe

wavelength was 437.5 nm. The ground state spectrum was

taken in the absence of a pump pulse. The concentration of

ZnOEP was 5 no3 M in THF.

Time-resolved resonance Raman difference spectra of ZnOEP in 101
pyridine in (a) low-wavenumber and (b) high-wavenumber range
The ground state features observed in Figure 3.7 at various delays
for high and low frequencies have been subtracted out such that
no negative peaks appear by using techniques identical to that
used in Figure 3.2. Solvent spectra acquired immediately after
the experiment were used to subtract out the strong solvent
features which obscure the 27 ns spectrum. The peak positions
of the vibrational modes are collected in Table 3.1. Asterisks
mark artifacts arising from subtraction of the pyridine solvent
features.

xiii

 

Figure 3.8

Figure 3.9

Figure 4.1

Figure 4.2

Figure 4.3

Figure 4.4

Figure 4.5

Time-resolved resonance Raman difference spectra of ZnOEP in 104
pyridine taken at 100 ps in (a) low and (b) high wavenumber

regions. Spectra are shown for all polarizations, and parallel

and perpendicular polarizations. Qualitative assignments to

symmetries of vibrational modes are labeled as either totally

symmetric (P), or nontotally symmetric (DP). Asterisks mark

artifacts arising from subtraction of the pyridine solvent features.

Time-resolved resonance Raman difference spectra of ZnOEP in 106
pyridine taken at 27 ns in (a) low and (b) high wavenumber

regions. Spectra are shown for all polarizations, and parallel

and perpendicular polarizations. Qualitative assignments to

symmetries of vibrational modes are labeled as either totally

symmetric (P), or nontotally symmetric (DP). Asterisks mark

artifacts arising from subtraction of the pyridine solvent features.

a) Metalloporphyrin absorption characteristics for ground and 118
excited states of zinc(II) octaethylporphyrin (ZnOEP) in methylene
chloride. The S1 and T1 spectra are taken with reference to results

of Rodriquez, J .; Kirmaier, C.; Holten, D. W. J. Am. Chem. Soc.

1989, 111, 6500-6501. b) Energy level diagram representing

observed energy levels for ZnOEP.

Ground state Raman spectra of ZnOEP (n.a.), ZnOEP-d4, 124
ZnOEP-”N4, ZnEPI (n.a.) and MgOEP (n.a.). The probe beam

was at 450 nm. Each porphyrin was dissolved in tetrahydrofuran.

The concentration for all porphyrins was 5010'3 M, except ZnEPI,
which was 1010'3 M.

Parallel and perpendicular(---) Raman spectra onnOEP (n.a.), 127
ZnOEP-d4, ZnOEP-”N4, ZnEPI (n.a.) and MgOEP (n.a.).

Other than the polarizer, conditions were the same as those

described in Figure 4.2.

Raman difference spectra of ZnOEP (n.a.), ZnOEP-d4, 130
ZnOEP-”N4, ZnEPI (n.a.) and MgOEP (n.a.). The pump pulse

was set at 572 nm, except for MgOEP, which used 575 nm, and
preceded the probe pulse (450 nm) by 100 ps. Porphyrins were
dissolved in tetrahydrofirran to approximately 5010'4 M.

Parallel and perpendicular(----) Raman difference spectra of 132
ZnOEP (n.a.), ZnOEP-d4, ZnOEP-”N4, ZnEPI (n.a.) and

MgOEP (n.a.). Other than the polarizer, conditions were

identical to those used in Figure 4.4.

xiv

 

Figure 4.6

Figure 5.1

Figure 5.2

Figure 5.3
Figure 5.4
Figure 5.5
Figure 5.6
Figure 5.7
Figure 5.8
Figure 5.9
Figure 5.10
Figure 5.11

Figure 5.12

Figure 5.13

Relative isotopic shifts in ground and excited states
(from Table 4.1).

Molecular orbitals for the HOMO, a2u and a1“, LUMO, eg, and

e”, and the next two lowest unoccupied states, b“, and b2“, of

porphine as defined by Longuet-Higgins, H. C.; Rector, C. W.;
Platt, J. R. J. Chem. Phys. 1950, 18(9), 1174-1181.

Labeling sequence used for natural orbital calculations.

Composite orbital diagrams for atom electron densities of
singlet states of metalloporphin.

Composite orbital diagrams for atom electron densities of
triplet states of metalloporphin.

Composite orbital difference diagrams for atom electron
densities of lowest excited states of metalloporphin.

Composite orbital difference diagrams for atom electron
densities of second lowest excited states of metalloporphin.

Composite orbital bond order diagram for the singlet ground
state of metalloporphin.

Composite orbital bond order diagrams for the two lowest singlet
excited states of metalloporphin.

Composite orbital bond order diagrams for the two lowest triplet
excited states of metalloporphin.

Composite orbital bond order difference diagrams for the lowest
singlet and triplet excited states of metalloporphin.

Composite orbital bond order difference diagrams for the second
lowest singlet and triplet excited states of metalloporphin.

Metalloporphin bond density changes in first and second excited
states. Positive values suggest bond compression, whereas
negative values suggest bond stretching.

Representative distortions in S], T], $2, and T2 excited
electronic states.

135

158

174

176

178

180

182

184

186

188

190

192

199

201

 

CHAPTER 1

INTRODUCTION TO TIME-RESOLVED RAMAN SPECTROSCOPY

ABSTRACT

The reaction center in photosynthesis initiates charge transfer processes while its
special pair chlorophyll dimer resides in an excited singlet electronic state (8;). The role
of structural effects in this charge transfer step has been elusive but can be obtained
through vibrational spectroscopy. In this thesis, a picosecond Raman system was
developed to study these photosynthetically active chlorophyll pigments, and porphyrins
were used to study structural changes in excited states of biologically relevant molecules.
Since two-pulsed picosecond Raman spectroscopy underwent large developments during
my tenure as a graduate student, and structural knowledge about excited states of
metalloporphyrins was relatively unknown, this chapter provides relevant background
information that will be helpful in perusing Chapter 2 through Chapter 5. Bacterial
reaction centers, the original focus of this project, are discussed in detail. A synopsis of
the relevant uses of ultrafast time-resolved Raman spectroscopy (picosecond) on
biological and model systems is discussed. Detailed understanding of structural changes in
the Sr state of model porphyrins requires a detailed understanding of dynamic and
structural behavior of ground state porphyrins. Thus, a section has been included in this

chapter that summarizes the extensive literature background on metalloporphyrins.

 

I. INTRODUCTION
Free energy consumed by biological systems is ultimately generated from solar
energy that is trapped by the process of photosynthesis. Photosynthesis in green plants

results in evolution of oxygen through a deceptively simple redox equation:

H20 + 002$. (CH20).. (sugars) + 02

The actual mechanism is quite complex and requires the interplay of many proteins and
smaller molecules, involving coupled proton and electron transfer reactions.‘ Purple
photosynthetic bacteria are a well studied example of photosynthesis that generates free
energy but does not evolve oxygen. The research project described in this thesis concerns
understanding the mechanism of primary charge separation within reaction centers such as
those of purple photosynthetic bacteria.2

The bacterial reaction center contains two closely associated bacteriochlorophyll
molecules that form a bacteriochlorophyll dimer (BChIz). This dirneric species can be
photoexcited into its singlet excited state (S1). The bacteriochlorophyll dimer is still in the
Sr state when it becomes the initial electron donor in rapid electron transfer to a
bacteriopheophytin (BPh).3‘6 This charge transfer initiates a series of chemical reactions
that eventually result in conversion of C02 into organic compounds and generation of free
energy. Protein x-ray crystallography has shown that reaction center components for
bacteria such as Rb. sphaeroides and Rps. viridis are highly organized structures.7 The

structures of the redox intermediates in this series of chemical reactions are known to

 

3
regulate electron transfer rates.7 Likewise, the initial electron transfer rate to BPh is

influenced by the structure of the BChlz dimer.7 However, the structure of the inital Sr
state of the dimer has not been found by using x-ray crystallography, and must be accessed
with other physical techniques. The method we have selected for this purpose is time-
resolved resonance Raman spectroScopy.

Resonance Raman spectroscopy is a powerful technique that probes molecular
vibrations coupled to electronic transitions. When the monochromatic laser frequency is
brought into resonance with an electronic transition of a species of interest, dramatic
enhancements in scattered intensity are observed for the selected molecular species. Time-
resolved resonance Raman spectroscopy is even more selective and can be used to
measure conformational and Structural changes of transient species and excited electronic
states.” This technique can be appropriate for bacterial reaction center studies, but its
application at Michigan State University first required development of a picosecond time-
resolved resonance Raman system capable of studying S. states.

Investigation of primary charge separation in reaction centers through application
of time-resolved resonance Raman spectroscopy is the overriding goal of this project.
However, the initial experiments with the newly developed picosecond Raman
instrumentation were carried out on porphyrin model compounds. This was done for
several reasons. Inherent in the resonance Raman technique is the fact that structural
changes observed are localized. In the case of 1m" resonance Raman within the porphyrin
ring, many structural similarities with the BChlz dimer exist. Moreover, porphyrins are

worthy of studies not only for their relevance to biological systems, 1’9"“ but also for

 

4

technology driven applications such as solar energy systems,“13 development of molecular
electronic switches,“ and medical utilization in photodynamic therapy. 15 The universal
feature of these applications is that all involve fundamental processes such as vibrational
and electronic relaxation, photodissociation, and photoinduced energy and electron
transfer. This introductory chapter, as well as the remainder of the thesis, demonstrates
that these fundamental processes can be investigated by Raman spectroscopy. Since, at
this point in time, very little is known about the vibrational properties of singlet excited
electronic states of metalloporphyrins, much less the actual reaction centers, it became
necessary to use the relative simplicity of model porphyrins to aid in the analysis of excited
state structures.

This chapter gives a brief introduction to the bacterial photosynthetic system that
will be studied by using the Raman instrumentation described in this thesis in fiiture
projects. Ultrafast time-resolved Raman spectroscopy has seen extensive development in
the last few years, so a brief resume of current results on biological and model systems are
included. The structural changes that are observed in the S1 state are best understood in
relation to the structural behavior of ground state porphyrins, so a section summarizes the
basic structure and dynamics of metalloporphyrins within the context of absorption and

Raman spectroscopies.

II. REACTION CENTERS
Photosynthesis is the process whereby light energy is transformed into chemical

energy. Two classes of organisms perform photosynthesis: green plants, which evolve

 

5
oxygen, and photosynthetic bacteria, which do not. Green plants contain two

photosystems, I (PSI) and II (PSII). Bacterial photosynthesis bears similarities to that of
green plants, but is simpler and contains only one photosystem. In purple photosynthetic
bacteria, a light-driven proton pump coupled to electron transfer drives the conversion of
light energy to chemical energy. The bacterial reaction center plays a critical role in this
process by absorbing photons, performing the initial rapid electron transfers, and providing
the initial site for proton uptake reactions. The coupling between electron transfer and
proton transfer is displayed in Figure 1.1. The initial photoexcitation of the special pair
dimer (BChlz), results in subsequent electron transfers through a bacteriopheophytin
(BPhA), and then quinone A (QA) to produce doubly reduced quinone B (QB), which is
bound near the cytoplasmic side of the membrane. Two protons bind to QBZ', and the
doubly reduced quinone then dissociates from the reaction center, and is reoxidized by the
cytochrome b/cl complex, releasing protons on the periplasmic side of the membrane. The
net result is proton transport across the membrane. The proton transfer is driven by
electron transfer occurring within the reaction center. The redox centers and
chromophores of the reaction center from Rps. viridis '6 (Figure 1.2) are typical of
bacterial reaction centers and include the following cofactors: four bacteriochlorophylls,
two bacteriopheophytins, two quinones, and one nonheme Fe”. These species are bound
into two homologous protein subunits, L and M, that are not shown in Figure 1.2””
The arrangement of the cofactors in Figure 1.2 imply an apparent C2 symmetry between
subunits L and M.

The L and M subunits each have five membrane-spanning helices that form the

core of the reaction center. A third protein subunit, H, contains only one

 

 

a
2“ , . -
“wk. ,
WLQIiu I ‘
2‘ dflaMummqlhmrroU'lea raining ."1 , 1.19. up)!
mbmqa'oril 10 min iiaxw?4.r.‘1 3 a; ' '34» o-Iq I. rm: .f
‘3er iii“! mu a'a- 'Ui".fl1‘5"‘.‘i mn'.
mol OOH“: ' menu uii'w ‘ .rf‘f‘f'i h «iiwmmmmm
-, mufluon. Interimti' agar 1 r1; -l\~ ,, 1‘, avian."
, 1.O.1erleli, " ‘J' an rr'.1’3 moi ~ 1.11 rr-r
0mm 130°i man? in m: “minim . 1
. '- . / ’ - ' ‘
‘ ' ' 211"

 

 

Figure 1.1

Representation of coupled electron and proton transfer in
bacterial photosynthesis. Photoexcitation of the special pair
dimer (BChlz) results in electron transfer through
bacteriopheophytin A (BPhA) and quinone A (Q A) to reduce

quinone (QB), which triggers proton transfer across the membrane.

Diagram taken from Okamura, M. Y.; Feher, G.
Annu. Rev. Biochem. 1992, 61, 861-896.

On?

REACTION CENTER

 

hv
2H+ l PERIPLASM
.
___,.___
2e' ' 'i
BPhB BPhA

3r '.
Qurnone pool - - - - -> ‘QB‘_2e-—- QA

 

 

CYTOPLASM

 

211+

Figure 1.1

 

 

941”}an

BAC 1'

   
   

lCRl/sl m...“

3.0 ‘yrv. :3; .‘ ‘13.,

lit/W» r'

;;{T

1 ”’3’,“- '4':

inf-{‘0 , “‘11... .-.-~
3

\
‘ I
c D
O
6 u
i
O I
o
I
O
a' \
u'r
\
. \
:HCl‘

\

\ 5 I‘
."‘ ' ,r I . : ii a .
«Q f ’ 1 ‘ . ‘., V
\ l _ r
P 1) 5" I 'x
I' I‘ .
. 3?»:
°. a. i
b .- t
N
“ I) ":x -. _
. . ¢~ ,:
(“-1 ft" s '4 N y ) ( I §'—‘ ‘ .F
V; f, - " .,
.~' - I {1 ‘3 ”fl " Q
6“ and, 53(“ i “a
"." o .13 ".0 “3‘ ‘_ P ._ .\ v
“A '~ . ' -' . ';
u a . ' .
r M w . a
fly.

7V! 15!

A

h
i I
1 t A 2
‘ ‘ 11
n .
’i‘-
J '
u .
.‘ 7‘9!
5.,
x '1
v
‘u

U
1‘59 "-1 ’rfi
”W34 «

liPl

tr

1

0

 

Figure 1.2

Reaction center structure of Rps. viridis. Structure taken from
Deisenhofer, 1.; Epp, 0.; Miki, R.; Huber, R.; Michel, H.
Nature(London) 1985, 318, 618-624.

BACTERIAL REACTION CENTER

 

Figure 1.2

 

mem'

comr

pror

enri

an

the

10

membrane-spanning helix and is asymmetrically located on the cytoplasmic side of the
complex overlaying the QA Fezi complex. Subunit H behaves as a cap over QA and
promotes proton transfer association with Q; by shielding QA from the aqueous
environment.2

The arrangement of the cofactors within the bacterial reaction center (RC) are very
symmetrical, yet electron transfer is favored almost unilaterally down the L branch and not
the M branch. Environmental and structural factors between the reaction center and the
surrounding protein are used to explain this selectivity. Figure 1.3 summarizes the series
of electron transfer steps that occur within the RC. The primary charge separation step is
initiated by photoexcitation of the special pair dimer (BChlz). Since X-ray crystallography
has shown a separation distance of about 7 A between the central bacteriochlorophyll ring
positions of the special pair bacteriochlorophylls in Rps. viridis,"5 excitonic effects are
presumed to influence the subsequent electron transfer (the average bacteriochlorophyll
separation distance between rings is ~3.0 A).'9'2° The special pair dimer then proceeds to
reduce the L-side bacteriopheophytin (BPhA) within 3 psec.

This short transfer time is surprising given the relatively large distance of 17 A
(between ring centers) between BChlz and 8%,, and their small electron density
overlaps. 7’“ The bridging monomer, BChlA, is believed to be involved in the electron
transfer mechanismzl'n’23’2‘ One proposal suggests a two-step sequential mechanism, in
which BChlA' serves as a transient intermediate state that has not been observed using

femtosecond spectroscopy. An alternative mechanism, called superexchange, proposes

 

‘
-——W— 'v

 

Figure 1.3

11

Electron transfer reactions in reaction centers of photosynthetic
bacteria. Transfer times for forward and recombination rates
are taken fi'orn Feher, G.; Allen, J. P.; Okamura, M. Y.; Rees,
D. C. Nature 1989, 339, 111-116.

el

 

1.5 l_.. *
(130102 BPhQA Fe2+QB
j} 10' l 2S

+ -
,0 __ (Ecru)2 BPh QAFeZ‘LQB
h 10105
v -9
10
eV 5 + -
BChl BPh Fe2+
10'88 ( b QA QB
0.5 _ N048

Chi): BPhQA Fe2+QB

ls

 

 

l

0 — (Ecru)2 BPhQA Fe2+QB

 

 

 

Figure 1.3

 

hr
1

BO

elec

hie

nit:

The

the

13
that electronic overlap between BCh12* and RPM is enhanced by a virtual state

BChlfBChlg.

Electron transfer from BPhA to QA proceeds in ~ 200 psec. QA transfers an
electron to Q3 within a millisecond to complete the charge separation process. Q; then
interacts in the proton transport process. The function of the non-heme iron appears to be
regulation of environmental differences between the two quinones. The iron is ligated to
nitrogen atoms on histidine residues that in turn are ligated to protein subunits L and M.
These polar interactions with histidine residues appear to stabilize the tertiary structure of
the L and M subunits.”

Interest is taken in the bacterial photosynthetic reaction center, particularly as a
result of its usefulness in analogies to the more complex green plant photosystem.26
Structural and functional similarities between oxygenic and non-oxygenic reaction centers
begin with the pseudo-C2 symmetric arrangement of chromophores, 27’” but are also
apparent in many other ways. For example, the simpler bacterial organisms obtain
reductants from organics and synthesize sugars through a single light reaction. The
oxygenic photosynthetic organisms not only fix the carbon, but also evolve molecular
oxygen in two light reactions associated with PSI and P811. The L and M protein subunits
of bacterial reaction centers have been shown to have significant homology with the dimer
of proteins that enclose the reaction center of PSII.29'3°'31 These stnrctural similarities
suggest that PSI may have evolved from green photosynthetic bacteria, whereas PSH
probably arose from purple photosynthetic bacteria.32 In any case, PSI and purple bacteria

appear to have a common ancestor.33 The differences that exist in comparisons of

A

qw

pho:

this

err.

183

b)

14

quantum yield and energy conversion efficiency have been speculated to be due to the
differing environments between bacterial reaction centers and those of oxygenic
photosynthetic organisms. The bacterial organisms operate near unity quantum yield, but
this is done at the expense of energy conversion efficiency (~ 30%).7 Plant systems have
higher energy conversion efficiency but a lesser quantum yield (~ 90%).34

These structural homologies suggest that calculated electron transfer kinetics are
sensitive to the structures of the associated intermediates. Additionally, experimental
evidence suggests that structural changes occur during charge separation.”36 For these
reasons, a time-resolved Raman spectroscopic apparatus that could monitor both time and
structure of BChlz was designed. As will be shown in the following chapters, this was

37'3“” Structural changes incurred

successfully applied to BChIz like porphyrin molecules.
by BPh (and BCth upon reduction can also be monitored via Raman spectroscopy. This
combined dynamical information should give quantitative understanding of the mechanism

of the initial electron transfer step.

III. ULTRAFAST RAMAN SPECTROSCOPY APPLICATIONS

Time-resolved Rarnan spectroscopy combines the vibrational and structural
information gained through the Raman process with the ability to probe transient species
in photophysical and photobiological processes. Resonance Raman spectroscopy takes
advantage of the high selectivity and increased sensitivity gained by tuning the probe laser
wavelength to an electronic transition of interest. While the time resolution can include '

time ranges from microseconds to sub-picoseconds, this review of experimental methods

 

Willi

to h

1115!

list

lith

C01

the

(£3

15

will be limited to recent applications that use picosecond time-resolved spontaneous
Raman spectroscopy and gives a general overview of the techniques, and the applications

to biological and model systems.

Instrumental Techniques

For time-resolved picosecond Raman spectroscopy, lamp-pumped solid state
lasers, such as neodynium—yttrium—aluminum—gamet (Nd:YAG) or neodynium-yttrium-
lithium-fluroborate (Nd:YLF) lasers are the most commonly used laser systems, often in
conjunction with dye lasers and/or amplifiers. Laser pumped solid state lasers, such as
titanium sapphire lasers, have been more recently developed and are becoming increasingly
used for picosecond pulse generation. Currently, the most common approach for
picosecond Raman experiments involves spatially and temporally overlapping two pulsed
lasers with pulse widths in the range 4-30 psec.8"°'“ Single pulse picosecond experiments
have also been done."2""z"'43

Pulse repetition rates and pulse energies used in ultrafast Rarnan spectroscopy vary
greatly among experimental setups. The choice of amplification scheme requires a
compromise between the need for higher peak power and the need for moderate average
power. High average power enhances high collection of Rarnan scattered light, but high
pulse energies can also destroy the sample through non-linear processes. The high
repetition rate (70 to 80 MHz) characteristic of source lasers like Nd:YAG or Nd:YLF
can generate 50 Hz or near KHz repetition rates when used with various amplifiers."°"‘2'44

However, amplified and slower repetition rate picosecond lasers bring higher pulse

 

Cliff

thos

picc

den

‘7!

l6
energies. Typical pulse energies for 50 Hz systems are on the order of 100’s of uJ, while

those with KHz amplification produce 5 to 20 11]. Some research groups have neglected
the use of low repetition rate amplifiers and retained the MHz repetition rates of
picosecond lasers. The MHz systems typically have pulse energies on the order of tens to
hundreds of nanojoules.‘l Single grating spectrometers and charge coupled device (CCD)
detectors or intensified photodiode array (IPDA) detectors are the most commonly
employed collection and detection devices used in these picosecond time-resolved
resonance Raman spectroscopy experiments. Difi'erent combinations of these lasers and
collection/detection devices have been used for various research projects. In order to
visualize the general impact of time-resolved Raman spectroscopy on research systems
more clearly, the following discussion will be divided into broadly based topic areas.
These areas include studies done on vibrational relaxation processes, photochemical

systems, and photobiological systems.

Vibrational Relaxation Processes

Vibrational relaxation is a molecular process whereby excess energy from a
vibrationally excited molecule is dissipated to its surroundings. Information on vibrational
relaxation can be obtained in different ways through time-resolved Raman spectroscopy.
Vibrational relaxation of vibrational modes is expressed as a sum of contributions from

population decay time, T1, and pure dephasing time, Tzf‘:

T2" = (21‘0" + (T2*)"

 

Tim

PCP

que

tell:

pie;

mol

1711)

the

Phi

Pit“

S‘s-e

Time-resolved spontaneous anti-stokes Raman spectroscopy can obtain information on the
population lifetime (T1) of vibrational modes. This is of paramount importance for
questions addressing state-selective chemistry. The decay time of vibrational modes are
reflected by their intrinsic linewidth. Raman bands of electronically excited molecules are
often much broader than those of ground state molecules.37'38'39'45'46'47 This indicates the
presence of additional mechanisms for vibrational dephasing in electronically excited
molecules. As an example, the vibrational dephasing in the C=C stretching mode of Sr
trans-stilbene was explained within a dephasing model that supported theories addressing

the isomerization process from trans-stilbene to cis-stilbene.48

Photochemical Systems

The ideal excited electronic state to be studied with time-resolved resonance
Raman spectroscopy should have the following traits: the excited state should be
produced in a photochemically simple process, the extinction coefficients of both the
ground and excited electronic state should be large, and preferably the excited state should
allow resonance Raman excitation in the near ultraviolet region. The strong absorption
and selective resonance Raman in the ultraviolet region would give high signal-to-noise
transient Raman spectra. Delocalized 1m" transitions have all these qualities and are
particularly pliable for time-resolved resonance Raman studies.

Studies of 1m" excitation of trans-stilbene are especially amenable since the

spectroscopic properties are reasonably well understoodwo The time and solvent

 

1 8
dependent changes of band shapes for C=C and C-Ph stretches have been noted for

stilbene.’o These experiments have also been done with structural analogues such as
paraphenylenes, biphenyl, and terphenyl.‘9'5"52
The 1m" transitions of polyenes are also amenable to Raman studies. In an attempt
to understand the pericyclic rearrangements relevant to vitamin D photochemistry,
Mathies and coworkers monitored several of these polyenes via picosecond ultraviolet
time-resolved stokes and anti-stokes Raman spectroscopy.““”53’5"’5”6 Their studies on
1,3,5-cycloheptatriene (CHT) first determined the kinetics of a sigrnatropic shift on the
CHT rings”s Later experiments on CHT characterized Raman scattering from an
electronically excited state populated by internal conversion from the initially prepared

electronic excited state, and vibrationally unrelaxed and relaxed ground states.“"'” In

56"“ they concluded that ring

conjunction with resonance Raman analysis results,
planarization of the polyene transpired prior to all pericyclic nuclear rearrangements.
Ensuing picosecond Raman results determined reaction dynamics on ring-openings of 1,3-
cyclohexadiene (CHD) and or-phellandrene (or-PHE).57 Photoexcitation of these polyenes
resulted in a ring-opening that occured in a conrotatory (i.e. with a twisting of tenninal
groups in the same direction) fashion. The conrotatory movements were shown to agree
with the orbital symmetry predictions of Hoffmann and Woodward.’8 Interestingly, the
time-resolved Raman results on 1,3,5-cyclooctatriene (COT) show disrotatory rotation
upon photoexcitation of the ring.” However, these results are also consistent with orbital
symmetry predictions.” Overall, their work is a good example of the utilility of time-

resolved stokes and anti-stokes Rarnan scattering for characterizing structural and

 

A

allt‘
sin.

“h

the
off:

him

l9
vibrational relaxation kinetics of these photoproducts. Furthermore, this linking of

experimental results to theoretical predictions was implemented within this thesis

(particularly within Chapter 4 and Chapter 5).

Photobiological Systems

Picosecond time-resolved resonance Raman studies have also been applied to
directly measure biological systems such as carotenoids, bacteriorhodopsin, and heme
proteins. As noted earlier, single pulse picosecond experiments have been done on
metalloporphyrins. Two-pulse picosecond Raman experiments on metalloporphyrins have
only recently been completed, and are discussed in detail in Chapter 2 through Chapter 5
within this thesis.

Time-resolved Raman spectra have been obtained for photobiological systems such
as B-carotene and other membrane-bound carotenoids. Carotenoids are of biological
importance since they perform as energy absorbers in the wavelength region 400-600 nm,
and subsequently transfer energy to chlorophylls and bacteriochlorophst through their
singlet excited states. Essentially, carotenoids complement the chlorophyll chromophores,
which are basically non-absorptive between the Soret and Q bands. Hashimoto er 01.6““
concluded from their time-resolved studies on carotenoids, that the lowest singlet excited
state was 21A,, not the I'Bu orbital. Subsequent studies of carotenoids have looked into
the relaxation of the 2'A. state to form a vibrationally excited 11Ag ground state. This is
of relevance since the vibrationally excited ground state is thought to transfer energy to

bacteriochlorophyll.‘52'63

 

20
Another photobiological system that has attracted interest involves photocyclic

reactions in retinal proteins, such as the protein bacteriorhodopsin. Bacteriorhodopsin
(BR) is an integral protein in the purple membrane of Halobacterium halobium that, upon
photoexcitation, undergoes isomerization at the all-trans retinal protonated Schiff base

“’65 This protein then behaves as a light-driven transmembrane proton

prosthetic group.
pump.“5 The reversible photocycle has generic labels for the numerous intermediates (1, J,
K, L, M, N, O) that are involved in the isomerization of all-trans BR and pump protons
across a membrane and drive ATP synthesis.67 Time-resolved picosecond Raman studies
were essential in identifying the different conformations of early intermediates such as J
and K. Atkinson and coworkers initially identified the resonance Raman features of these
species, and established that neither I nor K contained all-trans retinal, but were both
conformationally distinct species.68 They concluded that the primary BR photocycle event
must include a configurational change in the retinal chromophore. Subsequently, Mathies
and coworkers evaluated the structure and dynamics of the J and K intermediates of
bacteriorhodopsin by using stokes and anti-stokes Raman.69 They concluded that
isomerization induced several subtle changes that resulted in a greater u-electron
delocalization over the polyene. Strong mode intensities for the anti-stokes Raman
spectrum of the J intermediate were used to interpret J as a highly twisted chromophore
that is vibrationally excited, but cools upon formation of K. As judged by the out-of-plane
vibrational modes, K is a more planar intermediate and shows adjustment to protein

changes in the local environment within 3 picoseconds.

 

2 1
Oxygen binding to heme proteins, such as hemoglobin and myoglobin, are another

example of the successful application of picosecond Raman spectroscopy to biological
systems. Time-resolved studies by Findsen et aI.,7° Dasgupta et al.,-'1 and Rousseau et
a1.” all indicate time dependent Raman frequency shifts between photolyzed heme protein
and stable deoxy-heme protein. Interpretations of these changes were drawn mostly from
the v4 and iron-histidine stretching frequencies. They correlated these shifts to bond
distance changes in the central core of the heme. The collective results of these studies
suggest a recombination rate for the heme proteins that proceeds in the following manner.
The initial unligated photoexcited state of the heme protein forms within 50 fsec.‘ The
protein relaxes to its vibrationally excited ground state in less than 2.5 psec."3 The
vibrationally excited bands in deoxy-hemoglobin have been studied by Lingle et £11.74”
using stokes and anti-stokes time-resolved Raman

Afier vibrational cooling is complete, the v4 mode, which initially shifted down
upon photoexcitation, slowly increases in frequency from ~10 psec to 10 nsec.76 This
suggests a second slower relaxation process in the protein. This relaxation process has
been attributed to a more global protein motion that induces a tilt in the proximal histidine
relative to the heme plane."5 This theory is supported by correlation between iron-
histidine and v4 stretching frequencies for a variety of heme structures.7o Clearly, time-
resolved vibrational information such as provided by Raman is essential to perform this
type of structural analysis.

While the previous discussion shows that excited electronic states of

49-59

photochemical systems have been studied rather effectively, and certain

fev

life

22

photobiological systems like bacteriorhodopsin have also been analyzed,6469 at this time
few studies had been done with singlet excited states of porphyrins. Since excited state
lifetimes of iron porphyrins are on the order of a picosecond,73 heme protein studies were
limited to studying conformational changes of electronically relaxed proteins."'7M5 While
triplet excited states were accessible by nanosecond pulsed lasers," picosecond
experiments on metalloporphyrins had been limited to d,d excited states,”“‘43 Since this

37.38.39

thesis presents the first picosecond results on singlet excited electronic states, a brief

discussion of porphyrin structure and dynamics is in order.

V. METALLOPORPHYRIN STRUCTURE AND DYNAMICS

Porphyrins are aromatic macrocycles with four pyrrole rings connected by
unsaturated methine bridges at the four corners of a square (Figure 1.4a). The conjugated
macrocycle of a D41, metalloporphyrin shows three unique carbon sites that are
distinguished by labels as C., Cm, and C1,. This pemiits the assignment of vibrational
modes to observed Raman frequencies (Figure 1.4b). A typical porphyrin has 69
non-substituent vibrational modes, but five were of particular interest in this research
project. These five, the v4, v3, v2, V“), and v“ modes, predominantly reflect vibrational
stretching (and stretching force constants) between different combinations of the C., C»,
and Cm atoms. An empirical correlation has been found between the vibrational
frequencies of the five modes listed in Figure 1.4b and the overall size of the porphyrin

ring.78 Hence, these modes are commonly called the porphyrin “core size” skeletal mode

 

Figure 1.4

23

a) Structural Diagram of a D41. metalloporphyrin. X and Y
indicates positions for peripheral substitution. Pyrrole core
carbons are assigned to a, b, and m positions to permit
interpretation of Raman peaks. b) Compositions of porphyrin
in-plane “skeletal” modes deemed important in 81 state studies
within this thesis.

 

3|

a)
X\, b X
Y Y
\ a- m
r:,:\ j»
l /N .................. M .................. N i
X/ \\\/ / N§\\l/ _ < \X
/ . \\
Y _, __ Y
/’W* m—\_
X X
b)

04: 0(CaN),U(CaCb) 03: 0(Cbe),U(CaCm)
023 0(Cbe) U1030(CaCm)

0113 0(Cbcb)

Figure 1.4

 

up

59

fr:

re}

1’6}

lEi

Fig
Kl
re;
Sp
do
sh.

fee

Eie

25

fi'equencies. Moreover, these porphyrin resonance Raman features are very dependent
upon the absorption characteristics.

The individual vibrational frequencies typically observed in a resonance Raman
spectrum arise from normal modes in the ground electronic state." Shifts in these
frequencies can reveal information about ligation, oxidation, and spin states, as well as size
dimensions of the porphyrin ring (core size). The intensities of the Raman lines, and the
relation of Raman intensity between perpendicular and parallel polarized light excitation,
reflect the nature of the electronic excited states. In addition, most of the resonance
Raman bands have been assigned to metalloporphyrin normal modes on the basis of
isotope substitution and normal-mode coordinate analysis.”80 This permits application of
resonance Raman spectroscopy to a wide range of structural studies.

Absorption characteristics for zinc(II) octaethylporphyrin (ZnOEP) are shown in
Figure 1.5 and are typical of normal metalloporphyrin absorption characteristics.
Metalloporphyrins typically have two visible bands (Qo,o and QOJ) in the 500-600 nm .
region and have an intense band called the Soret near 400 nm. Transient absorption
spectroscopy of ZnOEP and many other metalloporphyrins in the S. and T1 states has been
done.73’"'” Typical absorption spectra of metalloporphyrin in the So, SI, and T1 states are
shown in Figure 1.5.73 The excited state metalloporphyrins have similar absorption
features, but peaks are usually shifted and broadened relative to the ground state spectra.73
These absorptions are normally interpreted within the context of Gouterman’s four-orbital
model (Figure 1.6).‘3'34

Gouterman assumed an idealized Du. symmetry for the porphyrin ring, with excited

electronic transitions fi'om the two nearly degenerate an. and an HOMO ‘s into the

 

U’

)-

 

r”

 
   
  

Q10; 2’54? Vt
may: or 152.3;

”’08" .i":
.mt 36.7.. eff)

 

ha 047' .obnohbfflp than u‘unvt

"Ai :L‘J A

A‘
231:1: Mtge-c :Z"l;.litlf'flii;m’t

1.1qu 3"

any Lad ma

(ii, 01" jflnfip"

»3301W‘im " 31:“ an: u, , 1:;

MK} .‘.N N\;,4m‘1,’
' 1
i - i-‘
l.
" ’vf '
H .
/ °.
\ . s.“
I "

J '19. 'flril
shaft"

.FW’JJ'IU ‘ (o M)

Figure I .5

r. I email

 

Figure 1.5

26

Metalloporphyrin absorption characteristics for ground and
excited states. The spectra are simulated to represent those of
zinc(II) octaethylporphyrin in methylene chloride. The S. and
T1 spectra are taken with reference to results of Rodriquez, J .;
Kirmaier, C.; Holten, D. W. J. Am. Chem. Soc. 1989, 111,
6500-6501.

27

B

zinc (II) octaethylporphyrin

 
 
    

13Eu Triplet Excited State

11Eu Singlet State

Q, Ground State
J “T L—

PROBE
PUMP

 

 

 

'350'450'5c'n‘ebo'7bo'8inofi
wavelength(nm)

Figure 1.5

 

 

.w
)- .-
.' \ i" i
“)5— r’ ‘ ’ < 'h ‘
‘ ‘l, . .' \._, r.‘ .
r ‘0‘- x I ." -.’ . ,7.-
I'.' ’3‘ 3‘03"
0 , - ' i I
, 5 ,‘ \~ " "
. . . l
. L (1.
\—-f--, ‘-‘ . '
x V ‘
t . ’ ‘
.- i. i
_ 4" I, \x’ t i "i\ r“
x ' i “I
‘ ,"\ ’ ‘ I ”t
, .I "I x
i
I")
G

'I f i \ I

   
  

-' {flit}, ‘ 3 - «i
has 0mm mm emigrate; «rhE‘F-Wd'yir fem"?! -
mutiiammqsiw “10‘.an W' m. 4‘ "Iv - @
-II.\\"?-.."\. iti);.r.?dcx . H 1"“ ’ 05¢ ‘ {I {in :34? V
at «mt-mt .. +-
4. ’ 1 ‘ ' - i
95;; r; . o .
’ 8-. l . ‘ -
I'M . 0‘ I |
”-11%; -- r , i i
m c --~-* at “(a ~
‘ 3 ”1r." _ . i
' ‘

 

 

 

   

m ' h

Figure 1.6

 

Figure 1.6

28

Molecular orbital coefficients for the HOMO and LUMO orbitals
involved in lower electronic state transitions (as calculated

ab initio by Sekino, H.; Kobayashi, H. J. Chem. Phys. 1987,
86(9), 5045-5052.

Ll

29

 

 

 

 

LUMO cg = F..— i...—

 

 

HOMO a1u
a2u

 

 

Figure 1.6

dl

30
degenerate e, LUMO pair. The lowest singlet excited state configurations l(ameg) and

l(amen) are of the same Ell symmetry and have nearly the same energies. As a result, these
transitions have strong electronic interactions and mix by configuration interaction. The
transition dipole moments add up for the Soret band, and nearly cancel for the weaker Q0
band transition. Vibronic mixing of the B and Q04, produces the Qo,1 sideband. The
symmetries of the vibrational modes that actively couple the E.I states are A1,, + A2B + B1,
+ 132,. A coordinate operator with any one of these symmetries will be active in Vibronic
coupling, and is quantified by the derivative of the electric dipole moment for the excited

state with respect to the normal coordinate:

23me = m. < s / SI-IISQ/e >/ h(u. - u.)

where SHISQ is the derivative of the Hamiltonian with respect to the normal mode, h is
Planck’s constant, e and s are excited and mixing state wavefunctions, u, is the excited
state frequency, 1).. and m. are the frequency and electric dipole moment of the excited
state, and ‘u. and m. are the fi'equency and electric dipole moment of the mixing state.

The ground state Raman spectra of metalloporphyrins are explained by using the

following formalism. Treating the substituent groups at the mesa and [3 positions as point

masses, the metalloporphyrin with N = 37 atoms has 71 in-plane vibrations (ZN-3), 35 of

which are Raman-active:85

PM = 9A1! + 8A28 + 9813 + 9323 + 18E“

31

Likewise, there are 34 out-of-plane vibrations (N-3):85

1“out-of-plane = 3Alu + 6A2u + 581:: + 43211 + 8Eg

The resonance Raman frequencies below 1000 cm'1 are dominated by the out-of-plane
bending or ring defamation porphyrin modes, and modes involving the heavy-atom metal
center. The resonance Raman bands in the 1000 to 1700 cm’1 range are the in-plane
stretching C-C and ON modes and are sensitive to the inner porphyrin environment,
responding according to Badger’s rule.“86
These modes are found to be affected by conditions such as the porphyrin core size

and metal complex spin state, metal oxidation state, porphyrin ring oxidation and

reduction, and 1t backbonding and doming effects. These modes are resonantly enhanced

by M. electronic transitions.

Since most of the spectra that are presented in Chapter 3 and Chapter 4 were taken
with Raman excitation wavelengths between 420 to 460 nm, consider the Raman spectrum
of Zn(II) octaethylporphyrin in tetrahydrofirran (Figure 1.7). The porphyrin was excited
at 413.1 nm with a krypton ion laser and illustrates effects of Soret enhancement upon
porphyrin Raman spectra. The totally symmetric bands become relatively more enhanced
when excited near the Soret. Specifically, the strongest band in resonance Raman spectra

excited near resonance with the Soret is D4, the C.-N stretching mode. This mode is often

called the 1: electron density marker or oxidation state marker because its frequency is

32

Figure 1.7 Raman spectrum of zinc(II) octaethylporphyrin in tetrahydrofuran.
Taken with glass nmr tube in a backscattering apparatus at room
temperature (1877 triple spectrometer, 1200 grooves/mm,

A = 413.1 nm, 10 mW, exposure time 60 seconds). Solvent peaks
are marked by an asterisk.

33

w
u 59 v
2 E 89

u 1 39 can.

1

u 3 Ev we:
*

.0

 

 

4
D

E 05

zinc(II) octaethylporphyrin

iv «8,

39 :m.

E 8:

E295 $25.9“.

E 82

5..

E 89

*

E NE

                        

 

1000 1200 1400 1600

800

Raman shift (cm'1)

Figure 1.7

res

0X

Sit”:

rel

re

34
responsive to shifts with electron density changes.” The U4 frequency is dependent on the

oxidation state of the central metal atom, and typically increases with increase in oxidation
state for a given metal. Macrocycle cation and anion radicals show frequency shifts
relative to neutral metalloporphyrins that agree with electron occupancy arguments using
Gouterman’s four orbital model (Figure 1.6).78b'8‘5 The HOMO for octaethylporphyrins

(substitution at the [3 position) is the an. orbital fi'om Figure 1.6, whereas mesa substituted
porphyrins, such as tetraphenylporphin, have a2ll as the HOMO."‘"’"”"90 Sirniliar effects on

the Raman spectra are observed with It backbonding but this occurs when d electrons on
the metal contribute to the 1t* orbitals of the porphyrin ring.84 Doming effects (ruffling)

are smaller but occur when the pyrrole nitrogens follow out-of-plane motion of the central
atom.“ Changing the metal complex spin state from low-spin to high-spin lengthens the
bond length to the metal, enlarging the core size and reducing the vibrational frequencies.
The aim of this work at the outset was to develop an appropriate picosecond time-
resolved apparatus capable of monitoring excited electronic states using Raman
spectroscopy. This introduction has shown that picosecond time-resolved Raman
spectroscopy has worked well for other chemical and biological systems, and porphyrins
have the optical characteristics that make them suitable for similiar studies. The intention
was to use this apparatus to characterize the vibrational and structural properties in St
states of metalloporphyrins. Since vibrational characterization of model porphyrins had
not been previously addressed, this was a necessary prerequisite to bacteriochlorophyll .
analysis. Once this characterization is complete, this experiment can be applied towards

understanding of the primary charge separation step in bacterial reaction centers.

(1)

(2)
(3)

(4)
(5)

(6)
(7)
(3)
(9)

(10)

(11)

LIST OF REFERENCES

Deisenhofer, J .; Norris, J. R. in The Photosynthetic Reaction Center; Academic
Press, Inc.: California, 1993, Vol 1.

Okamura, M. Y.; Feher, G. Annu. Rev. Biochem. 1992, 61, 861-896.

Breton, J .; Vermeglio, A. in The Photosynthetic Bacterial Reaction Center:
Structure and Dynamics; Plenum: New York 1988.

Norris, J. R.; Schiffer, M. Chem. Eng. News 1990, 68(31), 22-28.

Michel-Beyerle, M. E. in Reaction Centers of Photosynthetic Bacteria; Springer-
Verlag: New York, 1990, Vol I.

Mitchell, P. Nature 1961, 191, 144-48.
Feher, G.; Allen, J. P.; Okamura, M. Y.; Rees, D. C. Nature 1989, 111-116.
Hamaguchi, H. and Gustafson, T. L. Ann. Rev. Phys. Chem. 1994, 45, 593-622.

(a) Lutz, ; Robert, B., in Biological Applications of Raman Spectroscopy; Spiro,
T. G., Ed; John Wiley & Sons: New York, 1988, pp 347-413. (b) Rutherford, A
W.; Zimmerman, J.-L.; Boussac, A. in The Photosystems: Structure, Function and
Molecular Biology; Barber, J ., Ed; Elsevier: Amsterdam, 1992, pp 125-158.

(a) Wilkstrbm, M.; Krab, K.; Saraste, M. in Cytochrome Oxidase-A Synthesis;
Academic Press: New York, 1981. (b) Babcock, G. T. in Biological Applications
of Raman Spectroscopy; Spiro, T. 6., Ed; John Wiley & Sons: New York, 1988,
pp 293-346.

(a) Rousseau, D. L.; Friedman, J. M. in Biological Applications of Raman
Spectroscopy, Spiro, T. 6., Ed; John Wiley & Sons: New York, 1988, pp 133-
217. (b) Cartling, B. in Biological Applications of Raman Spectroscopy; Spiro, T.
G., Ed; John Wiley & Sons: New York, 1988, pp 217-248. (c) Champion, P. M.
in Biological Applications of Raman Spectroscopy; Spiro, T. 6., Ed; John Wiley
& Sons: New York, 1988, pp 249-292.

35

(12)

(13)

(14)

(15)

(16)

(17)

(13)
(19)
(20)

36

(a) Haniman, A.; Porter, G.; Richoux, M.-C. J. Chem. Soc. Faraday Trans 2
1981, 77, 833-844. (b) Kalyanasundaram, K.; Neumann-Spallart, M. J. Phys.
Chem. 1982, 86, 5163-5169. (c) Kay, A; Gratzel, M. J. Phys. Chem. 1993, 97,
6272-6277.

(a) Wasielewski, M. R Chem. Rev. 1992, 92, 43 5-461. (b) Wasielewski, M. R. in
The Photosynthetic Reaction Center; Deisenhofer, J .; Norris, J. R, Eds.;
Academic Press, Inc.: California, Vol H, 1993, pp 465-512. (c) Gust, D.; Moore,
T. A. in The Photosynthetic Reaction Center; Deisenhofer, J .; Norris, J. R, Eds;
Academic Press, Inc.: California, Vol II , 1993, pp 387-418.

(a) Carter, F. L. in Molecular Electronic Devices; Dekker: New York, 1982, Vol
I. (b) Hopfield, J. J .; Onuchic, J. N.; Beratan, D. N. Science 1988, 241, 817-823.
(c) Parthenopolous, D. A.; Rentzepis, P. M. ibid 1989, 245, 843. (d) O’Neil, M.
P.; Niemczyk, M. P.; Svec, W. A.; Gosztola, D.; Gaines, G. L.; Wasielewski, M.
R Science 1992, 257, 63-65.

(a) Jori, G. Radiat. Phys. Chem. 1987, 30, 375. (b) Borland, C. F.; Mcgarvey, D.
J .; Morgan, A R; Truscott, T. G. J. Photochem. Photobiol. B: Biology 1988, 2,
427. (c) Schermann, G.; Volcker, A.; Seikel, K.; Schmidt, R; Brauer, H.-D.;
Montforts, F.-P. Photochem. Photobiol. 1990, 51, 45-52. (d) Mironov, A. F.;
Seylanov, J. A.; Pizhik, V. M.; Deruzhenko, I. V.; Ju Nockel, A. J. Photochem.
Photobiol. B: Biology 1992, 16, 341.

Deisenhofer, J .; Epp, 0.; Miki, R; Huber, R; Michel, H. Nature(London) 1985,
318, 618-624.

F eher, G.; Okamura, M. Y. in The Photosynthetic Bacteria; Plenum: New York,
197 8.

Parson, W. W. in Photosynthesis; Elsevier: Amsterdam, 1987, pp 43-61.
Kasha, M. Radial. Research 1963, 20, 55-71.

(a) Gregory, A. R; Henneker, W. H.; Siebrand, W.; Zgierski, M. Z. J. Chem.
Phys. 1975, 63(12), 5475-5489. (b) Netzel, T. L.; Kroger, P.; Chang, C. K.;
Fujita, I.; Fajer, J. Chem. Phys. Lett. 1979, 67, 223-228. (c) Lalonde, D. E.; Petke,
J. D.; Maggiora, G. M. J. Phys. Chem. 1988, 92(16) 4746-4752. ((1) Hunter, C.
A.; Sanders, J. K. M.; Stone, A. J. Chem. Phys. 1989, 133, 395-404. (e) Won, Y.;
Friesner, R. A.; Johnson, M. R; Sessler, J. L. Photosynth. Res. 1989, 22(3), 201-
210. (f) Tran-Thi, T. H.; Lipskier, J. F.; Maillard, P.; Momenteau, M.; Lopez-
Castillo, J.-M.; Jay-Gerin, J.-P. J. Phys. Chem. 1992, 96, 1073-1082. (g)
Helennius, V. M.; Hynninen, P. H.; Korppi-Tommola, J. E. I. Proc. SPIE-Int. Soc.
Opt. Eng. 1993, 1921(Laser Spectroscopy of Biomolecules), 86-93.

(21)
(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(3 1)
(32)

(33)

(34)

(35)
(36)
(37)
(38)

37
Marcus, R A. Chem. Phys. Lett. 1987, 133, 471-477.

Bixon, M.; Jortner, J .; Michel-Beyerle, M. E.; Ogrodnik, A; Lersch, W. Chem.
Phys. Lett. 1988, 140, 626-630.

Fischer, S. F.; Scherer, P. O. J. Chem. Phys. 1987, 115, 151-158.

Creighton, S.; Hwang, J. K.; Warshel, A; Parson, W. W.; Norris, J. Biochemistry
1986, 27, 774-781.

Schifi‘er, M., Norris, J. R. The Photosynthetic Reaction Center; Deisenhofer, J .;
Norris, J. R, Eds; Academic Press Inc.: New York, 1993; Vol 1, pp 1-10.

Rutherford, A W. Biochim. Biophys. Acta. 1985, 14, 15-17.
Debus, R J .; Biochim. Biophys. Acta., 1992, 1102, 269-3 52.

Deisenhofer, J .; Epp, 0.; Miki, R; Huber, R; Michel, H. Nature (London), 1985,
318, 618-624.

Youvan, D. C.; Bylina, E. J .; Alberti, M.; Begusch, H.; Hearst, J. B. Cell, 1984,
37, 949-957.

Deisenhofer, J .; Epp., 0.; Mrochaix, J. D.; Dron, M.; Rahire, M.; Malnoe, P.
Plant. Mol. Biol. 1984, 3, 363-370.

Michel, H.; Deisenhofer, J. Biochemistry, 1988, 27, 1-7.

Seibert, M. in The Photosynthetic Reaction Center; Deisenhofer, J .; Norris, J. R,
Eds; Academic Press Inc.: New York, 1993; Vol 1, pp 319-347.

Nitschke, W.; Rutherford, A. W. Trends Biochem. Sci. 1991, 16, 241-245.

Babcock, G. T.; Ghanotakis, D. F .; Ke, B.; Diner, B. A Biochim. Biophys. Acta.
1983, 723(2), 276-286.

Kleinfeld, K.; Okamura, M. Y.; Feher, G. Biochemistry, 1984, 23, 5780-5786.
Kirmaier, C.; Holten, D. Photosynth. Res. 1987, 13, 225-260.
Kreszowski, D. H.; Deinum, G.; Babcock, G. T. Biophys. J. 1994, 66(2), a374. ‘

Kreszowski, D. H.; Deinum, G.; Babcock, G. T. J. Am. Chem. Soc. 1994, 116,
7463-7464.

(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

38

Deinum, G.; Kreszowski, D. H.; Chang, C. K.; Babcock, G. T. submitted to J. Am.
Chem. Soc.

(8) Reed, R. A; Purrello, R; Prendegast, K.; Spiro, T. G. J. Phys. Chem. 1991,
95, 9720-9727. (b) Phillips, D. L.; Rodier, M-J.; Myers, A. B. Chem. Phys. 1993,
175, 1-12. (c) Rodgers, K. R; Su, L.; Subramanian, S.; Spiro, T. G. Biochemistry
1993, 32, 4547-4551. ((1) Myers, A. B.; Philips, D. L.; Ci, X.; Westerfield, C.;
Rodier, J-M Proc. SPIE-Int Soc. 1993, pp 1-12.

(a) Gustafson, T. L.; Iwata, R; Weaver, W. L. in Ultrafast Phenomena VII;
Harris, C. B.; Ippen, E. P.; Mourou, G. A.; Zewail, A H., Eds; Springer-Verlag:
New York, 1990 (b) Hayashi, H.; Kolaczkowski, S. V.; Noguchi, T.; Blanchard,
D.; Atkinson, G. H. J. Am. Chem. Soc. 1990, 112, 4664-4670. (c) Noguchi, T.;
Kolaczkowski, S.; Gartner, W.; Atkinson, G. H. J. Phys. Chem. 1990, 94, 4920-
4926. (d) Hayashi, H.; Brack, T. L.; Noguchi, T.; Tasumi, M.; Atkinson, G. H. J.
Phys. Chem. 1991, 95, 6797-6802. (e) Noguchi, T.; Hayashi, H.; Tasumi, M.;
Atkinson, G. H. J. Phys. Chem. 1991, 95, 3167-3172. (f) Iwata, K.; Hamaguchi,
H. Chem. Phys. Lett. 1992, 196, 462-468.

(a) Findsen, E. W.; Shelnutt, J. A.; Ondrias, M. R, J. Phys. Chem. 1988, 92, 307-
314. (b) Apanasevich, P. A.; Kruglijk, S. G.; Kvach, V. V.; Orlovich, V. A. in
Time Resolved Vibrational Spectroscopy V; Takahashi, H., Ed; Springer-Veflag:
Berlin, 1992. (c) Courtney, S. H.; Jedju, T. M.; Friedman, J. M.; Alden, R. G.;
Ondrias, M. R. Chem. Phys. Lett. 1989, 164(1), 39-43. (d) Alden, R. G.; Sparks,
L. D.; Ondrias, M. R; Crawford, B. A.; Shelnutt, J. A J. Phys. Chem. 1990, 94,
1440-1443.

(a) Terner, J.; Spiro, T. G.; Nagumo, M.; Nicol, M.; El-Sayed, M. A. J. Am.

Chem. Soc. 1980, 102, 3238-3239. (b) Terner, J.; Strong, J. D.; Spiro, T. G.;
Nagumo, M.; El-Sayed, M. A. Proc. Natl. Acad Sci. 1981, 78, 1313-1317. (c)
Nagumo, M.; Nicol, M.; El-Sayed, M. A. J. Phys. Chem. 1981, 85, 2435-2438. (d)
Hseih, C.-L.; Nagumo, M.; Nicol, M.; El-Sayed, M. A. J. Phys. Chem. 1981, 85,
2714-2717. (e) Hseih, C.-L.; El-Sayed, M. A.; Nicol, M.; Nagumo, M.; Lee, J. H.
Photochem. Photobiol. 1983, 38, 83-94. (f) El-Sayed, M. A. Pure Appl. Chem.
1985, 57, 187-193.

(8) Reid, R J.; Doig, S. J.; Wickham, S. D.; Mathies, R. A J. Am. Chem. Soc.
1993, 115, 4754-4763. (b) Reid, P. J .; Lawless, M. K.; Wickham, S. D.; Mathies,
R A J. Phys. Chem. 1994, 98, 5597-5606.

Weaver, W. L.; Huston, L. A.; Iwata, K.; Gustafson, T. L.; J. Phys. Chem. 1992,
96, 8956-8961.

Butler, R. M.; Lynn, M. A.; Gustafson, T. L. J. Phys. Chem. 1993, 97, 2609-2617.

(47)
(43)

(49)

(50)
(51)
(52)
(53)
(54)
(55)
(56)

(57)

(53)

(59)

(60)

(61)
(62)

(63)
(64)
(65)

(66)

39
Ishibashi, T.; Hamaguchi, H. J. Spectrosc. Soc. Jpn. 1990, 40, 222-226.

Hamaguchi, H.; Iwata, K. Chem. Phys. Lett. 1993, 208, 465-470.

Gustafson, T. L.; Roberts, D. M.; Chemofi’, D. A. J. Chem. Phys. 1983, 79, 1559-
1564.

Hamaguchi, H.; Kato, C.; Tasumi, M. Chem. Phys. Lett. 1983, 100, 3-7.

Iwata, K.; Yamaguchi, S.; Hamaguchi, H. Rev. Sci. Instrum. 1993, 64, 2140-2146.
Kato, C.; Hamaguchi, H.; Tasumi, M. Chem. Phys. Lett. 1985, 120, 183-187.
Reid, P. J.; Doig, S. J.; Mathies, R. A. J. Phys. Chem. 1990, 94, 8396-8399.

Reid, P. J.; Doig, S. J.; Mathies, R. A Chem. Phys. Lett. 1989, 156, 163-168.
Reid, P. J.; “fickham, S. D.; Mathies, R. A. J. Phys. Chem. 1992, 96, 5720-5724.
Reid, P. J.; Shreve, A P.; Mathies, R. A. J. Phys. Chem. 1993, 97, 12691-12699.

Reid, P. J.; Doig, S. J.; Wickham, S. D.; Mathies, R. A. J. Am. Chem. Soc. 1993,
115, 4753-4763.

Woodward, R B.; Hoffmannn, R. in The Conservation of Orbital Symmetry;
Verlag Chemie International: Weinheirn, 1981, 3rd edition.

(8) Goldfarb, T. D.; Linquist, L. J. Am. Chem. Soc. 1967, 89, 4588. (b) Datta, P.;
Goldfarb, T. D.; Boikess, R. S. J. Am. Chem. Soc. 1969, 91, 5429.

Hashimoto, H.; Koyama, Y.; Hirata, Y.; Mataga, N. J. Phys. Chem. 1991, 95,
3072-3076.

Kuki, M.; Hashimoto, H.; Koyama, Y. Chem. Phys. Lett. 1990, 165, 417-422.

Hayashi, H.; Noguchi, T.; Tasumi, M.; Atkinson, G. H. Biophys. J. 1991, 60, 252-
260.

Hayashi, H.; Tasumi, M.; Atkinson, G. H. Chem. Phys. Lett. 1991, 178, 388-392.
Birge, R R Annu. Rev. Biophys. Bioenerg. 1981, 10, 315.
Smith, S. 0.; Lugtenburg, J .; Mathies, R. A. J. Membrane Biol. 1985, 85, 95.

Stockenheis, W.; Bogomolni, R. A. Annu. Rev. Biochem. 1982, 51, 587.

(67)

(63)

(69)

(70)

(71)

(72)

(73)

(74)

(75)

(76)

(77)

4o

Mathies, R. A; Lin, S. W.; Ames, J. B.; Pollard, W. T. Annu. Rev. Biophys. Chem.
1991, 20, 491-518.

Atkinson, G. H.; Brack, T. L.; Blanchard, D.; Rumbles, G. Chem. Phys. 1989,
131, 1-15.

(8) Smith, S. 0.; Pardoen, J. A.; Lugtenberg, J .; Mathies, R A. J. Phys. Chem.
1987, 91, 804-819. (b) Mathies, R. A.; Brito-Cruz, C. H.; Pollard, T. W.; Shank,
C. V. Science 1988, 240, 777-779. (c) Doig, S. J.; Reid, P. J .; Mathies, R. A. J.
Phys. Chem. 1991, 95, 6372-6379.

Findsen, E. W.; Scott, T. W.; Chance, M. R; Friedman, J. M.; Ondrias, M. R J.
Am. Chem. Soc. 1985, 107, 3355-3357.

Dasgupta, S.; Spiro, T. G.; Johnson, C. K.; Dalickas, G. A.; Hochstrasser, R. M.
Biochemistry 1985, 24, 5295-5297.

Rousseau, D. L.; Argade, P. V. Proc. Natl. Acad Sci. USA 1986, 83, 1310-1314.

Rodriguez, J.; Kirmaier, C.; Holten, D. W. J. Am. Chem. Soc. 1989, 111, 6500-
6501.

Lingle, R. J.; Xu, X.; Zhu, H.; Yu, S. 0; Hopkins, J. B.; Straub, K. D. J. Am.
Chem. Soc. 1991, 113, 3992-3994.

Lingle, R J.; Xu, X.; Zhu, H.; Yu, S. 0; Hopkins, J. B. J. Phys. Chem. 1991, 95,
9320-9331.

Petrich, J. W.; Martin, J. L.; Houde, D.; Poyart, C.; Orszag, A. Biochemistry 1987
26, 7914.

(a) Walters, V. A.; de Paula, J. C.; Babcock, G. T.; Leroi, G. E. J. Am. Chem. Soc.
1989, 111, 8300-8302. (b) Nam, H. H.; Walters, V. A.; de Paula, J. C.; Babcock,
G. T.; Leroi, G. E. In Proceedings of the filth Intemational Conference on
Raman Spectroscopy; Durig, J.R; Sullivan, J .F., Eds; Wiley and Sons: New
York, 1990, p 618. (c) Reed, R A.; Purello, R; Prendegast, K.; Spiro, T. G. J.
Phys. Chem. 1991, 95, 9720-9727. (d) Sato, S.; Asano-Someda, M.; Kitagawa, T.
Chem. Phys. Lett. 1992, 189, 443-447. (e) Kumble, R; Hu, 8.; Loppnow, G. R;
Vitols, S. E.; Spiro, T. G. J. Phys. Chem. 1993, 97, 10521-10523. (f) Kreszowski,
D. H.; Babcock, G. T. Biophys. J. 1993, 64, a276. (g) Bell, S. E. J.; Al-Obaidi, A
H. R; Hegarty, M.; Hester, R. E.; McGarvey, J. J. J. Phys. Chem. 1993, 97,
11599-11602.

(73)

(79)

(80)

(81)

(32)

(83)
(34)

(35)

(86)

(37)

(33)

41

(a) Sarma, Y. A. Spectrochimica Acta. 1989, 458(6), 649-652. (b) Oertling, W.
A; Salehi, A.; Chang, C. K.; Babcock, G. T. J. Phys. Chem. 1989, 93, 1311-1319.

Procyk, A. D.; Bocian, D. F. Annu. Rev. Phys. Chem. 1992, 43, 465-496.

(a) Abe, M.; Kitagawa, T.; Kyogoku, Y. J. Chem. Phys. 1978, 69, 4526-4534. (b)
Li, X.-Y.; Czernuszewicz, R S.; Kincaid, J. R; Stein, P.; Spiro, T. G. .1. Phys.
Chem. 1990, 94, 47-61.

(a) Tait, C. D.; Holten, D.; Barley, M.; Dolphin, D. D.; James, B. R. J. Am. Chem.
Soc. 1985, 107(7), 1930-1934. (b) Apanasevich, P. A.; Chirvonyi, V. S.; Kruglik,
S. G.; Kvach, V. V.; Orlovich, V. A. Proc. SPIE-Int. Soc. Opt. Eng. 1991,
1403(Laser Appl. Life Sci, Pt. 1) pp 195-211. (c) Apanasevich, P. A.; Chirvonyi,
V. S.; Kruglik, S. G.; Kvach, V. V.; Orlovich, V. A. Proc. SPIE-Int. Soc. Opt.
Eng. 1991, 1403(Laser Appl. Life Sci., Pt. 1) pp 240-242. (d) Loppnow, G. R;
Melamed, D.; Leheny, A. R; Hamilton, A. D.; Spiro, T. G. J. Phys. Chem. 1993,
97(35), 8969-8975. (e) Gentemann, S.; A;baneze, J .; Garcia-Ferrer, R; Knapp, S.;
Potzena, J. A.; Schugar, H. J .; Holten, D. J. Am. Chem. Soc. 1994, 116(1), 281-
289.

Irvine, M. P.; Harrison, R. J .; Strahand, M. A.; Beddard, G. S. Ber. Bunsenges.
Phys. Chem. 1985, 89, 226-232.

Gouterrnan, M. J. Mol. Spectroscopy 1961, 6, 138-163.
Sekino, H.; Kobayashi, H J. Chem. Phys. 1987, 86(9), 5045-5052.
Spiro, T. G.; Czernuszewicz, Li, X.-Y. Coord Chem. Rev. 1990, 100, 541-571.

(a) Spaulding, L. D.; Chang, C. 0; Yu, N.-T.; Felton, R H. J. Am. Chem. Soc.
1975, 97,2517-2524. (b) Choi, S.; Spiro, T. G.; Langry, K. C.; Smith, K. M. J.
Am. Chem. Soc. 1982, 104, 4337-4344. (c) Choi, S.; Spiro, T. G.; Langry, K. C.;
Smith, K. M.; Budd, L. D.; Lamar, G. N. J. Am. Chem. Soc. 1982, 104, 4345-
4353.

(a) Czernuszewicz, R. S.; Macor, K. A.; Li, X.-Y.; Kincaid, J. R; Spiro, T. G. J.
Am. Chem. Soc. 1989, 111, 3860-3869. (b) Pemg, J.-H.; Bocian, D. F. J. Phys.
Chem. 1992, 96, 4804-4811.

(a) Salehi, A.; Oertling, W. A.; Babcock, G. T.; Chang, C. K. J. Am. Chem. Soc.
1986, 108, 5630-5631. (b) Oertling, W. A; Salehi, A.; Chung, Y.-C.; Leroi, G. E.;
Chang, C. K.; Babcock, G. T. .1. Phys. Chem. 1987, 91, 5887-5898. (c) Salehi, A;
Oertling, W. A; Babcock, G. T.; Chang, C. K. Inorg. Chem. 1987, 26, 4296-
4298.

42

(89) (a) Donohoe, R J.; Duchowski, J. K.; Bocian, D. F. J. Am. Chem. Soc. 1988, 110,
6119-6124 (b) Czemuszewicz, R. S.; Macor, K. A.; Li, X.-Y.; Kincaid, J. R;
Spiro, T. G. J. Am. Chem. Soc. 1989, 111, 3860-3869. (c) Macor, K. A.;
Czernuszewicz, R S.; Spiro, T. G. Inorg. Chem. 1990, 29, 1996-2000. ((1)
Duchowski, J. K.; Bocian, D. F. J. Am. Chem. Soc. 1990, 112, 3312-3318. (e)
Pemg, J.-H.; Duchowski, J. K.; Bocian, D. F. J. Phys. Chem. 1990, 94, 6684-
6691. (t) Duchowski, J. K.; Bocian, D. F. Inorg. Chem. 1990, 29, 4158-4160. (g)
Duchowski, J. K.; Bocian, D. F. J. Am. Chem. Soc. 1990, 112, 8807-8811.

(90) (a) Gurinovich, G. P.; Gurinovich, I. F .; Ksenofontova, N. M.; Terekhov, S. N. J.
Appl. Spectrosc. 1985, 43, 758-763. (b) Teraoka, J .; Hashimoto, S.; Sugimoto, H.;
Mori, M.; Kitagawa, T. J. Am. Chem. Soc. 1987, 109, 180-184. (c) Donohoe, R
J.; Atamian, M.; Bocian, D. F. J. Am. Chem. Soc. 1987, 109, 5593-5599. ((1)
Gurinovich, G. P.; Gurinovich, I. F.; Ivashin, N. V.; Sinyakov, G. N.; Shulga, A
M. J. Mol. Struct. 1988, 172, 317-343. (e) Atamian, M.; Donohoe, R. J .; Lindsey,
J. S.; Bocian, D. F. J. Phys. Chem. 1989, 93, 2236-2243. (1) Reed, R. A. Purrello,
R; Prendergast, K.; Spiro, T. G. J. Phys. Chem. 1991, 95, 9720-9727.

CHAPTER 2

INSTRUMENTATION

ABSTRACT

Time-resolved resonance Raman techniques applied to electronically excited
porphyrins has thus far been limited mostly to time-resolution in the nanosecond range,
and, correspondingly, the techniques that have been used typically involve low repetition
rate pulsed lasers. The research project described here extends time-resolved resonance
Raman to the picosecond regime, and opens up the ability to probe more dynamic
processes, such as vibrational relaxation in electronically excited molecules. This project
involved taking advantage of the high repetition rate characteristics of picosecond lasers
described in the experimental apparatus section. A rapidly flowing glass free jet cell
similar to that developed by Atkinson et. al. (Atkinson, G. H.; Brack, T. L.; Blanchard,
D.; Rumbles, G. Chem. Phys. 1989, 131, 1-15) was used in conjunction with a microscope
objective to refresh the sample between individual flashes in a series of laser pulses.
Implementation of a “quasi-simultaneous” technique analogous in purpose to that of Sato
et. a1. (Sato, S.-I.; Karnogawa, K.; Aoyagi, K.; Kitagawa, K. J. Phys. Chem. 1992, 96,
10676-10681) was useful in dealing with the difference spectra associated with time-
resolved techniques. Spectrum Basic programs were written to process the data

accumulated through this technique.

43

44
I. EXPERINIENTAL APPARATUS

The experimental apparatus used for picosecond transient experiments is shown
schematically in Figure 2.1. An actively mode-locked Q-switched Coherent 76-8 Nd:YAG
laser with built-in second harmonic fi’equency-mixing capabilities uses a standard KTP
crystal to generate 70 psec firll-width-half-maximum, 76.5 MHz, 532 nm second harmonic
pulses. The output is directed through a Coherent third harmonic generator by using a
BBO crystal that mixes 1064 nm and 532 nm wavelengths to generate 355 nm. The 532
nm and 355 nm pulses are directed into two identical Coherent 702 tunable pulse dye
lasers with variable tuning ranges dependent upon dye selection. Photoacoustically driven
cavity dumpers on the 702 dye lasers are enabled by Coherent 7200 cavity dumper drivers
that permit selection of variable pulse repetition rates (38.25 MHz divisible by any integer
between 4 and 259). One Coherent 702 dye laser initially used rhodamine 6G (R6G)
dissolved in ethylene glycol and permitted a tuning range of 570-630 nm. Recent
characterization of pyrromethene laser dyes for ultrafast pulse generation suggests
exceptionally eflicient power generation.1 For this reason, in later experiments we
implemented pyrromethene 567, which was dissolved in propyl phenyl ether (PPH), and
has a tuning range of 550-600 nm. The other 702 dye laser used stilbene 420 dissolved in
ethylene glycol with tunability from 420-470 nm. All dyes and the PPH were purchased
from Exciton Corporation (PO. Box 3126 Overlook Station, Dayton, Ohio 45431).
Figure 2.2 displays relative efficiencies as a function of emission wavelength for the dyes
used in our time-resolved experiments. R6G and pyrromethene 567 were pumped with ~2

W of 532 nm, while the stilbene 420 was synchronously pumped with ~1 W of 355 nm.

Figure 2.1

45

Instrumentation used for time-resolved picosecond Raman
experiment SHG: second harmonic generator using KTP crystal;
THG: third harmonic generator using BBO crystal; Shutter driver:
homebuilt device to control mechanical shutters; PI/CCD:
princeton instruments charge coupled device detector; OPTICS:

a collimating lens (focus length = 2.5 inch, diameter = 2 inch),

and a focusing lens ( focus length = 12 inch, diameter = 2 inch).

DlCHROIC

 

DET.
CONTROLLE

Pl
LN/CCD
l 152UV

 

 

 

 

 

JOBIN YVON
HR 640

\/

 

 

 

46

 

COHERENT 76-S
MODE-LOCKED

 

 

 

 

 

 

 

 

 

 

 

 

NdYAG
SHG
1‘" 355 nm THG
M2 532 nm
M7 5 = M6
COHERENT COHERENT
702 702
MECHANICAL DYE DYE
' SHUTTERS LASER LASER
CAVITY
DUMPERS
M PUMP(SSO-600) 'j
L'l—PRbBl—ZMBO-WO) , 5
\ ‘ 5 :
CAVITY CAVITY
DUMPER DUMPER
DRIVER DRIVER

 

SAMPLE

 

 

HUTTER

S
DRlVER

 

 

 

 

 

 

386

 

PC/DOS

 

 

Figure 2.1

Figure 2.2

47

Laser dye efficiency curves for a) stilbene 420 in ethylene glycol
b) pyrromethene 567 in propyl phenyl ether (PPH), and

c) rhodamine 6G in ethylene glycol. Data were collected by
delivering 1 W 355 nm and 2 W 532 nm pulses fiom a Coherent
76-s Antares laser to a pair of Coherent 702 dye lasers with the
cavity dumper drivers operated at 3.825 MHz .

Relative efficiencies

0.18

0.16

0.14

0.12

0.10

0.08

0.06

0.04

48

ll

 

 

I

II

II

b)

 

l

 

 

 

450

II

550

Wavelength (nm)

Figure 2.2

600

650

49
The relative efficiency is defined as dye laser output power (mW) divided by pumping

power (mW). Readings were taken with a Coherent 210 power meter. It should be noted
that dye laser companies normally specify dye efiiciencies for continuous wave lasers,
whereas these efficiencies were taken at 3 .825 MHz and thus appear comparatively lower.
Brief comparisons with the dye lasers set in continuous wave mode agreed with
documented laser efficiencies. The “red” pulses generated from either R6G or
pyrromethene 567 act as the pump beam in promoting molecules into electronic excited
states. The “blue” pulses from the stilbene 420 were used as the probe beam and take
stokes Raman spectra of the resultant excited state. The pulse energy dependence upon
cavity dumper repetition rate for the 702 dye lasers is shown in Figure 2.3, where pulse
energy (in nJ) reaches a maximum and then plateaus around 1.27 MHz.

Two identical Inrad model 5-14B autocorrelators, (purchased from Coherent),
monitor the pulse shape coming out of the dye lasers with cross-correlation full-width-
half-maximum of 7 psec. The cross-correlation shape has the general appearance of a
gaussian profile.

The two beams were spatially overlapped at a dichroic filter (yellow subtractive,
520FD22-50S, 1.0 mm thickness) and directed into the sample compartment. The
dichroic filter was purchased fiom the Andover Corporation (4 Commercial Drive, Salem,
New Hampshire 03079), and its spectral characteristics are shown in Figure 2.4. The
temporal delay between the pump and probe beams can be controlled both electronically
through the cavity dumper drivers, and spatially through an optical delay line. The relative
temporal position of pump and probe beams can be altered in 26.4 or 13.2 ns steps using

electronic delay switches on either 7 200 Coherent cavity dumper driver. Finer delays can

Figure 2.3

50

Pulse energies from Coherent 702 dye lasers as a function of
repetition rate for a) pyrromethene 567 in propyl phenyl ether
b) rhodamine 6G in ethylene glycol and c) stilbene 420 in
ethylene glycol. The pulse energy dependence was done at the
efficiency curve maxima of each dye, with 2 W 532 nm and l W

355 nm pump pulses. The actual repetition rate equals
38.25 MHz/Repetition divisor.

Pulse energies (nJ)

Pulse energies (nJ)

120

100

8

20

51

 

 

 

 

 

 

 

 

I
I
‘ . a
A A A
l 1 n g l l I
100 150 200 250

Repetition divisor

Figure 2.3

 

Figure 2.4

52

Spectral response curves for optical filters 520DFD22-50S and
505FD64-50S. The responsivity of the 520FD22-SOS is shown
at incident beam angles of 0° and 45° to display the angle
dependence of transmission.

%Transrn'ttanoe

53

    
   

1m r'
A - :‘r A Fl, /" “‘1.
50513364508 520111224508 Art?
SHORT PASS
FILTER

%T'ansm'ttanoe
8 . 8

 

., .,
400 450 500 550

Wavelength (nm)

Figure 2.4

Cit
fas
we

all

an

trar

Was
15 Si
deia

by t]

54
be achieved through directing the pump beam down an optical delay line controlled by a

Klinger CC1.1 programmable motor-driven translation stage (Klinger Scientific, Garden
City, New York 11530). The time delay was monitored by sending the two beams into a
fast Hewlett-Packard 4220 photodiode detector with a rise time <1 11sec. The two pulses
were distinguished on a Tektronix DSA 602A Digitizing Signal Analyzer scope with
averaging and difference detection capabilities, and real time accumulation up to 1 GHz.
When finer time resolution near zero-time overlap was desired, it was obtained by
a technique we loosely term “zero-time measurements”. Essentially, we use a simplified
transient absorption experiment in which we pump a molecule into an electronically
excited state with a “red” pulse, and monitor the intensity of the “blue” pulse. This blue
pulse is absorbed by the time-dependent electronic excited state but not by the electronic
ground state. Initially, the experimental apparatus used was that described in Figure 2.5,
and DODC iodide was used as a reference sample (purchased from Exciton). DODC
iodide is a commonly used laser dye that has strong ground state optical absorption at 580
nm and broadband excited state absorption around 450 nm, similar to the simulated plot in
Figure 2.6. Its excited state lifetime is ~ 2 nsec.2 The two spatially overlapped beams are
tightly focused onto the sample, the pump beam intensity is filtered out, and the probe
beam intensity is directed into the Hewlett-Packard 4220 photodiode. The short pass filter
was purchased fi'om Andover (505FD64-50S, blue dichroic), and its spectral dependence
is shown in Figure 2.4. The probe pulse intensity can be monitored as a function of pump
delay on the optical delay line. A sudden decrease in probe intensity indicated absorption.

by the excited state and zero—time overlap. In later experiments, we monitored zinc(II)

55

Figure 2.5 Experimental setup initially used in determining zero-time
overlap through a modified transient absorption experiment.

56

SIGNAL

 

TEKTRONIX

 

DSA

 

HP 4220
PHOTODIODE

 

 

 

 

OSCILLSCOPE

 

 

 

 

COHERENT 702

TRIGGER CAVITY DUMPER

 

 

SANDOVER FILTER l

505FD64-5OS

 

 

 

5X _
MICROSCOPE
OBJECTIVE

 

 

. O
----‘

PUMP&PROBE
BEAMS

Figure 2.5

DRIVER

 

 

 

57

Figure 2.6 DODC iodide transient absorption. a) Simulated absorption
spectra representing optical behavior of DODC iodide.
b) Time-averaged voltage pulse of “blue” laser showing intensity
decrease when in temporal overlap with “r ” pulses
(as seen on Tektronix DSA 602 oscilloscope).

Relative Intensity

1))

58

 

 

excited state

 

j

l l l A l l

J

 

 

gL

5m 55) HI)
Wavelength (am)

2‘6"

7(1)

 

    

NOT OVERLAPPED

, TEMPORAL &
SPATIAL OVERLAP

 

 

 

 

 

Figure 2.6

59
octaethylporphyrin through either a similar transient absorption setup or the actual time-

resolved Raman experiment to determine zero time. Each of these protocols performed
this zero time determination well, and enabled determination to within ~ 18 psec (270
counts on the Klinger programmable counter).

Scattering from the sample was collected through lenses designed to match the f/#
of 5.6 for the Jobin Yvon I-1R640 single grating monochromator. The monochromator
used a 2400 gr/mm grating and is of 0.64 meter focal length with a spectral resolution of
approximately 7 cm'1 under typical experimental conditions. A charge coupled device
(Princeton Instruments, model LN/CCD 1152UV) was used to detect the Raman
spectrum.

Absorption spectra were recorded prior to and following the Raman studies to
confirm the integrity of the samples. UV/V is spectra were recorded on a Perkin-Elmer
Lambda 5 spectrometer (resolution of 0.25 nm, :1: 0.3 nm wavelength accuracy, and stray
light of less than 0.001% at 370 nm).

Solvents used in these experiments were usually organics such as tetrahydrofuran
or pyridine, which were of reagent quality, purchased fiom Aldrich, and purified by
distillation, if necessary. The Raman spectra were observed through 90 degree scattering
and the sample was replenished by using a rapid flowing liquid jet system that was kept at

a constant temperature by cooling the sample reservoir either in or above dry ice.

60
II. FLOWING LIQUID JET ASSEMBLY

Selecting an appropriate technique for monitoring time-resolved Raman
spectroscopy of liquid solutions can be critical. Several approaches exist. Dual-beam
flow experiments use continuous wave lasers which monitor intermediates on microsecond
and millisecond time scales.3 Nanosecond pulsed lasers operating at 10 Hz have also been
used for similar delay regions."5 These approaches do not have the time resolution we
desire, which can only be offered by picosecond lasers. These lasers can function at 50 Hz
or near KHz repetition rates when used with various amplifiers."7 Other research groups
have neglected the use of low repetition rate amplifiers and retained the MHz repetition
rates of picosecond lasers.8 We have selected to omit low repetition rate amplification and
use the megahertz repetition rates and nanojoule pulse energies which are intrinsic in
cavity-dumped picosecond dye lasers synchronously pumped by a mode-locked cw
Nd:YAG laser. High repetition rate lasers offer several advantages over comparable low
repetition rate systems9 The average power available is sufficient to generate resonance
Rarnan from intermediates with good signal-to-noise without the high pulse energies
associated with multiphoton processes. The high repetition rates utilized provide
improved signal averaging capabilities as long as the sample is not exposed to multiple
excitations by consecutive pulses. The stability of the picosecond laser system can be
optimized for high repetition rate, low peak power operation.

For time-resolved experiments on singlet and triplet electronic states of porphyrins
it is essential that the ground state does not bleach through multiple excitations by

consecutive pulses. To avoid multiple excitations we replenished the sample between

61
pairs of pump-probe pulses by using a small irradiated volume and a rapidly flowing

sample. A high speed rotary pump (Gelber Industries (1136 Lloyd Rd, Wickliffe, Ohio
44092), Micropump 415 motor with a 120 magnet drive gear pump) flowed sample at a
rate of ~1 .0 ml/sec through a 300 um diameter glass nozzle (Figure 2.7). This flow rate
was used since the jet quality for our chosen organic solutions were acceptable at this rate
and this rate was sufficient to avoid multiple excitations when the irradiated volume is
small enough. To reduce the volume of sample optically pumped and sampled, both laser
beams were spatially overlapped at a dichroic lens and then focused on the sample by
using a 5X microscope objective (569-199, by W. Nuhsbaum, 3918 West Main Street,
Mchenry, Illinois 60050, achromatically corrected at 546.1 nm, with a working distance of

6.2 mm and depth of focus ~280 um). The beam diameter of this microscope objective

can be measured by several techniques. 1° The knife-edge technique was selected for its
ease of application. "

The knife edge experimental setup is shown in Figure 2.8a. A razor blade was
attached to a differential micrometer and translated horizontally across the laser beam
spot. Resultant beam intensities (read in mV) were monitored at incremental positions by
using a Tektronix DSA 602 oscilloscope and a Hewlett-Packard 4420 photodiode. The
pulse intensity is plotted versus relative translation distance in Figure 2.8b. The pulse
intensity of the unobstructed beam (10), drops rapidly once the razor beam translates into

the beam pathway. The beam waist ((110) is determined by measuring the horizontal

translation difl‘erence between the 12% lo and 86.4% 10 distances (the He2 values). ' "‘2

Plots were generated with the razor blade at several different vertical positions so as to

62

Figure 2.7 Flowing jet cell assembly used in high repetition rate
picosecond Raman experiments.

GLASS NOZZLE

 

 

 

 

MOUNT

 

\ 7300um
NOZZLE ........ 0..

63

 

 

 

 

 

 

 

 

 

 

 

CATCHER
TUBE

 

 

 

7‘ 5x _

MICROSCOPE
OBJECTIVE

.....
...... s
0.- Q.‘

 

 

 

 

” \\

l l
‘.. -O’

 

 

 

 

 

 

 

 

RESERVOIR

Figure 2.7

 

 

 

 

 

 

 

 

SAMPLE \j

 

Figure 2.8

64

Knife-edge technique used for determining beam waist size.

a) Experimental setup used with 5X microscope objective(569-199).
b) Plots of beam intensity versus changes in razor edge translation
taken at four vertical positions. The vertical positions are listed

in reference to the position that gave the minimal beam waist

value of 13 um.

65

 

 

 

1
HP 4220
PHOTODIODE

SIGNAL JTEKTRONIX
, , DSA
I OSCILLSCOPE

 

 

5

ANDOVER FILTER
505FD64-50S

 

 

| COHERENT 702 7
CAVITY DUMPER

I Z

 

 

   

 

PUMP&PROBE
BEAMS

 

 

 

 

Figure 2.8

66

ensure convergence of the beam waist to the true beam diamter. This was reflected by the
strongest slope, which gave a measured beam waist value of ~ 13 um.

With this beam diameter the flow rates necessary for sample regeneration at

individual repetition rates can be calculated and are listed in Table 2.1.

 

 

 

 

 

 

 

Table 2.1 Flow Rate Requirements for Sample Replenishment.
+N Repetition Rate Pulse Train Gap meters/sec milsec at 100%
sample replenishment
10 3.825 MHz 261.4 nsec 38.25 2.704
30 1.275 MHz 783.8 nsec 12.75 0.901
40 0.956 MHz 1.051 usec 9.560 0.676
50 0.765 MHz 1.305 psec 7.650 0.541
100 382.5 KHz 2.614 usec 3.825 0.270
200 191.2 KHz 5.228 usec 1.913 0.135

 

 

 

 

 

 

 

Application of rapid flowing jet systems to spectroscopy requires consideration not
only of sample replenishment between pulses but also inherent restrictions on time delay
measurements owing to sample movement between the initial excitation by the pump pulse
and the delayed arrival of the probe pulse. Calculations in Table 2.2, using the typical
experimental flow rate of ~1 .0 ml/sec, show the feasibility of time-resolved experiments

out to a time delay of about 350 nsec.

67

Table 2.2 Flowing Jet Restrictions on Temporal Delay.

time between and nsec %Same
0. 1 99.99
0.8 99.89
10 98.58

26.4 96.26
50 92.92
200 71.70
350 ' 50.47

 

After establishing the presence of singlet and triplet excited electronic state spectra
with the glass nozzle, we alternatively used a Spectra-Physiks dye laser nozzle. This
nozzle had a rectangular opening 200 um by 2 mm which brings the liquid to a focus ~1.5
cm beyond the nozzle opening. We originally were concerned about possible lack of
sample replenishment between pulse pairs, but failed to observe any presence of
photoaccumulated product at the flow speeds and repetition rates used for our
experimentation (normally 1.27 MHz and ~1.0 ml/sec). The rectangular flowing cell gave
'us some advantages in minimizing background scattering fi'om surface reflections at the
air-liquid interface.

Operating with repetition rates of 1.27 MHz or less, pulse energies of the pump
and probe lasers were typically ~45 n] and ~15 nJ at the sample. The corresponding peak
powers are ~6.5 KW and ~2.1 KW. Under the experimental conditions that involved high
repetition rates lasers and a tightly focusing microscopic objective, as described above, the
flux densities are ~55 mJ/cm2 for the pump and ~ 20 mJ/cm2 for the probe. These flux
densities, particularly for the pump, are comparable with other published values for time-
resolved Raman spectroscopy on porphyrin-like molecules. ‘3 Under these conditions it is
reasonable to carry out a calculation to determine the number of molecules that are

excited, and to calculate what percentage of molecules are excited. Figure 2.9 shows that

68

Figure 2.9 Molecular excitations expected under rapid flowing jet setup.

69

 

l

 

 

300nm
13pm

 

 

 

 

\/

Under typical experiment conditions of pulse energy 40 ml and Aexc = 572 nm.

= ((Aexc)'(10)'(1'1O'SClWhC

= 572.10-9m . 40.10-91. 1-10-(30mM‘lcm'1)-(O.5mM)-(300-10'4cm)}
(6.6256-10-34J-s)-(3.00-108m/s)

   

 

= 7.43-1010 photons]

 

# MOLECULES = Volume enclosed - Concentration - Avagadro’s number

It-(beam radius)2-(l)-(C)-(NA)

= 1r.(6.5-10'5dm)2-(300-10'5dm)-(5-10'4M)-(6.022-10‘23)

 

=l1.20-1010 molecules

 

Figure 2.9

70

under the typical experimental parameters of the rapid flowing jet experiment there are
~10” molecules enclosed within the laser beam. A sufficient number of photons are
present in each laser pulse to excite virtually all molecules even at higher porphyrin

concentrations.

III. QUASI-SIMULTANEOUS ACCUMULATION

Even under optimal conditions, transient Raman spectra are dominated by ground
state and solvent features. Subtraction of ground state and solvent features ideally leaves
only transient features. However, for the long time exposures required for pump/probe
experiments that involve difference spectra determination, instabilities contribute noise to
the difference spectrum. Laser power fluctuations and temperature differences in optical
components become significant and reduce sensitivity and reproducibility of experimental
data. 1‘ A technique has been applied that addresses these problems and improves signal-
to-noise in the difference spectrum. A rapid series of short time sample accumulations are
subsequently summed up to generate the long time exposure, in an approach similar to
that applied by Sato et. al.,15 and called the “quasi-simultaneous” technique. ‘6'” As
illustrated in Figure 2.10, two external mechanical shutters (Uniblitz VS14 modelsfrom
AW. Vincent Associates, 125 5-T University Avenue, Rochester, New York 14607, with
total opening time 4.0 msec, closing time 3.0 msec, and capable of withstanding laser
energy up to 10 W/mmz), are placed in the pathway of the pump and probe beams. These
shutters are driven by a device constructed by the Michigan State Chemistry Department.
electronics shop and appropriately called the shutter driver. The schematic of the shutter

driver is given in Figure 2.11.

71

Figure 2.10 Simultaneous method implementing mechanical shutters
(SI and 811) within laser beam pathways.

 

72

 

 

 

 

SAMPLE
SHUTTER l '
DRIVER

 

 

 

 

 

 

 

trlgger
EROBE -='~:-2

DETECTOR
CONTROLLER

 

 

 

 

 

 

 

 

 

 

 

 

 

 

‘1
{ PUMP L—l

Figure 2.10

¥ +

 

 

 

73

Figure 2.11 Schematic of shutter driver. a) voltage transformer.
b) analog switching permitting shutter operation.

74

a)

BLK 36V 50VDC

‘_\/\/

 

 

 

 

 

 

120VAC
W.T. 5VDC
4 To
5.1.3
b)
SHUTTER #1
OPEN
CLOSE.
TI'L
'I'I‘L VCC
INPUT J
G i: CL
ICl-A

 

 

 

up

 

 

Figure 2.11

75
The shutter driver transforms incoming 120 VAC into 50 VDC and 5 VDC. The 5

VDC supplies two IC analog voltage switches while the 50 VDC is used to power the two
mechanical shutters.

A TTL pulse is inverted prior to entering ICl-A. The pulse is alternately
transmitted to ICl-B, an identical analog switch in series with IC1-A. The alternate pulse
fiom ICl-A also actuates transistor 1 (T1), which permits opening of shutter #1. ICl-B
performs a similar function for transistor 2 (T2) and shutter #2. Since IC l-B receives a
TTL pulse only every other cycle, the net result is that shutter #2 opens and closes half as
often as shutter #1. Thus, the shutter driver permits four possible combinations of shutter
switching, and these switchings are regulated by a 'I'I‘L input pulse. This TTL pulse
originates fiom the NOTSCAN outlet from the Princeton Instruments ST130 controller,
and changes with every closing of the internal mechanical shutter associated with the CCD
detector. The shutter driver is sequenced to produce four partitions of time intervals, such

as shown in Figure 2.12, which can be sequenced with the CSMA software.

SHUTTER #1 (PROBE) OPEN |‘__| m
CLOSE . . .

SHUTTER #2 (PUMP) OPEN I

 

CLOSE

 

1 [1111 IV

Figure 2.12 Partitions of time intervals generated by shutter driver.

76

These time intervals can be varied, but typically were run between 0.5-3 minutes.
Partition I has both shutters open and exposes the sample to both pump and probe beams.
Partition II has one shutter closed and consequently records the pump only spectrum. In
partition III the other shutter closes while the first one reopens and the probe only
spectrum is recorded. Partition IV has both shutters closed and dark counts are recorded.

Post data processing subsequently divides the accumulated spectra into
pump&probe, pump, probe, and dark spectra. The software associated with Princeton
Instuments detection systems is called CSMA (113.0), and uses a programming language
called Spectrum Basic, which has written programs that perform the operations essential
to data acquisition, spectral processing, image processing, and reviewing acquired data. ‘8
It was necessary to write three additional programs to assist in processing of data
accumulated through the “quasi-simultaneous” technique. These programs are written in
Spectrum Basic and are listed in Appendix A.

Initially a program called Separatepro was created to process spectra acquired by
using only one external mechanical shutter. The standard procedure is to first collect
continuously a series of short exposures independently stored into an even number of
frames selected by prompts from the CSMA software. The Separate. pro is a post-
processing program that separates even frames from odd frames, then adds up and stores
the individual frames into two newly created data files.

Quartilespro is a similar post-processing program that enables the use of two
external mechanical shutters and permits differentiation of consecutive data fi'ames into
four separate data files. The only restriction is that the total number of selected fiames
must be an integral number of four.

Additionally, it was useful to write a program that can edit the occasional bad
frame in an otherwise good series of spectra. Frmeditpro is a modification of Replaypi
which will perform this function. ‘8 The usual sequence after starting this program is to

first type “Alt-M” to assume manual control over replay of frames in the unprocessed

spa

and

I613
COll

sim'

pun

00H
lllCl

prol

ene-
exp
Wel
cha

Mic

77

spectra. Ifone wishes to save the current frame displayed, one types “3” to save. The up
and down arrows, or “Pgup” or “Pgdn”, will move the displayed spectra to subsequent
flames. The “3” key must be hit for each frame you wish to store, and it is important to
retain the proper sequence of frames stored. When editing is finished, typing “Esc”
completes the process and terminates the program. The application of the “quasi-
sirnultaneous” method, along with some post-data processing, was found to be especially
useful in taking difi‘erence spectra.

This combination of features, a tunable pair of high repetition rate synchronously
pumped picosecond pulsed dye lasers, a flowing liquid cell that adequately refreshes the
sample, and a data processing technique that minimizes fluctuations, represents the first
combination of these techniques applied to studying Raman spectroscopy of
metalloporphyrins. The dye lasers offer broad selectable wavelengths for both pump and
probe beams. Time resolution of intermediates can be applied to within ~18 picoseconds.
The post-processing programs in conjunction with the high average powers and low pulse
energies of the dye lasers permit favorable signal-to-noise levels. Completion of this
experimental setup provides the flexibility to study shorter lived electronic species,19 as
well as a wider range of processes including Vibrational relaxation. 1930'” The following
chapters detail some of the initial applications of picosecond Raman spectroscopy at

Michigan State University.

(1)
(2)

(3)

(4)

(5)

(6)

(7)

LIST OF REFERENCES

O’Neil, M. P. Optics Letters 1993, 18(1), 37-38.

(a) Schafer, F. P.; Schmidt, W.; Marth, K.; Appl. Phys. Lett. 1967, 24A, 280. (b)
Maeda, M.; Miyazoe, Y.; Jpn. J. Appl. Phys. 1972, 11(5), 692.

(a) Han, S.; Ching, Y-C.; Rousseau, D. L. J. Am. Chem. Soc. 1990, 112, 9446-
9451. (b) Ames, J. 3.; Rap, J.; Lugtenburg, J .; Mathies, R A. Biochemistry 1992,
31, 12546-12554. (c) Han, Y.; Ching, Y-C.; Rousseau, D. L. Nature, 1992, 348,
89-90. (d) Ogura, T.; Takahashi, S.; Hirota, S.; Shinzawa-Itoh, K.; Yoshikawa, S.;
Appelman, E. H.; Kitagawa, T. J. Am. Chem. Soc. 1993, 115, 8527-8536.

(a) Findsen, E. W.; Shelnutt, J. A.; Ondrias, M. R. J. Phys. Chem. 1988, 92, 307-
314 (b) Kaminaka, S.; Kitagawa, T. J. Am. Chem. Soc. 1992, 114, 3256-3260. (d)
Bell, S. E. J .; Al-Obaidi, A. H. R; Hegarty, M.; Hester, R E; Mcgarvey, J. J. J.
Phys. Chem. 1993, 97, 11599-11602.

(a) Babcock, G. T.; Jean, J. M.; Johnston, L. N.; Palmer, G.; Woodruff, W. H. J.
Am. Chem. Soc. 1984, 106, 8305-8306. (b) Walters, V. A; dePaula, J. C.;
Babcock, G. T.; Leroi, G. E. J. Am. Chem. Soc. 1989, 111, 8300-8302. (c)
Varotsis, C.; Babcock, G. T. Biochemistry 1990, 29(32) 7357-7361. (d) Varotsis,
C.; Woodrufl‘, w. H.; Babcock, G. T. .1. Biol. Chem. 1990, 265(19), 11131-11136.
(e) Varotsis, C.; Zhang, Y.; Appelman, E. H.; Babcock, G. T. Proc. Natl. Acad
Sci. USA 1993, 90, 237-241.

(a) Courtney, S. H.; Jedju, T. M.; Friedman, J. M.; Alden, R. G.; Ondrias, M. R
Chem. Phys. Lett. 1989, 164(1), 39-43. (b) Alden, R G.; Sparks, L. D.; Ondrias,
M. R; Crawford, B. A.; Shelnutt, J. A. J. Phys. Chem. 1990, 94, 1440-1443.(c)
Reid, R J .; Doig, S. J .; Wickham, S. D.; Mathies, R. A. J. Am. Chem. Soc. 1993,
115, 4754-4763. ((1) Reid, P. J.; Lawless, M. K.; Wickham, S. D.; Mathies, R A
J. Phys. Chem. 1994, 98, 5597-5606.

(a) Reed, R A; Purrello, R; Prendegast, K.; Spiro, T. G. J. Phys. Chem. 1991, 1
95, 9720-9727. (b) Phillips, D. L.; Rodier, M-J.; Myers, A. B. Chem. Phys. 1993,
175, 1-12. (c) Rodgers, K. R; Su, L.; Subramanian, S.; Spiro, T. G. Biochemistry
1993, 32, 4547-4551. (d) Myers, A. B.; Philips, D. L.; Ci, X.; Westerfield, C.;
Rodier, J -M Proc. SPIE-Int Soc. 1993, 1-12.

78

(8)

(9)

(10)

(11)

(12)

(13)

(14)
(15)

(16)

(17)

79

(a) Gustafson, T. L.; Iwata, R; Weaver, W. L. in Ultrafast Phenomena VII;
Harris, C. B.; Ippen, E. P.; Mourou, G. A.; Zewail, A H., Eds; Springer-Verlag:
New York, 1990. (b) Hayashi, H.; Kolaczkowski, S. V.; Noguchi, T.; Blanchard,
D.; Atkinson, G. H. J. Am. Chem. Soc. 1990, 112, 4664-4670. (c) Noguchi, T.;
Kolaczkowski, S.; Gartner, W.; Atkinson, G. H. J. Phys. Chem. 1990, 94, 4920-
4926. (d) Hayashi, H.; Brack, T. L.; Noguchi, T.; Tasumi, M.; Atkinson, G. H. J.
Phys. Chem. 1991, 95, 6797-6802. (e) Noguchi, T.; Hayashi, H.; Tasumi, M.;
Atkinson, G. H. J. Phys. Chem. 1991, 95, 3167-3172. (I) Iwata, K.; Hamaguchi,
H. Chem. Phys. Lett. 1992, 196, 462-468. (g) Weaver, W. L.; Huston, L. A.;
Iwata, K.; Gustafson, T. L.; J. Phys. Chem. 1992, 96, 8956-8961. (h) Butler, R.
M.; Lynn, M. A.; Gustafson, T. L. J. Phys. Chem. 1993, 97, 2609-2617.

Atkinson, G. H.; Brack, T. L.; Blanchard, D.; Rumbles, G. Chem. Phys. 1989,
131, 1-15.

(8) Yoshida, A.; Asakura, T. Opt. Laser T echnol. 1976, 8, 273. (b) Shayler, P. J .;
Appl. Opt. 1978, 17, 2673-2675. (c) Sasaki, A; Yamada, F.; Masui, T. Jpn. J.
Appl. Phys. 1978, 17, 1163-1164. (d) Dickson, L. D. Opt. Eng. 1979, 18, 70. (e)
Mccally, R. L.; Appl. Opt. 1982, 23, 2227-2229 (f) Liu, J. M.; Opt. Lett. 1982, 7,
196-198. (g) Kimura, S.; Munakata, C. App. Opt. 1988, 27, 84-88.

(a) Amaud, J. A.; Hubbard, W. M.; Mandeville, G. D.; de la Claviere, B.; Franke,
E. A.; Franke, J. M. Appl. Opt. 1971, 10, 2775. (b) Skinner, D. R; Whitcher, R
E. J. Phys. E. 1972, 5, 237. (c) Suzaki, Y.; Tachibana, A. Appl. Opt. 1975, 14,
2809. (d) Firester, A. H.; Heller, M. E.; Sheng, P. Appl. Opt. 1977, 16, 1971. (e)
Schneider, M, B.; Webb, W. W. Appl. Opt. 1981, 20, 1382.

Karim, M. A. in Electro-Optical Devices and Systems; Karim, M. A Ed; Kent:
Cambridge, 1990, pp 189-195.

(a) Sato, S-I.; Asano-Someda, M.; Kitagawa, T. Chem. Phys. Lett. 1992, 189,
443-447. (b) de Paula, J. C.; Walters, V. A.; Nutaitis, C.; Lind, J.; Hall, K. J. Phys.
Chem. 1992, 96(26), 10591-10594. (c) Kumble, R; Hu, 8; Loppnow, G.; Vitols,
S. E.; Spiro, T. G. J. Phys. Chem. 1993, 97, 10521-10523.

Diller, R; Stockburger, M. Biochemistry 1988, 27, 7641.

Sato, S-I.; Kamogawa, K.; Aoyagi, K.; Kitagawa, T. J. Phys. Chem. 1992, 96,
10676-10681. -

Nakagawa, M.; Maeda, A.; Ogura, T.; Kitagawa, T. J. Mol. Struct. 1991, 242,
221-234.

Muench, R; Schaack, G.;Vlf1nterfeldt, V. J. Phys. E. 1982, 15, 776-781.

(18)

(19)

(20)

(21)

80

See Princeton Instruments manual (CSMA User’s Manual, 113.0) for a complete
listing of programs that were available at that time.

(a) Alden, R. G.; Schneebeck, M. C.; Ondrias, M. R; Courtney, S. H.; Friedman,
J. M. J. Ramon Spectrosc. 1992, 23, 569-574. (b) Schneebeck, M. C.; Vigil, L. E.;
Ondrias, M. R. Chem. Phys. Lett. 1993, 215, 251-6. (c) Qian, J.; Schultz, 8.;
Bradbum, G. R; Jean, J. M. J. Phys. Chem. 1993, 97, 10638-10644. (d) Markel,
F.;Ferr1is, N. S.; Gould, I. R; Myers, A. B. J. Am. Chem. Soc. 1992, 114, 6208-
6219. (e) Iwata, K.; Hamaguchi, H. Springer Proc. Phys. 1994, 74(Time-Resolved
Vibrational Spectroscopy VI), pp 85-88. (1) Moore, J. N.; Matousek, R; Parker,
A. W.; Toner, W. T.; Towrie, M.; Hester, R. E. Springer Proc. Phys. 1994,
74(Time-Resolved Vibrational Spectroscopy VI), pp 89-92.

(a) Rodriquez, J .; Holten, D. J. Chem. Phys. 1989, 91, 3525-3531. (b) Rodriguez,
J.; Holten, D. J. Chem. Phys. 1990, 92, 5944-5949. (c) Bilsel, 0.; Rodriguez, J .;
Holten, D. J. Chem. Phys. 1990, 94, 3508-3512. ((1) Rodriguez, J .; Kirmaier, C.;
Holten, D. J. Chem. Phys. 1991, 94, 6020-6029.

Vos, M. H.; Rappaport, F.; Lambry, J.-C.; Breton, J.; Martin, J .-L. Nature 1993,
363, 320-325.

CHAPTER 3

RAMAN SPECTRA OF METALLOPORPHYRIN EXCITED STATES

ABSTRACT

Picosecond time-resolved resonance Raman measurements on Zn(II)
octaethylporphyrin are reported. The lowest excited singlet state (S1) is detected within
100 ps following laser excitation and intersystem crosses to the triplet (T1) state on the
nanosecond time scale. The Raman spectrum of S1 is dominated by totally symmetric
modes that can be correlated with ground state porphyrin modes. Depolarized modes are
only weakly enhanced in S1, which indicates that Jahn-Teller distortion in this state is
essentially absent. By contrast, depolarized modes are prominent in the spectrum of T1
and indicate significant Jahn-Teller distortion, in agreement with previous
metalloporphyrin T1 work fiom this (Walters, V. A, de Paula, J. C., Babcock, G. T.,
Leroi, G. E. J. Am. Chem. Soc. 1989, 111, 8300-8302) and other (Kumble, R, Hu, S.,
Loppnow, G. R, Vitols, S. E., Spiro, T. G., J. Phys. Chem. 1993, 97, 10521-10523)
laboratories. These observations confirm theoretical predictions that the excited states are
influenced by Jahn-Teller efi‘ects, but that configuration interaction minimizes the

distortion in the S1 state.

81

82
I. INTRODUCTION

Light induced charge separation in the photosynthetic reaction center is initiated by
photoexcitation of a chlorophyll complex into its first excited singlet (S1) state.1 To
understand this critical biological process better, more insight into the electronic structure
and conformation of singlet excited porphyrin-like molecules is essential. Time-resolved
resonance Raman scattering is an ideal technique for providing this insight and has been
used recently to characterize the Vibrational properties of electronically excited
porphyrins.2'4 However, due to a limited time resolution, this approach could only focus
on relatively long-lived excited states like the T1 and d,d states of metalloporphyrins,}4
and the St and TI states of free-base derivatives?“ Moreover, dynamic processes, such
as Vibrational relaxation of electronically excited macrocycles, could not be studied at all.
These latter issues are likely to be of firndamental importance in understanding the details
of the photosynthetic process.5 To increase the time resolution of the Raman technique
we have developed picosecond pump-probe methods6 that allow us to monitor the time
evolution of a metalloporphyrin in its excited state manifold. Here, we present our initial
results on Raman scattering fiom excited electronic states of Zn(II) octaethylporphyrin
(ZnOEP)?” In the St state the number and symmetry of the bands indicate weak or
absent J ahn-Teller distortion. ZnOEP in the T1 state, in agreement with previous work on
the ZnTPP T1 state,2 shows relatively strong enhancement of depolarized modes. The
latter observation indicates Jahn-Teller distortion and shows that metalloporphyrin excited

states of different spin multiplicity exhibit qualitatively different Jahn-Teller effects.

83
II. MATERIALS AND METHODS

To induce the 81 state, we excited ZnOEP dissolved in tetrahydrofirran (TI-IF) or
pyridine at the maximum of its Q(0,0) absorption band. The Raman spectrum of the
sample was probed in the blue with a second laser pulse. A pyrromethene 567 dye laser
(Coherent 702) generated pulses (7 ps firll-width-half-maximum) at 572 nm. The laser is
synchronously pumped by the second harmonic of a mode-locked Nd:YAG (Coherent
Antares 76-s). The laser setup used for these experiments is firlly documented in Chapter
2 under “Experimental Apparatus”. The third harmonic is used to pump a stilbene 420
dye laser with an output wavelength of 450 nm. The wavelength of 450 nm was chosen
because the porphyrin has a maximum near this wavelength in the S1 state, whereas the
absorption of the porphyrin in the ground state is very low. A cavity dumper on each dye
laser is used to create a time delay of 790 ns between subsequent pulses and to increase
the peak power of the individual pulses. Time delays between the pump and probe pulses
were adjusted electronically or spatially by adjusting the path length of the pump pulse.
The pulse energies of the pump and probe lasers used were 40 n] and 25 nJ at the sample,

respectively. A microscope objective was used to focus these beams to ~13 um spots on

the sample. A low-pass filter (Andover blue dichroic, 50% transmittance at 515 nm) was
used to prevent scattered light from the pump beam and fluorescence to enter the
spectrometer. The Raman spectrum was dispersed by a single grating monochromator
(Jobin Yvon HR640, 2400 grooves/mm grating) and detected with a CCD detector
(Princeton Instruments, model LN/CCD 1152UV). The spectral resolution was 7 cm“. I

To refresh the sample between laser pulses, we used a flowing liquid jet assembly

84
similar. to that of Atkinson et. al.,10 the experimental setup for which is firlly described in

Chapter 2 under “Flowing Liquid Jet Assembly”.

The nanosecond time-resolved pulsed Raman setup (Figure 3.1) used for triplet
state evaluation has the same basic setup as that used earlier in the ZnTPP experiment by
Walters et. at”, but will be briefly described below. "

Two Quanta Ray Q-switched Nd:YAG lasers (Quanta-Ray DCR2’s, 1250
Charleston Road, Mountain View CA, 94043) with pulse widths of 10 ns and repetition
rates of 10 Hz were controlled by a Stanford DG35 Delay Generator (Stanford Research
Systems, 1290-D Reamwood Avenue, Sunnyvale CA 94089) that provides programmable
triggering for the flash lamps and Q-switches to produce the desired time delays. The
delay was continuously monitored with a photodiode and a Tektronix oscilloscope. A
glass slide intersected the two spatially overlapped beams, and reflected approximately 5%
of the intensity into the photodiode. The pump wavelength was generated fi'om the
second harmonic (532 nm) of the first laser. The probe wavelength (435.7 nm) was
provided by pumping a H2 Raman shifter with 532 nm fi'om a second Quanta Ray
Nd:YAG laser and separating the first anti-stokes wavelength with a pellin-broca prism.‘l
A low temperature back-scattering arrangement was chosen to collect data with the
sample in a spinning EPR tube.12 A spherical lens ( fl = 2.5 inches) was used to focus
these beams to ~100 um spots on the sample. The Raman scattering was collected in a
135° orientation and focused by collection mirrors (5 times magnification) into a triple
grating spectrometer (SPEX 1877 Triplemate, 1800 grooves/mm grating blazed at 450 nm

(SPEX industries, Inc., 3880 Park Avenue, Edison NJ, 08820) with a spectral resolution

Figure 3.1

85

Instrumentation used for time-resolved nanosecond Raman
experiments. HG: hamonic generators giving 2“" and 3" harmonic
wavelengths; Dye Laser: optionally used to generate wavelengths
not accessible by Raman shifter; H2 Raman shifter: lst anti-stokes
of 532 nm used (435.7 nm); PBP: Pellin-Broca prism used to
separate the stokes lines coming fiom the Raman shifter; 1459
Illuminator: mirrors used to direct Raman scatter into Spex 1877
Triplemate.

 

 

86

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.1

 

 

1459. Flow
486 Illuminator Cell
PC/DOS
Spex 1877
p] p, Triplemate
Detector LN/CCD
Controller 1 152UV
Stanford
Digital PBP
F7 Delay DICHROIC
Generator 0 435-7 11111
\l\ H2 Raman Shifter . 5 32 11 M3
M4 m—>\
Quanta-Ray DYE
— Q-switched H .
Nd:YAG LASER M2
Quanta-Ray
__ Q-switched H 532 nm
Nd:YAG Ml

87
of 8 cm". The Raman intensity was detected with a CCD (Princeton Instruments, model

LN/CCD 1 152UV).

III. RESULTS

Figure 3.2 shows Raman spectra of ZnOEP in tetrahydrofuran obtained by using
the pump-probe arrangement described in Chapter 2. The excitation pulse preceded the
probe pulse by 100 ps, 800 ps, 2 ns, and 27 ns, respectively. The spectra of the solvent
THF and of the ground state are also shown in Figure 3.2. Figure 3.3 shows the time
dependent difference spectra after subtracting out ground state features with a scaling
constant and following procedures described in Chapter 2 under “Quasi-Simultaneous
Accumulation”. Polarization dependent measurements were carried out to determine the
symmetry species of the Raman modes. The Raman difference spectra in Figure 3 .4 show
the polarization dependence of ZnOEP obtained with the excitation pump pulse preceding
the probe pulse by 100 ps. Likewise, the Raman difference spectra in Figure 3.5 show
polarization dependence when the pump pulse is 27 ns prior to the probe pulse. The
vibrational mode assignments based on these polarization measurements are listed in Table
3 . l . At short time delays (100 ps, 800 ps), ZnOEP is in its S1 state and the spectrum is
dominated by totally symmetric modes, although some weak nontotally symmetric modes
are also visible. After 27 ns only the triplet state remains. At this time delay, the spectrum
resembles that of Kumble et. al.“ obtained in a nanosecond pump-probe experiment,
although some lineshape and intensity differences are apparent. These differences do not '

at‘ise fi'om the use of picosecond pulses in the present study. We have also taken excited

Figure 3.2

88

Time resolved resonance Raman spectra of ZnOEP in THF in

(a) low wavenumber region and (b) in the high wavenumber
region. The pump and probe beams were at 572 nm and 450 nm,
respectively. The time evolution of the Raman scattering is
represented by spectra obtained at time delays between the

pump and probe pulse of 100 ps, 800 ps, 2 ns, and 27 ns. The
concentration of ZnOEP was 5010" M. The spectra of the
ground state (gs) and the solvent (dotted line) were obtained

in the absence of the pump pulse.

lntensitv

lnfnnc “\I

89

 

9.8.
100 ps
\800 ps

 

 

 

 

 

2ns

 

 

27 ns

 

 

 

 

 

 

 

g/xm/
V}:

l

1200 I 139°
Raman shift (0111-1)

Figure 3.2

Figure 3.3

90

Time resolved resonance Raman difference spectra of ZnOEP

in THF in (a) low-wavenumber and (b) high-wavenumber range.
Experimental spectra fi'om Figure 3 .2 were acquired “quasi-
simultaneously”. The ground state features observed in Figure 3 .2
at various delays for high and low frequencies have been subtracted
out such that no negative peaks appear. The peak positions of the
vibrational modes are collected in Table 3.1. The low intensities
observed at 669 cm'1 and 753 cm'l suggest that the intensities of
ground state features in the excited state spectra are minimal.

Intensity

91

 

 

 

 

 

9.5.
100 ps
800 ps
2 ns
27 ns
A I A l A I A l A l J
700 800 900 1000 1100

Raman shift (cm’1 )

(b) l

 

 

 

‘ l A l A l A _1 LA I

1200 1300 1400 1500 1600

Raman shift (cm")

Figure 3.3

Figure 3.4

92

Time-resolved resonance Rarnan difference spectra taken at
100 ps in (a) low and (b) high wavenumber regions. Spectra
are shown for all polarizations, parallel polarization, and
perpendicular polarization. Qualitative assignments to
symmetries of vibrational modes are labeled as either totally
symmetric (P) or nontotally symmetric (DP). Solvent spectra
(dotted lines) are obtained in the absence of a pump pulse and
peaks are marked by an asterisk.

 

93

(a)

 

 

Intensity

 

 

.......... IIPERPEINDICU'AR 1 .L
700 800 900 1000 1100

Raman shift (cm'1)

(b) P

* DP *
g _ P DP
‘9 .....
5 ............ .‘.
.......... P , 'L'ELH- ......

 
 

 

.................... ‘- FEW ."'--:--..........'....
l . 1 . 1 . 1 . l
1200 1300 1400 1500 1600

Raman shift (cm“)

Figure 3.4

Figure 3.5

94

Time-resolved resonance Raman difference spectra taken at 27
nanoseconds in (a) low and (b) high wavenumber regions.

Spectra are shown for all polarizations, parallel polarization, and
perpendicular polarization. Qualitative assignments to symmetries
of vibrational modes are labeled as either totally symmetric (P)

or nontotally symmetric (DP). Solvent spectra (dotted lines)

are obtained in absence of a pump pulse and peaks are marked by
an asterisk.

.-‘-—‘:‘- -

95

 

,,,,,

 

PERPETIWCULAR
l . l r l

1300 1400 1500

Raman shift (cm’1)

Figure 3.5

-----
.........
01.-
......

96

 

 

 

 

Table 3.1 Raman shifts, polarizations and assignments of the observed
Raman modes onnOEP in its ground and Sr and T1 excited
states in tetrahydrofuran and pyridine.
tetrahydrofirran pyridine
S0 S1 T1 So 81 T1 pa assignmentb characterb
669 664 665 669 663 663 p V7 v(CaN), 6(CaCmCa),
acacia)
710c p
730° 731° 729° 733° 729° p
753 753 752° dp
765° 764° p
781 p
8170 dp
1025C1 997 996 e e e p v(C1C2) V(C1C2)
1096 1097 1097 1096 dp
1138 1135 1138 11351136 1135 p V5 V(CaCb),V(CbC1)
1211 1210 dp V13 8(CmH),v(CaCb)
1239 1240 dp
1261 1258 1262 1258 p CHz-twist 5(CbC1H),5(C2C1H)
1276 1276 dp
1320 1313 1319 1315 p CHz-wag 8(CbC1H),5(C2C1H)
1330 dp
1375 1359 1373 1353 p V4 V(CaN),v(CaCb)
1429f 1429 dp
1486 1450 1450 1484 1447 p V3 V(CaCm), V(Cbe)
1558 1582f 1582f 1558 dp V11 v(Cbe)
1584 1564 1564 1583 1564 1566 p V2 V(Cbe)
i613 1611 dp v19 v(C4Cm)

 

 

 

a P. Polarized; dp, depolarized. 9 From ref 12b. ° Weak. ‘1 Evident at higher porphyrin
concentrations. e Obscured by solvent modes. fIdentified by polarization measurements.

97
state spectra of the T1 state obtained by using nanosecond Q-switched Nd:YAG lasers as

pump and probe sources (Figure 3.6). The lineshapes and intensities from Figure 3.6 are
nearly identical with the picosecond T. results in Figure 3.4.
At a pump-probe time delay of 2 ns (Figure 3.2), modes characteristic of both S1
and Tl are seen. Overall, the temporal evolution of the singlet into the triplet, as
monitored by the Raman spectra in Figure 3 .2, shows excellent agreement with earlier
optical work in which the lifetime of the singlet state was determined to be 2.5 ns7

The S1 and T1 modes remain observable when we use pyridine as a solvent. Figure
3 . 7 shows Raman spectra of ZnOEP in pyridine with the excitation pulse preceding the
probe pulse by 100 ps and 27 ns. Since pyridine has analogous vibrational fi'equencies
with that of the pyrrole rings in porphyrins, the solvent features were subtracted out of all
spectra involving ZnOEP dissolved in pyridine.

The scaling criteria for subtraction of solvent were similar to those used for
subtraction of ground state features. Solvent modes that did not overlap in frequency with
Porphyrin modes were subtracted out such that no negative peaks appeared. This
subtraction was done to the ground state spectrum in Figure 3.7b by using the 1599 cm'1
solvent mode as the reference. This technique revealed ground state porphyrin modes

such as the 013(1211 cm"), 112(1582 cm"), and 113 (1486 cm'l), modes that were

pr eViously buried under strong pyridine modes. Comparison with known ZnOEP spectra
verified the validity of this approach. 13 This approach was then applied to the time-

d"Pendent high fi'equency spectra.

Figure 3.6

98

Time-resolved resonance Raman difference spectra taken using

nanosecond Q-switched Nd:YAG laser pulses as pump and probe.

Spectra are shown for (a) ground state (g.s.), (b) time delay
between pump and probe of 13 ns, (c) scaled difference spectrum,
((1) parallel polarization, and (e) perpendicular polarization. The
wavelength of the pump pulse was 532 nm, and the probe
wavelength was 437 .5 nm. The ground state spectrum was

taken in the absence of a pump pulse. The concentration of
ZnOEP was 5 o10'3 M in THF.

.n‘finnz‘g A

99

DP

 

  

 

 

DP
PERPENDICULAR
OPP DP
P
P P PDP
P 8
(e) P 5"- § PARALLEL
(d)
DIFFERENCE
%
E l 3
C
'_ "83
h 13ns
(c)l
9.3.
(b)
(a)rITJTLrIrIrlrlLJrITJTI

 

700 800 900 1000 1100 1200 1300 1400 1500 1600 1700
Raman shift (cm‘l)

Figure 3.6

Figure 3.7

100

Time-resolved resonance Raman difference spectra of ZnOEP in
pyridine in (a) low-wavenumber and (b) high-wavenumber range.
The ground state features observed in Figure 3.7 at various delays
for high and low frequencies have been subtracted out such that
no negative peaks appear by using techniques identical to that
used in Figure 3.2. Solvent spectra acquired immediately after
the experiment were used to subtract out the strong solvent
features which obscure the 27 ns spectrum. The peak positions
of the Vibrational modes are collected in Table 3.1. Asterisks
mark artifacts arising from subtraction of the pyridine solvent
features.

101

   

 

 

 

Q * g
:0?) § * Is 3:
c 8
at .—
g u 100 ps
('0
2 a
27 ns
A I 1 l 1 I L l n L l
700 800 900 1000 1100
Rama shift(cm“)
(b) i
ii . .: i.
a?! .................. .00.. ‘i """" 9.5.
2 ,, a I
g .5. g g g.s.-so
‘- 100 ps
27 ns
J r 1 l 1 l 1 I 1
1200 1300 1400 1500 1600

Raman shift (cm‘1)

Figure 3.7

102
This same approach was used for the low frequency spectra in Figure 3.7a using

the 991 and 1032 'cm'1 modes as the reference. However, the low frequency subtraction
left derivative type features in these regions. These artifacts were unavoidable, probably
due to the strong intensities of these solvent bands (~10 times the intensity of the strongest
porphyrin modes).

The 100 ps spectrum reproduces the modes observed in the 100 ps and 800 ps
spectra fi'om Figure 3 .2 quite closely. The slight differences in frequency and intensity
may be attributed to coordination differences between the two solvents. (The 27 ns high
frequency spectrum is somewhat obscured by strong pyridine scattering at 1219 cm"1 and
1 582 cm").

Polarization dependent measurements on ZnOEP in pyridine were carried out and
are displayed in Figure 3.8 and Figure 3.9. The Raman difference spectra in Figure 3.8
show the polarization dependence of ZnOEP with the excitation pump pulse preceding the
probe pulse by 100 ps. Likewise, the Raman difference spectra in Figure 3.8 show
polarization dependence when the pump pulse is 27 us prior to the probe pulse. All
Spectra were solvent and ground state feature subtracted. The vibrational mode
assignments based on these measurements are also listed in Table 3.1.

In Table 3.1, ZnOEP modes in the S1 and T1 states are labeled with reference to
those reported for NiOEP. ‘3 Assignments for excited state Vibrations were made by
analogy to ground state vibrations. For totally symmetric modes this was facilitated by the
One to one correlation that appears to exist between modes in the ground and excited 1

State, respectively. For the S1 spectra, the depolarized modes are almost all of low

Figure 3.8

103

Time-resolved resonance Raman difference spectra of ZnOEP in
pyridine taken at 100 ps in (a) low and (b) high wavenumber
regions. Spectra are shown for all polarizations, and parallel
and perpendicular polarizations. Qualitative assignments to
symmetries of vibrational modes are labeled as either totally
symmetric (P), or nontotally symmetric (DP). Asterisks mark
artifacts arising from subtraction of the pyridine solvent features.

104

 

 

a)
P P *
E * DP
3 DP
9; .
5
PARALLEL Ii}
1] l'
PERPENDICULAR '
_L l A l A I A l l l
700 800 900 1000 1100
(1)) Raman shift (cm‘1)
p
p
p
DP
- P
l 'P
E‘
m
c
.9
5.:
PARALLEL .
PERPENDICULAR
I 1 I A L l L l I
1200 1300 1400 1500 1600

Raman shift (cm")

Figure 3.8

Figure 3.9

105

Time-resolved resonance Raman difference spectra of ZnOEP in
pyridine taken at 27 ns in (a) low and (b) high wavenumber
regions. Spectra are shown for all polarizations, and parallel
and perpendicular polarizations. Qualitative assignments to
symmetries of vibrational modes are labeled as either totally
symmetric (P), or nontotally symmetric (DP). Asterisks mark
artifacts arising from subtraction of the pyridine solvent features.

'n‘an- :‘u a

106

 

P DP
DP
3?." PARALLEL
(D
C
.9.
s.
PERPENDICULA ’
I I A I A I I
700 000 900 1000 1100
Raman shift (cm’1)
(b) DP
DP
P P
.E‘
2
g PARALLEL
' ERPENDICULA '
I AAAAA I AAAAA I A g A I I #1
1200 1300 1400 1500 1600

 

Raman shift (cm'1)

Figure 3.9

107

intensity, and some could only be identified by polarization measurements. We discuss the

S1 spectrum below and then consider the T1 data briefly.

IV. DISCUSSION

The electronic configuration of metallooctaethylporphyrins in the ground excited
singlet state is known to have an al.,alu configuration. The lowest excited state
configurations, l(1110,63) and l(a2..,e,,), are of nearly equivalent energies and both have 13,.
symmetry. Hence, they strongly mix via configuration interaction. 1‘ Molecular orbital
calculations predict that the individual modes should show sensitivity to the electron
occupancy of the an. and e, orbitals.” Diagrams that depict the relative electron
occupancy of these three molecular orbitals can be seen in Chapter 1, Figure 1.6. Modes
involving primarily stretches of C.C..., 0.0,, or C.N bonds are predicted by basic
Gouterman molecular orbital theory to shift to lower frequencies in the S; state, whereas
modes comprising mostly €be stretches are predicted to shifi to higher frequencies.
Earlier work on both metalloporphyrin n-cation and 1t-anion radicals has demonstrated the
generality of these orbital occupancy arguments in the assignment of vibrational modes
and electronic state ordering. ‘6 In general, our assignments of the vibrational modes in 81
agree with predictions based on the orbital occupancy assignments above. An exception
to this is the behavior of 02, which apparently shifis to lower, rather than higher, frequency
in the 81 state. Pemg and Bocian16c observed similar behavior for some modes in the
ZnOEP u-anion radical, that is, frequency shifts that were counter to those predicted for e,

orbital occupancy, and suggested that structural changes of the macrocycle in the anion

108

radical state were responsible. From the behavior of 02, it appears that similar phenomena
may occur to some extent in the S; state.

As opposed to our previous work on the T; state of ZnTPP,2"" and that described
below for T; of ZnOEP, Jahn-Teller distortion occurs to only a minor extent, if at all, in
the S; state. This conclusion follows fiom our observations in Figure 3.3 and the
frequency assignments in Table 3.1 that only weak enhancement of depolarized modes
occurs, and that the number of totally symmetric modes does not increase relative to the
ground state. This conclusion agrees with theoretical predictions that Jahn-Teller effects
in the S; state will be minimized by configuration interaction. 17

In the T; state, the Raman shifts of many modes coincide with those in the S; state.
The depolarized modes though, are significantly stronger than in the S; state (Figure 3.3,
Figure 3.4, and Table 3.1). The broad feature at 1239 cm'1 with a shoulder at 1276 cm'1
may be comprised of several underlying modes. The appearance and shape of this spectral
feature was independent of whether tetrahydrofirran or pyridine was used as the solvent.
Ethyl modes are expected to occur in this fi'equency range, and although they are not
strongly enhanced in the ground state,13 they may gain intensity in the excited state and
account for the spectral congestion. The shifi of u; in the triplet state, as in 8;, doesn't
follow the predictions based on the molecular orbital calculations. As in HZOEP and
ZnTPP,2 1).; is not strongly enhanced in the T; state of ZnOEP. The enhancement of
depolarized modes in the T; state strongly suggests that J ahn-Teller distortion is present in
this state, in agreement with the now general observation of structural rearrangement in

triplet states of metalloporphyrin chromophores."18

109
To conclude, the observations presented here on the S; and T; states of ZnOEP

test and confirm theoretical predictions that Jahn-Teller distortion is significantly more
pronounced in the triplet than in the singlet state. This conclusion has possible
ramnifications for the initial steps in the photosynthetic process. Previous work on
metalloporphyrin and metallochlorin cation radicals has shown that these are formed with
little change in the normal coordinate composition and hence little nuclear rearrangement
of the macrocycle.'6"16b"9 Our demonstration here that the light-induced formation of the
singlet state also proceeds in the absence of large nuclear rearrangements means that
charge separation from this state to form the cation radical product state may occur with
minimal reorganization energy.”21 Accordingly, the rate of the initial energy conversion

step can be fast, as it is observed experimentally to be.1

(1)

(2)

(3)

(4)

(5)

(6)

LIST OF REFERENCES

see, e.g.,: Research in Photosynthesis, Proceedings of the IX th International
Congress on Photosynthesis; Murata, N., Ed.; Kluwer Academic Publishers:
Dordrecht, 1992.

(a) Walters, V. A.; de Paula, J. C.; Babcock, G. T.; Leroi, G. E. J. Am. Chem. Soc.
1989, 111, 8300-8302. (b) Nam, H. H.; Walters, V. A; de Paula, J. C.; Babcock,
G. T.; Leroi, G. E. In Proceedings of the XIIth International Conference on
Raman Spectroscopy; Durig, J. R; Sullivan, J. F., Eds; Wiley and Sons: New
York, 1990, p 618. (c) Reed, R. A.; Purello, R.; Prendegast, K.; Spiro, T. G. J.
Phys. Chem. 1991, 95, 9720-9727. ((1) Sato, S.; Asano-Someda, M.; Kitagawa, T.
Chem. Phys. Lett. 1992, 189, 443-447. (e) Kumble, R; Hu, 8.; Loppnow, G. R;
Vitols, S. E.; Spiro, T. G. J. Phys. Chem. 1993, 97, 10521-10523. (0 Kreszowski,
D. H.; Babcock, G. T. Biophys. J. 1993, 64, a276.

(a) Findsen, E. W.; Shelnutt, J. A.; Ondrias, M. R., J. Phys. Chem. 1988, 92, 307-
314. (b) Apanasevich, P. A.; Kruglijk, S. G.; Kvach, V. V.; Orlovich, V. A. in
Time Resolved Vibrational Spectroscopy V; Takahashi, H., Ed.; Springer-Verlag:
Berlin, 1992.

Bell, S. E. J.; Al-Obaidi, A. H. R.; Hegarty, M.; Hester, R. E.; McGarvey, J. J. J.
Phys. Chem. 1993, 97, 11599-11602.

(a) Rodriguez, J.; Holten, D., J. Chem. Phys. 1989, 91, 3525-3531. (b) Rodriguez,
J.; Holten, D., J. Chem. Phys. 1990, 92, 5944-5949. (c) Bilsel, 0.; Rodriguez, J.;
Holten, D., J. Phys. Chem. 1990, 94, 3508-3512. (d) Rodriguez, J .; Kirmaier, C.;
Holten, D. J. Chem. Phys. 1991, 94, 6020-6029. (e) Vos, M. H.; Rappaport, F.;
Lambry, J .-C.; Breton, J.; Martin, J .-L. Nature 1993, 363, 320-325.

For other or comparable methods to measure resonance Raman with picosecond
time-resolution see, e. g., (a) Gustafson, T. L.; Iwata, K.; Weaver, W. L. in
Ultrafast Phenomena VII; Harris, C. B.; Ippen, E. P., Mourou, G. A.; Zewail, A.
H., Eds; Springer-Verlag: New York, 1990. (b) Iwata, K.; Hamaguchi, H. Chem.
Phys. Lett. 1992, 196, 462-468. (c) Weaver, W. L.; Huston, L. A; Iwata, K.;
Gustafson, T. L. J. Phys. Chem. 1992, 96, 8956-8961. ((1) Butler, R M.; Lynn, M.
A; Gustafson, T. L. J. Phys. Chem. 1993, 97, 2609-2617.

110

(7)

(3)
(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17)

111

Excited state absorption spectra and lifetimes on a picosecond time scale have
been measured by Rodriguez, et. al.: Rodriguez, J .; Kirmaier, C.; Holten, D. W. J.
Am. Chem. Soc. 1989, 111, 6500-6501. The S; state of ZnOEP has an absorption
maximum at 450 nm; the ground state absorption at this wavelength is low. With
high yield, the S; state converts into the T; state, which has an absorption

. maximumat435 nm.

Kreszowski, D. H.; Deinum, G.; Babcock, G. T. Biophys. J. 1994, 66(2), a374.

Kreszowski, D. H.; Deinum, G.; Babcock, G. T. J. Am. Chem. Soc. 1994, 116,
7463-7464.

Atkinson, G. H.; Brack, T. L.; Blanchard, D.; Rumbles, G. Chem. Phys. 1989,
131, 1-15.

For a description of the nanosecond time-resolved Raman apparatus see the thesis
of Varotsis, Constantinos, Diss. Abstr. Int. B 1991, 51(12) pp 52-54.

For a description of the low temperature back-scattering dewar see the thesis of
Kean, Robert, Diss. Abstr. Int. B 1988, 48(7), pp 42-43.

(a) Abe, M.; Kitagawa, T.; Kyogoku, Y. J. Chem. Phys. 1978, 69, 4526-4534. (b)
Li, X.-Y.; Czernuszewicz, R. S.; Kincaid, J. R; Stein, P.; Spiro, T. G. J. Phys.
Chem. 1990, 94, 47-61.

(a) Spellane, P. J .; Gouterman, M.; Antipas, A.; Liu, Y. C. Inorg. Chem. 1980, 19,
386-391. (b) Shelnutt, J. A; Ortiz, v. J. Phys. Chem. 1985, 89, 4733-4739.

Gouterman, M. J. Mol. Spectroscopy 1961, 6, 138-163. We used the
coeflicients of the fi'ontier molecular orbitals in: Prendergast, K.; Spiro, T. G. J.
Phys. Chem. 1991, 95, 9728-9736.

(a) Oertling, W. A.; Salehi, A.; Chang, C. K.; Babcock, G. T. J. Phys. Chem.
1989, 93, 1311-1319. (b) Czemuszewicz, R. S.; Macor, K. A; Li, X.-Y.; Kincaid,
J. R; Spiro, T. G. J. Am. Chem. Soc. 1989, 111, 3860-3869. (c) Pemg, J.-H.;
Bocian, D. F. J. Phys. Chem. 1992, 96, 4804-4811.

(a) Gouterman, M. Ann. NY. Acad Sci. 1973, 206, 70-83. (b) Shelnutt, J. A;
Cheung, L. D.; Chang, R. C. C.; Yu, N.-T.; Felton, R H. J. Chem. Phys. 1977,
66, 3387-3398. (c) Cheung, L. D.; Yu, N.-T.; Felton, R. H. Chem. Phys. Lett. .
1978, 55, 527-530.

(13)

(19)

(20)

(21)

112

(a) Langhoff, S. R; Davidson, E. R; Gouterman, M.; Leenstra, W. R; Kwiram,
A. L. J. Chem. Phys. 1975, 62, 169-176. (b) van der Waals, J. H.; can Dorp, W.
G.; Schaafsma, T. J. in Ihe Porphyrins; Dolphin, D., Ed.; Academic Press; New
York, 1978, Vol 4, pp 257-312.

Salehi, A.; Oertling, W. A.; Fonda, H. N.; Babcock, G. T.; Chang, C. K.
Photochem. Photobiol. 1988, 48, 525-530.

Dutton, P. L; Moser, C. C. Proc. Natl. Acad Sci. USA 1994, 91, 10247-10250
(a) Vos, M. H.; Lambry, J. C.; Robles, S. J.; Youvan, D. C.; Breton, J .; Martin, J.

Proc. Natl. Acad Sci. USA, 1991, 88, 8885-8889. (b) Vos, M. H.; Rappaport, F .;
Lambry, J .-C.; Breton, J.; Martin, J .-L. Nature 1993, 363, 320-325.

CHAPTER 4
VIBRATIONAL ANALYSIS OF ZINC(II) OCT AETHYLPORPHY RIN IN THE

LOWEST EXCITED SINGLET STATE

ABSTRACT

Resonance Raman spectra of the lowest excited singlet state (S;) are reported for
ZnOEP and its isotopomers ZnOEP-d4 and ZnOEP-”N4, as well as for ZnEtio and
MgOEP. The Raman active vibrational modes in the 8; state of ZnOEP were assigned
from depolarization ratios, isotopic shifts, and intensity correlations with ground state
features. Downward frequency shifts of totally symmetric S; porphyrin modes of ZnOEP
reflect structural changes of the porphyrin macrocycle. These changes occur along totally
symmetric normal modes as indicated by weak enhancement of depolarized modes and a
l: l correlation between totally symmetric modes in the S;/So states. This indicates the D4;
Symmetry of the 8;; molecule is retained and confirms our earlier conclusion that Jahn-
Teller effects are negligible in the S; state (Kreszowski, D. H.; Deinum, G.; Babcock, G.
T. J. Am. Chem. Soc. 1994, 116, 7463-7464). Based on frequency/core size relationships
obtained for porphyrins in the ground state, we estimate that, upon excitation to S;, the
core Size increases by about 0.1 A. Calculations that include different excited state
configllrations predict small b;, distortions in the S; state that may be related to increased

linewidths of the core modes that indicate inhomogeneity in the macrocycle structure.

113

1 l 4
L INTRODUCTION

The primary step of photosynthetic energy conversion involves charge separation
in the reaction center where a chlorophyll complex is oxidized after excitation into its

lowest excited 1c-1r“ singlet state (S;) by light absorption or by energy transfer from

neighboring pigments.l The charge separation is stabilized by transfer of the electron to
subsequent acceptors and by reduction of the oxidized chlorophyll by a secondary electron
donor. Most critical in this process are the electronic and redox properties of the reaction
center chlorophst that are mediated by the protein environment and by exciton
interactions between different chromophores. 1'2 More insight in the electronic and
structural properties of chlorophylls and related macrocycles is a prerequisite for a better
better understanding of the structure and functioning of the reaction center. Resonance
Raman spectroscopy has proven to be an excellent tool for studies of chlorophylls and the
small structural changes of their macrocycles in the transient states, e. g., vide infi'a the S;,

tt-anion radical and n-cation radical.

However, not only the static properties of the reaction center are important.
According to Marcus theory, the rates of electron transfer in photosynthesis3 strongly
depend on the reorganization energy associated with deformations of the pigments. Large
deformations will reduce the rate of electron transfer, leading to less efficient
photosynthesis.

Transient absorbance measurements on isolated photosynthetic reaction centers .
reveal that reorganization associated with low frequency vibrations up to about 100 cm"

persist on the time scale of charge separation.‘ These vibrations could not be clearly

115

related to chlorophyll modes and were tentatively attributed to modes involving the
surrounding protein. Measurements by Zhu et al. 5 showed that, for the NO binding
reaction to the heme of myoglobin, similiar vibrations can be ascribed to the out-of-plane
iron/histidine stretching mode at 220 cm". These latter results indicate that low fiequency
vibrational modes are coupled to reactions in these pigment-protein complexes.

To fit measurements of electron transfer rates in reaction centers using
reorganization mode frequencies as a parameter,6 high frequency vibrations are also
expected to influence the electron transfer rate.“ Higher frequencies, however, cannot be
conveniently studied by time-resolved absorbance measurements because of the limited
bandwidth of the laser flash. To study the coupling of high fi'equency vibrations to
reactions directly, measurements should be performed in the fi'equency domain, i.e., using
time-resolved Raman or infrared spectroscopy. Picosecond time resolved resonance
Raman measurements8 performed on, e.g., S; excited trans-stilbene and analogues have
shown that depolarization ratios, linewidths, mode intensities and frequency positions can
give valuable information about the dynamic processes occuring in transient excited states.

Some Raman experiments have been performed with intact photosystems,9 but
most work to date has been done with model porphyrins with a symmetry higher than that
of chlorophyll. The higher symmetry of, e. g., Zn(II) octaethylporphyrin (ZnOEP) and
Zn(II) tetraphenylporphyrin (ZnTPP), leads to Raman spectra that can be interpreted more
easily and can provide bench marks against which analogous data for the lower symmetry

macrocycles may be compared.

1 16
The ground state configurations of ZnOEP and ZnTPP retain a molecular

symmetry of D4; and the spectral properties can be explained by using the four orbital
model of Gouterrnan. 1° According to that model two electronic transition states, the Q-
and B—(Soret) states, arise from transitions fiom two nearly degenerate orbitals with a;..
and an. symmetry, respectively, to a degenerate pair of orbitals with ex symmetry. Strong
configuration interaction leads to a more intense band in the blue (Soret) and a less intense
band in the red (Q). Gouterman orbitals depicting the relative electron occupancy of these

10,46
.6.

three molecular orbitals can be seen in Chapter 1, Figure 1 The absorption
spectrum of ZnOEP and an energy level diagram are shown in Figure 4.].

Because they are models for intermediates in the electron transport chain in the
photosynthetic reaction center, a number of experiments have been performed on
porphyrin tc-cations and tt-anions. It was shown that, upon oxidation of the porphyrin, an
electron is removed fiom the a;.. or a2u orbital, respectively, depending on whether the
porphyrin is a metallo-OEP or a metallo-TPP.”’12 Oxidation has little effect on the overall
geometry of the macrocycle. 13’” Measurements performed by Salehi et al. '5 on n-cations
of metallomethyloctaethylchlorin (metallo-MeOECs), which have one reduced pyrrolic
ring, showed a significant mode composition change due to the presence of this reduced

pyrrolic ring, but they concluded, similiar to the earlier conclusions for porphyrins,“l4 that

upon tt-cation formation the general structure of the macrocycle remains the same.
Large frequency shifis have been reported for the modes of u-anions of metal- .

TPPs due to Jahn-Teller distortion of the ground state anion. '6’" This distortion

originates from the presence of one electron in degenerate orbitals. '8 In metallo-OEP 1t-

Figure 4.1

117

a) Metalloporphyrin absorption characteristics for ground and
excited states of zinc(II) octaethylporphyrin (ZnOEP) in methylene
chloride. The S; and T; spectra are taken with reference to results
of Rodriquez, J.; Kirmaier, C.; Holten, D. W. J. Am. Chem. Soc.
1989, 111, 6500-6501. b) Energy level diagram representing
observed energy levels for ZnOEP.

118

a)

 

 

 

 

 

 

 

 

 

 

 

 

 

B
zlne(ll)oebetlrylporphyrln
i
PROBE PL1MP
300 ‘ 41'» ' 51'» I 000 E 000
wavelength (nm)'
b)
eV
6.0 - 1
4 F... (L)
1
4.0 -—3'E‘ (N)
2‘13; (B)
, Q00)
20 , IE1; 9&0)
'33,, T;
3.11eV
2.33eV
0.0

 

 

 

Figure 4.1

1 19
anions the shifts are only small and consequently Jahn-Teller effects are assumed to be

much weaker. ‘9 This further indicates that the details of the macrocycle structure and
electronics are influenced by the type and location of peripheral substituents.
Resonance Rarnan experiments performed with nanosecond lasers pemritted

studies of triplet states of metalloporphyrins?“21 As for the rt-anion, the macrocycle of

ZnTPP is Jahn-Teller distorted in the triplet state.20 Because the lifetimes of the S; excited
states of free-base porphyrins are longer, nanosecond lasers could also be used to study
their characteristicszz’23 However, the results of the latter experiments have been
interpreted only to a limited extent. Experiments with even higher time-resolution,

24'25'26 of nickel porphyrins and recently by

picoseconds, have been performed on d-d states,
us on the singlet excited state (S;) of ZnOEP.27'2“'29 Such types of experiments are
especially important for unravelling photosynthesis since femto- or picosecond time
resolution is necessary for more insight in dynamic processes of electronically excited
macrocycles. Previous transient absorbance experiments showed that vibrational
relaxation processes like those in reaction centers take place on a time scale of up to about
20 ps in model porphyrins.30

Here, we present results from extensive picosecond transient Raman experiments
done on the model system ZnOEP, its isotopomers ZnOEP-d4 , ZnOEP-"N4, and Zn(II)
etioporphyrin I (n.a.), and MgOEP. We especially focused our experiments on the

structural changes, e.g., J ahn-Teller distortions, and will test whether these are important

for the lowest excited singlet states (8;) of ZnOEP.

120

Combination of the isotopomer data and depolarization measurements was
essential to assign the Raman active vibrational modes and to estimate changes of the
macrocycle structure of ZnOEP in the S; state. The Raman data on that state reveal
frequency changes that indicate core size expansion but retention of D4; symmetry, and
suggest the relative absence of Jahn-Teller distortion in S;, The directionality of the
vibrational shifts in the S; state of ZnOEP indicates that excitation to this state cannot
simply be described by one electron promotion, but that several configurations need to be
taken into account. To rationalize the shifts of the modes more quantitatively, we
calculated natural orbitals, which are combinations of molecular orbitals that can describe
electronic transitions involving configuration interaction. These calculations predict a
small b;, distortion of the porphyrin in the S; excited state, which may be related to the

experimentally observed increased width of the porphyrin core modes.

11. MATERIALS AND METHODS

Zn(II) octaethylporphyrin (ZnOEP) and Mg(II) octaethylporphyrin (MgOEP) were
purchased from Aldrich and used without further purification. Exchange of the hydrogen
for deuterium in ZnOEP-d4 was accomplished by using the DZSOJDZO method.31 The
N‘s-labeled compound (ZnOEP-”N0 was prepared from Nal’NOZ and ethylpropionyl
acetate.32 Zinc(II)-Etioporphyrin-I (ZnEPI) was purchased from Midcentury Chemical
Company (Posen, Illinois) and used as received. The solvent tetrahydrofirran (Baker

Analyzed reagent) was used without further purification. The integrity of the porphyrins

121

were confirmed before and after the measurements by UV-visible absorption spectroscopy
on a Perkin Elmer Lambda 5 spectrometer.

The experimental setup used for these experiments is that shown in Figure 2.1
(Chapter 2). Raman scattering was collected in a standard 90° orientation and focused by
refi’active optics into a single grating spectrometer (Jobin Yvon HR 640, 2400
grooves/mm grating blazed at 400 nm) with a spectral resolution of 7 cm". A charge-
coupled device (Princeton Instruments, model LN/CCD 1152UV) cooled to -100°C was
used for detection. A dye laser (Coherent 702) tuned to 450 nm with Stilbene 420 as the
dye was used to generate the Raman spectra. This dye laser was synchronously pumped
with the third harmonic of a mode locked Nd:YAG laser (Coherent Antares 76-s). The
wavelength of 450 nm was chosen because the porphyrin has a maximum near this
wavelength in the S; state}3 whereas the absorption of the porphyrin in the ground state is
very low. To excite the porphyrins at the maximum of the Q(0,0) band a second dye laser
(Coherent 702) with pyrromethene 567 as the dye34 was tuned to 572 nm for the zinc
porphyrins, and 575 nm for MgOEP. This laser was pumped by the second harmonic of
the Nd:YAG. Both dye lasers produce pulses of about 7 ps (FWHM). Cavity dumpers on
the dye lasers were used to create a time delay of 790 ns between subsequent pulses and
to increase the peak power of the individual pulses. The pulse energies of the pump and
probe lasers were 80 nJ and 30 11], respectively. Most experimental setups use repetition
rates between 50 Hz and several KHz rather than MHz.8’35 Even though MHz repetition
rates complicate the sample setup, an advantage is that high average laser powers can be

used resulting in high signal-to-noise ratios. At low repetition rates, high average powers

122
lead to peak powers that are high enough to produce undesirable non-linear effects in the

sample. Chapter 2 ( Flowing Liquid Jet Assembly) gives a more complete discussion
about the relationship between pulse repetition rates and laser powers. A microscope
objective (Newport M-SX) was used to focus these collinear beams to 15 um spots on the
sample. A low-pass filter (Andover blue dichroic, 505FD64505, 50% transmittance at
515 nm) was used to prevent scattered light fi'om the pump beam and fluorescence from
entering the spectrometer. To refresh the sample between individual pulses a rapid
flowing jet similar to that of Atkinson et al.36 was used. Spectra were recorded with
shutters in the probe and pump beam pathways, respectively, and were accumulated for
two minute time periods. The shutters for the probe and pump beams were used as
described in Chapter 2 under “quasi-simultaneous technique”. This permitted easy
elimination of temporal instabilities fi'om the laser or the sample jet during postprocessing
of the data. Spectra of S; excited state porphyrins were determined by a scaled
subtraction of the contribution of ground state porphyrins from the raw data. Solvent
contributions were also eliminated in a similiar fashion. The uncertainties in most of the
peak positions are about 2 cm". To determine the polarization of the Raman active
vibrations, the Raman spectra were recorded with a dichroic sheet polarizer in front of the

spectrometer.

III. RESULTS
Raman spectra onnOEP (n.a.), ZnOEP-d4, ZnOEP-”N4, ZnEPI (n.a.) and
MgOEP (n.a.) in the ground state, without a pump pulse preceding the probe pulse at 450

nm, are shown in Figure 4.2. Raman spectra taken under the same conditions but detected

123

Figure 4.2 Ground state Raman spectra of ZnOEP (n.a.), ZnOEP-d4,
ZnOEP-”N4, ZnEPI (n.a.) and MgOEP (n.a.). The probe beam
was at 450 nm. Each porphyrin was dissolved in tetrahydrofuran.
The concentration for all porphyrins was 5010'3 M, except ZnEPI,

which was 1010‘3 M.

124

 

 

 

 

 

 

 

 

 

 

 

    

 

«Few
SS :3 88 $8 1
£3 83 82 89 em!
mvmp Pump 5mm? mmmw mmmw .
mm: mm: vm: an: mmvw
:2 .22 89 £2 1 mm: 1
“111 A” “111
29 oz: can.
r.
FmN— hmmw N4 owNF 4
5 d
1. P
D:
:m.. cm. NON_v E x
em 0 n
2 .m 2

 

 

memcmE m>=£wa

1300 1400 1500 1600

1200

Raman shift (cm")

Figure 4.2

125

with a polarizer in front of the spectrometer are given in Figure 4.3. Raman frequencies
and vibrational assignments for ZnOEP-d4, ZnOEP-”N4, ZnEPI (n.a.) and MgOEP (n.a.)
are given in Table 4.1, with Raman frequencies listed as isotopic shifts relative to
ZnOEP(n.a.). Assignments for ground state (So) Raman modes of ZnEPI, MgOEP, and
ZnOEP and its isotopomers were made by analogy to those of NiOEP37 and previously
published results on ZnOEP.19 The ground state Raman spectra of these porphyrins are
very similar to previously published spectra obtained in resonance with the Soret

absorption band. They are dominated by totally symmetric modes with only a few

nontotally symmetric modes visible.
Raman difference spectra obtained with the pump pulse preceding the probe pulse

by 100 ps are shown as difference spectra in Figure 4.4. These spectra were generated by

subtracting out ground state and solvent contributions, which both are of about the same
size as the difference spectra obtained. Previously, we showed that these spectra evolve in
time parallel to transient absorbance changes that reflect the transition fi'om the S; state to
the lowest triplet state (T;).27’28’33 We also showed that the Raman spectrum is
independent of whether pyridine or tetrahydrofirran is used as the solvent. At a delay time
of l 00 ps the modes observed belong exclusively to the S; state. Determination of the
polarization of these excited state modes was done with the spectra shown in Figure 4.5.
It shows that the spectra taken at 100 ps, similar to the ground state spectra, are
dominated by totally symmetric modes with only weakly enhanced nontotally symmetric
modes visible. i

In our earlier work we observed a 1:1 correlation between totally symmetric

mOdes in the ground and S; states of ZnOEP. This made vibrational assignments

126

Figure 4.3 Parallel and perpendicular(---) Raman spectra of ZnOEP (n.a.),
ZnOEP-d4, ZnOEP-"N4, ZnEPI (n.a.) and MgOEP (n.a.).
Other than the polarizer, conditions were the same as those
described in Figure 4.2.

127

 

 

ZnEtio

 

Relative intensities

ZnOEP-d4

 

 

 

 

I I I I I
1200 1300 1400 1500 1600

Raman shift (cm'1)

Figure 4.3

A; 05 .3 @8330 £ 5:32 8695 m: :5 acumen fl 3008 egg—8% m .

A.

6055820 .5 “Sufi-Eon mm 356

05 85oz? 820 8: fl : $588388 05 co woe-mp a “cumin—ace do Howie—om d fluxes-5 E 53w 2a 858 cos-NEAR” QE-

 

 

 

 

 

 

 

 

 

 

 

 

2> :3 _+ ae $23 N- A5322
A
22-20., Nov _- AN: _+ Ate TANNC 2?:me ANS o a N- Am: a- AN: IAN: $2
ass? a 7- a N- a N1. a N- a 2:83A- a 7&3_-AN.N.:+ANN.V22
A
m a, Re m- AN: N- AN: a-AAmNe m-Azewmmzfe «LAN: N- Ae: TAN: _- ANBWE
1 > e 8 N- as N- an e- an N:
A
9 NOE-AN: 0 A0: A- 83 N-_A2.vom3 a SAN: 1AA: TAEe NAAEXN:
__> Ao.::-AN3 N+Ao.: N- Aenv A- Adina
9 a N- a N+ Ate N- Av: NAAoNeSEAwoe e- AA: .1 Am: A- AN: A- ABC-N2
2> ANeeNA-ASe o Ass N- ACES-A0322
is .3 Ass .22 as Ade
abscess awe»: eemam amofi awe»: enmeN amoeN
A o
m m

 

 

 

68 .88..” 835.». 3:88 «832 05 28 Aomv
.Bfim 250% 05 E mmOwE e5 0:ch .mmO=N mo mcoumNA-BBQ
28 A758 mafia 390mm A783 3653on 3:23.55

3a 3an

Figure 4.4

129

Raman difl‘erence spectra of ZnOEP (n.a.), ZnOEP-(la,
ZnOEP-”N4, ZnEPI (n.a.) and MgOEP (n.a.). The pump pulse
was set at 572 nm, except for MgOEP, which used 575 nnr, and
preceded the probe pulse (450 nm) by 100 ps. Porphyrins were
dissolved in tetrahydrofirran to approximately 5010" M.

130

 

 

vaw

o mom.
oem. Am.
ill-azvv
on: 8:
NNv_ NNeF
mm? mm? ommw 1‘ me
can. PAN. m.m.
NmN. emN. mN.
mu Aw
w. B
um hm

mmmw

aver

mmmw

mew

thP
4

Romp

omvw
mva

 

mmN

j

V

 

 

 

moEmcoE m>=m_m

0: ZnOEP-"SN

Raman shift (cm'1)

lfigme4A.

131

Figure 4.5 Parallel and perpendicular(---) Raman difference spectra of
ZnOEP (n.a.), ZnOEP-d4, ZnOEP-"N4, ZnEPI (n.a.) and
MgOEP (n.a.). Other than the polarizer, conditions were
identical to those used in Figure 4.4.

 

Relative intensities

 

ZhCEHD-cl4

ZhOEElD
1200

1

r
1300 1400 15H00

Raman shift (cm'1)

Figure 4.5

 

1600

 

133
relatively straightforward by using direct analogy to ground-state vibrations. Shifts are

observed for the v2 (-17cm"), v3 (-35 cm"), v4 (-17 cm'l), vcmms (-6 cm”), and the
vcmm (-3 cm!) bands. The relative intensity of individual peaks differ in the ground and

excited states. In principle, peak intensities contain information about the higher excited
state that is used for the resonance enhancement of the Raman scattering. However, a
complicating factor is that the resonance enhancement depends on the wavelength if the
excitation is not precisely at the 0,0 transition of the S; state to the higher excited state.38
The location of that transition is unknown for S; excited ZnOEP and our excitation
wavelength may differ significantly from it. Further, for the ground state, we are not in
resonance with the 0,0 transition of the Soret band.

The difference spectra of ZnOEP-d4, ZnOEP-”N4, ZnEPI (n.a.), and MgOEP
(n.a.) in Figure 4.4 all preserve the same general features of the S; state spectrum of
ZnOEP. The small changes of the Raman shifts due to isotopic substitution are consistent
with isotopic substitution and can be compared to those of the ground state. The Raman
shifts of the isotopes facilitated the vibrational assignment of the modes in the S; state
(Table 4.1). The intensities of most of the isotopomer, ZnEPI and MgOEP modes are
also very similar. Noteworthy is the low intensity of the v; of MgOEP which is much
weaker in both the ground and excited state than the v:- of ZnOEP. The same is true for
the ethyl mode of ZnEPI at 1259 cm'1 which is also relatively weak in both states. The
preservation of the intensity features in both states indicates that the central atom or the
peripheral substituents can perturb the porphyrin n—electron system in a similar fashion in

difl‘erent electronic states. Figure 4.6 shows the isotopic shifts in the excited state as a

134

Figure 4.6 Relative isotopic shifts in ground and excited states
(from Table 4.1).

135

 

Vs

- V
8 v4CH2-wag W

0’ ——Y=-0.72+1.03*x
I=O.823

'E
3
mV-
.5 '5 "
SE
.C
(D
m D
Q.
.9.
8
- -10 -

 

-15 J 4 I A I A I A I A J A I

   
     

 

 

-12 -10 s -6 -4 -2 0
Isotope shift in 80 (cm")

Figure 4.6

136
firnction of the isotopic shifts in the ground state. The magnitudes of the shifts are of the

same size in both states. Some small differences exist in those of ZnOEP-d4. The shift for

the v2 mode of this isotopomer is 7 cm’1 larger in the S; state than in the ground state.

This suggests a small normal mode change in the S; state. It should be mentioned,
though, that in all S; state spectra the v; is very broad and at the high frequency side may
comprise a depolarized mode (v;o?). A comparable argument is true for the MgOEP
mode that shows a 4 cm'1 larger shift for the S; state v3 mode. This mode is clearly
obscured by a depolarized mode at about 1425 cm". The v4 mode of ZnOEP-d; is the
only other mode for which the isotopic shift in the excited state differs fi'om that in the
ground state. In the former state the isotopic shift is 5 cm", whereas it is only 1 cm" for
the latter state. In general, though, the shifts in both states indicate that there are no
significant normal mode changes upon excitation of the porphyrin to 8;.

Table 4.1 shows that the depolarization values for excited state modes of the
isotopomers are very comparable to the ground state values. Under D4;- symmetry the
totally symmetric modes are a;, and have a depolarization ratio of 1/8. If the symmetry is
lowered to Dzh, as in HZOEP,39 the depolarization ratios of a3 modes, corresponding to a;,
under Drh, can be anywhere between 1/8 and 3/4.“0 Because the polarization values are
almost unchanged in S;, our data do not support such a significant molecular distortion.
If, e. g., the symmetry is reduced to D2;., the b;,, modes become totally symmetric and can
show reduced depolarization values. All the depolarized modes are, however, weak in S;
and partly overlapping with other modes. Therefore, we were not able to determine

depolarization values for these modes.

1 3 7
IV. DISCUSSION

Upon excitation at 450 nm the resonance Raman scattering of ZnOEP in the S;
state resembles the Soret excited resonance Raman spectrum of the So state.37 Totally
symmetric modes dominate both spectra. These modes gain enhancement via the Franck-
Condon mechanism and their intensity is directly related to displacements along these
modes within the excited state that is used for the resonance enhancement. The weakness
of other modes besides the totally symmetric modes suggests that there is only a single

electronic transition in the blue. Other transitions would lead to substantial Herzberg-

 

Teller scattering. The presence of strong polarized modes only in the spectrum of the ;
ZnOEP rc-anion was explained analogously by Pemg and Bocian. '9

There are differences in intensities of the modes in the So and S; state spectra, e. g.,
the v4 dominates the 8;; state whereas its intensity is comparable to that of the other totally
symmetric modes in the S; state. This mode mainly comprises C.N stretching and
indicates that, when going fi'om the So to the 8; state, there are significant electron density
changes along this mode. When going to higher states from S; the changes are smaller
along this mode. The increase of the v3 and the vcmm in S; suggest that in the latter
case there are more electron density changes along the periphery of the macrocycle.
Another explanation for the increase of the v3 in the S; state can be found in its potential
energy distribution (PED). Li et. al. 37" used normal mode calculations to explain the

intensities of the v3 and v2 in the ground state by the mixing of QC, and C.CIn bond

stretches in their PEDs. For the v2 this would lead to a relatively intense peak, but

138
apparently to weak v3 peak intensity. A small change of the normal mode in S; could

change the mixing and explain the more intense v3 in the S; state.

The vibrational fi'equencies of the ethyl modes are almost unchanged in the S;
state. This indicates that upon excitation to S; there is little change in electron density at
the position of the ethyl groups and that for the So and S; states the porphyrin localized
orbitals of Gouterman’s four orbital model are appropriate. '0 As mentioned above, for
totally symmetric modes the resonance enhancement is directly related to the displacement
along that normal coordinate. Our data indicate that transitions to higher excited states
coincide with electron density changes at the ethyl groups. Similar observations were
made with other peripheral substituents. The triplet spectrum20 of Zn(II)
tetraphenylporphyrin is dominated by a phenyl mode at 1595 cm'l. This mode occurs at
almost the same fi'equency as in the ground state and is also unchanged upon addition of

'6’“ or removal of an electron from the a2.- orbital

an electron to the e,3 orbital (anion)
(cation).ll These results indicate that for the lowest excited singlet or triplet states a
porphyrin core localized model is valid, but that, for electronic transitions to higher
excited states, the substituents have to be considered explicitly since there are significant
electron density changes at the periphery of the macrocycle.

The depolarized modes appear to have lower intensity in the S; state. These
modes are enhanced by Jahn-Teller effects in higher excited states.“’42 As mentioned
above, we do not know the resonance enhancement spectrum and do not have a good

calibration of the intensity of the S; state Raman spectrum. It is therefore difficult to

quantitatively compare the intensities in the ground and excited state. The decreased

139

intensity of depolarized modes is similar to the dimunition observed by Pemg and Bocian
on the ZnOEP rt-anion,l9 but a relationship between these observations is not clear, as for
both cases it is not known which excited states are used for the resonance enhancement.
We can use the frequency shifts of ZnOEP in S; to calculate the core size of the
porphyrin in that state if we assume that the experimentally obtained fi'equency/core size
relationships for ground state porphyrins still hold in the excited state. For many modes
there is a linear relationship between their frequency and the core size. ”flit“ The latter
is defined as the distance between the central metal and the pyrrole nitrogen. The
relationship for metal octaethylporphyrins obtained by Anderson et al. ‘5 predicts for
ZnOEP in the S; state a core size of 2. 14, 2.13, or 2.19 A, respectively, depending on

whether the v2, v3, or v; is used as the core size marker. It is noted that these values are
outside the calibrated range and that the error in the core size obtained with the v4 may be

especially substantial, because, for this mode, the core size dependence is only weak. The
core size of ZnOEP in the ground state is about 2.05 A. ” Excitation to the singlet excited
state thus increases the core size by about 0.1 A This value is more than 10 times higher
than that obtained by Prendergast and Spiro46 from semiempirical calculations. The latter
small core size increases were reflected by only small predicted frequency shifts. The 5 %
increase of the core size in the S; state is of similar order to core size increases reported
previously when the oxidation state of the central ion is changed. Findsen et. al.”
reported a 3 % core size change upon photoinduced change of the metal oxidation state

in nickel protoporphyrin DC Furthermore, there is about a 3 % core size difference

between high and low spin irons in six-coordinate heme complexes. 1‘ In these two

140

examples there is no redistribution of the porphyrin rt—electron system. Therefore, a 5 %
increase of the core size in the S; state, where there is significant electron redistribution, is

only small. According to our estimated core size increase the we would shift to about
1575 cm". It is therefore not surprising that no clear v;o was found in the data as this

mode would probably be obscured by the v2 mode. Previously obtained linear

relationships between core size and the frequencies of several modes were also found for
metal-OEP cations, ‘3 although these relationships difl‘er from those of ground state
porphyrins. X-ray crystallography data did not indicate core size difl‘erences between
cations and their neutral parents, whereas there are significant shifts in frequency. This
indicates that fiequency shifts may not necessarily reflect core size changes, but that they
can also be caused by mode composition changes. In that case, however, we do not
expect different modes to predict similar core size changes.

Transient absorption measurements by Rodriguez et al.33 showed that the S; state
absorbs strongly between 400 nm and 500 nm and suggests similarity with the ground
state absorption spectrum. They proposed that the absorption spectrum of S; results fi'om
similar transitions as that of the So state and not fiom promotion of an electron in the es
orbital to a higher orbital. Even though the intensities of the modes in the ground and
excited states are different, the correlation that seems to exist for the intensities of some
peaks corroborates their suggestion.

17'” showed an increase of

Measurements previously performed on ZnTPP triplets
the number of Raman active modes relative to the ground state and depolarized ground

state modes that become polarized in the triplet state. This was attributed to a distortion

141
of the macrocycle along a nontotally symmetric coordinate due to Jahn-Teller effects. In

general, macrocycle distortion will reduce the symmetry and increase the number of
modes, because nontotally symmetric modes can become totally symmetric under reduced
symmetry. ZnOEP in the S; state does not show an increase of polarized a“, modes nor a
change of polarization of the nontotally symmetric bu, modes. This leads to the conclusion
that in the S; state there is no significant change of symmetry of the porphyrin and that the
D4; symmetry of the ground state porphyrin is retained. A static J ahn-Teller distortion, as
previously reported for the ZnTPP triplet, is therefore absent or negligible in S; excited
ZnOEP. It was shown before that J ahn-Teller effects in singlet excited states can be
quenched by configuration interaction. ‘8

Previously Sato et al.23 published resonance Raman spectra of free-base OEP
(HZOEP). Because of the long lifetime of the singlet state of HZOEP, they performed their
experiments with intense nanosecond laser flashes. These data were not analyzed
extensively, but there are similarities between the Raman spectra of the ground states of
ZnOEP and HZOEP, even though this porphyrin has a lower symmetry (02;) than ZnOEP.
Differences are mainly in the modes around 1600 cm". These modes have become totally
symmetric in the reduced symmetry of H20EP and consequently are more enhanced. In

the S; state there is a downshift of v; by 23 cm'1 which is comparable in size to the 18 cm'1
downshift that we found for ZnOEP. In HZOEP the intensity of the v3 is increased in the

S; state and is located at 1445 cm'1 (in ZnOEP at 1450 cm") and has probably shifted

down significantly. It is noted that the authors did not identify this mode as the v3, but

this mode bears clear resemblance with the v; of S; excited ZnOEP. Sato et al. observed

142
a mode at 1596 cm'1 in the S; state and attributed this mode to v2. Consequently they

concluded that v; shifts up in S; excited H20EP, but judging from their data, it can also be
argued that the v2 shifts down in frequency. We attribute modes at 1255 cm'1 and 1316

cm'l to ethyl modes, that correspond to the 1258 cm'1 and 1318 cm'l modes in S; excited
ZnOEP. As in ZnOEP these modes do not change in fi'equency in going from Soto S;.

The v;3 mode that is present in the 8;; states of ZnOEP (1211 cm") and HzOEP (1210

cm") cannot be identified in the S; state of H20EP. Here, we do not present modes at
frequencies lower than 1200 cm'l, but previously!"28 we reported Raman spectra obtained
at lower frequencies. Modes at 1138 and 1096 cm'1 in S; excited ZnOEP seem correlated
to modes at 1121 and 1076 cm'1 of HzOEP. It is noteworthy that for both HZOEP and
ZnOEP the modes at 1076 and 1096 cm", respectively, cannot be clearly related to
ground state modes. It thus seems that the S; excited states of H20EP and ZnOEP have a
number of common characteristics. This is important because S; excited ZnOEP is
potentially subject to Jahn-Teller effects,20 whereas H20EP is not because the degeneracy

ll,16

of the e, orbitals is lifted in the lower symmetry. As shown for ZnTPP n—anion and for

ZnTPP triplet,20 Jahn-Teller effects can lead to large frequency shifts. Similar shifts for
both ZnOEP and H20EP firrther indicate that Jahn-Teller effects are small for S; excited
ZnOEP. They also suggest that the influence of the central metal on the shift is only small.
The two highest occupied molecular orbitals (a;.- and a2.) of metalloporphyrins are
almost accidentally degenerate. It was shown that for metallo-OEPs the energy of the 3;.-
orbital is slightly higher than that of the a2.- orbital. The HOMO of ZnOEP is therefore a;.-.

Previously,”28 we explained our data by assuming that the electronic configurations of

143
ZnOEP changes from a;.a;.. to a;..e8 upon excitation into the first excited singlet state. For

a more accurate description, several configurations should be taken into account for the S;
state. Since the a;.. and a2" orbitals are almost accidentally degenerate, excitation to S;
cannot be described by one electron promotion, but is more properly described by
promoting about 0.6 electron fi'om the a;.. orbital to the e8y orbital and 0.4 electron from
the a2.- orbital to the es, orbital."’“ To gain more quantitative insight into electron
redistribution in the porphyrin macrocycle upon excitation to S;, we used results fiom
calculations on metalloporphin by Sekino and Kobayashi"7 to generate plots of porphyrin
natural orbitals.” These natural orbitals are combinations of molecular orbitals that
describe the electron populations in a given electronic transition. The calculations and
application theory for these natural orbitals are discussed firlly within Chapter 5.

The relative sizes of the atomic orbital coefficients used for calculations on the
molecular orbitals a;.., a2“, and the degenerate orbitals eg- and eKy are shown in Figure
1.6." The composite orbital diagrams for singlet electronic states are shown in Figure
5.3. Electron density changes after excitation to S; and S2, respectively, are shown in
Figure 5.5 and Figure 5.6. Likewise, bond strengths (Figure 5.7 and Figure 5.8) and
changes therein (Figure 5.10 and Figure 5.11) can be estimated with atomic orbital.
coefficients. It should be noted that these electron distribution changes that occur upon
promotion of a single electron are small compared to the the absolute electron population.
As indicated by the overall weakening of the bonds, the calculations predict an overall
expansion of the core, in agreement with our experimental observations. More A

specifically, the calculations predict the largest changes at the C.C... bonds. Of the

144
observed modes, v; has the largest contribution of QC... stretching, and also shows the

largest downshift in S;. The full benefits of natural orbitals as a tool in interpreting excited
state Raman spectra are discussed in Chapter 5. Interestingly, after excitation there is a
small b;, distortion of the porphyrin. The distortion is a result of the different
contributions of the a;-. and a2-- orbitals to the electronic transition. Ifthese orbitals are
accidentally degenerate, the transition can be described as originating fi'om 0.5 electron
promotion fi'om each of these orbitals, and there will be no distortion in that case. All our
data indicate that the distortion is not reflected by a significant symmetry change of the
molecule as this would have increased the number of Raman active modes. Based on
resonance enhancement profiles of depolarized modes Shelnutt et al.50 argued that the S;
state of nickel etioporphyrin I is weakly distorted by dynamic J ahn-Teller effects. This
distortion may well be related to predicted geometry change as calculated.

Besides the calculations, we can also use results obtained on rt-cation and rt-anion
radicals of ZnOEP and ZnTPP, the latter with the a2.- orbital as the HOMO, to predict
fiequency shifts of the vibrational modes in S; based on the structures of the a;-, azu, and
e, orbitals. In Table 4.2 we test whether the vibrational shifts in the S; state of a porphyrin
cannot simply be described as a summation of shifts induced by addition of an electron to
the e, orbital (anion) and shifts induced by loss of an electron fi'om the a;-. or a2.- orbital

(cation). Assuming that the orbital occupancies of ZnTPP and ZnOEP n—cations are
purely a2; and a;.I respectively, we can predict the shifts in the S; state to be 0.6(AZnOEP'
+ AZnOEP’) + 0.4(AZnTPP+ + AZnTPP'). The results are presented in Table 4.2. Since

no proper data for ZnTPP" rt-cation has been reported, we used data of CuTPP+ instead.

145

.Efiefi + 2553 + Emofia + hmofis a 32:28
. .A $18- .8 .NSA .523 53m x .ad .553 EA .Neea ace e93 . Nae--53 .3 .52 .526 exam .5 Ho 1833
$5 .maanu ewe .mefi no.» .35 m.< .203 ”.55 .3900 see 5 .NeNNeNNN .8 .Naa ease exam .5 m d .555 em

.A see-ea: : .m .0285 ”.2 gas... see . .eeNN-oewm ._ : sea 5% see-o as. .N. .o N .eam em .1285 crux .5 is. .m

.882 em 1m #0533580 88w 5 68: mm 838 :> new 2> 05 me $25 05 mo owflog 05 98 .9 New eaten»: noon 8: 02.? o: a

 

 

 

 

 

 

 

 

 

 

 

 

2. cm. N- 4N- a- A:- 9
em- a- N- N- AN- .02 9
8. N- o 2+ 2. Nm- 9
Area Aves Arse Arse Aves Aves see
.2 see .54 .28 .amofiq Eamofia 9.52% tam-:52 Rupees;
.036 _m 05 5 32m mo 82? 3.5938 05 5:5
Ago-89:8 £558 NEH-:0 e5 mmO:N new .325 nAmacN ecu
NEH-AN no 36532... 3:23.53 mo NEAR :o woe-B 25:23an me 2an

 

146
Earlier data for ZnTPP+ by Yarnaguchi et al. 5 ‘ seems not completely reliable, because the

frequency shifts of their cations are very different from data published later on. 11 For

CuTPP+ there is no observed 1); and we used the averaged values of u;;; and 1);;- modes

with similar stretching contributions. Gouterman’s model does not address the substituent
modes. Therefore, they are not included in the table. Table 4.2 shows that, though there
are quantitative differences between predicted and measured fiequencies, the directions of
the shifts agree. However, if configuration interaction is neglected and it is assumed that
for ZnOEP one electron is promoted from the a;-. orbital to the e, orbital as has been
suggested in literature,46 the shift of the u; would be opposite to what is found
experimentally. It should be noted that the metallo-TPP anion radicals are subject to Jahn-
Teller distortion,ll whereas these effects are only small in S; excited porphyrins.18 This
will partly be responsible for the large shifts of some of the anion modes. Predictions
based on semiempirical calculations of porphyrin structures in excited states by
Prendergast and Spiro“S predict frequency shifts much smaller than we observed. They
included geometry optimization but no configuration interaction in their calculations, and
calculated only minor changes of the porphyrin core geometry with consequently only
small predicted shifts of the normal modes.

We will now discuss the change of the linewidths in the S; state. An increase in
linewidth can be caused by different dephasing or by structural inhomogeneity of the
porphyrin sample in S;. In the former case, modes are expected to be Lorentzian, whereas
in the latter case the lines have Gaussian shapes. The differences between these shapes are

mainly in the wings of the modes, and are difficult to assess if several modes overlap. The

147

v4, however, is reasonably isolated with a shape that resembles a Gaussian most. The

linewidth broadening is thus due to inhomogeneous effects. Due to anharrnonicity, an
inhomogeneity can be caused by vibrational population relaxation in the S; excited state
manifold. It then, however, should disappear on the same time picosecond timescale as
the excited vibrations. Spectra obtained at 100 ps, 800 ps, and 2 ns all have the same
linewidths (see Chapter 2, Figure 2.1),28 and indicate that the latter relaxation processes,
which occur within tens of picoseconds,30 are not responsible for the inhomogeneity. The
broader modes, therefore, originate from a structural heterogeneity. Courtney et al.26
observed narrower linewidths in the d,d excited state of nickel(II) protoporphyrin IX
dirnethyl ester and attributed this to nonplanar distortions in the ground state of nickel
porphyrins}2 They suggested that, in the excited state, the conformational heterogeneity
of the macrocycle was less, leading to narrower modes. Our data indicates that for
ZnOEP the conformational heterogeneity increases when going to the excited state. It is
possible that this is a reflection of the theoretically predicted nontotally symmetric

distortion of the porphyrin macrocyle.

(1)

(2)

(3)
(4)

(5)
(6)
(7)
(8)

LIST OF REFERENCES

see e.g., Research in Photosynthesis, Proceedings of the Ixth International
Congress on Photosynthesis; Murata; N.; Ed.; Kluwer Academic Publishers:
Dordrecht, 1992.

Lin, X.; Murchison, H. A; Nagarajan, V.; Parson, W. W.; Allen, J. P.; Williams, J.
C. Proc. Natl. Acad Sci. USA 1994, 91, 10265-10269.

Dutton, P.L; Moser, C. C. Proc. Natl. Acad Sci. USA 1994, 91, 10247-10250.

(a) Vos, M. H.; Lambry, J. C.; Robles, S. J.; Youvan, D. C.; Breton, J .; Martin, J.
Proc. Natl. Acad Sci. USA, 1991, 88, 8885-8889. (b) Vos, M. H.; Rappaport, F.;
Lambry, J.-C.; Breton, J.; Martin, J.-L. Nature 1993, 363, 320-325.

Zhu, L.; Sage, J. T.; Champion, P. M. Science 1994, 266, 629-632
Jortner, J. J. Chem. Phys. 1976, 64, 4860-4867.
Gunner, M. R; Dutton, P. L. J. Am. Chem. Soc. 1989, 111, 3400-3412.

(a) Gustafson, T. L.; Iwata, K.; Weaver, W. L. in Ultrafast Phenomena VII;
Harris, C. B.; Ippen, E. P.; Mourou, G. A.; Zewail, A. H., Eds; Springer-Verlag:
New York, 1990. (b) Hayashi, H.; Kolaczkowski, S. V.; Noguchi, T.; Blanchard,
D.; Atkinson, G. H. J. Am. Chem. Soc. 1990, 112, 4664—4670. (c) Noguchi, T.;
Kolaczkowski, S.; Gartner, W.; Atkinson, G. H. J. Phys. Chem. 1990, 94, 4920-
4926. (d) Hayashi, H.; Brack, T. L.; Noguchi, T.; Tasumi, M.; Atkinson, G. H. J.
Phys. Chem. 1991, 95, 6797-6802. (e) Noguchi, T.; Hayashi, H.; Tasumi, M.;
Atkinson, G. H. J. Phys. Chem. 1991, 95, 3167-3172. (f) Iwata, K.; Hamaguchi,
H. Chem. Phys. Lett. 1992, 196, 462-468. (g) Weaver, W. L.; Huston, L. A.;
Iwata, K.; Gustafson, T. L.; J. Phys. Chem. 1992, 96, 8956-8961. (h) Butler, R
M.; Lynn, M. A.; Gustafson, T. L. J. Phys. Chem. 1993, 97, 2609-2617. (i) Reid,
R. J.; Doig, S .J.; Wickham, S. D.; Mathies, R. A. J. Am. Chem. Soc. 1993, 115,
4754-4763. (j) Reid, P. J.; Lawless, M. K.; Wickham, S. D.; Mathies, R. A J.
Phys. Chem. 1994, 98, 5597-5606. (k) Hamaguchi, H.; Gustafson, T. L. Ann. Rev.
Phys. Chem. 1994, 45, 593-622.

148

(9)

(10)
(11)

(12)

(13)

(14)

(15)

(16)

(17)

(13)
(19)
(20)

(21)

(22)

149

(a) Donohoe, R. J .; Dyer, R. B.; Swanson, B. 1.; Violette, C. A.; Frank, H. A.;
Bocian, D. F. J. Am. Chem. Soc. 1990, 112, 6716-6720. (b) Moénne-Loccoz, P.;
Robert, B.; Ikegarni, I.; Lutz, M. Biochemistry 1990, 29, 4740-4746. (c) Mattioli,
T. A; Hofl‘mann, A.; Robert, B.; Schrader, B.; Lutz, M. Biochemistry 1991, 30,
4648-4654. (d) Shreve, A P.; Cherepy, N. J .; Franzen, S.; Boxer, S. G.; Mathies,
R A Proc. Natl. Acad Sci. USA 1991, 88, 11207-11211. (e) Palaniappan. V.;
Aldema, M. A; Frank, H. A.; Bocian, D. F. Biochemistry 1992, 31, 11050-11058.
(f) Cherepy, N. J .; Shreve, A. R; Moore, L. J .; Franzen, S.; Boxer, S. G.; Mathies,
R A J. Phys. Chem. 1994, 98, 6023-6029.

Gouterman, M. J. Mol. Spectr- 1961, 6, 138-163.

Czernuszewicz, R S.; Macor, K. A.; Li, X.-Y.; Kincaid, J. R; Spiro, T. G. J. Am.
Chem. Soc. 1990, 111, 3860-3869.

Oertling, W. A.; Salehi, A.; Chung, Y. C.; Leroi, G. E.; Babcock, G. T. J. Phys.
Chem. 1989, 93, 1311-1319.

Oertling, W. A; Salehi, A.; Chung, Y. C.; Leroi, G. E.; Chang, C. K.; Babcock, G.
T. J. Phys. Chem. 1987, 91, 5887-5898.

Spaulding, L. D.; Chang, C. C.; Yu, N.-T.; Felton, R. H. J. Am. Chem. Soc. 1975,
97, 2517-2524.

Salehi, A.; Oertling, W. A; Fonda, H. N.; Babcock, G. T.; Chang, C. K.
Photochem. Photobiol. 1988, 48, 525-530.

Atamian, M.; Donohoe, R J.; Lindsey, J. S.; Bocian, D. F. J. Phys. Chem. 1989,
93, 2236-2243.

Reed, R A.; Purello, R; Prendergast, K.; Spiro, T. G. J. Phys. Chem. 1991, 95,
9720-9727.

Gouterman, M. Ann. N. Y. Acad Sci. 1973, 206, 70-83.
Perng, J .-H.; Bocian, D. F. J. Phys. Chem. 1992, 96, 4804-4811

Walters, V. A; de Paula, J. C.; Babcock, G. T.; Leroi, G. E. J. Am. Chem. Soc.
1989, 111, 8300-8302.

Kumble, R; Hu, 8.; Loppnow, G. R; Vitols, S. E.; Spiro, T. G. J. Phys. Chem.
1993, 97, 10521-10523.

Bell, S. E. J.; Al-Obaidi, A H. R; Hegarty, M.; Hester, R. E.; McGarvey, J. J. J.
Phys. Chem. 1993, 97, 11599-11602.

(23)

(24)

(25)

(26)

(27)

(28)

(23)

(30)

(31)

(32)

(33)
(34)
(35)

(36)

150

Sato, S.; Asano-Someda, M.; Kitagawa, T. Chem. Phys. Lett. 1992, 189, 443-
447.

Apanasevich, P. A; Kruglik, S. G.; Kvach, V. V.; Orlovich, V. A. in Time
Resolved Vibrational Spectroscopy V; Takahashi, H., Rd; Springer-Verlag: Berlin,
1992.

Findsen, E. W.; Shelnutt, J. A.; Ondrias, M. R. J. Phys. Chem. 1988, 92, 307-314.

Courtney, S. H.; Jedju, T. M.; Friedman, J. M.; Alden, R. G.; Ondrias, M. R.
Chem. Phys. Lett. 1989, 164, 39-44.

Kreszowski, D. H.; Deinum, G.; Babcock, G. T. Biophys. J. 1994, 66(2), a374.

Kreszowski, D. H.; Deinum, G.; Babcock, G. T. J. Am. Chem. Soc. 1994, 116,
7463-7464.

Deinum, G.; Kreszowski, D. H.; Chang, C. K.; Babcock, G. T. submitted to J- Am.
Chem. Soc.

(a) Rodriquez, J.; Holten, D. J. Chem. Phys. 1989, 91, 3525-3531. (b) Rodriguez,
J .; Holten, D. J. Chem. Phys. 1990, 92, 5944-5949. (c) Bilsel, 0.; Rodriguez, J .;
Holten, D. J. Chem. Phys. 1990, 94, 3508-3512. (d) Rodriguez, J.; Kirmaier, C.;
Holten, D. J. Chem. Phys. 1991, 94, 6020-6029.

Fuhrhop, J .-H.; Smith, K. M. In: Porphyrins and Metalloporphyrins; Smith, K.
M., Ed.; Elsevier: Amsterdam, 1975, pp 765-766, 816-817.

Paine, J. B.; Kirshner, W. B.; Maskowitz, D. W.; Dolphin, D. J. Org. Chem. 1976,
41, 3857.

Rodriguez, J .; Kirmaier, C.; Holten, D. J. Am. Chem. Soc. 1989, 111, 6500-6506.
O’Neil, M. P. Optics Letters 1993, 18(1), 37-38.

(a) Iwata, K.; Hamaguchi, H. Chem. Phys. Lett. 1992, 196, 462-468. (b) Butler, R
M.; Lynn, M. A; Gustafson, T. L. J. Phys. Chem. 1993, 97, 2609-2617. (c) Iwata,
K.; Yamaguchi, S.; Hamaguchi, H. Rev. Sci. Instrum. 1993, 64, 2140-2146. ((1)
Qian, J.; Schultz, S. L.; Bradbun, G. R; Jean, J. M. J. Phys. Chem. 1993, 97,
10638-10644. (e) Yamaguchi, S.; Hamaguchi, H. Chem. Phys. Lett. 1994, 227,
255-260.

Atkinson, G. H.; Brack, T. L.; Blanchard, D.; Rumbles, G. Chem. Phys. 1989,
131, 1-15.

(3 7)

(38)

(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(43)

(49)

(50)

(51)

151

(a) Abe, M.; Kitagawa, T.; Kyogoku, Y. J. Chem. Phys. 1978, 69, 4526-4534. (b)
Li, X.-Y.; Czernuszewicz, R. S.; Kincaid, J. R; Stein, P.; Spiro, T. G. J. Phys.
Chem. 1990, 94, 47-61.

Babcock, G. T.; Callahan, P. M.; Ondrias, M. R; Salmeen, 1. Biochemistry 1981,
20, 959-966.

Merchén, M.; Orti, E.; Roos, B. 0. Chem. Phys. Lett. 1994, 221, 136-144.

(a) McClain, W. M.; J. Chem. Phys. 1971, 55, 2789-2796. (b) Siebrand, W.;
Zgierski, M. Z. in Excited States; Lim, E. C. Ed.; Academic Press: New York,
1979, Vol 4., pp 1-136.

Yamaguchi, H.; Soeta, A.; Toeda, H.; Itoh, K. J. Electroanal. Chem. Interfacial
Electrochem. 1983, 159, 347-359.

Cheung, L. D.; Yu, N.-T.; Felton, R. H. Chem. Phys. Lett. 1978, 55, 527-529.
Shelnutt, J. A.; Majunder, S. A.; Sparks, L. D.; Hobbs, J. D.; Medforth, C. J .;
Senge, M. O.; Smith, K. M.; Miura, M.; Luo, L.; Quirke, M. E. J. .1. Raman.
Spectr. 1992, 23, 523-529.

Sparks, L. D.; Anderson, K. K.; Medforth, C. J .; Smith, K. M.; Shelnutt, J. A.
Inorg. Chem. 1994, 33, 2297-2302.

Anderson, K. K.; Hobbs, J. D.; Luo, L.; Stanley, K. D.; Quirke, J. M. E.; Shelnutt,
J. A. J. Am. Chem. Soc. 1993, 115, 12346-12352.

Prendergast, K.; Spiro, T. G. J. Phys. Chem. 1991, 95, 9728-9736.

Sekino, H.; Kobayashi, H. J. Chem. Phys. 1987, 86, 5045-5052.

(a) Pettke, J. D.; Maggiora, G. M.; Shipman, L. L.; Christofferson J. Mol. Spectr-
1978, 71, 64-84. (b) Edwards, W. D.; Zemer, M. C. Can. J. Chem. 1985, 63,
1763-1772

Levine, I. N. in Quantum Chemistry; Levine, I. N., Ed; Allyn and Bacon:
Massachusetts, 1983, Vol 111, pp 383-384.

Shelnutt, J. A.; Cheung, L. D.; Chang, R C. 0; Yu, N.-T.; Felton, R. H. J. Chem.
Phys. 1978, 66, 3387-3398.

Yamaguchi, H.; Nakano, M.; Itoh, K. Chem. Phys. Lett. 1982, 1397-1400.

152

(52) Shelnutt, J. A.; Majunder, S. A.; Sparks, L. D.; Hobbs, J. D.; Medforth, C. J.;
Senge, M. O.; Smith, K. M.; Miura, M.; Luo, L.; Quirke, M. E. J. J. Raman.
Spectr- 1992, 23, 523-529

CHAPTER 5

STRUCTURAL PREDICTION S FROM NATURAL ORBITAL CALCULATIONS

ABSTRACT

The Gouterman approach has been successfully used to interpret bond
displacements observed in triplet excited state Raman spectra of metalloporphyrins. This
was done through direct comparison of frequency shifts with atomic electron density
changes in individual molecular orbitals. However, configuration interaction between
nearly degenerate electronic states of metalloporphyrins makes the Gouterman approach
inadequate for discussing excited singlet state Rarnan spectra. Therefore, natural orbitals
have been constructed to account for this configuration interaction. This chapter briefly
discusses the selection of the ab initio study used to generate these natural orbitals. The
quantum mechanical calculations necessary for generation of these orbitals are presented.
Pictorial representations for the S;, T;, S;, and T2 states, kindred to the original
Gouterman molecular orbital diagrams, are given and represent both atom electron
densities and bond orders of adjacent atoms. Both diagrams are necessary for more
quantitative interpretation of the spectra presented in Chapter 4. The S; and S2 states are
estimated to undergo a minimal b;3 distortion that disappears only under accidentally .
degenerate conditions. Relative to singlet excited states, the T; and T2 states undergo
large nonuniform changes in electron densities. These composite orbitals illustrate the

inhibition of Jahn-Teller distortions by significant configuration interaction.

153

1 54
I. INTRODUCTION

Once the picosecond Raman experiments detailed in Chapter 2 through Chapter 4
were completed, complications arose in interpreting the Raman features observed for
singlet excited states of zinc(II) octaethylporphyrin and its associated isotopomers.

Triplet excited state studies of metalloporphyrinsl'5 frequently invoked Gouterman
molecular orbitals (M0) to explain the Raman shifts and intensity enhancements observed
in the T; state.6 These arguments were based upon theoretical changes in atomic orbital
coefficients when a single electron from the HOMO state (a;-. or a2“) is promoted into the
LUMO (ea, or en) (Figure 1.6). This approach was particularly convenient when the atom
electron densities were expressed in diagrams. These Gouterman MO’s were simple
pictorial descriptions of atomic orbital coefficients (atomic orbital electron densities)
originally adapted from MO calculations on porphine done by Longuet-Higgins et al. (see
Figure 1.6 for Gouterman molecular orbital diagrams).7 Research studies such as those
done by Walters et al.,1 and Reed et al.2 effectively used these orbitals to make qualitative
arguments about directional shifts in the T; state of zinc tetraphenylporphyrin. To a lesser
degree, other studies by Sato et al.3 and Bell et al.5 have rationalized T; frequency shifts in
lower symmetry free base porphyrins HZOEP and HzTPP by invoking Gouterman orbitals.
This approach becomes clearly inappropriate for interpreting excited singlet state Raman
spectra, since configuration interaction between the the awe, and 32;;83 transitions results in
mixing of the molecular orbitals. Ideally, a fictional orbital could be envisioned that would
include this configuration interaction, while retaining the same basic simplicity of i

Gouterman orbitals. Therefore, by knowing the extent of configuration interaction for

1 55
metalloporphin, we constructed so called “natural” orbitals8 that enabled predicting

electron occupancy changes that occur upon transition to the S;, S;, T;, and T2 excited
states. Natural orbitals are defined strictly in quantum-mechanical terms and their

characteristics are summarized as follows. A composite wavefirnction (1y) can be written

that describes a given electronic state of a porphyrin (in this case metalloporphin). This

wavefiinction can be expressed in terms of linear combinations of MO wavefirnctions (41;).
These MO wavefirnctions are called natural orbitals when the density matrix of 11!, 7(1/ 1'),

has the form

1(1/1') = 21.151010)

where the b;’s reflect relative contributions of each natural orbital to the composite
wavefirnction. Essentially, the natural orbitals reflect the individual electronic transitions
that combine to form the electronic state, and their relative contributions to the composite
wavefunction density matrix are detemrined by the extent of configuration interaction. If
calculated values for atomic orbital coefficients are known, composite orbital diagrams can
not only predict directional frequency shifts of Raman bands, but also relative magnitudes
of electron occupancy changes. Hence, these orbitals also have quantitative predictive
value. We foresee firture applications for rationalizing Raman intensity patterns as well as
inhomogeneities and symmetry changes in excited states.

13-16,17,l8

Since several semi-empirical“;19 and ab initio molecular orbital

calculations have already been done on D;;. porphyrins, it made sense to choose one such

156
study and apply its results towards the generation of these natural orbitals. Our selection

criteria for an appropriate study involved convenience in using the data as well as accuracy
and complete information. Essentially, we needed a complete set of atomic orbital
coeflicients that can be used to generate molecular orbitals, and wave functions that
describe the extent of configuration interaction between the molecular orbitals.

The initial molecular orbital calculations done on porphine by Longuet-Higgins et
al. (Longuet-Higgins H. C.; Rector, C. W.; Platt, J. R. J. Chem. Phys. 1950, 18(9), 1174-
1181) in 1950 were initially very appealing since their calculations contain atomic orbital
coefficients for b;.. and b2; orbitals as well as the a;.., a2“, e;x and eKy orbitals and were the
original basis of the Gouterman MO’s.‘5’7 The molecular orbital diagrams, kindred to the
original Gouterman orbitals, but including the lowest unoccupied b ;.- and b2u orbitals, are
shown in Figure 5.1. These additional orbitals would have been very useful in discerning
the influence of higher unoccupied molecular orbitals upon excited state Raman spectra.
However, this study was rejected for natural orbital generation since several weaknesses
became rapidly apparent. Longuet-Higgins et al. recognized the presence of configuration
interaction in the electronic absorption spectrum but did not account for it in the
calculations, and therefore swayed electron occupancy values at the C. and C;- positions.
Furthermore, the electronegativities of the carbon and nitrogen atoms were treated as
equivalent. That likewise influenced the relative electron occupancies at the C“- and N
positions. Hence, we needed to involve more recent calculations.

In 1978, Petke et al. (Petke, J. D.; Maggiora, G. M.; Shipman, L. L.;
Christofferson, R E. J. Mol. Spectrosc. 1978, 71, 64-84) did more comprehensive ab

initio calculations on magnesium porphine that considered the configurational

157

Figure 5.1 Molecular orbitals for the HOMO, a2.- and a;.., LUMO, es. and
en, and the next two lowest unoccupied states, b;-. and b2“, of
porphine as defined by Longuet-Higgins, H. C.; Rector, C. W.;
Platt, J. R. J. Chem. Phys. 1950, 18(9), 1174-1181.

158

 

 

 

 

 

 

 

Figure 5.1

159

compositions of degenerate rm“ states. ‘7 Interestingly, their calulations for electron

populations and bond orders for the S; and T; states agreed very well qualitatively with
our experimental results for ZnOEP. Bond electron density changes between ground and
excited states did agree with observed fi'equency shifts in the S; state. However, there is
no documentation of the atomic orbital coefficients used, so we could not use these
results. Additionally, the study did not extend into higher excited states, which limits
comparisons to only the lowest singlet and triplet excited states.

Sekino and Kobayashi ( Sekino, H.; Kobayashi, H. J. Chem. Phys. 1987, 86(9),
5045-5052) carried out similiar ab initio calculations on metalloporphin in 1987. Unlike
earlier studies, even those including configuration interaction,“’" Sekino and Kobayashi
found good reproduction of energy levels for excited states. ‘8 They attributed their
success to a screened potential term that incorporates instantaneous polarization effects in
their MO calculations. This term can vary appreciably between distinct electronic states. '8
Additionally, their study presented complete atomic orbital coefficients for the a;.., a2“, es,-
and e8y orbitals, and potential wave functions that expressed the extent of configuration
interaction for the five lowest singlet and triplet excited states. Their molecular orbital
coeficients for the a;-, 82“, es,- and e8y orbitals were previously displayed in Figure 1.6
(Chapter 1). For these reasons, their results were optimal for creating natural orbitals. It
should be noted that semi-empirical studies have also been done, notably a MNDO/3 study
on zinc porphine by Prendergast and Spiro (Prendergast, K.; Spiro, T. G. J. Phys. Chem.
1991, 95, 9728-9736), where geometry optimization was used without inclusion of

configuration interaction. '9 Their calculated atomic orbital coefficients did agree to within

160
10% of Sekino and Kobayashi values, except for the C; values which were generally

higher than the ab initio calculations. However, Prendergast and Spiro’s MNDO/3
calculations predicted core size changes of 0.025 A in the S; state, whereas our
experimental results show core size expansion of 0.1 A (see Chapter 4 for calculation of
core size expansion)?“21 Consequently, as a result of the core size and atomic orbital
coefficients differences, the frequency shifts predicted by Prendergast and Spiro for
electronic states of zinc porphine strongly disagree with our experimental values for
ZnOEP. Therefore, the ab initio results of Sekino and Kobayashi were selected to modify

Gouterman molecular orbitals.

II. CALCULATIONS

The composite diagrams compare electron distributions of specific atoms in the 8;;
ground state of a metalloporphyrin with electron distributions of the same atoms in
electronically excited states. To draw these diagrams, the composite wavefunction
representing the relevant electronic state must be comprised of linear combinations of
natural orbitals that depict individual electronic transitions. Natural orbital contributions
to the composite wavefunction depend upon the extent of CI. Table 5.1 shows MO
orbital assignments and energies for metalloporphin as calculated by Sekino and
Kobayashi.18 Since electronic symmetry rules permit ungerade-gerade or gerade-
ungerade electronic transitions, configuration interaction (CI) can take place between

several molecular orbitals of appropriate symmetries. The two lowest 1m" excited states

of the D.;;I metalloporphin are known to come almost exclusively from la;..4e8 and 3a2u4e8

l 6 1
Table 5.1 Molecular orbital energies of metalloporphin.‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Orbital E/eV
432,, 2.93
6eg 2.80
3b;,_l 2.31
2319 2.11
59, 1.10
3b2uL -0.28
2b;u -1.15
4eg -3.21
332;; '9.25
1a1L '9.58
3eg -11.51
2b2u -12.08
2&2}; -12.19
2eg -12.23
1b;u -12.44
1b2_u -15.63
leg -15.88
13A; '16.10

 

 

 

'calculations by Sekino, H.; Kobayashi, H. J. Phys. Chem. 1987, 86, 5045-5052

162
orbital interactions,22 although higher transitions could be included in the CI with proper

atomic orbital coefficients. The two electronic transitions considered for CI calculations
are both of E; symmetry and form two degenerate pairs of orbital transitions that can be

expressed in the following manner:

Eu(x, y) Ia;.,4egy (0,0) X (e, o) = (o, e)
lam/cgJr (0, 0) X (o,e) = (e, o)
Eu(x, y) 3a2-,egy (e, e) x (e, o) = (e, 0)
3a2uegx (e, e) X (0, e) = (0, e)

For our calculations we considered only the Ew- complement of the degenerate
transition pair, and the appropriate screened potential wavefunctions are shown in Table
5.2. The parentheses represent the symmetry of the orbitals under consideration in (x,y)
directions. For example, Ian-4%, involves promoting one electron from the a;-. orbital,
which is antisymmetric about both the x and y axes, to the e8y orbital, which is symmetric
about the x-axis, but antisymmetric about the y-axis (Figure 5.1). The transition dipole is
along the y-axis and is pemritted due to its symmetry change. The degenerate complement
to this transition is the 101-Meg, which utilizes the x-axis transition dipole. Configuration
interaction for ungerade-gerade and gerade-ungerade transitions can be expressed in

simple bra-ket form when we introduce A as the configuration parameter.

1013) = lalwegv) + A lake-ex) |e,o) : ialuegx) + ’1 lake-git)

163

Nam-meow em .52 see-o 3m x .m images em 55% B 882528.

 

+ A320 AI 22 VN emmvd + meow TaNmmvm Emme- H 55mm N

 

+ Anna 1 seem NAN: - Anne 1 33. NA; 1 5mm s

 

+ Ass. .1 see. $3.0 + Aaee T .89 New; 1 an: N

 

 

+ @on I an; fiend - came Al :3sz mnmed n Vsin: A

 

 

mafia-“82308 036
3:88 wars-32 mo 3030:3332 13:88 eocoouom Nd 2ns-A-

164

Since these two transitions are degenerate, we did calculations based only on the

la, e) transition with the additional limitation that we were considering configuration
interaction (CI) only within the four orbitals a1“, a2", egx and egy. In other words, this is a
two state CI model, where the la, e) transition is composed of a mixture of two electron
configurations, I!) and |2), that express electron population among the four molecular

orbital wavefunctions:

I!) = 1a1u23agu’4egx l2) = [al./302.34%y

Fundamentally, the two possible electron occupancies shown above, ll) and |2),
are the natural orbitals that sum up to form the overall singlet excited state wavefunction
(’Eu). Their relative electron contributions to IE“, are determined by the configuration
parameter (1). Before tracing the procedure derived by Professor J. F. Harrison of
Michigan State University Department of Chemistry and used for generating natural
orbitals to express lEm, remember what we were looking for: molecular orbital diagrams.
The Gouterman orbital diagrams that we wished to emulate typically display relative
electron occupancy amidst individual pz (out-of-plane) atoms in the delocalized porphyrin
ring. Hence, to draw our diagrams effectively, ’E.“ must be written explicitly in terms of

atomic orbital wavefunctions. This required breaking down the natural orbitals, 4), and (fig ,

into linear combinations of MO wavefunctions, (p1, (p2, rp3, (p4 that can be expressed as

slater determinants. The slater determinants are themselves products of the molecular

165

orbital wavefimctions 1a,”, 3a;.,, 4eg,, and 4e”, which in turn are merely linear
combinations of the pz atomic orbitals. The pz atomic orbital electron densities can then be
drawn in Gouterman-type diagrams. These diagrams, which are called composite
diagrams since they are essentially composed of natural orbitals, reflect relative
delocalized 1t electron occupations at and between individual atoms. The following
equations trace the generation of natural orbitals for the lowest excited singlet state.
Other excited electronic states can be derived using identical methodology.

The lowest excited singlet state wavefimction can be expressed as a combination

of natural orbital wavefimctions due to transitions ll) and |2) and then written in

normalized form:

{Eqn 5.1}

IE =|1)+A|2>

"x J1+1V

The natural orbital wavefunctions for the two possible electron promotions are expressed

completely when A, the antisymmetry operator, and spin functions a and [3 are taken into

consideration.

{Eqns. 5.2}

= 2 I afl‘fia = l 2 afl-fla
II) Ala,“ 3a,, 4eg,( f2— ) |2) Ala,“ 3a,, 4eg[——‘/§ ]

166

ll ) and [2) are wavefunctions that can be expanded into slater determinants (to) to fulfill

the antisymmetry requirement for a many-electron molecule,23 but first let:

1a,“ = a 361214 = b 498x = x 4e” =y

Then the determinants are:

{Eqns. 5. 3}

(P: = Aaaaflbwcfi (p; = Aaaafibfixa (P3 = Ababflaayfi (p. = Ababfiafiya

For indistinguishability of promoted electrons to remain true, the following must hold:

 

 

Abaxfl = (bwfljixflba) Abfixa = (bflan-Exabfl).

These two terms can be summed or subtracted and the wave functions separated from the

spin functions.

 

Abaxfl e Abpxa = [bx(aB + 30‘) ‘ ”4043 + [30)]

J3

 

Abmfl_Abfim=[bx(aB-Ba)-xb(Ba-afi)]

J3

167

_ [ bx + xb)(aB - 5a)]
— J3

 

These determinants are incorporated into the individual natural orbitals (d); and 4:2), and

the electronic state wavefunction (w) (’13,, in this case).

{Eqns. 5. 4}

(q’z-‘P2)=|1> (¢3“P4)=|2)

J3 J3

V,=I1>+AI2>___¢,+x¢.
J1+k2 Jl+lz

The electron occupancy probability (7(1/ 1’)) ( or density matrix) is simply the integrated

conjugate product of the electronic state wavefunction fiom equation 5.4.
{Eqns 5. 5}
y(1/1’) = 4 J w * (1,2,3,4) w(1’,2,3,4)dr(2,3,4)

J(¢, + A¢2)(¢, + A¢,)dr(2,3,4)
= 4 1 + A2

= (%+ A)” Mudr +A’J¢2¢2dr}

 

168

where all the cross terms cancel out since natural orbitals d), and 4); always differ by two
orbitals. Expansion of the integral product of natural orbital 4)] into its slater determinant

wavefunctions (rp ) yields:

{Eqn 5. 6}

J‘Mdr: J‘GPIJE (p2){<p, J.;-Walt 0..5J‘<p,qr,dr+05J‘(p,(p2dr

 

The cross terms for the determinants (p, and (p2 also cancel out since they differ by two

orbital spins. Expansion of these determinants into their composite spin firnctions yields
ojjegpm = F{7(aoz(1)aoz(1') + afl(1)afl(l’) + ba(1)ba(1’) + xB(1)x/3(1'))
Integration over the spin variables produce.
0.5.9- <p,<p,dr = 3—{7(2a(1)a(1’)+ b(1)b(1’) + x(1)x(1’))

Similar results can be obtained for the determinant (p; and the results used to solve

equation 5.6.

169

0.5] mar = flaws» b<1>b<o+x<1>x<r>l

 

I¢I¢1d= (2a(1)a(1')+b(1)b(1)+x(1)x(1'))

Likewise, the integral 4); *4); can be calculated and incorporated into equations 5.5 to

express electron probability, 7(1/ 1’) , solely in terms of MO wavefunctions and the

configuration interaction parameter.

{Eqns. 5. 7}

l ¢2¢2d WW >+a(1)a(1 )+ yum: ))

 

 

(be + aa + ){fl}

          

, _ 1
7(1/1)"[1+)3

 

=(1+ W){(2+A"aa+(2/12+1)bb+xx+l’ }

The probability distribution should reflect the n: electron population of the a1“, a2“, eg, and

e33, orbitals in metalloporphin, as indeed it does upon normalization.

 

0.517(1/1)dv)=(1+1fi ]{(2+A’) +(2/‘l’ +1)+1+A’}

r 2 _
(1+1, )4+4A }_

 

170

We next employ the coefficients of equations 5.7 to show electron occupation
changes in the S; state of this natural orbital. The normalization shows that the natural
orbitals do not change in total electron count, rather, relative populations shift according

to the configuration parameter.

{Eqn 5. 8}
1a,, (2 + 112)/(1 +12) 302., (212+1)/(1+l’)
4eg, 1/(1 + )3) 4eg, 112/(1 +12)

The configuration parameter can be calculated by taking the ratio of the transition
coefiicients for the screened potential wavefirnctions derived by Sekino and Kobayashi.18

For example, when we consider the ’Eu, state

_ -0.7612
0.6373

 

= —1.1944

the molecular orbital occupancies from equations 5.8 become 1.4121 electrons for 1a,“,
1.5879 electrons for 3a;.., 0.4121 electrons for 4e“, and 0.5879 electrons for 4e”. This
contrasts with the ground state, where the electron occupancy resides completely in 1a,“

and 3a)., orbitals (2.0 electrons each).

171

The total electron densities can be expressed among the molecular orbitals for both

ground and excited states where R represents coordinates x,y,z.

p“ (R) = 1.4121p(1a,, ) + 1.5879p(3a,, ) + 0.4121p(4eg,) + 0.58 79p(4egv)

ps, (1'?) = 2p(1a,..)+ 29(3a2.)

Apsrso = —o.5879p(1a,, ) - 0.4121p(3a,, ) — 0.4121p(4eg,) — 0.5879p(4egy)

p represents the molecular orbital summation of electron densities over the 24 atomic
orbitals in metalloporphin. Remember that the four electrons we have been discussing are
the 1: electrons involved in Mt“ transitions, so naturally the atomic orbitals comprising
these MO’s come fiom p, orbitals on each individual carbon atom. These MO densities

can be written explicitly to express every atomic orbital. For example, consider the la,“

MO.

24 2
plan = [Eng-101.]

i=1

24

plan = E l ci,la“cj,la,.pr‘pj

l'.j

The c, values are the atomic orbital coefficients provided by Sekino and Kobayashi, and

are the familiar values associated with Gouterman-type MO pictorials (Figure 1.6). ”'6

172

The overall expression for the electron density matrices for the ground (p30) and singlet
excited (ps,) states can be written as the summation of electron densities from an“, azu, egx,

and em, molecular orbitals. Contributions fi'om the MO’s are weighted by the

configuration parameter as expressed in equations 5.7 and 5.8.

{Eqns. 5.9}
24 24
P s, (R) = 22 2 (ci.lalucj.lalu + cr.3a2ucj,3a.2u )pr'pj
i=1 1:1
* 4 1'41210i.101.cj.101. +1-5879cr.3a,.cj.3a,. +
Ps,(R) = 2§{05879c c +04121c c p,p,
i=1 j=l '34“: 134%, ‘ i,4¢” [.400

III. RESULTS AND DISCUSSION

The approach used to express natural orbital electron density for the 81 state
(equations 5.9) was also applied to the S;, T1, and T2 states. Gouterman originally
displayed his four-orbital model pictorially with spheres that reflected the size of atomic
orbital electron densities.6 I have taken a similiar approach, under the following
conditions. Atoms are labeled in Figure 5.2. In all figures, spheres are arbitrarily scaled
with respect to the largest value in a given set. The numbers to the right of the figures
represent arbitrary scaling up to 1. The independent atom electron densities for the So, S;,
S;, T;, and T2 orbitals are one set (Figure 5.3 and Figure 5.4). Figure 5.5 and Figure 5.6
are atomic orbital density differences with respect to ground state So and define a second

set. The bond densities of individual electronic states comprise another set (Figure 5.7,

173

Figure 5.2 Labeling sequence used for natural orbital calculations.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

175

Figure 5 .3 Composite orbital diagrams for atom electron densities of
singlet states of metalloporphin.

176

SINGLET STATE ATOM ELECTRON DENSITIES

 

p52

N 1 0.30
N; 0.27
C“ 0.66
Ca 0.68
Cu 0.13
Cu 0.14
Cu 0.82
N. 0.30
N; 0.32
C... 0.63
Cu 0.61
Cu 0.14
C52 0.13
C... 0.90
N 0.30
C. 0.67
Cb 0.11
C... 0.88

 

Figure 5.3

177

Figure 5.4 Composite orbital diagrams for atom electron densities of
triplet states of metalloporphin.

178

TRIPLET STATE ATOM ELECTRON DENSITIES

 

N. 0.30
N2 0.21
C... 0.69
Caz 0.76
C... 0.12
Cu 0.17
C... 0.73
N. 0.30
N. 0.39
C... 0.60
C... 0.53
C... 0.14
Cu 0.10
C... 1 ‘

 

Figure 5.4

179

Figure 5.5 Composite orbital difference diagrams for atom electron
densities of lowest excited states of rrretalloporphin.

180

LOWEST EXCITED STATE ELECTRON DENSITY

CHANGES

pTl'pSo

 

 

Figure 5.5

-0.01
+0.58
-0.42

-0.87
+0.24
-0.07
+0.74

-0.01
+0.15
-0.19

-O.32

+0.22
+0.10
+0.09

181

Figure 5.6 Composite orbital difference diagrams for atom electron
densities of second lowest excited states of metalloporphin.

182

SECOND LOWEST EXCITED STATE ELECTRON
DENSITY CHANGES

pTz'pSo

N. -0.03
N; -0.57
Ca] +0.19
C32 +0.61
C... +0.1 1
C52 +0.40
(1... -1

N. -0.02
N 2 -0. 18
C31 -0.02
Caz +0.1 1
C51 +0.17
Cb: +0.25
Cm -0042

 

Figure 5.6

183

Figure 5.7 Composite orbital bond order diagram for the singlet ground
state of metalloporphin.

184

GROUND STATE BOND ELECTRON DENSITIES

p50

CaN -0.48
C,C., +0.63
CaCm +0.82

 

'''''

.....
3:2? :5:

Figure 5.7

185

Figure 5.8 Composite orbital bond order diagrams for the two lowest singlet
excited states of metalloporphin.

186

SINGLET STATE BOND ELECTRON DENSITIES

 

Figure 5.8

C31N1
cazNz
Caerr
Ca2cb2
Calcm
CaZCrn
Cberr
CbZCbZ

CalN 1

C:aZNZ

Caerr
Ca2Cb2
CalCm
CaZCm
Cblcb 1
Cb2cb2

-O.57
-O.51
+0.63
+0.73
+0.57
+0.37
-O.27
-0.33

-O.55
-O.62
+0.66
+0.55
+0.44
+0.65
-O.30
-O.23

187

Figure 5.9 Composite orbital bond order diagrams for the two lowest triplet
excited states of metalloporphin.

188

TRIPLET STATE BOND ELECTRON DENSITIES

 

Figure 5.9

CarNr

Ca2N2

Cal Cb1
CaZ-Cbz
CalCrrr
CaZCrn
Cblcbl
Cb2Cb2

Ca.N.
Ca2N2
Calcbl
CaZCbZ
Cal Cm
Cazcm
Cberr
Cb2Cb2

-O.61
-O.37
+0.59
+0.95
+0.72
+0.05
-O.23
-O.45

-0.51
-O.76
+0.70
+0.31
+0.27
+1
-O.35
-O.1 1

189

Figure 5.10 Composite orbital bond order difference diagrams for the lowest
singlet and triplet excited states of metalloporphin.

LOWEST EXCITED STATE ELECTRON DENSITY
CHANGES

pTl'pSo

 

Figure 5.10

C3 1N1
C32N2
Caerr
Cazcbz
Calcrn
C:aZCm
Cberr
Cb2cb2

CalN l

CaZNZ

Calcbl
CaZCbZ
Crrlcm
CaZCm
Cblel
CbZCbZ

-0.04
-O.36
+0.09
-O.42
-O.71
+0.23
-O.14
+0.13

-0.09
-0.18
+0.04
-0.1 1
-O.49
-0.23
-O.13
-0.04

191

Figure 5.11 Composite orbital bond order difference diagrams for the second
lowest singlet and triplet excited states of metalloporphin.

SECOND LOWEST EXCITED STATE ELECTRON
DENSITY CHANGES

pTz'pSo

 

Figure 5.11

Ca.N.
cazNz
Calcbl
Ca2cb2
Calcm
CaZCm
Cberr
Cb2Cb2

CalNl

CazNg

Calcbl
Ca2cb2
Ca1 Cm
CaZCm
Cblel
CbZCbZ

-O.17
+0. 14
-0.05
+0.42
-O.13

-0.03
-0.32

-0. 12
-0.04
-0.01
+0.14
-O.32
-0.58
-0.08
-0.17

193
Figure 5.8, and Figure 5.9). Figure 5.10 and Figure 5.11 are bond density differences with

respect to the ground state (So) and are considered a separate set. Each set provides
slightly different information about the electronic states so we consider each group
separately.

The atom electron densities in Figure 5.3 and Figure 5.4 were essential for
generation of Figure 5.5 and Figure 5.6, but also provided some insight on their own. For
all natural orbitals, electron population resides mainly within the inner ring, chiefly at the
C. and C... positions. Electron distribution changes upon promotion to the S., T., S2, and
T2 states are small in comparison to the absolute electron population. The changes that do
occur, however, suggest a redistribution of electrons to the periphery, primarily Cb
orbitals. The S. and T. states have distribution changes that are associated with the
dominant contribution of electron promotion fi'om the 1a... to 4e.y orbitals. Likewise, the
82 and T2 states have comparable directional changes that can be correlated to their
primary 3a2.. to 4e... electron promotion.

These changes become more apparent in Figure 5.5 and Figure 5.6. The S. natural
orbital in Figure 5.5 shows relatively uniform electron redistribution from C. atoms to
surrounding atoms. For the S. state increased electron density was observed at the N2,
C..., and both C. positions, while the largest decreases in density occur at the C. positions.
The T. state observed similiar changes at the same positions, but the changes in atomic
orbital densities are much larger, particularly at the C.n and N2 positions. The S2 and T2
states in Figure 5.6, however, show decreased populations at the N2 and C... positions,

with increased populations at the C. and C. positions. A small b... distortion is present in

194
the singlet excited states, that only disappears in cases of accidental degeneracy between

the a2. and a... orbitals. This distortion becomes more dominant in the triplet states.
Clearly, absence of Jahn-Teller effects in singlet excited states is strictly true only for
accidentally degenerate a... and a2n orbitals, although the Raman data on
metallooctaethylporphyrins suggest no direct observation of Jahn-Teller modes (see
Chapter 4).2°’2"25

Despite this clear correlation with experimental evidence describing the impact of
configuration interaction upon Jahn-Teller distortions?” these atom electron density
diagrams do not directly address actual bond order changes. So at this point we have no
more insight than we have with the frequently used Gouterman orbitals as applied to
triplet excited states. Indeed, atom electron densities indicate increased electron
populations at the C., positions, which could suggest upward frequency shifis of C..C.,
modes in Raman spectra. Yet, as mentioned previously, our experimental results show
negative fi'equency shifis for the 02 (Chapter 3 and Chapter 4).2°’2"25 For these reasons,
natural orbitals were next applied to bond orders.

As seen in the calculation section, computation of bond densities involved atomic
orbital coemcient products of adjacent porphyrin atoms. In order to relate the atom
electron density results to bond orders, we recall that the coefficient products imply
nonequal sharing of 1t electrons between adjacent atoms for a given natural orbital. Thus,
comparison of predictions from atom electron densities and bond electron densities can be

difficult. As an example of how to interpret these results, we consider the C...C.,. bond in

the S. state. The C...C.,. bond was chosen as an example as this eliminates nonequal

195
sharing between atoms, since the two atomic orbital coefficients magnitudes are

identical.18 The electron bond density for So in Figure 5.7 is negative, -O.20, thus this is
antibonding in nature. Bond order changes for S.-So in Figure 5.10 show a value of -O. 13.
This suggests an increase in antibonding character, which can occur by increased electron
population of an antibonding orbital. Now refer back to Figure 5.5. We see that the atom
density change at the C... position in the S.-So orbital is positive and describes an increased
electron occupancy. Thus, the atom electron density and bond density diagram
predictions are consistent. However, in the absence of these bond order diagrams, we
cannot predict whether the increased electron occupancy at C... implies a bond lengthening
or bond shortening at the C.,C., position. Thus, bond order orbitals became a necessity.
The relative electron bond orders in Figure 5.7, Figure 5.8, and Figure 5.9 were
used for generation of Figure 5.10 and Figure 5.11. The so natural orbital has equal
contributions to bond electron density fi'om the la... and 3a2.. molecular orbitals.
Consequently, bond densities at the C.C., and C..C., positions have predominantly 3...
characteristics, and show bonding and antibonding bonds, respectively. Bond densities at
C.N and C.C... are predominantly a2. in character, and are antibonding and bonding. Note

that this bonding/antibonding label refers only to the four electrons within the 1t orbitals,

not the entire porphyrin. Consistent with atom electron densities, these figures show
relatively small electron redistribution upon promotion of a single electron. Similiar to
atom electron density results, larger changes are predicted for T. and T2 states, and the
relative changes are best seen with difference diagrams. Figure 5.10 is a good example of
the utility of these calculations as a predictive tool for interpreting experimental data. The

S.-So diagram forecasts bond length expansion at the C.C..., C.N, and C.,C., positions, and

196

contrasting changes at the C.C., position. This diagram also estimates relative bond
expansions where C.C... > C.N > C.,C.,. These predictions can be compared to observed
Raman frequency shifts for the 03(C.C...,C.,C.,), u2(C.N,C.C.,), and 04(C.,C.,) modes in the
S. state of ZnOEP (Chapter 4).20 To do this, we must assume the normal coordinate
compositions of these ground state modes are similiar to those calculated for NiOEP.29
Since the isotope data suggests no great mode composition changes in the 8. state
(Chapter 4 and Figure 4.6), the correlation between observed experimental fiequency
shifts of - 35 cm", -20 cm", and -17 cm'1 and natural orbital predictions is valid. The b...
distortion noted earlier, when discussed in terms of nonuniform bond
expansion/contraction, can explain the source of line broadening in the Raman spectra of
higher excited states?"31 This distortion becomes stronger once the porphyrin enters the
triplet state. Similiar to the atom electron density results, the T.-So diagram in Figure 5.10
shows large bond order changes characteristic of Jahn-Teller distortion. The C.2N2,
C.2C.,2, and C..C... bond orders decrease appreciably, indicating a general bond lengthening
in these three regions of the macrocycle. Conversely, the C.2C... and C.,2C.,2 bond orders
become positive, suggesting bond compression as the porphyrin elongates in the x
direction. These predictions suggest that the modes most affected by stretching and

compressing in the T. state are the '04 and 1);. modes. Unfortunately, time-resolved Raman

results for the T. state of ZnOEP could not clearly resolve these peaks (Chapter 3).”"25

Similiar defomrities of the porphyrin are predicted for the S2 and T2 states (Figure 5.11),
but in the y-axis direction. The relative deformities of porphyrins in S., T., S2, and T2

states are drawn schematically in Figure 5.12. The lowest and second lowest excited

197

Figure 5.12 Metalloporphin bond density changes in first and second excited
states. Positive values suggest bond compression, whereas
negative values suggest bond stretching.

 

 

 

 

Bond Changes Relative to Ground State

 

 

 

Figure 5.12

I 0.4
0.2
0.0
-0.2
-0.4
-0.6

-0.8

199

states are related in somewhat of a rrrirror image pattern. This is a direct reflection of the
excited state configurations used within the screened potential wavefunctions (Table 5.1
and Table 5.2).18 For S., the largest implied stretch is at C...Cm positions, whereas for S2
the largest stretch is at C.2C.... Transition to the triplet state causes large changes in bond
densities at specific positions that deviate noticeably fi'om the singlet state bond densities.
For the T. state, bond compression appears particularly strong at the C..2Cm and C.,2C.,2
positions, while bond stretching appears at C.2N2, C.2C..2, and C..C.... The T2 state
observes compression at C..C..., C.2C.,2, and C.2N2 positions, and bond stretching at C.2C...
and C.,2C.,2. The relative distortions in the S., T., S2, and T2 states can be visualized with

the porphyrin structures in Figure 5.13.

IV. FUTURE WORK

Since this is a project that I undertook only at the very end of my graduate career,
I will discuss two areas where these natural orbitals can be applied. One such application
would be experimental studies of D... porphyrins with more or less CI relative to ZnOEP.
The amount of CI in porphyrins can be easily gauged by the relative intensity of Q
band/Soret band absorption peaks in the ground electronic state spectrum. Correlations
could then be drawn between the predicted b .8 distortions in singlet states of different
porphyrins and observed line broadening effects in S. Raman spectra. For example,
metallotetraphenylporphyrins (MTPP) are porphyrins with generally stronger CI than their
metallooctaethylporphyrin (MOEP) counterparts. The Raman spectra of MTPP excited
electronic states can be accessed in a manner similiar to that done for the MOEP’s.1’2’5

However, the initial experiments to detect excited singlet states of zinc and tin

200

Figure 5.13 Representative distortions in S., T., S., and T2 excited
electronic states.

 

 

 

 

 

Figure 5.13

 

 

 

 

 

 

 

202
tetraphenylporphyrin via time-resolved Raman spectroscopy have given incomplete

results.32 Another feasible application is using these natural orbitals to predict
chlorin/chlorophyll and metallochlorin transitions. Marcus theory predicts that the rate of
electron transfer in photosynthesis is strongly dependent on reorganization energy, so
strong deformations in singlet excited states could intuitively reduce the rate of electron
transfer. Knowing the amount of CI, natural orbitals could be used to predict
deformations in these lower symmetry excited states. Since CI is noticeably weaker in the
electronic absorption spectra of chlorins (and chlorophylls), due to symmetry, CI may
perhaps even be ignored. Preliminary experiments on chlorins suggest that singlet state
Raman spectra of metallochlorins may well be very different than those of

metalloporphyrins.33’3"3’

(1)

(2)

(3)
(4)

(5)

(6)
(7)

(3)

(9)
(10)

(11)

(12)

LIST OF REFERENCES

Walters, V. A.; de Paula, J. C.; Babcock, G. T.; Leroi, G. E. J. Am. Chem. Soc.
1989, 111, 8300-8302.

Reed, R A.; Purello, R.; Prendergast, K.; Spiro, T. G. J. Phys. Chem. 1991, 95,
9720-9727.

Sato, S.; Asano-Someda, M.; Kitagawa, T. Chem. Phys. Lett. 1992, 189, 443-447.

Kumble, R.; Hu, 8.; Loppnow, G. R.; Vitols, S. E.; Spiro, T. G. J. Phys. Chem.
1993, 97, 10521-10523.

Bell, S. E. J .; Al-Obaidi, A. H. R.; Hegarty, M.; Hester, R. E.; Mcgarvey, J. J. J.
Phys. Chem. 1993, 97, 11599-11602.

Gouterman, M J. Mol. Spectroscopy 1961, 6, 138-163.

Longuet-Higgins, H. C.; Rector, C. W.; Platt, J. R. J. Chem. Phys. 1950, 18(9),
1174-1181.

Levine, I. N. in Quantum Chemistry; Levine, I. N., Ed; Allyn and Bacon:
Massachusetts, 1983, Vol III, pp 383-384.

Kobayashi, H. J. Chem. Phys. 1959, 30, 1362-1373.

(a) Weiss, C. Jr.; Kobayashi, H; Gouterman, M.; J. Mol. Spectroscopy 1965, 16,
415. (b) Rawlings, D. C.; Davidson, E. R.; Gouterman, M. Theor. Chim. Acta.
1982, 61, 227.

(a) Maggiora, G. M. J. Am. Chem. Soc. 1973, 95, 6555. (b) Loew, G. H.; Rohmer,
M.-M. J. Am. Chem. Soc. 1980, 102, 3655. (c) Loew, G.; Goldblum, A. J. Am.
Chem. Soc. 1980, 102, 3657.

(a) Edwards, W. D.; Zemer, M. C. Int. J. Quantum. Chem. 1983, 23, 1407-1419.
(b) Edwards, W. D.; Zerner, M. C. Can. J. Chem. 1985, 63, 1763-1778.

203

(13)

(14)

(15)

(16)

(17)

(13)
(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

204

(a) Dedieu, A.; Rohmer, M.-M.; Bernard, M.; Veillard, A. J. Am. Chem. Soc.
1976, 98, 3717. (b) Rohmer, M.-M.; Dedieu, A.; Veillard, A. Chem. Phys. 1983,
77, 449.

(a) Spangler, D.; Mckinney, R.; Christofferson, R. E.; Maggiora, G. M.; Shipman,
L. L. Chem. Phys. Lett. 1975, 36, 427. (b) Spangler, D.; Maggiora, G. M.;
Shipman, L. L.; Christofferson, R. E. J. Am. Chem. Soc. 1977, 99, 7470-7478. (c)
Christofi‘erson, R E. Int. J. Quantum Chem. 1979, 16, 573.

(a) Obara, S.; Kashiwagi, H. J. Chem. Phys. 1982, 77, 3155. (b) Saito, M.;
Kashiwagi, J. J. Chem. Phys. 1985, 82, 15.

(a) Petke, J. D.; Maggiora, G. M.; Shipman, L. L.; Christofferson, R. E.
Photochem. Photobiol. 1979, 30, 203 (b) 1980, 31, 243 (c) 1980, 32, 399 (d)
1981, 33, 663 (e) 1982, 36, 383. (f) 1978, 73, 311.

Petke, J. D.; Maggiora, G. M.; Shipman, L. L.; Christofferson, R. E. J. M01.
Spectroscopy 1978, 71, 64-84.

Sekino, H.; Kobayashi, H. J. Chem. Phys. 1987, 86(9) 5045-5052.
Prendergast, K.; Spiro, T. G. J. Phys. Chem. 1991, 95, 9728-9736.

Deinum, G.; Kreszowski, D. H.; Chang, C. K.; Babcock, G. T. submitted to J. Am.
Chem. Soc.

Anderson, K. K.; Hobbs, J. D.; Luo, L.; Stanley, K. D.; Quirke, J. M. E.; Shelnutt,
J. A J. Am. Chem. Soc. 1993, 115, 12346-12352.

Gouterman, M. in The Porphyrins; Gouterman, M., Ed; Academic: New York,
1978, Vol III, p 1.

Levine, I. N. in Quantum Chemistry; Levine, I. N., Ed; Allyn and Bacon:
Massachusetts, 1983, Vol 111, pp 249-250.

Kreszowski, D. H.; Deinum, G.; Babcock, G. T. Biophys. J. 1994, 66(2), a374.

Kreszowski, D. H.; Deinum, G.; Babcock, G. T. J. Am. Chem. Soc. 1994, 116,
7463-7464.

Perrin, M. H.; Gouterman, M.; Perrin, C. L. J. Chem. Phys. 1969, 50, 4137.\

Gouterman, M. Ann. N. Y. Acad Sci. 1973, 206, 70-83.

(23)

(29)

(30)

(31)

(32)

(33)

(34)

(35)

205

Shelnutt, J. A; Cheung, L. D.; Chang, R. C. C.; Yu, N.-T.; Felton, R. H. J. Chem.

Phys. 1977, 66, 3387—3398.
Cheung, L. D.; Yu, N.-T.; Felton, R. H. Chem. Phys. Lett. 1978, 55, 527-530.

(a) Abe, M.; Kitagawa, T.; Kyoguku, Y. J. Chem. Phys. 1978, 69, 4526-4534. (b)
Li, X.-Y.; Czernuszewicz, R. S.; Kincaid, J. R.; Stein, P.; Spiro, T. G. J. Phys.
Chem. 1990, 94, 47-61.

Courtney, S. H.; Jedju, T. M.; Friedman, J. M.; Alden, R. G.; Ondrias, M. R.
Chem. Phys. Lett. 1989, 164, 39-43.

Unpublished data for singlet excited states of zinc(II) tetraphenylporphyrin and
tin(IV) tetraphenylporphyrin (C12) showed strong interference in Raman spectra
fi'om broadband fluorescence centered around 442 run, that coincided with
excitation into excited electronic states. Preliminary fluorescent studies with
variable pump laser intensities suggested possible two-photon processes.

Salehi, A.; Oertling, W. A.; Fonda, H. N.; Babcock, G. T.; Chang, C. K
Photochem. Photobiol. 1988, 48, 525-530.

Procyk, A D.; Bocian, D. F. Annu. Rev. Phys. Chem. 1992, 43, 465-496.
Unpublished data for singlet and triplet excited states of zinc(II)

methyloctaethylchlorin suggests different features and more similarity between the
two excited states than was true for metalloporphyrins.

 

APPENDIX

APPENDIX A

SPECTRUM BASIC SOURCE CODES USED FOR DATA PROCESSING

SEPARATE.PRO

rem Separate.pro

rem Created: 06/08/93 by Doug Kreszowski

rem This program reads a file of frames, separates
rem even frames from odd frames, and stores each separately.
print "SEPARATE",0,0,0,15

let dcolor = 12

rem new c.g.r.

rem suspend macro if running

let tmp = suspendmacro

let tmp = setbasicrunflag(1)

float starttime

float timefactor

int errorflag

rem errorflag set to 1 if error occurs, otherwise 0

let errorflag = 0

let startframe = getpivar(49)

let startwindow = getpivar(50)

let repcont = getpivar(74)

remzz start at first unused curve in window if flag is set
to do so

remr: replay “1:2” with flag when available.

let curvenumber = 1

STRING infile, outfileOdd, outfileEven

let infile = repname

let outfileOdd = " "

let outfileEven = “ "

inp infile, outfileOdd, outfileEven

open infile for input as 1

open outfileOdd for output as 2

open outfileEven for output as 3

long strips

long numframes

rem read strips per frame from file header

let strips=header(1,6)

rem read total number of strips from header

let numframes=header(l,5)

rem calculate total number of frames = total strips/strips
rem.per frame

let numframes=numframes/strips

206

207

rem read from header the type of data stored O—float, 1—
rem long, 2-int
let datatype=header(1,3)
rem read from header whether file is scrambled (scram=0) or
rem unscra
let scram=header(1,7)
rem read in x-axis
let linecount=header(1,4)
rem:: Read spectral direction from header.
let specdir = header( 1, 8 )
let errorexit = 0
if strips <1 then
let errorexit = 1
beep
print "Error: strips < 1",0,0
endif
if numframes <= 0 then
let errorexit = 1
beep
print “Error: Number of frames <= O",0,0
endif
if errorexit = 0 then
rem check for a wavelength calibration file
int wcfile
let wcflag getwcfilename(l, wcfile )
let ytflag ytheader(l, timefactor, starttime)
if wcflag+ytflag = 0 then
rem CLEAR Wavelength Calibration
let tmp initwavcal
let tmp wavcaltype( l, 1.0 )
let tmp makewavcal( 1 )
endif
print “ “,0,0,0,14
print ' "
print ' ",95,40,0,14
rem set default color when in 256 mode
if maxcolors > 16 then
let tmp = setcolor(dcolor)
endif
rem Don't allow more than 9 strips to be replayed
let totalwindows = 2
if totalwindows > 9 then
print ”too many strips to replay in this program.",0,0
let errorflag=1
endif
rem validate startframe
if startframe > numframes then
let currentframe = numframes
else
if startframe < 1 then
let currentframe = 1
else
let currentframe

startframe

r

208

endif
endif
if errorflag = 0 then
int EvenData
int OddDAta
let tmp = setarray( EvenData, datatype, strips,
linecount, O, O, 0 1
let tmp = setarray( OddData, datatype, strips,
linecount, O, 0, O )
float SumOddData(strips, linecount)
float SumEvenData(strips, linecount)
let SumOddData = 0.0
let SumEvenData = 0.0
rem check if spectral direction is right to left
if specdir = 1 then
int rs
let tmp = setarray(rs, datatype, strips,
linecount, O, 0, 0)
endif
rem set the number of windows to the number of
rem strips per window

if iswindow(totalwindows) = 0 then
let tmp = setscreen(1,totalwindows)
endif

for i=startwindow to totalwindows
if getwindowsize(i) # linecount then
while iscurve(i,1) = 1
let tmp = delcurve(i,1)
wend
rem if windows size changes then we must start
rem at curve 1
rem if you want another result use recallm.rep
let curvenumber = 1
let tmp=windowsize(i,linecount)
endif
next i
let tmp=setscreen(4,0)
print “name = ”,0,0,0,14
print repname
print ”total frames = ",9S,40,0,14
print numframes
print ” strips 2 "
print strips
print " lines = "
print linecount
rem for 1 frame/1 strip and set to MANUAL MODE
if numframes = 1 then '
Print “ Number of frames=1"

endif
rem make sure frames is even
let frames = (numframes/2) * 2

for index = 1 to frames step 2
let tmp=seekframe(1,index)

 

209

read 1, OddData
let tmp=seekframe(1,index+1)
read 1, EvenData
if scram = 0 then
let OddDatazunscramble(OddData)
let EvenDatazunscramble(EvenData)
endif
rem check if spectral direction is right to
rem left
if specdir = 1 then
let tmp = reverse(OddData,rs)
let OddData = rs
let tmp = reverse(EvenData,rs)
let EvenData = rs
endif
rem Sum up odds and Evens.
let SumOddData = SumOddData + OddData
let SumEvenData = SumEvenData + EvenData
print “frame = “,0,30,0,14
print index,0,40,0,15
next index

let w(1,1) = SumOddData

let w(2,1) = SumEvenData

write 2, SumOddData

print ' ",95,40,0,14
write 3, SumEvenData

print ' ".95,4o,0,14

let tmp=transfer(1,2)
rem set x-axis size
let a=setheader(2, 0, linecount)
rem set accumulation to # stripes added to program
let a=setheader(2, 2, 1)
rem set data type to float (0)
let a=setheader(2, 3, O)
rem set x-dimension
let a=setheader(2, 4, linecount)
rem set total number of stripes in file
let a=setheader(2, 5, l)
rem set the number of stripes in the new frame
let a=setheader(2, 6, 1)
rem set the unscramble flag
let a=setheader(2, 7, 1)
rem set the reverse flag
let a=setheader(2, 8, 0)
endif rem: End of “if errorexit = 0 then“
ENDIF
close 1
close 2
ram restart macro if it was running
let tmp = restartmacro

i'ifii'r

 

210

QUARTILES.PRO

rem
rem
rem
rem
prin
let
rem
rem
let
let
floa
floa
int
rem
let
let
let
let

rem::
rem::
rem::

let
STRI
let
let
let
let
let
inp
open
open
open
open
open
long
long
rem
let
rem
let
rem
rem
let
rem
rem
let
rem
rem
let

Quartiles.pro

Created: 08/11/93 by Doug Kreszowski

This program reads a file of frames, separates

and stores these frames into four different files

t 'QUARTILES:DHK“,0,0,0,15

dcolor = 12

new c.g.r.

suspend macro if running

tmp = suspendmacro

tmp = setbasicrunflag(1)

t starttime

t timefactor

errorflag

errorflag set to 1 if error occurs, otherwise 0
errorflag = 0

startframe = getpivar(49)

startwindow = getpivar(50)

repcont = getpivar(74)

start at first unused curve in window if flag is set

to do so

replay “1:2“ with flag when available.

curvenumber = 1

NC infile, outfilel,outfile2,outfile3,outfile4

infile = repname

outfilelz" "

outfile2=“ "

outfile3=" "

outfile4=' "
infile,outfilel,outfile2,outfile3,outfile4
infile for input as 1
outfilel for output as
outfile2 for output as
outfile3 for output as
outfile4 for output as
strips

numframes

read strips per frame from file header
strips=header(1,6)

read total number of strips from header

numframes=header(l,5)

calculate total number of frames = total strips/strips

per frame

numframes=numframes/strips ‘

read from header the type of data stored O-float, 1—
long, 2-int

datatype=header(1,3)

read from header whether file is scrambled (scram=0) or

unscra

scram=header(l,7)

U'IIbUJN

rem
let

let
let

211

read in x-axis

linecount=header(1,4)

rem:: Read spectral direction from header.
specdir = header( 1, 8 )

errorexit = 0

if strips <1 then

let errorexit = 1
beep
print "Error: strips < 1",0,0

endif
if numframes <= 0 then

let errorexit = 1
beep
print “Error: Number of frames <= 0”,0,0

endif
if errorexit = 0 then

rem

rem check for a wavelength calibration file
int wcfile
let wcflag = getwcfilename(1, wcfile )
let ytflag = ytheader(1, timefactor, starttime)
if wcflag+ytflag = 0 then
rem CLEAR Wavelength Calibration
let tmp initwavcal
let tmp wavcaltype( 1, 1.0 )
let tmp makewavcal( 1 )
endif
print " ”,0,0,0,14
print " "
print ' “,95,40,0,14
set default color when in 256 mode
if maxcolors > 16 then
let tmp = setcolor(dcolor)
endif
rem Dont't allow more then 9 strips to be replayed
let totalwindows = 4
if totalwindows > 9 then
print "too many strips to replay in program.",0,0
let errorflag=1
endif
rem validate startframe
if startframe > numframes then
let currentframe = numframes
else
if startframe < 1 then
let currentframe = 1
else
let currentframe = startframe
endif
endif
if errorflag = 0 then
int Firstdata
int Seconddata
int Thirddata

212
int FourthData

let tmp:
setarray(Firstdata,datatype,strips,linecount,
O, 0, 0)
let tmp:
setarray(Seconddata,datatype,strips,linecount, O,
O, 0)
let tmp:
setarray(Thirddata,datatype,strips,linecount, O,
0, 0 )
let tmp:
setarray(Fourthdata,datatype,strips,linecount, 0,
O, 0 )

float Sumfirstdata(strips, linecount)
float Sumseconddata(strips, linecount)
float Sumthirddata(strips, linecount)
float SumfourthData<strips, linecount)
let Sumfirstdata = 0.0
let Sumseconddata = 0.0
let SumThirddata = 0.0
let SumfourthData = 0.0
rem check if spectral direction is right to left
if specdir = 1 then
int rs
let tmp = setarray(rs, datatype, strips,
linecount, O, O, 0)
endif
rem set the number of windows to the number of
rem strips per win

if iswindow(totalwindows) = 0 then
let tmp = setscreen(1,totalwindows)
endif

for i=startwindow to totalwindows
if getwindowsize(i) # linecount then
while iscurve(i,1) = 1
let tmp = delcurve(i,1)
wend
rem if windows size changes then we must start
rem at curve 1
rem if you want another result use recallm.rep
let curvenumber = 1
let tmp=windowsize(i,linecount)
endif
next i
let tmp=setscreen(4,0)
print ”name = ',0,0,0,14
print repname
print "total frames = “,95,40,0,14
print numframes
print ' strips = ”
print strips
print ' lines = ”
print linecount

213

rem for l frame/1 strip and set to MANUAL MODE
if numframes = 1 then
Print ” Number of frames=1"

endif
rem.make sure frames is even
let frames = (numframes/4) * 4

for index = l to frames step 4
let tmp=seekframe(1,index)
read 1, Firstdata
let tmp=seekframe(l,index+l)
read 1, Seconddata
let tmp=seekframe(l,index+2)
read 1, Thirddata
let tmp=seekframe(1,index+3)
read 1, Fourthdata
if scram = 0 then
let Firstdatazunscramble(Firstdata)
let Seconddatazunscramble(Seconddata)
let Thirddatazunscramble(Thirddata)
let Fourthdatazunscramble(Fourthdata)
endif
rem check if spectral direction is right to
rem left
if specdir = 1 then
let tmp = reverse(Firstdata, rs)
let Firstdata = rs
let tmp = reverse(Seconddata,rs)
let Seconddata = rs
let tmp = reverse(Thirddata, rs)
let Thirddata = rs
let tmp = reverse(Fourthdata, rs)
let Fourthdata = rs
endif
rem Sum up odds and Evens.
let Sumfirstdata = Sumfirstdata + Firstdata
let Sumseconddata = Sumseconddata + Seconddata
let SumThirddata = Sumthirddata + Thirddata
let Sumfourthdata = Sumfourthdata + Fourthdata
print “frame = ",O,30,0,14
print index,0,40,0,15
next index

let w(1,1) = Sumfirstdata

let w(2,1) = Sumseconddata

let w(3,1) = Sumthirddata

let w(4,1) = Sumfourthdata

write 2, Sumfirstdata

print “ ",95,40,0,14
write 3, Sumseconddata ‘
print " ",95,40,0,14
write 4, Sumthirddata

print " ",95,40,0,l4
write 5, Sumfourthdata

print ' ",95,40,0,14

 

214

let tmp=transfer(1,2)
rem set x-axis size
let a=setheader(2, O, linecount)
rem set accumulation to # stripes added to program
let a=setheader(2, 2, 1)
rem set data type to float (0)
let a=setheader(2, 3, O)
rem set x-dimension
let a=setheader(2, 4, linecount)
rem set total number of stripes in file
let a=setheader(2, 5, 1)
rem set the number of stripes in the new frame
let a=setheader(2, 6, 1)
rem set the unscramble flag
let a=setheader(2, 7, 1)
rem set the reverse flag
let a=setheader(2, 8, 0)
endif rem: End of "if errorexit = 0 then“
ENDIF
close 1
close 2
rem restart macro if it was running
let tmp = restartmacro

FRMEDIT.PRO

REM FRMEDIT

rem Adaptation of REPLAY.PI modified by Doug Kreszowski
rem 08/11/93.

PRINT “FRAMEEDITORzDHK/08/11/93“,0,0,0,15

rem This program replays frames, displaying each stripe in
rem its own window. Thus it CAN NOT be used with frames
rem larger than 10 strips. It takes control of your keyboard
rem and redefines the following keys. This program then
rem selection of frames to save in a second output file.
rem PgDn - backup frames by 10.

rem PgUp - advance frames by 10.

rem HOME - Goto lst frame.

rem END - Goto last frame.

rem UP ARROW — advance 1 frame.

rem DOWN ARROW - backup 1 frame.

rem ESC - leave program.

rem filename — name of file to read data from

rem startframe - frame to start at

rem S— command to save frame

rem set a default color if display is in 256 mode

let dcolor = 12

rem new c.g.r.

215

rem suspend macro if running
let tmp = suspendmacro
let tmp = setbasicrunflag(1)
float starttime
float timefactor
int errorflag
rem errorflag set to 1 if error occurs, otherwise 0
let errorflag = 0
let startframe = getpivar(49)
let startwindow = getpivar(SO)
let repcont = getpivar(74)
rem:: Start at first unused curve in window if flag is set
rem:: to do so
rem:: replay '1
let curvenumber =
if 1 = 2 then
let curvenumber = 1

= 2” with flag when available
1

while iscurve( startwindow, curvenumber ) = 1
let curvenumber = curvenumber + l
wend
endif

STRING infile,outfile
let infilezrepname
let outfile=' “
inp infile,outfile
open infile for input as 1
open outfile for output as 2
rem open file as input assign filr to file handle number
long strips
long numframes
rem read strips per frame from file header
let strips=header(1,6)
rem read total number of strips from header
let numframes=headerll,5)
rem calculate total number of frames = total strips / strips
per frame
let numframes=numframes/strips
rem read from header the type of data stored O—float, 1-
rem long, 2-int
let datatype = header(1,3)
rem read from header whether file is scrambled (scram=0) or
rem unscram
let scram:header(1,7)
rem read in x axis
let linecount = header(l,4)
rem:: Read spectral direction from header

let specdir = header( 1, 8 )
let errorexit = 0
if strips <= 0 then

let errorexit = 1

beep

print ”Error: strips <= 0“,0,0
endif

216

if numframes <= 0 then
let errorexit = l
beep
print “Error: Number of frames <= O”,0,0
endif
rem:: Don't do anything else if we have a strips or
rem:: numframes error.
if errorexit = 0 then
int wcfile
rem chk for a wavelength calibration file
let wcflag = getwcfilename(1, wcfile )
rem chk for a YT_FILE and set Wavelength Calibration
let ytflag = ytheader(1, timefactor, starttime)
rem reset x-axis if no wavelength calibration or YT
rem involved
if wcflag+ytflag = 0 then
rem CLEAR Wavelength Calibration
let tmp initwavcal
let tmp wavcaltype( l, 1.0 )
let tmp makewavcal( 1 )
endif
print ' ",0,0,0,l4
print " "
print ' ",95,40,0,14
rem set default color when in 256 mode
if maxcolors > 16 then
let tmp = setcolor(dcolor)
endif
let totalwindows = strips + startwindow — 1
rem Don't allow more then 10 strips to be replayed
if totalwindows > 10 then
print ' too many strips to replay this program.',0,0
let errorflag=1
endif
rem validate startframe
if startframe > numframes then
let currentframe = numframes
else
if startframe < 1 then
let currentframe = 1

else
let currentframe = startframe
endif
endif
if errorflag = 0 then
int 3
let tmp = setarray( s, datatype, strips, linecount,
O, O, 0 )

rem check if spectral direction is right to left.
if specdir = 1 then
int rs
let tmp = setarray(rs, datatype, strips,
linecount, O, O, O)

217

endif
rem set the number of windows to the number of
rem strips per window.

if iswindow(totalwindows) = 0 then
let tmp = setscreen(1,totalwindows)
endif

for i=startwindow to totalwindows
if getwindowsize(i) # linecount then
while iscurve(i,1) = 1
let tmp = delcurve(i,1)
wend
rem if window size changes then we must start at
rem curve 1.
rem if you want another result use recallm.rep
let curvenumber = 1
let tmp=windowsize(i,linecount)
endif
next i
let tmp=setscreen(4,0)
print “name = ",0,0,0,14
print repname
print "total frames = ",95,40,0,14
print numframes
print ' strips 2 "
print strips
print ' lines = “
print linecount
rem initialize some variables for later use
let lastkey=0
let done = 0
let inkey = 77
let actionflag = 1
rem for 1 frame/1 strip and set to MANUAL MODE
if numframes = 1 then
if strips = 1 then
let repcont = 0
endif
endif
getkbrd
while done = 0
if repcont = 1 then
rem look for ALT-M manual mode
if inkey = 12800 then
let repcont = 0
rem: currentframe was incremented
rem: before checking for ALT—M so
rem: decrement it before going into
rem: manual mode.
if currentframe = 1 then
let currentframe = numframes
else
let currentframe = currentframe - 1
endif

218

else
rem continous replay is picked
rem read in frame
let tmp=seekframe(1,currentframe)
read 1, s
if scram = 0 then
let s=unscramble(s)
endif
rem check if spectral direction is right
rem to left.
if specdir = 1 then
let tmp = reverse(s,rs)
let s = rs
endif
rem chk if YT file and update
rem calibration.
if ytflag # 0 then
if currentframe = 1 then
let tmp = ytheader(l,timefactor,

starttime)
else
let tmp =ytaddoffset(timefactor)
endif
endif

rem set each window to a stripe in frame
for j=1 to strips
let w(startwindow + j — 1,
curvenumber) = s(j)
next j
print “frame = ',O,30,0,14
print currentframe,0,40,0,15
print ” '
if currentframe < numframes then
let currentframe = currentframe + 1
else
let currentframe = 1
endif
rem chk for only 1 frame and set to
rem.manual.
if numframes = 1 then
let repcont = 0
endif
endif
else
rem manual replay picked
if inkey > 0 then
rem look for ALT-A auto mode
if inkey = 7680 then
let repcont = 1
endif
Rem look for S key
If inkey = 8051 Then
let tmp=seekframe(1,currentframe)

219

let tmp=writeframe(2,s)
endif
REM check for DOWN ARROW
IF inkey = 20480 THEN
LET currentframe = currentframe—1
IF currentframe < 1 THEN
let currentframe = numframes
ENDIF
LET actionflag = 1
ENDIF
rem look for END
if inkey = 20224 then

let currentframe = numframes
let actionflag = 1

endif

rem look for HOME

if inkey = 18176 then
let currentframe 1

let actionflag = 1
endif
rem look for PgDn
if inkey = 20736 then
let currentframe = currentframe — 10
if currentframe < 1 then
let currentframe = numframes
endif
let actionflag = 1
endif
rem look for PgUp
if inkey = 18688 then
let currentframe = currentframe + 10
if currentframe > numframes then
let currentframe = 1
endif
let actionflag = 1
endif
rem check for UP ARROW
if inkey = 18432 then
let currentframezcurrentframe+1
if currentframe > numframes then
let currentframe = 1
endif
let actionflag = 1
endif
rem if a function was requested
if actionflag # 0 then
rem read in frame .
let tmp=seekframe(1,currentframe)
read 1, s
if scram = 0 then
let s=unscramble(s)

endif

220

rem check if spectral direction is
rem right to left.
if specdir = 1 then
let tmp = reverse(s,rs)
let s = rs
endif
rem chk if YT file and update
rem calibration.
if ytflag # 0 then
let tmp = ytheader(1,timefactor,
starttime)
if currentframe # 1 then
let tmp =ytaddoffset
(timefactor*
(currentframe—1))
endif
endif
rem set each window to a stripe in
rem frame.
for j=1 to strips
let w(startwindow + j - 1,
curvenumber ) = s(j)
next j
print "frame = “,O,30,0,14
print currentframe,0,40,0,15
print ” "
print "name = ',0,0,0,14
print repname
print "total frames = ",9S,40,0,l4
print numframes
print " strips = "
print strips
print " lines = "
print linecount
let actionflag = 0
endif
endif
endif
rem get key from keyboard
let inkey=getkey
rem check for ESCAPE
if inkey = 283 then
let done = 1
endif
rem look for ?, Q or q — and display“replay.hlp"
if (inkey=1363l)or(inkey=4177)or(inkey=4209)then
let tmp=helpdisplay(1)

endif
wend
if getpivar( 23 ) = 1 then

let tmp = cpyimageptr( 4, s, getcurrentimage )
endif

endif

221

rem:: End of "if errorexit = 0 then"
endif
rem:: File was opened before errorexit check, so must be
rem:: closed here.
close 1
rem:: If we're NOT error exiting, we've locked the
keyboard, so release it
if errorexit = 0 then
rem:: check if spectral direction is right to left.
if specdir = 1 then
erase rs
endif
releasekbrd
endif
print " ",95,40,0,14
rem new c.g.r.
rem restart macro if it was running
let tmp = restartmacro
close 2

 

 

”11111111111111111“