r".
.
'7‘

.L‘ Y-.
v

3:

< : c 4‘: A
‘t‘a'a'auk‘ai

. 4-

" u)

 

M?W

4".mult
V
‘ u

'3
"6
. __
‘ 7‘.
1,:
E:
_ a

v

"V Hund'nmi

.—
H

.-

'7

-‘

   
      
  
   
    

 

.23.: ’0» .‘A. ,
6:“: “)3"? '.-.Yf’{‘-'
"‘Wmfifi ~‘

at l1:
7?; ‘

*3. 3M

‘1’?

u. k

#1.. ,

\ 1‘ '
.1 t '33”: .
$11,,“

a." “uh I v‘fi‘r‘v‘pfi‘
‘ “5:“- - 33g 4

V
I"

v
x
>
b

     
        

 
 
    
    
 

    

 

. . ‘ l A _ o 4:12;!
'- ' ‘ . i ‘ " ,.I'-' 1%":-
C. , A' . M ".4 k h w ‘
V93 .. _ ‘ . .- ’_ j)"; m g" , - . H x .' 3“,

- O- ':: \ ‘ ~
- Z ’ I ; I
. , '4‘ ‘ H.
— -’ ' '5 .‘ ,
_ ' "H" 1‘
V - ’ f ‘ ‘ '-‘ . , ‘L ‘

I‘ - - . _ t - -‘ h. ' 55$" . ‘Ih [unhmai

‘6 - ‘ ll. ‘ L‘ ‘7 ' “‘- nfiff.‘ u.

“as.“ .3

 

LIBRARY
Michigan State
University

 

This is to certify that the

dissertation entitled

Vibrational, Electronic and Structural Properties of High

Valent Catalytic Intermediates of Horseradish Peroxidase

presented by
W. Anthony Oertling

has been accepted towards fulfillment
of the requirements for

[40 degreem CAQM/s/‘v/

M/ 7, Mm

Major professor

 

"(II-$.- T— .:_._ A - .- In , I “Ln—l 0.1277,

 

 

 

 

 

bViSSI,J RETURNING MATERIALS:
Place in book drop to

LJBRARJES remove this checkout from

u. your record. FINES will

 

 

be charged if book is
returned after the date
stamped below.

 

 

 

 

VIBRATIONAL, ELECTRONIC AND STRUCTURAL PROPERTIES OF HIGH

VALENT CATALYTIC INTERMEDIATES OF HORSERADISH PEROXIDASE

By

W. Anthony Oertling

A DISSERTATION

Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1987

ABSTRACT

VIBRATIONAL, ELECTRONIC AND STRUCTURAL PROPERTIES OF HIGH

VALENT CATALYTIC INTERMEDIATES OP HORSERADISH PEROXIDASE

By

W. Anthony Oertling

The active site of the two-electron oxidized compound I intermediate of
horseradish peroxidase (HRP-I) is usually described as an oxoferryl porphyrin
n cation radical. In order to assess separately the effects of each oxidation
equivalent on the structural, electronic and vibrational properties of this heme
complex we present a resonance Raman (RR) characterization of a series of
model compounds. From comparison of the FeO stretching frequency, v(Fe0),
of synthetic oxoferryl hemes and oxoferryl heme proteins the predominant
determinant of the v(FeO) is identified as a trans-ligand effect, with a lesser
effect coming from H-bonding of the oxo ligand. Through a separate RR study
of metalloporphyrin Tr cation radicals (MP+°) the vibrations of the oxidized
porphyrin macrocycle are analyzed and compared to those of the parent
metalloporphyrin (MP). The frequency of the vibrational modes in the 1450-1700
cm"1 range of the MP“ are found to be an inverse linear function of porphyrin
center-to-pyrrole nitrogen distance, similar to that which describes the
vibrational frequencies of the parent compound. Thus, as is the case for the

MP, the high frequency vibrations of the MP4” reflect porphyrin core geometry.

W. Anthony Oertling

In the enzymatic cycle, HRP-I is reduced by substrate to the more stable
HRP—ll. These two intermediates are thought to contain similar oxoferryl
structures. Based on RR measurements of the v(FeO) by others, the oxo ligand
of HRP-II is thought to be H-bonded at pH 7 but not as pH 11. We present RR
scattering from a flowing HRP-I sample prepared by rapid mixing. The v(FeO)
is very similar to that of HRP-II at high pH. This is taken to suggest that, unlike
HRP-II, there is no H-bond to the oxo ligand of HRP-I at pH 7. A mechanism,
consistent with these RR observations and past kinetic studies, is proposed
for the reduction of HRP-I by p—cresol. The frequencies we measure for the
in—plane vibrations of HRP-I, however, are not characteristic of a MP4". Possible

explanations for this result are discussed.

ii

ACKNOWLEDGMENTS

For their various contributions to this work, my thanks is given to: the
members of my family for support and encouragement; G.T. Babcock, G.E.
Leroi and C.K. Chang for critical reading and guidance; Jerry Babcock, Pat
Callahan and John McMahon for instruction in the measurement of Raman
scattering; the entire supporting staff of the Chemistry Department, in particular
Ron Haas and Deak Watters, for technical services; Dwight Lillie, Bob Kean,
Harold Fonda and Ralph Thiim for computer programming. Special thanks is
extended to my "partner", Asaad Salehi, for his sincerity, enthusiasm and
perserverance. Thanks also goes to Carol Smith, Sharon Corner, Margy Lynch
and Bill Draper for their efforts in the preparation of this manuscript. Kris
Babcock and Mari’a Centeno are acknowledged for their fine cooking and
Anheuser-Busch and Miller Breweries for their fine products, without which

this work would have been much more difficult.

iii

TABLE OF CONTENTS

Page
LIST OF TABLES ...................................................................................... vi
LIST OF FIGURES .................................................................................... viii
Chapter 1 - GENERAL INTRODUCTION . .................................................... 1
Significance of the Oxoferryl and Porphyrin n Cation ,
Radical Structures ................... ......... ................................................. 3
Aim and Strategy of the Thesis ........ . ................................................... 5
Metalloporphyrin 1r Cation Radicals and Compound I ............................. 5
References ....................................................................................... 14
CHAPTER 2 - CHARACTERIZATION OF SIX-COORDINATE
OXOFERRYL PROTOHEME BY RESONANCE
RAMAN AND OPTICAL ABSORPTION SPECTROSCOPY
Summary ............................ ...................................................... 16
Acknowledgments ............................................................................... 21
References 6C Notes .................................... . ...................................... 22
CHAPTER 3 - RESONANCE RAMAN SPECTROSCOPIC DETECTION
OF DEMETALLATION OF METALLOPORPHYRIN 1r
CATION RADICALS
Summary ............................... . ............................................................ 24
Acknowledgments .............................................................. . ................ 31
References 6C Notes ............................................................................ 32
CHAPTER 4 VIBRATIONAL, ELECTRONIC AND STRUCTURAL
PROPERTIES OF COBALT, COPPER AND ZINC
OCTAETHYLPORPHYRIN n CATION RADICALS
Summary .................................................... .... ................................... 34
Introduction ...................................................................................... 35
Experimental ........................................... . ................ . ....................... 38
Results
Effects of oxidation on the high frequency
porphyrin core vibrations ............................................................. 39
One-electron oxidation of CuOEP and ZnOEP ................................ 40

iv

P88!

Porphyrin diacid impurities . ........... ..... . ...... ............. ..... 46
One- and two-electron oxidation of CoOEP ......... ............ ...... 52
Neal. uv RR Spectra 00000000000 oooooooooooooo o oooooooooooooooooooooooooooooooooooooo 58
Visible RR spectra ..... o ................. o ooooooooooooooooooooooooooooooooooooooooo 64
Low temperature effects ............................................................. 70
Discussion ............. . ..................................... ................ 74
Structural correlations ....................... . ........................................ 75
Porphyrin core geometry ............................................................. 80
Ring buckling effects ................................................................... 84
Cobaltic octaethylporphyrin 1r cation radicals ............................... 85
Effects of one-electron, metal centered oxidation
and axial ligation ..................... . ......................... . ........................ 86
Comoar‘ 20104 .............. ............. . ..................... 89
001110131)+ 23: ..... ....... .................................... 90
Conclusions .............. . ......................................................................... 92
Acknowledgments ................................................... 4. ........................... 9 4
References and Notes ............................................... .. ........................ 95

CHAPTER 5- DETAILS OF PEROXIDASE CATALYSIS SUGGESTED
BY RESONANCE RAMAN MEASUREMENTS OF IRON-OXYGEN
STRETCHING FREQUENCIES OF HORSERADISI-I PEROXIDASE

INTERMEDIATES
Introduction ...................................................................................... 1 0 1
Experimental
Materials ........ . ....... . ................... . ............................................... 104
HRP-I ................... ........... . ............................ .............. 104
HRP-II ....................................................................................... 106
Model compounds ........................................................................ 106
Instrumentation ............................................. . ......................... 106
Results
High frequency RR scattering from HRP intermediates .................. 110
Model compounds ......................... . .............................................. 1 1 3
Frequency shifts .. ...... ..................................... . ...................... 115
Low frequency RR scattering from HRP intermediates .................. 116
Excitation profiles ...... . ................ . ............... . ....... . ...................... 120
Discussion
Prediction of HRP-I vibrations from structural
correlations ......... ............................................................ 130
HRP-II ...................................................................................... 131
Photochemistry ............................ ........ ..... ............. .. ..... 131
Spin delocalization . ....... .. ........ . ........ ...... ...... 132
HRP-1" ......... ........... ........ ........... ....... . ..... 133
Kinetics and mechanism of HRP-I reduction .................................. 135
Conclusions ....................... ................................... .............. 136
Future Work ............... . ............. . ............ . ................. ........ . ..... 137
Acknowledgments ............... ......... ......... . ......... ........................ 138
References ....................................................................................... 139

2-]

LIST OF TABLES

Comparison of v (FeO) for Various Oxoferryl Species .....................

Resonance Raman Frequencies (cm‘l), Depolarization

Ratios and Optical Absorption Maxima (nm) for Parent

MOEP Compounds and their Corresponding n Cation

Radicals, MOEP+ CIO4‘ ..............................................................

Core Size Parameters from Crystal Structures of Metallo-
porphyrin 17 Cation Radicals and Related Compounds .....................

Resonance Raman FrequenciesI andO tical Absorption
Maxima for Axially Ligated CuII , N 1 and Co11 IOEP
Complexes ...... ............ . ............. ........... .. .......................

Comparison of Resonance Raman Frequencies (cm 1) and
of Depolarization Ratios for Cobaltic OEP" Complexes .................

vi

Page
2 0

83

88

91

1-1

1-2

1-4

LIST OF FIGURES

(a) Ferriprotoporphyrin IX. (b) Metallooctaethylporphyrin.
Carbons a (or a) are bonded to the nitrogen atoms. Car-
bons b (or B ) are bonded to the peripheral substituents. The
methine, CH, positions are meso. The metal center, M, may
be further ligated in the axial positions normal to the

macrocycle plane ......................................................................

Optical absorption spectra of native HRP (N), HRP-I (I)

and HRP-II (11). Reference 31. ..................................................

The atomic orbital (AO) structure of two HOMOs of porphine.
The A0 coefficients are proportional to the size of the
circles; solid lines indicate positive values, dashed lines
negative. The view is from the positive 2 axis. The straight
dashed lines indicate the nodes of the alu ( 1!) orbital.

Reference 14. .................. . ............ ..... . ............. ..............

Optical absorption spectra of (a) cobaltic octaethyl-
porphyrin 11 cation radicals, (b) compound I enzyme
transients. 2A2u type spectra (—). 2A1u type spectra (--).

Reference 16. ..........................................................................

Optical absorption spectra of compound I intermediates

(2Alu type).
(a) native horse erythrocyte catalase (CAT) and CAT-I.

Reference 26. .......... ...........................................................
(b) native chloroperoxidase (CPO) and CPO-I. Reference 24. ......

Optical absorption Spectra of compound I intermediates.

(a) native ferric bromoperoxidase (BPO) and BPO—I.

Reference 27. ..... . ............. ........ . .................................
(b) native lignin peroxidase (MP) and LiP-I. Reference 29. .........

(c) Rapid scan spectrophotometric measurements for the
reactions of intestinal peroxidase (IPO) and lactoperoxidase
(LPO) with hydrogen peroxide at pH 7.1. Dashed lines repre-
sent native enzyme. the numbers show time in ms from the
stop flow to the end of the wavelength scan. Thus, early times
represent compound I spectra, and late times compound 11. At

intermediate times, mixtures were obtained. Reference 28. .......

vii

Page

6

10
10

11
ll

..... 12

2—1

3-1

3-2

4-2

Page

Absorption spectra of native cytochrome c peroxidase (CC?)
and CcP-I. Reference 30. CcP-I contains an amino acid
radical rather than a porphyrin radical. ......... .. ............................. 13

Resonance Raman s ectra of 18O=FeIV(Im)PPDME (upper

trace), and 16O=FeI (Im)PPDME (lower trace). Vertical

line denotes v (FeO); * denotes solvent peak. Spectra were

obtained in toluene-d3 at -130 C with 15 mW (406.7 nm)

incident on the sample. Inset: optical absorption spectrum

of O=FeIV(Im)PPDME in toluene at -90 C. The shoulder at

619 nm is due primarily to u—oxo dimer contamination. ................... 18

Resonance Raman spectra excited at 406.7 nm.

(a) CoIIOEP+‘ClO4‘ prepared with AgClO4;

(b) CoIHOEP+'ZClO4‘ prepared with Fe(ClO4)3;

(c) CoInOEP+'2ClO4", prepared with Fe(ClO4)3 and a trace
amount of HCIO4;

(d) H4OEP2+2CIO4", prepared from HZOEP + HCIO4. All

samples were dissolved in CH2C12. A spinning quartz cell

and 'v 20 mW laser power were used. Solvent bands are

labelled with an *. ......................................... ................... 27

RR spectra of ZnOEP+'ClO4. (a) xex = 390 nm; (b) - (d)

show a time course at A ex = 406.7 nm for 'v 1 ml of sample

in a quartz spinning cell with a laser power of 16 mW. The

time values indicate total irradiation time at the end of the

scan. ........... . ............................................................................ 30

Electronic absorption spectra of (a) CuOEP+'ClO4" and
(b) ZnOEP+’ClO4' (0.2 mM) in dry CHZCIz. Extinction
coefficients were taken from references 29 and 4, respectively. ..... 42

RR spectra of Cu and ZnOEP and their corresponding

cation radicals. (a) CuOEP; (b) CuOEP+'ClO4'; (c) ZnOEP;

(d) ZnOEP+'ClO4‘, excitation at 390 nm (1.5 mJ/pulse,

[ZnOEP+'l 20.2 mM) reflects vibrations of the monomer.

Cchlz bands are marked with an *. CW laser power 20-35

mW. at 363.8 nm for (ch). ........................................................ 45

Resonance Raman spectra excited at 406.7 nm.

(a) CoIIOEP+'ClO4";

(b) ComOEP+'2ClO4" containing small amounts of porphyrin
diacid impurity;

(c) H4OEP2+ZCIO4‘;

(d) a mixture of H30EP+CIO4’ and H4OEP2+2CIO4‘
resulting from treatment of OEPHZ with AgClO4;

viii

Page

(e) H40EP2+2Br-.

RR bands of the solvent, dry CH2C12, are marked with
an". Vertical dashed lines mark the vibrations of the
diacid, H4OEP2+ZBr‘. Laser power, 20 mW. .................................. 48

Electronic absorption spectra of porphyrin free base acids.
(a) H4OEP2+2CIO4";

(b) a mixture of H30EP+CIO4' and H4OEP2+2CIO4"
corresponding to Fig. 4-3d.

Solvent, dry CHZCIZ. ...... . ............ ...... . ................... 51

Electronic absorption spectra of CoOEP and its oxidation
products.

(a) CoOEPH;

(b) CoIIOEP+'ClO4’(—-);

(c) ComOEP”'2ClO4‘("°).

Solvent, dry CHZCIZ. ....................................................... , .......... 5 4

Electronic absorption spectra of oxidation products of
CoOEP.

(a) Com(MeOH)20EPBr'(-);
(b) ComOEPBr' (——);

(c) ComOEP+'ZBr (m), the small feature at 401 nm is due
to 1% H4OEP2+ZBF contamination as discussed in the text.

Solvent, dry CH2C12, except for (a) which contains W 5%
methanOl (M9011). 00000000000000 oooooeoeeo 0000000000000 e ooooooooo oooooeooeo oooooooooooooo 57

RR spectra excited at 363.8 nm ( ~35 mW) of CoOEP and
its oxidation products.

(a) CoOEP;

(b) Com(MeOH)20EPClO4‘;

(c) CoHOEP+'CIO4‘;

(d) CoIIIOEP+’2ClO4‘; ................................................................ 60
(e) Com(MeOH)20EPBr‘;

(f) ComOEPBr';

(g) CoIIIOEP+'2Br-.

Solvent, dry CH2C12 except (b) and (e) which contain 'h 596
methanol. Solvent bands are marked with an *. ............................ 63

RR spectra excited at 514.5 nm (100 mW).

(a) CoOEP;

(b) Com(MeOH)20EPBr";

(c) ComOEPBr".

CH2012 bands marked *. ............................................................. 66

ix

4-9

4-10

4-1 1

4-12

5-1

5-2

5-4

Page

RR spectra of ComOEP+'2Br'. xex = (a) 620 nm, 300 mW;

(b) 647.1 nm, 250 mW; (c)—(d) 676.4 nm, 100 mW. CHZCIZ
bands marked *. Spike (**) in (b) is Rayleigh scattering at
676.4 nm. .......................................................... . ....................... 69

Electronic absorption 3 ectra showing the conversion of

CoHOEP+'ClO4‘ to Co (H20)20EPCIO4‘ as the temperature

is lowered. Solvent, "dry" CH2C12; path length 2 2mm (EPR

tube). ..................................................... . ......... ....... . ........... 73

Porphyrin core vibrational mode frequencies (for Raman

allowed bands, 1450-17 00 cm‘l) v_s_. porphyrin core size for

the indicated MOEP complexes and their corresponding

cation radicals, MOEP+’CIO4'. Mode assignments are accord-

ing to reference 33a. Open symbols correspond to parent

MOEP frequencies; filled symbols correspond to MOEP+°CIO4'
frequencies. The regression analysis does not include the Ni

(ng) points. Ct-N distances are from reference 45. ....................... 78

Electronic transition energies vs. porphyrin core size for
the indicated MOEP and MOEP‘F‘ClOf complexes. Open
and filled symbols correspond to absorption band maxima
of MOEP and MOEP“'CIO4‘, respectively. .................................... 81

End view (a), longitudal section (b) and cross-section (c) of
the eight-jet, tangential mixers. Ref. 48. ..................................... 107

Instrumental configuration used for pulsed resonance Raman
measurements of flowing HRP-I samples prepared by rapid
mixing. Laser irradiation at 390 nm (5 ns pulses, 1 mJ/pulse,
10 Hz) was provided by pumping Exciton LD390 laser dye

in the Quanta Ray Pulsed Dye Laser (PDL-l) with the third
harmonic (355 nm) from the Nd:Yag DCR-lA laser. Raman
scattering from the sample (at position X) was collected
with a Spex 1459 Illuminator, dispersed in a triple mono—
chromator, and detected and analyzed by using the EGétG
PAR 1420 diode array detector and associated OMA II
electronics. RR scattering excited at 390-435 nm from
HRP-II samples in a cooled (2-5 C) cylindrical quartz spin-
ning cell was also measured using this instrumentation. The
10 Hz rep rate was employed on all occasions. ............................... 109

High frequency RR spectra of HRP samples at 2-10 C,

1 mJ/pulse. Sample conditions and total accumulation

times: (a) spinning cell, 20 min; (b) flowing in the capillary

of the rapid mixer at 0.65-0.85 ml/min, 40 min;

(c) spinning cell, 22 min. ......... . ............................................... 112

RR spectra of model compounds in CHZCIZ. Conditions:

40-50 mW incident on sample in spinning cell at 25 C.
Accumulation times: (a) 5.3 min; (b) 10 min. .............................. 114

X

5-6

5-7

5-8

5-9

5-1 0

Page

Low frequency RR spectra of HRP samples obtained under
390 nm excitation. Total accumulation times: (a) 85 min;
(b) 160 sample: 144 min, 180 sample: 78 min; (c) 52 min;
((1) 160 sample: 48 min, 180 sample: 48 min. The

feature at 981 cm‘1 in (b) is due to sulfate ion in the

H21802 preparation. ................................................................. 1 1 8

High frequency RR scattering from HRP-II, pH 10.8.

Accumulation time for each spectrum was 10 min using

1 mJ/pulse. Excitation wavelengths were as indicated.

The 1049 cm‘1 peak is from 0.2 M NO3' used as an

internal standard to measure relative intensities. The

984 cm"1 feature is from 804‘. ................................................... 122

High frequency RR scattering from HRP-I, pH 7.

(a) 1.9 mJ/pulse, 4 min; (b) 15 mW, 0.5 s/cm'l; sample

flow rate, 0.22 m1/min; similar results were obtained

flowing at 0.7 ml/min; (c) 1 mJ/pulse, 40 min. .............................. 125

Low frequency RR scattering from HRP-II, pH 10.8.

Accumulation time was 10 min for (a)-(d), 20 min for

(er), using 1 mJ pulses. The 983 cm“1 peak is from

0.2 M SO ‘ used as an internal intensity standard. The

1049 cm‘ feature is from NO3‘. ............................................... 127

Excitation profiles of v7 = 680 cm‘l, v (FeO) = 787 cm'1
and v 4 = 1379 cm‘l. 129

Schematic diagram of the active site of HRP catalytic

intermediates, showing HRP-I, the postulated photo-

product HRP-1*, the reaction of HRP-I with a phenolic

substrate, and HRP-II. ............................................................... 134

xi

CHAPTER I

GENERAL INTRODUCTION

Aerobic organisms reduce molecular oxygen (02) to water (H20). Partial
reduction of dioxygen can result in the formation of hydrogen peroxide (H202)
and superoxide (02"). These compounds are potentially dangerous to the organism
and are removed from the cell by enzyme catalases, peroxidases, and superoxide
dismutases. While catalase enzymes convert hydrogen peroxide directly to
water and dioxygen, peroxidases utilize hydrogen peroxide to oxidize various
organic and inorganic compounds.1

The active site of native peroxidases and catalyses most often contain
a five-coordinate ferriprotoporphyrin IX prosthetic group as shown in Figure
1-1a. The proximal ligand to the iron may vary and is typically a histidine
nitrogen (e.g. horseradish peroxidase, HRP),2‘4 cysteine sulfur (e.g.
chloroperoxidase, CPO),4’5 or tyrosine oxygen (e.g. bovine catalase, CAT).6

Horseradish peroxidase isolated from plants is the best characterized and
most readily available of the peroxidase enzymes. The general catalytic
mechanism of plant peroxidases is depicted by the following scheme proposed

originally for HRP:7

CH2

a u
cu CH3
H3C ca=ca2
asc CH3
C'Hz EH2
CH2 CH2

l I
COOH COOH

b c'H3 C|H3
CH2 C‘Hz
3 2
HBC‘HZC CHZ-CHS
9:2 9‘2
CH 3 CH3
Figure 1-1 . (a) Ferriprotoporphyrin IX. (b) Metallooctaethylporphyrin.

Carbons a (or a ) are bonded to the nitrogen atoms. Car-
bons b (or 8 ) are bonded to the peripheral substituents. The
methine, CH, positions are meso. The metal center, M, may
be further ligated in the axial positions normal to the
macrocycle plane.

Native (Fem) + H20 + compoundI(Fe"V") + H20
compoundI(Fe"V") + AH + compound II(FeIV) + A'
compound II(FeIV) + AH + native (Fem) + A' + H20

2A' + products

In the first step of this sequence, the heme reacts with hydrogen peroxide
to form compound I (green), which is two oxidation equivalents above the native
ferric state. Compound I rapidly oxidizes the substrate in the second step,
forming the more stable, second intermediate, compound 11 (red). In the final
step of the reaction, compound II relaxes to regenerate the native'ferric state
of the enzyme (brown).

The structural, electronic and magnetic properties of peroxidase compounds
I and II have been investigated for a number of years.8 Most studies have used
HRP owing to the ease of isolation and purification9 as well as the stability
of its compound I (HRP-I) relative to other peroxidases.1 The absorption spectra
of native HRP, HRP-I and HRP-II appear in Figure 1-2. HRP-I is currently
described as a spin-coupled oxoferryl porphyrin 1r cation radical complex.4 In
this structure one electron has been removed from the ferric center to yield
FeIV and the other electron has been removed from the porphyrin ring to yield
a delocalized porphyrin 1r cation radical. The FeIV center is six-coordinate,
being bound on the proximal side by histidine nitrogen and on the distal side
by an oxime group, yielding a O=FeIV-N(His) configuration axial to the heme
ring, reducing the 1r cation radical but leaving the oxoferryl nitrogen ligation

unchanged.2‘4

Significance of the Oxoferryl and Porphrin 1r Cation Radical Structures

The oxoferryl n cation radical structure is considered common to catalytic

Absorbonce

 

 

 

 

 

 

 

 

' 7
0.08
0.06
0.04
0.02
0’1 l l l l" J l 4 l " O
250 300 350 400 450 500 550 6CD 650
Wavelength (nm)
Figure 1-2 . Optical absorption spectra of native HRP (N), HRP-I (I)

and HRP-II (II). Reference 31.

intermediates of a range of heme peroxidase and catalase enzymesl‘4 and possibly
to cytochrome P450.10 The 1r cation radical structure has also been implicated
in the photosynthetic process of both higher plants and bacteria11 and in the
catalytic cycles of nitrite and sulfite reductases.12 The oxoferryl structure
has also been proposed for an intermediate in the reaction of cytochrome oxidase
with dioxygen.13 Thus the significance of both the oxoferryl and the porphyrin
ncation radical structures to the redox and electron transport functions of

metalloporphyrins in nature is clear.

Aim and Strategy of the Thesis

The aim of this work is to characterize the vibrational, electronic and
structural properties of the first intermediate of the catalytic cycle of HRP,
HRP-I, by using resonance Raman spectroscopy in conjunction with rapid mixing
techniques applied to the enzyme in vitro. The difficulties associated with
this measurement, the means adopted to circumvent them, and the results
obtained by the technique are discussed in Chapter 5. In addition to the work
with the enzyme, we utilize systematic spectroscopic studies of various model
compounds to explore separately the properties of the oxoferryl-nitrogen linkage

and the porphyrin n cation radical. This work is described in Chapters 2-4.

Metalloporphyrin I Cation Radicals and Compound I

Chapters 2 and 3 describe a spectroscopic study of highly symmetric (D4h)
metallooctaethylporphyrin n cation radicals (see Figure 1-1b). Oxidation of
the porphyrin ring occurs by the removal of an electron from either of the nearly
degenerate highest filled nmolecular orbitals (HOMOs). In D4h symmetry these
oribtals are designated a1u(1r) and a2u(1r). Their structure is shown in Figure

1--3.14 As representative examples for discussion we choose two cobaltic

 

HOMO'I "32“ (17)"

Figure 1—3 . The atomic orbital (A0) structure of two HOMOs of porphine.
The A0 coefficients are proportional to the size of the
circles; solid lines indicate positive values, dashed lines
negative. The view is from the positive 2 axis. The straight

dashed lines indicate the nodes of the alu ( 1!) orbital.
Reference 14 .

octaethylporphyrin 1r cation radicals, CoIIIOEP+'ZBr‘ and CoIIIOEP+'2CIO4‘.
These two compounds are thought to occupy the 2A1u and 2A2u electronic ground
states, respectively. Much emphasis has been given to characterization of the
spectroscopy of these (and other) metalloporphyrin 1r cation radicals (MP+')
in terms of these two electronic configurations. While this approach has been
successful in the first approximation, especially for interpreting the EPR and
ENDOR spectra of these compounds,15116 some controversy has emerged based
on NMR results.17 The original extension of the model compound studies to
the enzyme intermediates is presented in Figure 1-4. Based on the similarity
of the optical absorption spectra of HRP-I and CAT-I to that of ComOEP+‘ZCIO4'
and CoIIIOEP+'ZBr‘, respectively, Dolphin at 21.16 proposed porphyrin 1r cation
structures for the enzyme intermediates. It is further suggested that both
electronic states are represented by HRP-I (2A2u) and CAT-I (2Alu)- EPR and
ENDOR measurements have confirmed the if cation radical formulation for
HRP-1.13119 However, the question of electronic ground state for these and
other compound I structures is unclear. The results for EPR and Mossbauer
studiesZI‘23 often seem to contradict classifications suggested by optical
absorption spectra.20:24 More specifically, the EPR and Mossbauer measurements
of compound I type intermediates often give evidence of magnetic coupling
between the 3:1 oxoferryl and the S=l/2 porphyrin n cation radical spin systems.
Such interaction is not expected if the porphyrin radical resides in the a1u(1r)
orbital owing to the lack of spin density on the nitrogen. atoms (see Figure 1-3).
However, chloroperoxidase compound I (CPO-I) and certain reconstituted HPR-I
species exhibit both 2A1u type absorption spectra and magnetic coupling. Thus
a contradiction arises. Furthermore, the magnetic coupling between
paramagnetic metal centers, such as the oxoferrryl structure and porphyrin

1r cation radicals, is not well understood“,25 Part of this confusion is caused

 

   

  

 

 

 

 

 

I I j I A I
oilmOEPz’ZCI " i:
1.0 1.0
713 0.5- ‘ 0.5
7" Z Co1m1oeeg’zar" ,
I 1 1 l g 1 .0
‘3 O I f 1 I l O .-
9 x
x U
U
1.0- ~ 1.0
0.5r “0.5
0 0
Figure 1-4.

Optical absorption Spectra of (a) cobaltic octaethyl-
porphyrin 11 cation radicals, (b) compound I enzyme
transients. 2

A2u type spectra (——). 2A1u type spectra (- -).
Reference 16.

by the lack of knowledge of the effects of symmetry lowering on the spectral
properties of metalloporphyrin 1r cation radicals.22 In the natural compound
I systems the molecular symmetry is lowered (from D4h) by asymmetric peripheral
substitution of the ring, mixed axial ligation and possibly further by porphyrin
core distortions. Neglecting core distortions (i.e. doming or buckling of the
porphyrin ring system) the symmetry of compound I is CS or lower. An
understanding of the relationship between the different mechanisms of symmetry
lowering and the electronic ground state as well as an understanding of how
each influence of spectroscopy of the MP‘" are necessary to resolve these
questions. It has been suggested that S4 type ruffling of the porphyrin ring
may provide an overlap pathway for antiferromagnetic coupling in ferric porphyrin
«cation radicals.25 Thus, with our MP‘" model work we take a structural
approach which focuses on understanding the effects of the porphyrin core
geometry on the vibrational and electronic properties of the MP“. It is
anticipated that the resulting correlations can be used to suggest core geometries
for unknown structures. Below, in Figures 1-5, 1-6 and 1-7 optical absorption
spectra of various catalase and peroxidase compound I intermediates are

collected.

10

 

 

 

Figure 1-5 . Optical absorption spectra of compound I intermediates

(zAlu type). .

(a) native horse erythrocyte catalase (CAT) and CAT-I.
Reference 26.

 

 

2
w40- I
301—

'0b 1
clllllllllill IILIILIIILIIIIIIILIIIII

CPO-I

 

 

 

 

 

340 380 420 475 SIS 555 595 635 675

(b) native chlor0peroxidase (CPO) and CPO-I. Reference 24 .

lode ( qu")

11

 

 

   
   

 

l _L l l l l l L l
360 400 «0 400 520 so 500 1540 600
m1.»
Figure 1-6. Optical absorption spectra of compound I intermediates.

(a) native ferric bromoperoxidase (BPO) and BPO-I.
Reference 27.

 

    

 

 

  

 

140 1 1 1 35
.5.
:: L1?
120 - :3 — 30
E:
,1
'i
100 - i: — 25
:1
A ' I
... ' l
1 , :
g 00-— 5 I -« 20
7 1' :
2 " :
E 60b ! 1 ‘ '5
I
l
m 1'
1'
4O 1
20 ’
\‘~--I~-"~+ o
400 500 600 700
k (nm)

(b) native lignin peroxidase (LiP) and LIP-I. Reference 29.

Absorbonce

 

12

Absorbonce

 

 

 

 

1.00 450 500 550 600 650
Wavelength (nm)

(e) Rapid scan spectrophotometric measurements for the
reactions of intestinal peroxidase (IPO) and, lactoperoxidase
(LPO) with hydrogen peroxide at pH 7 .1. Dashed lines repre-
sent native enzyme. the numbers show time in ms from the
stop flow to the end of the wavelength scan. Thus, early times
represent compound I spectra, and late times compound II. At
intermediate times, mixtures were obtained. Reference 28.

l3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:Euyne
150 w ———-. uzoz
jg A
;‘ . \I
g IOC ., \I
r M
2
E
V 50
a, .N‘.
250 300 350 400 450 500 550 600 650 700
wave LENGTH (mp)
Figure 1-7. Absorption spectra of native cytochrome c peroxidase (CcP)

and CcP-l. Reference 30. CcP-I contains an amino acid
radical rather than a porphyrin radical.

10.

11.

12.

REFERENCES

Bolscher, B. Ph.D. Thesis.
Dunford, H.B.; Stillman, J.S., Coord. Chem. Rev. 1976, _1_9_, 187-251.
Frew, J.E.; Jones, P., Adv. Inorg. Bioinorg. Mech. 1984, _3_, 175-213.

Hewson, W.D.; Hager, L.P. in "The Porphyrins"; Dolphin, D., Ed.; Academic
Press: New York, 1979; Vol. VII, pp. 295-332.

Cramer, S.P.; Dawson, J.H.; Hodgson, K.O.; Hager, L.P., J. Am. Chem.
Soc. 1978,_1_0_0_, 7282-7290.

Murthy, M.R.N.; Reid, R.J.; Sicignano, A.; Tanaka, N.; Rossmann, M.G.,
J. Mol. Biol. 1981, _1_§_2_, 465-499.

(a) Chance, B., Arch. Biochem. Biophys. 1952, 4_1, 416-424; (b) George,
P., Nature 1952, .122, 612-613.

(a) Theorell, H.; Ehrenberg, A, Arch. Biochem. Biophys. 1952, 4_1_, 442- 461;
(b)Brill, A..;S Williams, R.J.P, Biochem. J. 1961, L8, 253- 263;
(c)Blumberg, W.E.; Peisach, J.; Wittenberg, B.A.; Wittenberg, J. B., J.
Biol. Chem. 1968, _2_43, 1854-1862; (d)Wittenberg, B..A; Kampa, L.;
Wittenberg, J. B.; Blumberg, W. E.; Peisach, J., J. Biol. Chem. 1968, _2_43
1863- 1870; (e) Peisach, J, Blumberg, W. E.; Wittenberg, B. A.; Wittenberg,
J.B., J. Biol. Chem. 1968, m, 1871- 1880.

(a) Keilin, D.; Hartree, E. F., Biochem. J. 1951, 4_9, 88-109; (b) Shannon,
L..;M Kay,F.., LeW,J..,Y J. Biol. Chem. 1966, _2__41, 2166- 2172.

Groves, J.T.; Haushalter, R.C.; Nakamura, M.; Nemo, T.E.; Evans, B.J.,
J. Am. Chem. Soc. 1981, 103, 2884-2886.

Parson, W.W.; Ke, R. in "Photosynthesis: Energy Conversion by Plants
and Bacteria"; Govindjee, Ed.; Academic: New York, 1983.

Chang, C.K.; Hanson, L.K.; Richardson, P.F.; Young, R.; Fajer, J., Proc.
Natl. Acad. Sci. USA 1981, _7_8, 2652-2656.

14

13.
14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.
28.
29.
30.

31.

15

Blair, D.F.; Witt, S.N.; Chan, 8.1., J. Am. Chem. Soc. 1985, 1_0_7_, 7389-7399.
Maggiora, G.M., J. Am. Chem. Soc. 1973, _95, 6555-6559.

Fajer, J.; Davis, MS. in "The Porphyrins", Dolphin, D., Ed.; Academic
Press: New York, 1979; Vol. IV, pp. 197-256.

Dolphin, D.; Forman, A.; Borg, D.C.; Fajer, J.; Felton, R.H., Proc. Natl.
Acad. Sci. USA 1971, Q8, 614-618.

Morishima, l.; Takamaki, Y.; Shiro, Y., J. Am. Chem. Soc. 1984, 106,
7666-7672.

Schulz, C.E.; Devaney, P.W.; Winkler, H.; Debrunner, P.G.; Doan, N.;
Chiang, R.; Rutter, R.; Hager, L.P., FEBS Lett. 1979, 103, 102-105.

Roberts, J.E.; Hoffman, B.M.; Rutter, R.; Hager, L.P., J. Biol. Chem.
1981, 256, 2118-2121.

DiNello, R.K.; Dolphin, D., J. Biol. Chem. 1981, 256, 6903-6912.

Rutter, R.; Valentine, M.; Henrich, M.P.; Hager, L.P.; Debrunner, P.G.,
Biochemistry 1983, _2_2, 4769-47 74.

Schultz, C.E.; Rutter, R.; Sage, J.T.; Debrunner, P.G.; Hager, L.P.,
Biochemistry 1984, 2_3_, 4743-47 54.

Rutter, R.; Hager, L.P., J. Biol. Chem. 1982, 257, 7958-7961.

Palcic, M.M.; Rutter, R.; Araiso, T.; Hager, L.P.; Dunford, B.H., Biochem.
Biophys. Res. Comm. 1980, 93, 1123-1127.

Gans, P., Buisson, G.; Duee, E.; Marchon, J.-C.; Erler, B.S.; Scholz, W.F.;
Reed, C.A., J. Am. Chem. Soc. 1986, 108, 1223-1234.

Schonbaum, G.R.; Chance, B. in the "The Enzymes"; Boyer, P.D., Ed.;
Academic Press: New York, 1976, Vol. 13, pp. 363-408.

Manthey, J.A.; Hager, L.P., J. Biol. Chem. 1985, _2_§_0, 9654-9659.
Kimura, S., Yamazaki, 1., Arch. Biochem. Biophys. 1979, 1_9_§_, 580-588.
Renganathan, V.; Gold, N.H., Biochemistry 1986, _2_5_, 1626-1631.
Yonetoni, T., J. Biol. Chem. 1965, _240, 4509-4514.

Hewson, W.D.; Hager, L.P., J. Biol. Chem. 1979, 153, 3182-3186.

CHAPTER 2

CHARACTERIZATION OF SIX-COORDIN ATE OXOFERRYL PROTOHEME BY

RESONANCE RAMAN AND OPTICAL ABSORPTION SPECTROSCOPY.

SUMMARY

As a model for oxoferryl intermediates of heme enzymes, we have
synthesized (l-methylimidazole) oxoferryl protoporphyrin IX dimethyl ester,
O=FeIV(Im)PPDME, and characterized it by using optical absorption and resonance
Raman spectroscopy. We observe an FeO stretching frequency, v(FeO), of
820 cm‘1 for O=FeIV(Im)PPDME, which is assigned by its shift to 784 cm"1
upon substitution of 16O with 180. While the optical absorption spectrum of
the six-coordinate protoheme model is very similar to that of oxoferryl myoglobin,
its v(FeO) is 23 cm'1 higher. Based on results from five- and six-coordinate
oxoferryl porphyrin models, we attribute this higher frequency to weaker

imidazole ligation and an absence of protic environmental effects.

*Robert T. Kean, W. Anthony Oertling, and Gerald T. Babcock, J. Am. Chem.

 

_S_o_c., 1987, in press.

16

17

Oxoferryl species, O=FeIV, have been postulated in the catalytic cycle
of cytochrome c oxidase,1 as the oxygen donating species in cyctochrome P-450,2
and as intermediates in the reactions of catalases and peroxidases.3 Given
the diverse chemistry catalyzed by these various enzymes, heme pocket
modulation of the chemical reactivity of the O=FeIv unit seems likely. Resonance
Raman detection of the v (FeO) in various protein species and model compounds
supports this notion. In a comparison of oxoferryl peroxidase species,4 oxoferryl
myoglobin,5 and five- and six-coordinate heme model compounds,617 the
frequency of v(FeO) varies by '1: 85 cm‘1 (see Table 2-1). This is in strong
contrast to v(Fe—Oz) which varies by only '1: 10 cm“1 in protein species and
heme model compounds.8 In addition, there are distinct differences in the optical
spectra of the various oxoferryl protein species.9"11 Previously reported
oxoferryl model compounds have given insight into the factors affecting the
v(FeO) frequency, but since they were made on non-physiologically active
hemes, they do not give specific information about the optical spectra or the
other Raman active vibrations of the protoheme-containing oxoferryl protein
species. To address these points, we present here optical and Raman data for
a six-coordinate, imidazole-ligated, oxoferryl protoheme model compound.

The (l-methylimidazole) oxoferryl protoporphyrin IX dimethyl ester, O=FeIV
(Im)PPDME, was prepared according to reference 12. The optical absorption
spectrum138 is inset in Figure 2-1. The shoulder at 619 nm is due primarily
to u-oxo dimer contamination. The peak positions compare favorably with those
of oxoferryl hemoglobin,9 and oxoferryl leghemoglobin,10 ('1: 418, 545, and 575
nm) but are most similar in both wavelengths and relative intensities to oxoferryl
myoglobin (m 420, 550, and 580 nm at pH 6.8).11 The spectrum of O=FeIv
(Im)PPDME, like the oxoferryl globin species, is distinct from that of horseradish

peroxidase compound 11 (HRP-II; m 418, 527, and 555 nm).14 In Figure 2-1 we

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ao‘z406.7 nm “6 x5
5 5
3 511 “
677 I 3
n g 619
I e
E 350 V ' I 111161111011? (my I ' ' I100
z
t!
_2_
z e
s . I
g 7114
751 W
'80 ~Av=36-‘ ‘
cm"
820
..0 Wfl/V
1 ' 1 ' I r 1 . ' 1 ' 1 . I '
soo FREQUENCY (cm") 900
Figure 2-1. Resonance Raman s ectra of 18O=FeIV(Im)PPDME (upper

trace), and 16O=FeI (Im)PPDME (lower trace). Vertical
line denotes v (FeO); * denotes solvent peak. Spectra were
obtained in toluene—d3 at -l30 C with 15 mW (406.7 nm)
incident on the sample. Inset: Optical absorption spectrum
of O=FeIV(Im)PPDME in toluene at -90 C. The shoulder at
619 nm is due primarily to u—oxo dimer contamination.

19

present the intermediate frequency region of the Raman spectrum13b of O=FeIv
(Im)PPDME. The peak at 820 cm‘1 shifts to 784 cm'1 upon substitution of 16O
by 18O, and is assigned to FeO stretching vibration. The 36 cm"1 shift is expected
for an oxoferryl structure. The v(FeO) frequency observed for O=FeIv
(Im)PPDME does not vary with temperature over the range -90 C to -190 C
or with porphyrin ring substituents (TPP or OEP).15 This frequency is compared
with the u (FeO) frequencies of other oxoferryl species in Table 2-1. The high
frequency region of the O=FeIV(Im)PPDME spectrum (not shown) contains features
of both HRP-II16 and oxoferryl myoblobin5 yet it is not identical to either.15
The similarity of the optical spectrum of oxoferryl myoglobin with that of O=FeIv
(Im)PPDME indicates that the two are electronically very similar. The difference
in the v (FeO) frequencies suggests environmental perturbations. Differences
in the v (FeO) frequency between the peroxidase species and oxoferryl myoglobin
have been discussed in terms of hydrogen bonding effects“:e and out-of-plane
iron effects.5 It has been suggested that 11 charge donation from the protein
into the porphyrin may be important in the mechanism of horseradish
peroxidase,17 but it is not known whether this has any affect on the v(FeO)
frequency. Comparison of these systems with the model compounds gives further

insights into the variables which affect the v (FeO) frequency.

A trans ligand effect is seen very clearly in the model compounds (see
Table 2-1). The five—coordinate models display the highest 0 (FeO) frequency,
while the strong ligand (Im), six-coordinate samples display the lowest frequency.
Electron density from the sixth ligand, along the z axis (normal to the heme
plane), may compete with the ferryl oxygen for o-bondingl8 and weaken the
FeO bond. Alternately, if the iron is initially displaced toward the oxygen in

the five-coordinate species, the presence of a strong sixth ligand may pull the

20

Table 2-1

Comparison of v(FeO) for Various Oxoferryl Species

Species 11 \1(Fe0)b Ref.
HRP-II pH 6.0 776 4c,e
HRP-II pH 11.0 787 4b,c
CcP-I 767 4d
oxoferryl Mb 797 5
O=FeIVTMP 843 21
o=FeIVTPP . 852 6
o=FeIV0EP 852 6
O=FeIV(THF)TpivPP 829 7
O=FeIV(Im)TpivPP 807 7
O=FeIV(Im)PPDME 820C
O=FeIV(Im)TPP 820
O=FeIV(Im)OEP 820

aAbbreviations: HRP-II, horseradish peroxidase com-
. pound II; CcP-I, cytochrome c peroxidase compound 1;
Mb, myoglobin; TMP, tetramesitylporphyrin; TPP,
tetraphenylporphyrin; OEP, octaethylporphyrin; THF,
tetrahydrofuran; Im, l-methylimidazole; TpivPP,
"picket fence" porphyrin, tetra(o-pivaloylphenyl)-
porphyrin; PPDME, protoporphyrin IX dimethyl ester.
b Frequency in cm‘l.

C This work and reference 15.

21

iron into plane causing greater 1! interaction between iron and porphyrin orbitals,
and a weakening of the FeO 1! -bond.5 The difference in the v(FeO) frequency
of our models (in toluene) versus the (TpivPP)7 species (in THF) appears to results
from solvent effects since temperature and ring substituent effects were ruled
out above. This may reflect stronger imidazole binding in the more polar and
non-aromatic THF. Although hydrogen bonding of the oxo ligand to the distal
histidine could be involved, the lower v(FeO) frequency of oxoferryl myoglobin
may indicate stronger proximal imidazole ligation in myoglobin than in the
model compounds. This seems reasonable since the imidazole is protein bound
in myoglobin and may not easily move or rotate to less strongly ligating
configurations available to the free solution models. Similar trans ligand effects
may contribute to the difference in v(FeO) frequencies observed for HRP-II
and oxoferryl myoglobin. For the respective five-coordinate ferrous enzymes,
the higher u(FeH-imidazole) frequency for HRP indicates stronger imidazole
ligation.19920 For the oxy complexes, v(FeH-imidazole) is also higher for HRP
than for myoglobin but its \1(Fe-02)8c is lower. A similar inverse relationship
between (Fen-CO) and trans ligand strength has been reported for monomeric

insect hemoglobins18a and heme model compounds.18b

ACKNOWLEDGMENTS

We thank Dwight Lillie for data handling and graphics software. This

research was supported by NIH grant GM 25480.

10.

REFERENCES AND NOTES

(a) Wikstrom, M., Proc. Natl. Acad. Sci. USA 1981, _7_8, 4051-4054; (b)
Blair, D.F.; Witt, S.N.; Chan, 8.1., J. Am. Chem. Soc. 1985, 107, 7389-7399.

Groves, J.T. in "Cytochrome P-450: Structure, Mechanism and
Biochemistry"; Ortiz de Montellano, P. Ed; Plenum Press: New York, 1985;
Chapter I.

Hewson, W.D.; Hager, L.P. in "The Porphyrins", Vol. VII; Dolphin, D. Ed.;
Academic Press: New York, 1979, pp 295-332.

(a) Terner, J.; Sitter, A.J.; Reczek, C.M., Biochim. Biophys. Acta 1985,
_8_28_, 73-80; (b) Hashimoto, S.; Tatsuno, Y.; Kitagawa, T., Proc. Japan
Acad. 1984, §_0_(B), 345-348; (c) Sitter, A.J.; Reczek, C.M.; Terner, J.,
J. Biol. Chem. 1985, _2_60, 7515-7522; ((1) Hashimoto, S.; Teraoko, J.;
Inubushi, T.; Yonetani, T.; Kitagawa, T., J. Biol. Chem. 1986, 261,
11110-11118; (e) Av(FeO) value of 775 cm'1 (HRP-II, pH 7, H20 outfit?)
is reported which shifts to 777 cm‘1 in D20 buffer. Hashimoto, S.; Tatsuno,
Y.; Kitagawa, T., Proc. Natl. Acad. Sci. USA 1986, _83, 2417-2421.

Sitter, A.J.; Reczek, C.M.; Terrier, J., Biochim. Biophys. Acta 1985, 828,
229-235.

(a) Bajdor, K.; Nakamoto, K., J. Am. Chem. Soc. 1984, 106, 3045-3046;
(b)Proniewicz, J.M.; Bajdor, K.; Nakamoto, K., J. Phys. Chem. 1986,
_9_0_, 1760-1766.

Schappacher, M.; Chottard, G.; Weiss, R., J. Chem. Soc., Chem. Commun.
1986, _2_, 93-94.

(a) Spiro, T.G. in "Iron Porphyrins", Part II; Lever, A.B.P.; Gray, H.B.
Ed.; Addison Wesley: Reading, Mass, 1983; pp. 107-133; (D) Kean, R.T.,
unpublished results; (c) Van Wart, H.B.; Zimmer, J., J. Biol. Chem. 1985,
_2_06, 8372-8377.

Dalziel, K.; O'Brien, J.R.P., Biochem. J. 1954, 26, 648-659.

Aviram, 1.; Wittenberg, B.A.; Wittenberg, J.B.; J. Biol. Chem. 1978, 253,
5685-5689.

22

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

23

George, P.; Irvine, D.H., Biochem. J. 1952, _5_2_, 511-517.

LaMar, G.N.; de Ropp, J.S.; Latos-Grazynski, L.; Balch, A.L.; Johnson,
R.B.; Parish, D.W.; Cheng, R., J. Am. Chem. Soc. 1983, 105, 782-787.

(8) Optical absorption spectra were obtained by using a house build optical
dewar mounted in a Perkin-Elmer Lambda 5. The sample was contained
in an EPR tube which was cooled to -90°C by flowing cold nitrogen gas;
(b) Raman spectra were obtained with a Spex 1401 scanning monochromator
(with PMT detection) by using 15 mW incident power at 406.7 nm
(Spectra-Physics model 164 Kr ion) in a backscattering geometry. The
samples, contained in EPR tubes, were spun continuously in a dewar while
the desired temperature was maintained by flowing cold nitrogen gas.

Blumberg, W.E.; Peisach, J.; Wittenberg, B.A.; Wittenberg, J.B., J. Biol.
Chem. 1968, 243, 1854-1862.

Kean, R.T.; Babcock, G.T., manuscript in preparation.
Terner, J.; Reed, D.E., Biochim Biophys. Acta 1984, 789, 80-86._

Shelnutt, J.A. ; Alden, R.G.; Ondrias, M.R., J. Biol. Chem. 1986, 261,
1720-1723.

(a) Gersonde, K.; Kerr, B.A.; Yu, N-T.; Parish, D.W.; Smith, K.M., J. Biol.
Chem. 1986, 261, 8678-8685; (b) Kerr, B.A.; Mackin, H.C.; Yu, N-T.,
Biochemistry 1983, 2_2_, 4373-437 9.

Teraoko, J.; Kitagawa, T., Biochem. Biophys. Res. Comm. 1980, _9_3,
694-700.

Teraoka, J.; Kitagawa, T., J. Biol. Chem. 1981, 256, 3969-3977.
Hashimoto, S.; Tatsuno, Y.; Kitagawa, T. In "Proceedings of the Tenth

International Conference on Raman Spectroscopy; Petiolas, W.L.; Hudson,
B. Ed.; Univ. of Oregon: Eugene, 1986, pp. 1-28, 1-29.

CHAPTER 3

RESONANCE RAMAN SPECTROSCOPIC DETECTION OF

DEMETALLATION OF METALLOPORPHYRIN s CA'HON RADICALS.

SUMMARY

Soret resonance Raman (RR) spectrum of samples of several different
metalloporphyrin 11 cation radicals (MP“) prepared by both chemical and
electrochemical oxidation in CH2C12 reveal the presence of free base diacid
salts produced by demetallation of the complexes. The large extinction
coefficient of the diacid salt allows its selective enhancement by the RR
technique at concentration levels not always evident in absorption spectra but
sufficient to cause serious artifacts in the RR spectra of these samples, making
recognition and analysis of the scattering from the MP“ impossible. Proper
control of experimental conditions, particularly choice of excitation frequency,
can eliminate these complications and produce RR spectra of MP“ free of

contributions from the diacid salt.

*

W. Anthony Oertling, Asaad Salehi, Chi K. Chang, and Gerald T. Babcock,

J. Phys. Chem., 1987, in press.

 

24

25

Resonance Raman (RR) spectroscopy has been utilized to extract structural
and electronic information from naturally occurring metalloporphyrin systems
and their models, and has recently been applied to synthetic metalloporphyrin
n cation radicals (MP“).1"3 The preparative techniques for MP“ generation,
however, rarely produce a homogeneous sample.4 Owing to the selective
enhancement afforded by Soret excitation RR, it is possible to detect trace
amounts (less than 396) of porphyrin free base diacid salts in samples of MP“
prepared by chemical and electrochemical oxidation in CH2C12. If unrecognized,
the presence of the diacid salt may cause serious artifacts in the RR spectra
of these samples that prevent accurate analysis of scattering from the MP“.1
Proper control of experimental conditions, particularly the choice of excitation
frequency, can produce RR spectra of MP“ free of contributions from the
diacid.3’5

One-electron oxidation of Co and Zn octaethylporphyrin (OEP) in dry CHZCIZ
leads to divalent metal porphyrin 1r cations.3v6 Laser excitation at 363.8 nm
selects against resonance enhancement of possible residual starting materials
and produces good quality RR spectra.3 RR spectra of these MP“ samples
excited at 406.7 nm, however, are often dominated by bands not present in
spectra of 363.8 nm and not assignable to parent MP modes.7 The frequencies
of these bands (most noteworthy are those at 1394 and 1558 cm'l) show no
metal dependence. These facts suggest the presence of an impurity which absorbs
strongly near 406.7 nm.

Figure 3-1a and 3-1b show RR scattering excited at 406.7 nm from samples
of ConOEP“ClO4‘ prepared with AgClO4 and ComOEP“2ClO4‘ prepared with
Fe(ClO4)3, respectively.8 Although the RR spectra of these two species excited
at 363.8 nm are very similar3, with excitation at 406.7 nm additional bands

appear in the spectrum of the cobaltic OEP“ relative to the cobaltous sample.

Figure 3-1 .

26

Resonance Raman spectra excited at 406.7 nm.

(a) CoHOEP“ClO4' prepared with AgClO4;

(b) ComOEP“2ClO4‘ prepared with Fe(ClO4)3;

(c) ComOEP“2ClO4‘, prepared with Fe(ClO4)3 and a trace
amount of H0104;

(d) H4OEP2+2CIO4‘, prepared from H2039 + H0104. an
samples were dissolved in CHZCIZ. A spinning quartz cell

and 'v 20 mW laser power were used. Solvent bands are
labelled with an *.

27

*MNVT

I;

2CK)

c)ComOEPI

+HCIO4

   

>._._mZm:.Z_ Z<§<m

 
      

d) H.051:>2+

r

 

RAMAN SHIFT (cm")

28

Figure 3-1c shows scattering from a sample of ComOEP“ZClO4‘ prepared

with Fe(ClO4)3 and a trace amount of HClO4. The added intensity of the spurious
bands in this spectrum) suggests that the impurity is caused by acid promoted
demetallation of the MP“ and protonation of the resultant free base porphyrin.
Figure 3-1d shows the RR spectrum of the diacid salt H4OEP2+2CIO4', confirming
this as the interfering species.

Thus, RR spectra of MP“ samples exicted at 406.7 nm may contain artifacts
due to free base diacid salts present at concentrations too low to be obvious
in the optical absorption spectrum yet high enough to dominate the RR scattering
at 406.7 nm. This is clear from the Soret optical properties of the impurity:
the Soret band sharpens dramatically and red-shifts to 406 nm (e = 430m M‘lcm'l)
upon formation of the diacid from OEPH2.9 Thus, the narrowed bandwidth,
the increased extinction coefficient a and the coincidence of the Soret transition
energy with the laser line at 406.7 nm, explain the selective scattering from
the impurity.

Figure 3-2 shows scattering excited at 390 and 406.7 nm of ZnOEP“ClO4‘.
The absence of bands at 1394 and 1558 cm"1 indicates no contributions from
the impurity with 390 nm excitation.10 Spectra excited at 406.7 nm show bands
at 1394 and 1559 cm'1 as depicted in Figure 3-2b. Unlike the ComOEP“2ClO4‘
sample, which appears to be stable to prolonged laser irradiation, the
contributions from the impurity (marked by the dashed lines) increase
dramatically with repeated scanning of the ZnOEP“ClO4' sample until they
dominate the spectrum as shown in Figure 3-2c-d. The solvent bands at 704
and 1423 cm‘1 show the opposite trend and decrease in intensity, indicating
stronger absorbance at 406 nm as the impurity concentration increases. Indeed,
for this MP“ sample with the Soret band at 387 nm, the Soret absorbance at

406 nm of the protonated porphyrin was easily seen to increase in parallel with

29

Figure 3—2. RR spectra of ZnOEP“ClO4. (a) Aex = 390 nm; (b) - ((1)
show a time course at Aex = 406.7 nm for 'v 1 ml of sample
in a quartz spinning cell with a laser power of 16 mW. The
time values indicate total irradiation time at the end of the
scan.

30

ZnOEP“C1021

* e051

 

    

IIIIIIIIII 111111 ..l111lll 1- .Ilmn:

 

m
m m n
n 7.11. "W h
7h
m 63 %2 @3
n m0 4 \I
O to d
w m lllll I..I.1|. .Illlmmm
1m

 

 

 

 

 

 

. 11 III 111111 8a
mom- 1 l INIJ1 I I lfiom

>._._mzw._.z_ Z<§<m

 

RAMAN SHIFT 1cm")

31

the changes in the RR spectrum.11

These results suggest a number of conclusions: (i) oxidation of MP samples
in CH2C12 often results in the formation of free base diacid impurities which
can be detected by RR at 406.7 nm; (ii) owing to trace amounts of aqueous
acid impurities in reagents, oxidation with Brz or Fe(ClO4)3 is more likely
to cause demetallation and diacid formation than oxidation with AgClO4 or
electrochemical techniques; (iii) MOEP“ complexes prepared by electrochemical
or AgClO4 oxidation show detectable diacid formation for M = Mg or Zn but
not for CoII or Ni, suggesting that the ease of demetallation is core size
dependent for the MP“ as it is for the parent MP;12 (iv) demetallation of
ZnOEP“ClO4" is enhanced by laser light in the Soret region of the spectrum;
and (v) though diacid formation may be limited to chlorinated solvents and
use of another solvent may minimize demetallation, the utility of CH2012 for
RR spectroscopy and porphyrin oxidations may not be easily replaced.
Recognition and elimination of these artifacts is essential to characterizing
the vibrational spectra of the MP“ species. Excitation at 363.8 nm produces
RR spectra free of these artifacts and illustrates vibrational similarity of 2A1u
and 2A2u radical types.5113 This similarity was obscured by the occurrence
of diacid impurities in RR samples of the presumed 2A1u species of previous

studies.1

ACKNOWLEDGMENTS

This research was supported by the NIH grants GM-25480 (G.T.B.) and

GM-36520 (C.K.C.).

REFERENCES AND NOTES

Kim, D.; Miller, L.A.; Rakhit, G.; Spiro, T.G., J. Phys. Chem. 1986, _9_9_,
3320-3325.

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

Salehi, A.; Oertling, W.A.; Babcock, G.T.; Chang, C.K., J. Am. Chem.
Soc. 1986,_l_0_§, 5630-5631.

Huet, J.; Gaudemar, A.; Boucly-Goester, C.; Boucly, P., Inorg. Chem.
1982, _2_1_, 3413-3419.

Oertling, W.A.; Salehi, A.; Chung, Y.C.; Leroi, C.E.; Chang, C.K.; Babcock,
G.T. J. Am. Chem. Soc. 1987, submitted.

(a) Fuhrhop, J.H., Struct. Bonding 1974, 1_8_, 1-67; (b) Dolphin, D.; Mulijiani,
Z.; Rousseau, K.; Borg, D.C.; Fajer, J.; Felton, R.H., Ann. N.Y. Acad.
Sci. 1973, _2_0_6_, 177-200.

Our results Show the phenomenon to be widespread. In addition to examples
presented here, we have observed Similar aritfacts when using Brz to
oxidize CoOEP and MgOEP. Oxidation of NiTPP, CuOEP, and FeOEPCl
by Fe(ClO4)3 in CH2C12 also produces significant amounts of analogous
impurities.

OEPHZ was prepared according to Wang, C.B.; Chang, C.K. Synthesis
1979, 548-549. Insertion of metals was carried out by standard methods:
Falk, J.E. "Porphyrins and Metalloporphyrins"; Elsevier: New York, 1964,
798. The metalloporphyrins were purified by recrystallization and/or
chromatography to eliminate traces of free base. CHZCIZ freshly distilled
from CaHz was used as a solvent for all preparations. Details of oxidations
are presented in references 3 and 5. Raman measurements on spinning
samples at 406.7 nm were made as described by Ondrias, M.R., Babcock,
G.T., Biochim. Biophys. Res. Comm. 1980, 93, 29-35.

(a) Corwin, A.H.; Chivvis, A.B.; Poor, B.W.; Whitten, D.G.; Baker, E.W.,

J. Am. Chem. Soc. 1968, 90, 6577-6583; (b) Ogoshi, H.; Watanabe, E.;
Yoshida, ZaTetrahedron 1973, _2_9, 3241-3245.

32

10.

11.

12.

13.

33

Excitation at 390 nm was provided by a pulsed laser system described
by Oertling, W.A.; Babcock, G.T., J. Am. Chem. Soc. 1985, 107, 6406-6407.
The RR spectral contribution from the diacid impurities aFé—less obvious
with pulsed excitation than with CW excitation at Similar wavelengths,
most likely reflecting a long-lived Soret excited state of the diacid species
not in resonance at 390 nm.

ZnOEP“ClO4‘ was prepared by oxidation with AgClO4. Use of Fe(ClO4)3
as an oxidant for ZnOEP is more difficult to control as an excess will
rapidly oxidize the sample to the dication. Along with this a copious amount
of the diacid impurity is produced as is evidenced by the rapid growth
of a Sharp absorption band at 405 nm. Three control RR experiments
at 406.7 nm were performed: degassed ZnOEP in CHZCIZ showed no Sign
of the impurity (ref. 1), whereas anaerobic samples of ZnOEP“ClO4"
exhibited behavior like that shown in Fig. 2, and OEPHZ in CHZClz rapidly
became protonated during the same time period of laser irradiation. The
band at 1377 cm'1 in Figure 3-2b is from unoxidized starting material
and indicates the presence of the diacid under conditions of incomplete
oxidation. Samples containing no detectable trace of neutral ZnOEP showed
the same behavior. Samples of ZnOEP“ClO4‘ not exposed to laser light
showed no Sign of diacid formation detectable by optical absorption spectra.

(a) Buchler, J.W.; Puppe, L.; Rohbock, K.; Schneehage, H.H., Ann. N.Y.
Acad. Sci. 1973, _2_06, 116; (b) Shelnutt, J.A., J. Am. Chem. Soc. 1983,
105, 774-778. Free base porphyrin diacid impurities (ca 1%) have been
reported in preparation of Zn and Mg porphyrin dications by Fajer, J.;
Borg, D.C.; Forman, A.; Dolphin, D.; Felton, R.H., J. Am. Chem. Soc.
1970, 2_2_, 3451-3459. It is possible that under certain conditions
demetallation might occur via the dication species (e.g. footnote 11),
especially for the Mg complex. However, because the second oxidation
potential of ZnOEP (1.02 V, ref. 6a) is beyond the oxidizing power of our
AgClO4/CH2012 system (0.9 V, ref. 3), demetallation via a ZnOEP2+ species
is unlikely in this case.

Shimomura, E.T.; Phillippi, M.A.; Goff, H.M., J. Am. Chem. Soc. 1981,
103, 6778-6780. "

CHAPTER 4

VIBRATIONAL, ELECTRONIC AND STRUCTURAL PROPERTIES OF

COBALT, COPPER AND ZINC OCTAETHYLPORPHYRIN I CATION RADICALS.

SUMMARY

Optical and resonance Raman (RR) spectroscopic characterization of the
oxidation products of several metallooctaethylporphyrins has been carried out.
One-electron oxidation of the macrocyclic yields a series of divalent metal
substituted octaethylporphyrin 11 cation radicals, MHOEP“CIO4' (M = Ni, Co,
Cu and Zn). The porphyrin core vibrational frequencies above 1400 cm"1 of
these complexes are described by a linear function of center-to-pyrrole nitrogen
distances. A comparison of these structural correlations and of Raman
depolarization ratios with those of the parent MOEP compounds is used to
establish vibrational mode assignments for the cations. The agreement between
the correlation parameters of the MOEP and MOEP“ClO4‘ suggests similar

potential energy distributions in the normal modes of both species. We find

*W. Anthong Oertling, Asaad Salehi, Young C. Chung, George E. Leroi, Chi

K. Chang, and Gerald T. Babcock, J. Am. Chem. Soc., 1987, submitted.

34

35

that the frequencies of the stretching modes with predominantly Cbe character
increase, whereas those with CaCm and CaN character decrease in the cation
radical relative to the neutral metalloporphyrin. Similar trends in Soret band
maxima for the 11 cation radicals and their parent compounds reflect changes
in the relative energy of the a2u(11 ) orbital. These structural correlations seem
to be essentially insensitive to 2A2u y_s_. 2A1u radical designation. With the
vibrational mode correlations as a guide to evaluation of porphyrin core geometry,
we have carried out a detailed analysis of the oxidation products of ConOEP
and we suggest structures for the two-electron oxidized species,
ComOEP“ZClO4' and ComOEP“ZBr". Differences in the high frequency
vibrations of these two compounds are interpreted in terms of expansion or
possible ruffling of the porphyrin core in the latter relative to the former
compound. RR excitation in the 600-680 nm region of the ComOEP“ZBr‘
absorptions shows a lack of anomalously polarized scattering and produces spectra
Similar to those obtained with near uv excitation. This suggests the absence
of strong Herzberg-Teller (HT) coupling between the excited electronic states

of the 11 cation radical.

INTRODUCTION

Oxidized states of metalloporphyrins (MP) participate in the redox chemistry
of a vareity of biological structures including light-harvesting photosynthetic
systems and heme proteins. Oxidation of the MP may occur at the central metal,
at the porphyrin ligand, or at both locations. The Few/Fe3+ redox chemistry
characteristic of the electron transfer function of cytochromes is well known
and metalloporphyrin structures oxidized at both the central metal and the
porphyrin ligand have been implicated in the catalytic cycles of heme peroxidases,

P450 monooxygenases, and catalases.1 Because of the extensive 11 system,

36

oxidation at the porphyrin yields a delocalized 11 cation radial; indeed, in the
case of the special pair bacteriochlorophyll a dimer in the photosynthetic
bacterial reaction center, radical character may be Shared between two porphyrin
rings.2 Characterization of metalloporphyrin 11 cation radicals is thus of interest
because of their widespread occurrence in nature and because of the often unusual
chemistry in which they participate.

Both chemical and electrochemical preparations of a variety of
metallooctaethylporphyrin (MOEP) and metallotetraphenylporphyrin (MTPP)
‘11 cation radicals have been reported.3‘6 These oxidized complexes have been
characterized principally by optical absorption,7 EPR,8 MCD,9 NMR“),11 and
IR12 Spectroscopies, and X-ray crystallography.13 Systematic resonance Raman
(RR) studies of synthetic metalloporphyrin 11 cation radicals (MP“) can identify
structural and electronic factors which influence their redox properties and
chemical reactivity. Although RR spectroscopy has been applied to a few MP“
systems, these studies are limited in scope for various reasons. Attempts to
measure RR scattering from the oxoferryl porphyrin ‘11 cation radical of
horseradish peroxidase compound I (HRP-I) are complicated by
photochemistry,14115 and Special precautions are required to obtain its
spectrum.16 Analysis of the RR scattering from the bacteriochlorophyll cation
radical iS obscured by the complexity of the parent system.17 The work of
Yamaguchi it. 31.18 on MTPP“ complexes is not strictly relevant owing to the
strong vibronic coupling and altered substituent pattern of the TPP which reduces
its significances as a biological model for RR studies.19 The MOEP“ system
studied by Kim _e_t_ _a_l.20 is an ideal choice for systematic RR investigation;
however, recent work21 has revealed that some of the spectra reported by these
authors were contaminated by artifacts from porphyrin diacid salt impurities

produced by demetallation of their MOEP“ samples. This led to incorrect

37

assignments of the principal vibrational modes and to erroneous conclusions
concerning the nature of metalloporphyrin 11 cation radical Raman scattering.
This circumstance necessitates a re-examination of the RR spectra of these
systems.

The high symmetry (D411) and saturated Cb substituents make OEP the
most simple biologically relevant porphyrin ligand for RR study, and the oxidation
products of cobaltous octaethylporphyrin (CoOEP) in particular are of interest.
One-electron oxidation of CoOEP in CHZClg may occur at either the metal
or the porphyrin.22 Depending on the counterion present, the CoOEP
two-electron oxidation product displays one of two distinct optical absorption
spectra. These are pr0posed as typical of the two possible electronic ground
states which result from loss of an electron from one or the other of the two
nearly degenerate HOMOS, alu (11 ) or 82“ (11), of the porphyrin 11 system.9123124
The similarity of the optical Spectra of ComOEP“2ClO4‘ and ComOEP“ZBr'
to those of HRP-I and catalase compound I (CAT-I) led to the early assignment
of the electronic states of these enzyme transients as 2A2u and 2A1“,
respectively.24 The versatility of the CoOEP system thus allows us to compare
metal _v_s_. porphyrin centered oxidation products for the one-electron case, and
2A2u _v_s_. 2A1u radicals for the two-electron case, all with the same MOEP
species. Comparison of the cobaltous 11 cation radical, CoHOEP“ClO4‘, with
other MHOEP“CIO4‘ species illustrates structural trends in both the RR
frequencies and Soret absorption maxima which are consistent with recent
calculations23 and very similar to those of the parent MOEP species. Exclusion
of the artifacts arising from the diacid salt impurities reveals a closer vibrational
similarity of 2A2u and 2A1u radical types than was thought previously.20 RR
excitation in both the Soret and visible absorption bands of cobaltic OEP“

samples provide spectra which suggest the absence of strong Herzberg-Teller

38

coupling of electronic states and a similar scattering mechanism in both spectral

regions.

EXPERIMENTAL

Preparation of OEPHZ was carried out according to the recent method
of Wang and Chang25 and metal insertion was achieved by standard methods.26
The metalloporphyrins were purified by recrystallization and/or chromatography
to eliminate traces of free base. CHZCIZ freshly distilled from CaHz was used
as a solvent for all preparations.

Oxidations were performed chemically at room temperature by using solid
anhydrous AgC104, solid Fe(ClO4)3 or dilute Br2 in CH2C12. CoHOEP“C104‘,
ZnHOEP“C104‘, and CuHOEP“ClO4' were prepared by stirring dry 0112012
solutions of CoOEP, ZnOEP, and CuOEP, respectively, with a 3-fold excess
of AgC104 for 1-3 hours.22 CoIHOEP“2CIO4“, was prepared by stirring excess
Fe(C104)3 in a solution of CoOEP for «:10 minutes. ComOEPBr‘ and
ComOEP“ZBr' were prepared by dropwise addition of dilute Br2 to CoOEP
in CH2C12. MgIIOEP+'CIO4‘ and Mg“0EP“Br" were prepared Similarly. from
MgOEP.

Dropwise addition of methanol (MeOH) to CH2C12 solutions of
CoIIOEP+'ClO4‘ and ComOEPBr‘ resulted in the formation of species with
Similar absorption Spectra, presumably Com(Me0H)20EPCIO4' and
Com(Me0H)20EPBr', respectively.27 Identical samples were made by oxidation
of CoOEP with HC104 and HBr, respectively, in 3:1 CH2C12:Me0H. All oxidations
and ligand changes were monitored by optical absorption measurements with
either a Cary 219 or a Perkin Elmer Lambda 5 spectrophotometer.

Raman spectra excited at 351.1, 363.8 and 390 nm were measured with

a Spex 1877 triple monochromator and Spex 1459 collection optics in conjunction

39

with an EGch PAR 1420 diode array detector and associated OMA II electronics.
The 351.1 and 363.8 nm laser emissions were from a Coherent 90-05 Argon
ion laser; the 390 nm line was the output of a Nd:YAG (Quanta Ray DCR-IA)
pumped pulsed dye laser (PDL-l) in which LD390 (Exiton) laser dye was used.
Instrumentation used to measure Raman scattering at 406.7, 514.5, 620, 647.1
and 676.4 nm included a Spex 1401 double monochromator with photon counting
detection interfaced to 8 DEC LSI 11-2 computer. The lasers were powered
by a Spectra Physics Model 265 Exciter; the lines at 406.7, 647.1 and 676.4
nm were from a Model 164 Krypton ion laser, the 514.5 nm line was from a
Model 165 Argon ion laser, and the 620 nm line was the output of an Argon
ion pumped Spectra Physics Model 375 dye laser in which Rhodamine 590 (Exciton)
laser dye was used.

Samples were spun in a cylindrical quartz cell for Raman measurements,
although static samples in cuvettes gave identical spectra. Absorption spectra
were checked before and after each Raman measurement. 0f the oxidized
species examined, only ZnOEP“ClO4’ and MgOEP“Br' showed any Significant
laser induced change.21 Samples of ComOEP“2CIO4‘ and ZnOEP“ClO4‘
prepared and studied under anaerobic conditions exhibited RR scattering identical
to samples which were not deoxygenated, and for this reason samples were

not routinely degassed.28

RESULTS
Effects of oxidation on the high fretnlency porphyrin core vibratiom.

Although the cobaltic OEP“ system is of principal interest owing to its relevance
to the compound I type enzyme intermediates, results from the Cu and ZnOEP“
systems are presented first. The chemistry of the cobalt system is complicated

by the two easily obtainable redox states of the metal center and by the diversity

40

of one— and two-electron oxidation products possible. The CuII and ZnnOEP
systems are simpler in that, under the conditions utilized here, oxidation occurs
exclusively at the porphyrin ligand. Since the porphyrin dication is currently
of little physiological interst, we consider only the singly oxidized state here.
Thus, we can establish the effects of oxidation on the porphyrin core vibrational
modes before addressing the more complex cobalt system.

One-electron oxidation of CuOEP and ZnOEP. The optical absorption spectra
of ClO4‘ complexes of CuOEP“ and ZnOEP“ are shown in Figure 4-1. The
absorption spectrum of CuOEP“ClO4' is like that reported earlier29 and has
features usually attributed to the 2A2u state.9123124 The Soret band is
blue-shifted and diminished in intensity with respect to that of the parent
compound and is without noticeable structure. The absorption in the visible
region is broad and diffuse, extending through the range 500-640 nm. The actual
electronic ground state of CuOEP“C104" is difficult to establish, however,
owing to the lack of an EPR signal from this compound under most conditions.30
Indeed, recent N MR measurements suggest a 2A1u state for CuOEP“ClO4’,10a
seemingly in contradiction with the absorption spectrum. The ground electronic
configuration of the ZnOEP“ species, however, is determined by EPR
measurements to be 2A1u.8 The absorption spectrum of ZnOEP“ClO4" presented
in Figure 4-1b agrees well with earlier measurements.4131 Although previous
work indicates that ZnOEP“ClO4‘ in CH2012 solutions is monomeric, Figure
4-1b shows that Significant dimerization of this compound takes place at the
concentrations used here (0.2 - 0.5 mM). The high-energy Shoulder m 350 nm
and the absorbance at 573 nm are attributed primarily to the dimer species.31
Dilution of ZnOEP“C104‘ samples results in reduced intensity of these features,
consistent with decreased dimerization, as Shown earlier.4

Figure 4-2 shows the RR spectra (1000-1700 cm'l) obtained with near uv

41

Figure 4‘1 - Electronic absorption spectra of (a) CuOEP“ClO4' and
(b) ZnOEP“ClO4’ (0.2 mM) in dry 0112012, Extinction
coefficients were taken from references 29 and 4, respectively.

 

 

 

 

 

42

 

 

 

 

 

 

 

 

1' Y I r 1 T I I I I I I T I l I I T
80-
60-
‘ a) Cu0EP“C|02
40- '
x5
7
E 20-
U
T - _
E 04 \ --50
‘6
~40
+0 .-
b)ZnOEP c104 _30
x5
-20
r10
I f 1 V I V I T V FT 1 1 U T I I j T o

 

300 400 500 600 . 700
WAVELENGTH (nm)

43

excitation of complexes of CuOEP“ClO4" and ZnOEP“C104‘ and their respective
neutral parent compounds. RR enhancement of vibrations of ZnOEP“C104"
dimers was avoided by using laser excitation at 390 nm, well to the red of the
dimer absorption. Under these excitation conditions, most of the resonance
enhanced vibrations shown in Figure 4-2 correspond to totally symmetric
fundamentals. The bands above 1400 cm‘1 are known to reflect porphyrin core
geometry32 and are of principal interest to this work. From the mode
designations established by Abe gt a_l_.,33a the v 3, 011, v2 and 010 modes
of CuOEP are identified at 1502 , 1563, 1591 and 1636 cm‘l, respectively, in
Figure 4-2a. The intense high frequency enve10pe from 1570-1650 cm'1 is poorly
resolved in the CuOEP“C104' spectrum in Figure 4-2b; however, three distinct
bands with different depolarization ratios are evident from polarized scattering
(Spectra not shown). Thus, together with the feature at 1497 cm'l, there are
four bands present above 1400 cm"1 and 1:1 correlation with the modes of the
parent compound is possible. We recognize the v3, v11, v2 and V“) frequencies
of CuOEP“ClO4‘ at 1497, 1600, 1610 and 1631 cm-1, respectively. While caution
should be exercised in the vibrational assignment of the MOEP“ based on analogy
to the parent M0EP,34 our mode assignments are based on structural trends
and depolarization ratio measurements discussed in detail below. The intense
band at 1378 cm'1 in the CuOEP spectrum in Figure 4-2a is v 4, the "oxidation
state marker" of heme proteins.35 This band is identified in the corresponding
11 cation radical Spectrum (Fig. 4-2b) at 1360 cm'l. Thus, in the high frequency
region, the modes involving primarily CaCm stretching character ( V3 and v10)33
decrease in frequency upon oxidation of the porphyrin ring, while modes involving
primarily Cbe stretching character (v 11 and 112) increase in frequency. The
frequency of v4, primarily a CaN stretch, decreases upon formation of the

porphyrin 11 cation radical. These trends are common to all of the MOEP systems

Figure 4-2 .

44

RR spectra of Cu and ZnOEP and their corresponding
cation radicals. (a) CuOEP; (o) CuOEP“ClO4‘; (c) ZnOEP;
(d) ZnOEP“ClO4", excitation at 390 nm (1.5 mJ/pulse,
[ZnOEP“]‘-‘-’0.2 mM) reflects vibrations of the monomer.
CH2C12 bands are marked with an ‘. CW laser power 20-35
mW, at 363.8 nm for (ch).

 

 

45

mun. 1

Xe, = 363.8nm
o) CuOEP

   

own—1

  

b) CuOEPI'CIOZ

ObnT

 

 

 

>H_mzwi_.z_

c)ZnOEP

Z<2<m

 

d) znoep*'C1o:,

 

 

RAMAN SHIFT (cm"')

46

studied here.

The RR spectra of ZnOEP and ZnOEP“ClO4‘ are shown in Figure 4-2c
and d. The vibrational mode assignments for the parent compound are analogous
to those of CuOEP. For example, the v4, v3, 011, v2 and 010 modes of
ZnOEP appear at 1376, 1485, 1559, 1581 and 1618 cm‘l, respectively, in Figure
4-2c. For the oxidized species (see Fig. 4-2d), these vibrations occur at 1342,
1477, 1587, 1600 and 1617 cm'l, respectively. The 1110 frequency of
ZnOEP“ClO4" is estimated to be 1617 cm"1 from depolarization ratio
measurements and spectra obtained under 363.8 nm excitation (not shown).
Inspection of these vibrational frequencies reveals shifts in the Zn cation radical
spectrum similar to those observed upon oxidation of CuOEP.

Porphyrin diacid impurities. Resonance Raman excitation of ZnOEP“ClO4‘
at 406.7 nm (spectrum not shown) reproduces the features present in the 390
nm spectrum (Fig. 4-2d). However, additional peaks are present with 406.7
nm excitation which cannot be correctly assigned to any ZnOEP“ClO[ species.
AS we established earlier,21 these vibrations arise from porphyrin diacid salt
impurities produced by demetallation of the ZnOEP“ClO4' complex. Because
the intensity of the diacid vibrations may totally obscure those of the
meta110porphyrin 11 cation radical under certain excitation conditions, it is
essential that these artifacts be recognized and eliminated from RR spectra
of these compounds. We present here additional information on this subject.

Figure 4-3a and b show RR scattering under 406.7 nm excitation from
samples of CoIIOEP“C104' prepared with AgClO4 and of ComOEP“C104"
prepared with Fe(ClO4)3, respectively. Although the RR spectra of these two
samples excited at 363.8 nm are similar22 (cf. Fig. 4-7c and d), excitation at
406.7 nm produces additional features in the spectrum of cobaltic OEP“ relative

to the cobaltous sample (Fig. 4-3a and b). These added vibrational contributions

Figure 4-3 .

47

Resonance Raman spectra excited at 406.7 nm.
(a) CoIIOEP“ClO4';

(b) ComOEP“2C104‘ containing small amounts of porphyrin
diacid impurity; ‘
(c) H40EP2+2C104‘;

(d) a mixture of H30EP+C104‘ and H40EP2+2C104"
resulting from treatment of OEPHZ with AgClO4;

(e) H40EP2+ZBr‘. ~

RR bands of the solvent, dry CHZCIZ, are marked with
an". Vertical dashed lines mark the vibrations of the
diacid, H40EP2+ZBr". Laser power, 20 mW.

48

 

a nNQT

 

Xex = 406.7 nm

  

 

 

 

 

 

 

 

 

    

 

a me..-
a v9.1 1
.6
.4 m .4 .f
m C 0 B
C .2 a ”I.
e e e
P
a .. a ....
0 mo 0 o
«0 I0 3 4
C C H .31 H
o as? 5
room- a

 

>._._mzm._.z_ 24.24.".

RAMAN SHIFT (cm")

49

are from the diacid salt impurities. Figure 4-3c shows the RR Spectrum of
the interferring Species, H4OEP2+2C104'. Addition of AgClO4 to a solution
of HZOEP in CH2C12 results in a mixture of mono- and diacid species which
produces the RR spectrum shown in Figure 4-3d. Whereas the vertical lines
identify the vibrations of the diacid, the additional peaks labelled in Figure
4-3d (e.g. those at 1505, 1582 and 1634 cm‘l) most likely arise from the monoacid
species. This suggests that the mono- and diacid species have different
vibrational properties, possibly owing to the lowered symmetry resulting from
the inequivalence of the x and y axes in the monoacid structure. The formation
of diacid in the sample of ComOEP“2CIO4‘ prepared with Fe(C104)3 probably
results from aqueous acid (i.e. HC104) present in the reagent. Thus, acid
demetallation of the MP followed by protonation of the resultant free base
could compete with the oxidation process. The lack of artifacts in the spectrum
in Figure 4-3a illustrates that this process does not occur when anhydrous AgClO4
is used to oxidize CoOEP. The spectrum in Figure 4—3d is of interest because
it suggests the possibility of both monoacid and diacid contamination of MP“
samples. Furthermore, the absence of aqueous acid impurities in the AgClO4
reagent suggests that porphyrin acid formation is a true by-product of the
oxidation process in certain cases, and not merely the result of "wet" oxidant.
Figure 4-3e illustrates that changing the counterion for the diacid from C104”
to Br‘ decreases the frequencies of the intense vibrational bands above 1300
cm“1 by approximately 3 cm’l. The above observations are all pertinent to
the interpretation of the RR Spectrum of ComOEP+'ZBr‘ obtained with 406.7
nm excitation by Kim 3 31:,20 which we suggest results from H40EP2+2Br'
(possibly mixed with H30EP+Br‘) impurities present in the sample.

The strong enhancement of the diacid vibrational modes with 406.7 nm

excitation, even when the species occurs as a minor impurity (less than 396)

50

Figure 4-4 . Electronic absorption spectra of porphyrin free base acids.
(6) H40EP2+20104‘;
(b) a mixture of H30EP+CIO4’ and H40EP2+2C104‘
corresponding to Fig. 4-3d.
Solvent, dry CH2C12.

 

e(mM"' cm")

51

 

U T 1 I l I I I I r I I I I l I T I I

400« -
300

2001 1
o) 114 051:“ 2610;

I001' '
Lm

 

 

 

 

 

 

 

 

o. 1 2, .38.
b) H40EP 2CIOZ
H30EP+ Clo;
- ~200
" x5 I . ' PI00
I K“—“/—Jd>““—:====ho

 

VrIiITIIU'UI

300 400 500 600 700
WAVELENGTH (nm)

52

in MP‘“ samples, is rationalized by its optical properties. Figure 4-4a shows
the optical absorption spectrum of the H4OEP2+ZCIO4‘ sample used for the
RR measurement shown in Figure 4—3c. Figure 4-4b shows the absorption
spectrum of a mixture of mono- and diacids produced by the addition of AgClO4
to OEPHZ in CH2012 and corresponds to the sample used for the RR spectrum
in Figure 4-3d. Contributions to the absorption spectrum from the monoacid
are evidenced by the high energy shoulder (N390 nm) in the Soret and by the
absorption maximum at 518 nm.36 While this sample is predominantly (>8096)
monoacid, the RR spectrum (Fig. 4-3d) is dominated by the diacid vibrations,
owing to the large extinction coefficient (430 mM‘lcm‘l) and close
correspondence of the Soret band (406 nm) of the latter species .to the laser
line (406.7 nm). The extent of demetallation and diacid formation in MOEP“
sample is dependent on a number of factors, including the oxidation method,
solvent and core size of the MOEP’“. In some cases it is enhanced by laser
irradiation in the Soret band region.21 By exciting well to the blue of the
absorption maxima of these potential contaminants, we avoid the complications
imposed by these species upon RR measurements.

One- and two—electron oxidation of CoOEP. Figure 4-5 shows uv—visible
absorption spectra of CoOEP and products of its oxidation in dry CH2C12 in
the presence of ClO4‘. The spectrum of the starting material, CoOEP, and
the product of ring centered, one-electron oxidation, CoIIOEP+'ClO4', are shown
in Figure 4-5a and b, respectively. Figure 4-5c shows the spectrum of the
two-electron oxidation product, CoIIIOEP+'ZClO4'. As we recently
demonstrated,22 cobaltous OEP‘“ is extremely sensitive to the presence of
axial ligands which coordinate more strongly than CIO4“. Such coordination
promotes electron transfer from CoII to the electron—depleted porphyrin.

Addition of methanol (MeOH) to CHZCIZ solutions of ConOEP+'CIO4' produces

53

.9038 iv .Emzom

.A...V-vo_o~.+mmo=_oo 3

xiv-¢o_o.+mmo=oo 3V

mTESS 3

duos—V95

:oSwExo m: 98 mmOoO mo 23on :23..on 8.8.533

.3 95mm.

54

2:5 Iszmqm><>>

 

 

 

 

 

Om.

CON

 

OmN

('_w3,_ww) 9

55

a cobaltic species, CoIH(MeOH)20EPClO4‘, with an absorption spectrum similar
to, but red-shifted with respect to that of the parent CoOEP species. Because
water coordinates and produces a cobaltic species similar to the dimethanol
adduct (see Fig. 4-10, below), rigorously anhydrous conditions are necessary
to provide a homogeneous preparation of CoIIOEP+'ClO4‘. The absorption
spectrum of this species (Fig. 4—5b) closely resembles that of NiOEP+'CIO4',29
and has features usually attributed to the 2A2u state (as does CoIIIOEP+'ZClO4",
Fig. 4-5c).

Figure 4-6 shows absorption spectra of three species obtained by Br2
oxidation of CoOEP in CHZCIZ. One-electron oxidation in this case also produces
two distinct species, but both exhibit metal centered oxidation. One-electron
oxidation by Brz followed by dropwise addition of MeOH produces the
six-coordinate Com(MeOH)20EPBr" shown in Figure 4-6a. The absorption
spectrum of this species is typical of six-coordinate cobaltic porphyrins37 and
similar, although slightly red-shifted, to that of the dimethanol adduct,
CoH(MeOH)20EPC104', described previously22 (see Table 4-1, below).
One—electron oxidation of CoOEP by Brz in dry CHZCIZ produces a presumably
five-coordinate cobaltic porphyrin which displays the spectrum shown in Figure
4-6b. For this compound, ComOEPBr', the Soret band (373 nm) blue-shifts
with respect to that of CoOEP (391 nm), and the extinction coefficient ratio
of the visible to the Soret band increases. The spectrum of the two-electron
oxidized 1r cation radical, CoIIIOEP+'ZBr‘, is shown in Figure 4-6c. The near
uv region contains three distinct bands, at 290 (not shown), 345 and 410 nm,
and there are at least two bands in the visible region at m 600 and 670 nm. These
features (split Soret with an intense red-shifted visible band) are considered
typical of the 2A1u electronic state.9923924 The feature at 401 nm in Figure

4-6c is attributed to the diacid salt (H4OEP2+ZBr'), the presence of which

 

 

 

56

.8035 3552.
*m a 953:8 :02; A3 ..8 3088 .SONIO En £528

.38“ 2.: E comm-.36 mm 53538350 ....muémmOv: 8H 3
26 m_ E: 2:. «a 9538 :25 2: xiv ammimmoaoo A8
xiv (ammo—=8 3v

x¢-..mmm0£:o§=_oo 3

.mmOoO
no 32695 53258 go 950on .8390me 3.8.5005

. E 2&5

57

ASE Ihozm4m><>>

 

 

 

 

 

fiOON

(,_u13'_wu1)9

58

explains the subtle variations in band shape and maximum of the 410 nm transition
evident in different preparations of this compounds”24 We estimate the diacid
salt contamination in the preparation represented by Figure 4-6c to be «196.38
Near uv RR spectra of one- and two-electron oxidation products of CoOEP.
Figure 4-7 collects RR spectra produced by 363.8 nm excitation of various
products of CoOEP oxidation. The effects of metal centered, one-electron
oxidation accompanied by axial metal ligation are demonstrated by comparing
the spectra of CoOEP and Com(MeOH)2OEPClO4‘ depicted in Figure 4-7a and
b. The oxidation state marker, v 4, increases from 1379 to 1383 cm‘l, reflecting
depopulation of the porphyrin 1r* orbitals caused by metal oxidation.39 There
is little systematic change in frequency of the modes above 1400 cm'l, indicating
that the core size of the porphyrin ring does not change significantly upon metal
oxidation and addition of axial MeOH ligands.32940 The v2 mode, which occurs
at 1599 cm'1 in CoOEP, is very weak in this spectrum of Com(MeOH)20EPClO4‘,
but it is seen at 1596 cm‘1 in spectra excited at 406.7 nm (not shown). The
remaining high frequency bands of CoOEP (Fig. 4-7a) at 1512, 1575 and 1647
cm“1 are assigned to v3, VII and v10, respectively. The weak mode at 1476
cm"1 may correspond to v23. Although this mode is not observed in Soret
excited spectra of other MOEP complexes, it is reported for NiPP
(PP=protoprophyrin IX) at 1482 cm'l.40 These frequencies are consistent with
the CaCm stretching character of this mode. Assignments for
Com(MeOH)20EPCIO4 are similar. Figure 4-7c and (I show the spectra of the
cobaltous and cobaltic OEP+° complexes. Neglecting differences in relative
intensity produced by the differences in Soret absorption, the RR spectra of
CoIIOEP+'ClO4' and CoIHOEP+'2ClO4' are essentially identical and (above
1300 cm'l) radically different from those of the neutral ring compounds in

Figure 4-7a and b. These observations provide a strong fingerprint basis for

59

Figure 4-7 . RR spectra excited at 363.8 nm ( ~35 mW) of CoOEP and
its oxidation products. _

(a) CoOEP;

(b) Com(MeOH)20EPClO4‘;
(c) CoIIOEP+'ClO4‘;

(d) ComOEP+'ZClO4‘;

60

Xe, = 363.8 nm

  

   

be

>._._mzw._.ZH Z<_>_<m

oept' zero;

d) Col"

     

  

|500 ITOO

|300

IDO

 

RAMAN SHIFT (cm")

61

identifying the former species as an authentic 1T cation radical.22 Our
assignments identify v4, v23, v3 and v10 at 1361, 1466, 1505, and 1642 cm‘l,
respectively, for ComOEP+'2ClO4‘ (Fig. 4-7d). Compared to the ring-neutral
CoII and Co111 species, all four of these modes have decreased in frequency,
consistent with the generalizations made above. The Cbe stretching modes,
v11 and v2, again show the opposite trend, as they are assigned to the intense
features at 1605 and 1617 cm‘l. The assignments for CoIIOEP+'ClO4‘ (Fig.
4-7c) are similar.

The RR spectra excited at 363.8 nm of the bromide adducts of oxidized
CoOEP are shown in Figure 4—7e-g. The RR spectrum of Com(MeOH)20EPBr',
Figure 4—7e, is similar to that of CoOEP and Com(MeOH)20EPClO4' in Figure
4-7a and b. The RR spectrum of ComOEPBr' is shown in Figure 4-7f. While
”4 (1378 cm'1 is slightly lower than in the other CoIII compounds, the core
sensitive bands V23 (1478 cm'l), V3 (1515 cm‘l), V11. (1577 cm'l), V2 (1601
cm'l) and particularly v 10 (1657 cm‘l) have increased in frequency, suggesting
a slight core contraction in this species.32:40 The spectrum of the 1r cationic
radical species, ComOEP+'ZBr‘ (Fig. 4-7g) is dominated by two bands around
1600 and 1611 cm"1 (presumably VII and v2, respectively), similar to that
of the other it cations discussed above. There is, however, no band easily
assignable to v4. The features at 1460 and 1648 cm‘1 are assigned to V23
and um, respectively. The RR spectrum of this compound with laser excitation
at 351.1 nm is essentially identical to the spectrum at 363.8 nm shown here.
The v3 vibration, which is enhanced in Soret RR spectra of neutral
metalloporphyrins, does not appear in these spectra. We assign this mode to
the polarized feature at 1497 cm“1 present in the spectra obtained in resonance

with the 670 nm transition of CoIHOEP+'ZBr‘ (see below).

62

(e) Com(MeOH)20EPBr‘;
(f) ComOEPBr';
(g) ComOEP+'ZBr‘.

Solvent, dry CH2012 except (b) and (e) which contain 'b 596
methanol. Solvent bands are marked with an ‘.

63

_mm_ I

 

Xe, - 363.8 nm

   

e) Co"'(MeOH)z OEP Br“

 

 

 

 

 

f) Co'"0EP Br‘

>.:mzm._.ZH 232$”.

 

 

IIOO ISOO ISOO - I700

900

RAMAN SHIFT (cm")

64

Visible RR spectra of one- and two-electron oxidation products of CoOEP.
While Soret (or B band) resonance Raman excitation of neutral metalloporphyrins
enhances totally symmetric modes (A1g in D4h symmetry) via a Franck-Condon
(FC) scattering mechanism, excitation in the visible absorption (Q band) of
these species enhances primarily nontotally symmetric modes (Blg, 82g and
Azg) via a Herzberg—Teller (HT) mechanism.41 In particular, excitation in the
B or Q(0-1) band strongly enhances A2g modes of the MP,42 well known for
their inverse polarization.43 Figure 4-8a shows both polarization components
of the spectrum of CoOEP obtained in Q(0-1) resonance under 514.5 nm
excitation. The A2g modes are easily identified in the I 1 scan at 1123, 1310,
1394 and 1598 cm'1 and are assigned as V22, V 21, V 20 and V19, respectively.
Although theory predicts that the value of the depolarization ratio, p 5 IL /I.. ,
will be on for these modes, the measured values are usually finite for various
reasons.43:44 The large p values obtained from Figure 4-8a indicate that in
CHZCIZ solutions CoOEP assumes a planar configuration and D4h symmetry
as does NiOEP in solution.45 In agreement with Soret RR data for these
ring-neutral species, Figure 4-8b shows that little change occurs in the spectrum
as the metal center is oxidized and ligated axially by MeOH; the A2g modes
of Com(MeOH)20EPBr' are apparent at 1128, 1315, 1394 and 1597 cm'l. The
RR spectrum of ComOEPBr“ (Fig. 4-8c), however, is different from that of
other ring-neutral MOEP compounds excited at 514.5 nm, of which the previous
two examples are typical. While the Agg modes in the spectra of CoOEP and
Com(MeOH)20EPBr’ are absent in the spectrum of ComOEPBr", the appearance
of ap modes at 1054, 1077, 1252, 1274 and 1572 cm"1 is clear from the 11 scan
in Figure 4—8c. The 1572 cm"1 band is reported in the early study by Spaulding
_e_t_ _a_l_.,45 and this frequency is noted as an exception to the trends presented

in that work. This band may not be assignable to v19 since it is 91:30 cm'1 too

65

Figure 4-8 . RR spectra excited at 514.5 nm (100 mW).
(a) CoOEP;
(b) Com(Me0H)20EPBr";
(c) ComOEPBr‘.
CH2012 bands marked *.

 

t VON-

66

ewowa

OEP

a) C0“

 

 

 

 

 

b) Co"'(MeOH)2 OEP 8r“

  

 

 

 

>thmHZ_ Z<§<m

C) Co'”OEP Br'

 

 

 

 

 

 

 

RAMAN SHIFT (cm")

67

low. It is possible that none of the ap modes appearing in the spectrum of this
compound is assignable to an Azg fundamental. Eg Raman active vibrations
should also exhibit anamolous polarization in D4}, symmetry. However, owing
to their relatively high frequencies it is improbable that these modes correspond
to Eg vibrations. Furthermore, activation of F.g modes (out-of-plane) would
presumably require electronic transition moments normal to the in-plane
porphyrin transitions.43 The appearance of these modes thus indicates a change
in symmetry from D4h, resulting in anamolous polarization of other vibrational
species.41a

In contrast to the cobaltic derivatives obtained by one-electron oxidation
of CoOEP with Br2, the two-electron oxidized cobaltic 1r cation radical,
ComOEP+'ZBr‘, fails to exhibit ap modes when excited in the intense visible
transition at 670 nm. Figure 4-9 shows RR spectra of CoIIIOEP+‘2Br‘ excited
at 620, 647.1 and 676.4 nm. The high frequency spectrum from 1000—1700 cm'1
produced with 620 nm excitation (Fig. 4-9a) is similar to the 363.8 nm spectrum
(Fig. 7g), except that the 1611 cm"1 (v 2) band observed with near uv excitation
is absent from the 620 nm spectrum. The similarity of these spectra suggests
that the same scattering mechanism(s) are effective in both Soret and visible
RR spectra of this compound. As the excitation frequency is moved into 0-0
resonance with the 670 nm electronic transition, the enhancement of polarized
modes at 160, 343, 355, 1126 and 1497 cm'1 (tentatively assigned to V3) is
apparent in the spectra presented in Figure 4-9b-d. Figure 4-9d shows both
parallel and perpendicular components to the scattering at 676.4 nm. The modes
below 700 cm'1 are clearly polarized, while the p values for those above 900
cm‘1 range from 0.4 - 0.75 with the exception of the mode at m 1497 cm"1

0.3. Clearly there are no ap modes present in these spectra.

(V3) with D

The p values obtained at 647.1 nm are the same as those measured from

68

Figure 4-9 . RR spectra of CoIHOEP+’2Br‘. Aex = (a) 620 nm, 300 mW;

(b) 647.1 nm, 250 mW; (CHd) 676.4 nm, 100 mW. CHZClz
bands marked *. Spike (**) in (b) is Rayleigh scattering at
676.4 nm.

69

Co'" ospt‘zsr'

 

m
n
O
2
6
.m

b)641lnm

  

an

 

c) 676.4 nm

nOVl

nvnl

_w_l

>._._mzm._.z_ Z<_2<m

   

at) 676.4 nm

 

can
nvnl

woo

I

' [500

”do ISCO

560 ' 760 960
RAMAN SHIFT (cm")

1

360

 

loo

7O

scattering at 676.4 nm in Figure 4-9d. These observations suggest an absence
of strong H-T coupling of the excited states of CoIIIOEP+'ZBr‘. We believe
the band at 160 cm'1 to be a mode involving the metal and the Br‘ ligand,
although no experiments to confirm this suggestion have been attempted. This
mode is also present in spectra excited at 406.7 nm (not shown).

For the 2A1u radical, CoIIIOEP+'2Br", RR excitation within the intense,
strongly red-shifted electron transition at 670 nm produces RR spectra of
reasonable quality which are free of contributions from other absorbing species.
RR excitation within the much weaker, diffuse visible absorption bands of 2A2u
type radicals is much more difficult owing to the smaller RR scattering
cross-section. Room temperature scattering from CoIIOEP+'ClO4‘ and
CoIIIOEP+'2C104‘ with 514.5 nm excitation (not shown) is similar and consists
of weak, poorly-resolved Raman bands on a high fluorescence background, and
is obtained only at high laser power. Our data for CoIIOEP+'ClO4' show
anomalously polarized features at 1312 and 1597 cm‘1 and a polarized feature
at 1383 cm‘l. However, these are all attributed to the ring-neutral cobaltic
species, Com(H20)20EPClO4'. Other weak features which we attribute to
the cobaltous 11 cation radical seem to correspond to modes present in the
spectrum of this sample excited at 363.8 nm (Fig. 4-7c). There is possibly an
ap feature between 1560-1570 cm"1 arising from the “Ir cation radical; howver,
further experiments are required to reach firm conclusions concerning the
scattering mechanisms operative in visible region RR spectra of this and other
2A2u type 1! cation radicals.

Low temperature effects. The results of our room temperature visible
excitation RR work described above are in contrast to those obtained by Kim
$11.20 at low temperature (77K). While our work suggests a lack of anomalously

polarized scattering from the W cations (at least for CoIIIOEP+'ZBr‘), the earlier

71

study assigns AZg modes of the 17 cations only slightly changed in frequency
from those of the parent compounds. We suggest that these features most likely
arise from ring—neutral contaminants in the preparations. This is particularly
feasible for visible Raman measurements on 2A2u type radicals, which must
be excited in the diffuse 500—600 nm absorption, near the Q-bands of the
ring-neutral compounds, in order to obtain resonance conditions. (Raman
excitation in the 500-570 nm region of the 2A1u radicals, e.g. ComOEP+'2Br',
will not achieve resonance conditions owing to the absence of absorption in
this range by the Ir cation radical.) For example, RR scattering from
CoHOEP+'ClO4' samples may contain contributions from the ring-neutral species,
Com(H20)20EPClO4', owing to the difficulty in obtaining completely anhydrous
conditions and the coincidence of common laser lines (e.g. 406.7 and 514.5 nm)
with its absorption bands (409 and 523 nm). The technical difficulties in obtaining
scattering from these 2A2u type 1r cation radicals is magnified at low
temperatures. Figure 4-10 shows optical spectra which document the formation
of increased amounts of Com(H20)20EPClO4' in a sample of CoIIOEP+'ClO4‘
as the temperature is lowered. As the 376 nm Soret band of the cobaltous 1r
cation decreases in intensity, absorptions characteristic of the cobaltic complex
at 409, 523, and 556 nm become apparent. Thus, simply lowering the temperature
of this sample has much the same effect on the absorption spectrum as the
addition of methanol,22 presumably due to increased diaxial ligation by H20
with decreasing temperature. Freeze-pump-thawing to remove water vapor
over the sample prior to cooling results in minimized conversion to
Com(H20)20EPCIO4‘, consistent with diaquo ligation.

Several other examples of the formation of ring-neutral species in
preparations of metalloporphyrin 1r cation radicals have been observed. Addition

of methanol at room temperature to CHZCIZ solutions of the cobaltic 1r cation

 

72

.323

meme semi 59.2 53 “£038 :29. .2838 69832 s
232858 2: as -eozofioaxoamz oo 3 -eo~o.+amo=oo
no 956.8250 2: 95505 atom m .5398me 2.558%

.c Te 239m

73

ASE Ikozm4m><>>
00» com com

b b L L . . P

i I;

« mMm
6mm e

(I . .
/

woo amo «8.3528 «In 3.0.38.8

 

 

m0¢

mhm

BONVSHOSSV

74

CoIIIOEP+'2ClO4‘ will also produce Com(MeOH)2OEPCIO4‘, although much
more methanol is required than in the previous case. The redox chemistry
involved here is unknown, although it may be similar to the room temperature
reduction of FeIn(C1')TPP+°ClO4' by imidazole (1m) which produces
Fem(Im)2TPPCl‘.10b Thus, strong axial ligation of the metal center may
destabilze the porphyrin Tr cation radical. Similarly, Soret RR measurements
at 406.7 nm on frozen samples of CoIHOEP+'2ClO4‘ (not shown) show increased
contributions from a CoHI(H20)20EPClO4' type species with repeated scanning.
Subsequent warming of these samples causes conversion back to cobaltic 1r
cation radicals, as evidenced by the room temperature absorption spectrum,
indicating reversible reduction in the low temperature case. Reversible
conversion of NiHTPP+' to NimTPP at low temperatures has also been reported;
however, no axial ligation was implicated.5t46 In general, low temperature
RR studies are limited by elevated fluorescence levels and by the inability to
measure depolarization ratios of frozen samples. Further complications are
imposed by the increased aggregation of MOEP‘" compounds which occurs upon
cooling.4929t31947 We note that under anhydrous conditions at room temperature
both Collosr+°Clo4' and ComOEP+'2ClO4" exhibit striking stability to relatively
high power laser irradiation at 514.5 nm. Thus, in view of the above, attempts
to utilize cryogenic RR techniques to stabilize these compounds is of questionable
expediency and will not easily produce reliable spectra. Room temperature
visible excitation RR studies of these 2A2u type cation radicals will be limited,
not by sample instability, but by the weak scattering cross-section offered by

the samples.

DISCUSSION

In this section the four orbital model of the neutral metalloporphyrin

75

electronic states is introduced briefly in order to provide a framework for our
analysis of the RR and uv—visible absorption spectra of the oxidized MP species.
Next, spectra of the one-electron, ring-oxidized, MHOEP+°CIO4" systems are
shown to exhibit structural trends in both the vibrational frequencies and
electronic transition energies which are similar to those of the parent MOEP
systems. The one-electron, metal-oxidized cobaltic OEP system is then discussed
in order to establish the electronic and structural effects of metal oxidation
accompanied by axial ligation. Finally, .the above correlations and analysis
are used to examine structural and electronic properties of the two-electron
oxidized ComOEP+'2X‘ species.

In the four-orbital model of Gouterman48 the optical absorption bands
of the MP originate from transitions involving the nearly degenerate HOMOs
(highest occupied molecular orbitals) of the porphyrin 1r system, 81u(TI') and
a2u(1r), and the degenerate LUMOs (lowest unoccupied molecular orbitals), eg(1r*).
Transitions to the CI (configuration interaction) mixed excited states, Q and
B, produce the a and Soret bands, respectively. The red-shifts in the Soret
and a band maxima produced by metal variation in the series Ni, Co, Cu, Zn,
MgOEP are explained by variations in metal electronegativity producing a totally
symmetric perturbation to the electronic states of the porphyrin. Red-shifts
accompanying axial coordination are explained similarly, as an effective variation
in electronegativity of the metal resulting in charge donation into the a2u(n)
orbital of the ring system“.

Structural correlations. The frequencies of many of the vibrations of the
porphyrin macrocyclic are linear functions of the center-to-pyrrole nitrogen
distance, d, or the core size.3:’3t45 This dependency is described by the expression
v = K(A-d) where K and A are parameters characteristic of the porphyrin

macrocyclic.49 For a given vibrational mode, the K value, i.e. the slope of

76

the line, is indicative of the P.E.D. (potential energy distribution), and increases
with the percentage of Cacm stretching character.32t50 Correlations of
vibrational frequency to core size for a series of metal substituted porphyrins
can be used to establish normal mode assignments. This approach has been
used to assign vibrations of metallochlorophyll 3 complexes51 and is particularly
useful when combined with vibrational data from isotopically substituted
porphyrins.52 In the following, we use depolarization ratios and core size
correlations from RR measurements to assign vibrational modes in MOEP+'
species. Similar analysis has been applied to establish RR band assignments
of iron octaethylchlorin complexes.53

Table 4-1 collects RR frequencies from 1300-1700 cm'1 obtained with
Soret excitation of divalent metal OEP+° complexes with C104" counterions
and juxtaposes them with known frequencies of the corresponding starting
materials. Depolarization ratios measured with 363.8 nm excitation and
wavelengths of the Soret band maxima are included. For MOEP+'C104' complexes
with core sizes of 2.00 A or less (i.e., M = Ni, Co and Cu), our Soret RR results
( )ex = 363.8 nm) agree well with those of Kim _e_t_§t_l_.20 ”ex = 406.7 nm). This
is because, for oxidation using AgClO4 or electrochemical techniques, diacid
impurities occur only for complexes with core sizes larger than 2.00 A (i.e.,
M = Zn and Mg).21 Here we use RR frequencies for NiOEP+’CIO4" taken from
Kim gt. 91.20; however, we employ alternative vibrational mode assignments
consistent with those we have made for other MOEP“ species. Figure 4-11
shows plots of these vibrational frequencies as a function of core size, assuming
that there is no appreciable change in d upon oxidation of the porphyrin ring.
The core size measurements are those used by Spaulding _e_t_ _a_1_.45 Given the
assumptions inherent in the plot (see below), Figure 4-11 indicates that the

high frequency RR bands of the MOEP’” in the 1400-1700 cm‘1 range are metal

.Ibn 00:93.2... cm genome» coma—33.0.: Iona voaaaaonco 0.53 3.04.2592. am < .23 a .«o '05:: unfit
6030393 I: 9an sum: $0.530- 0503 {moose—.05:— 5 sea“ 30:0... cemuantoaoaono

. ON OUBULUHOG . SM" UUCOLUMUG‘

77

 

 

 

 

 

 

 

 

 

 

 

A

is

awn Hon umm ~mm I

5" man man in is

ooc pan ~mn ~mm «when

I I
an.m ppm Ah.ovp~m~ Am.cv~mw~ Am.ovuvw~ ummu code

ch.wv cm.m Anovvmav Am.ovm~m~ Am.ovmnw~ Ap.ov>vw~ mmmu a
«9.0 can Ac.ovocw_ Av.ovo~m~ Av.ovcsm~ «Nan puma

Amm.pv mm.m Amwuvneu Am.ov~mm~ Av.°v~mm~ an.ovamm~ «cum 3
8.» 2" 8.882 8.882 3832 82 noun".

Aoa.mv mm.a Apuuvpmn Am.ovmmm~ Am.°vmmm~ Aw.ovvpm~ when a
I I
om.n can Av.oV>pv. “v.9vbmvn Av.ovncm~ -m~ one

Amm.mv mm.m vfimamvnmn Av.cvnm¢~ An.ov~on~ Am.oV-m~ mung a
.-
an.ovuva_ an.ovowm~ an.ovnmn~ cpnu zvo

an.ovm>m~ An.cvmpn~ 0A«.ovm>n~ «and a

3; ExTBC . +265 1.596 .188 emcee a. name; .322
< a tween A5050 0
.wc~o.+.—m9‘ .3825: .8300 : ucmvcoaaouhou :93 van nuance—loo mug econ...
you AIS algal! comanbouaz 1.03.5 van 0036: comacntuaaoa .Aullov gmocosaoun cal-E 00:93:3— eHIQ gnaw“.

78

I650 -

VIBRATIONAL FREQUENCY u(cm")

Figure 4-11 .

|600

I550

l500

 

    

5 11+.
.NiiDZd) Ni Co Cu Zn 3

l J l J l l— L 4 l

j l J I
l.940 |.960 I300 2.000 2.020 2.040 2.060

 

d Ct-N(Z\)

Porphyrin core vibrational mode frequencies (for Raman
allowed bands, 1450-17 00 cm'1)13_. porphyrin core size for
the indicated MOEP complexes and their corresponding

cation radicals, MOEP+'CIO4'. Mode assignments are accord-
ing to reference 33a. Open symbols correspond to parent
MOEP frequencies; filled symbols correspond to MOEP+'C104'
frequencies. The regression analysis does not include the Ni
(02d) points. Ct-N distances are from reference 45.

79

dependent and correlate to core size in the same manner as the analogous bands
of the neutral parent species. The close agreement between the K values obtained
for the neutral MOEP and the MOEP“ for each mode suggests that little change
occurs in the normal mode composition upon oxidation of the porphyrin ring.54
Together with the relative agreement of the depolarization ratio measurements
for modes of the MOEP and MOEP‘", these core size correlations provide support
for our band assignments in the MOEP“ spectra. Further confirmation will
be provided by RR measurements on MOEP“ incorporated with isotopic
substitutions, which are in progress. Our results thus far suggest that, in general,
normal mode calculations for the parent MP will remain useful for interpreting
the vibrations of the MP+° species.

Additional evidence of the correlation of vibrational frequencies of the
porphyrin core stretching modes above 1400 cm'1 to core size is provided by
IR studies of MOEP“ compounds.12 If we use complexes which presumably

exhibit D4h solution geometries,55,56 i.e. CoInOEP+'2ClO4', FeIIIOEP+'2C104’

and FemOEP+'2CF3803', we can correlate their diagnostic IR band frequencies12
to the core sizes provided by Spaulding _et 11:45 Although the number of points
is too few to establish a reliable correlation, we can estimate K and A from
these data to be 420 cm’I/A and 5.67 A, respectively. They compare remarkably
well to K and A values calculated for V38 from IR measurements of MOEP
complexes obtained at 15K in argon matrices.52 While there is some controversy
as to the normal mode composition of v33,32’33952 these IR frequencies yield
K and A values of 421.3 cm'l/A and 5.69 A, and imply predominantly CaCm
stretching character for this mode. For CoOEP, v33 is reportedi’2 at 1565
cm‘l, whereas the diagnostic IR band of ComOEP“'2ClO4' is reported at 1554
cm'l. The ~10 cm‘1 differences in these frequencies is in good agreement

with the frequency decreases in CaCm stretching modes resulting from porphyrin

80

ring oxidation observed in our Raman work. This, and the agreement between
K and A values, suggest that the diagnostic band of MOEP“ complexes
corresponds to v 33 of the parent M OEP.

In addition to RR frequencies, Table 4-1 also lists Soret maxima for MOEP
and MOEP“ compounds. These values, along with a band energies for the
parent MOEP, are plotted as a function of a core size in Figure 4-12. Absorption
data for MgOEP (408 nm Soret, 580 nm a band) and MgOEP+'ClO4' (392 nm
Soret) in CHZCIZ have been added. Our spectrum of the MgOEP+'ClO4" resembles
the previous measurement by Dolphin _e_t_ al.4 The Soret maximum for
MgOEP+°ClO4‘ used in Figure 4-12 represents the lowest energy transition
in the near uv envelope of the absorption spectrum. The qualitative similarity
of the trends for all transition energies is striking and can be explained by the
dominance of the a2u(1r)+ eg(1r*) character of the transitions in determining
the core size dependencies of these values. For the neutral MOEP the aguUT)
orbital energy may be considered to vary linearly with core size while the eg(TT*)
and a1u(1T) orbitals remain relatively constant.48’57‘59 Thus, the a2u( 1' ) +eg(" *)
character of the excited state determines the core size dependency of the
and Soret band energies. Calculations by Edwards and Zerner23 predict that,
in the case of the 2A2u cobaltic porphin radical, the dominant transition in
the Soret region has 3096 agu(")+eg(1T *) character, while for the 2A111 cobaltic
species the least energetic transition in the Soret envelope has 70%
a2u( 1r)->eg(1r"') character. Thus the correlations in Figure 4-12 suggests that
the a2u( 1r)+eg(1r*) character of these transitions is responsible for the core
size dependency of the Soret band energies in both the MOEP and MOEP‘"
compounds.

Porphyrin core geometry. The key assumption made in establishing the

parameters K and A in the relation v = K(A-d) for the MOEP‘" series is the

81

 

 

 

 

27
O
O
Q 26
_\
I
E 25
).
0 A
0: n
LIJ
2
Lu l9 -
Z -
9 l8 *-
[_—
<75 L 0
Z o (X
<12 I7 *-
E L- Ni Co Cu Zn Mg

'6 1 I L L 4 I 1 l 4 J L I l

l.960 l.980 2.000 2.020 2.040 2.060
0
Ct-N (A)

Figure 4-12 . Electronic transition energies vs. porphyrin core size for

the indicated MOEP and MOEP-F C104" complexes. Open
and filled symbols correspond to absorption band maxima
of MOEP and MOEP+‘CIO4", respectively.

82

invariability of the core size (d) upon oxidation of the parent MOEP. Table
4-2 summarizes the relevant parameters which describe the four X-ray crystal
structures of metalloporphyrin 1r cation radicals determined to date, and
compares them to the structures of the analogous parent compounds. The table
also compares perchlorate ligated ferric TPP and OEP as a control to establish
possible steric effects limited to TPP complexes. In each case the agreement
between the structures of MP and MP“ is good in terms of the Ct-N distances
and the metal displacements from the plane of the pyrrole nitrogens. However,
the porphyrin core structures of MgTPP’" and ZnTPP‘“ display larger deviations
from planarity than do the model parent structures. This is not considered
to represent a significant general structural deviation between MP and MP4”
and is explained rather as a result of steric interaction of the bulky C104' counter
ions and the mesophenyl groups resulting from crystal packing forces.l3 Two
other examples support this idea: 1) While crystalline CuTPP+'SbC16‘ exhibits
"unusually large ruffling of the porphyrin core", the CHZCIZ solution structure
is planar;55 2) Crystalline Fem(ClO4')TPP exhibits core ruffling almost as
large as the perchlorate ligated MTPP‘“ examples, whereas Fem(C104")OEP
has a planar core (see Table 4-2). Thus, we conclude that CH2012 solution
structures of the type MOEP+'CIO4' do not exhibit significant ruffling of the
porphyrin core, and the lines described by Figure 4-11 and 4-12 are useful
references, uncomplicated by the effects of macrocyclic distortions. The success
of the correlation attests to this and suggests that in general we can assume
that in solution the Ct-N distance of the MP‘“ and probably other aspects of
core geometry are comparable to those of the parent MP. Thus, it is assumed
that no significant changes in geometric structure of. the porphyrin ring
accompany the abstraction of an electron from the a2u(1r) or a1u(1r) orbital

of the metalloporphyrin in solution.

.Amv .muu oonv <wm.~uv
sum: Luzuoad .d co.” ho uocdunmv ziuu a £uml oco .uouauoauan Huum>uu aucmaumv 02a umam;xo ca luau msmuhzauon Lumaoof

.o:~o> bag“ nag“ «nag on flaw: oucduumv Ziao on» ondna o~u>oouool 0:0 loam vouoaanmv am so 0:» Mn

.OOGCUOMU Zia-Dv

.uousuuauun ovw~ag u x auaommu>a loam ouczo>o ca nanououaou nwsbu

.000:¢auwv ZIUE IOLM UUCfldOMU Zlao OHIAQUHCO 0H LUVHO Cm Uvflfiflflfl um USMC) Imfih

a

.amuhgauoauhvwzhnnuuou «human

 

 

 

 

 

 

 

 

 

 

n.o« m.o mam." mm aaaxwozuvuuuoa
vo.on m~.o sea." no emoxwozovuuuou
gauged . mam." mm mHopm guano
~mm.~ em unease suzaaeeeag o e u .+
. - . a e
a o
«no a. a wee N am aeauxomuvuuuou sueeaz g o nee N no .+eaa Aiozovuuaou
v_.oa _m.o _oo.« we aesx xv on gnomes. m.o “a.“ no m~onm amaa ~00 on
u - HH~ eHeeuag e i .+ i a“.
e_.ou an.o eeo.~ Hm easapasaceu e.ou pen.e eeo.~ m2 .+aaexwo~oveu
a np~.o emo.~ ow amp.o~=.uz e.o» ne.o amo.~ a” .+maaxwo~ovuz
fi¢ Adv Adv a(u EM Afiu
ozeeau .eum a: ofieesm :Teo 2:50 ..em a:
OLOU Ziuo ziuo ouoo .+
.nvczoaloo vououvm _vco uflcumvam c0300 I ambasaLaOAAqux .«o nonaaoauam 13280 5.5 aLOuOfluuum ouwm 0.30 .NIQ 0H£GH

84

Ring buckling effects. Distortions of the porphyrin core are known to weaken
the bonding in the macrocyclic, thus lowering the frequencies of stretching
vibrations of the core and causing negative deviations from the relation v =
K(A-d).67 The classic example is shown in Figure 4-11 by the vibrational
frequencies of the ng form of NiOEP,45 which displays deviations from planarity
in the macrocyclic of :I: 0.5 A.“ Spiro gt _a_l_.67 conclude that the effects of
core distortions on vibrational frequencies, although significant, are minor in
comparison to changes in Ct-N distance, demonstrating that core size is the
main determinant of the vibrational frequencies of the porphyrin macrocycle.
Our results indicate that the relationship of core size to vibrational frequency
is similar for the oxidized porphyrin ring. It is expected that the effects of
core distortions, such as S4 ruffling, will be manifest also in the vibrational
frequencies of the MP‘“, producing negative deviations from the correlations
of Figure 4-11.

This completes our structural analysis of the one-electron, ring—oxidized
MHOEP+'CIO4' system. Before continuing, it is useful to summarize our
conclusions thus far. If one assumes macrocyclic planarity of the CH2C12 solution
structures of the MOEP series considered here, the close correspondence of
the K values in the expression describing the porphyrin core vibrational
frequencies v = K(A-d) for both the MOEP and MOEP“ implies: 1) the
corrections of our mode assignments, 2) the essential invariance in the P.E.D.
of the normal modes of the porphyrin ring upon oxidation, and 3) the absence
of core ruffling in oxidized structures of type MIIOEP+'CIO4‘ (for M = Ni, Co,
Cu and Zn). The key assumption that the porphyrin core geometry (especially
Ct-N distance) remains intact upon oxidation is based on comparison of crystal
structures of MP45’60‘66 and MP+'13’55956 species and has already been

suggested.13’55 The success of the correlation of vibrational frequencies to

85

core size presented here suggests that the assumption is valid for the
MOEP+'CIO4' species. Since they seem independent of 2A2” y_s_. 2A1u electronic
state character, these linear correlations establish a reference point useful
to structurally based vibrational analysis of all MP’" species. As an example,
we predict structures for the two-electron oxidized cobaltic systems,
CoIIIOEP+'2X'. It is anticipated that a similar approach can be applied to both
ferric and oxoferryl porphyrin 1r cation radical models and ultimately to
compound I type intermediates of heme enzymes.

Cobaltic octaethylporphyrin I cation radicals. The CoIIIOEP+'2X' system
is particularly interesting owing to the variety of spectrally distinct two-electron
oxidation products possible as a function of counterions X'. Characterization
of the effects of different counterions in terms of the extent of axial ligation
and the relationship of axial ligation to the electronic ground and excited states
is essential to understanding the rich variety of absorption spectra displayed
by this system upon variation of X".69 In order to compare the electronic spectra
of ComOEP+'2X’ to the parent system, ConOEP, it is necessary to determine
how oxidation of the metal and addition of axial ligands influence the electronic
states of the OEP“ system. For example, the agreement between the Soret
maxima of CoHOEP (391 nm) and its two-electron oxidation product
ComOEP+'ZClO4’ (393 nm), would be misleading without acknowledging the
counterbalancing effects present. Recognition of the blue-shifted Soret maximum
of the cobaltous species CoHOEP+'C104' (376 nm) isolates the effect of ring
oxidation upon the Soret absorption maximum, and reveals a blue-shifted band
of diminished intensity compared to the parent CoOEP. Axial ligation by the
C104“ ion is unlikely in the cobaltous OEP+' because of the presence of an
electron in the metal dzz orbital. Thus, we consider this and the other

MHOEP+'C104' complexes discussed here to contain four-coordinate metal

86

centers, devoid of the effects of metal oxidation and axial ligation. Empirically

we see that oxidation of the ring causes a 1000 cm'1 blue shift in the Soret
maximum with reSpect to the parent CoOEP. This is compensated for in
CoInOEP+’2ClO4' by a 1150 cm-1 red-shift produced by oxidation at the metal
center accompanied by weak axial ligation by the C104‘ ions. Compared to
the cobaltous case noted above, weak ligation of the cobaltic center by both
ClO4" ions in the latter complex is more likely because the d6 configuration
of CoIII leaves the dzg orbital empty. Thus, we consider CoIIIOEP+'2ClO4'
to represent a six-coordinate complex, albeit with weak axial ligands. This
example illustrates the importance of recognizing the combined effects of metal
oxidation and axial ligation on the electronic states of the OEP’" systems in
order to interpret the absorption spectra correctly.

Effects of one-electron, metal centered oxidation and axial ligation. In
order to assess these effects on the electronic transitions of the OEP’“ system,
a good starting point is to consider the effects of metal oxidation and ligation
in the parent OEP system. For this analysis we utilize the ring-neutral
Com(MeOH)20EPX' as an example of a weakly ligated, six-coordinate cobaltic
complex. Comparison of the spectra of this species to those of CoOEP will
establish the effects of ligation and oxidation of the metal. These effects will
be compared to the effects of axial ligation alone on Cu and NiOEP.

Shelnutt _e_t_ 11.53 have shown that the red-shifts in the a and Soret bands
of Cu and Ni porphyrins upon axial ligation are accompanied by expansion of
the porphyrin ring as revealed by decreases in RR frequencies of the core size
sensitive modes. The core expansion accompanying axial ligation in Ni porphyrins,
confirmed by similar RR measurements by Kim gt a_l.69, is much greater than
in Cu porphyrins and involves a change in spin state and subsequent occupancy

of the dx2_y2 orbital of the Ni.70 These spectroscopic results are explained

87

by a relative increase in the porphyrin 8211M) orbital energy level caused by
the interaction of the metal atom with its axial ligands, and are thus in agreement
with the model for axial ligation proposed by Gouterman.48 Table 4-3 collects
RR frequencies and absorption maxima for Cu, Ni and CoOEP and their axially
ligated systems. We emphasize that with these examples we consider the effects
of ligation alone for Cu and N iOEP, and ligation accompanied by metal oxidation
for CoOEP. Inspection of Table 4-3 reveals that the red-shifts in the optical
absorptions of Com(L)20EPX‘ (where L = methanol or l-methylimidazole) with
respect to those of CoOEP are similar to those encountered upon ligation of
CuOEP or NiOEP. Indeed, the magnitude of the Soret absorption red-shift for
the cobalt system (1180 cm'l) is between that of the copper (970 cm'l) and
nickel (1800 cm’l) systems. However, the RR data clearly show that axial
ligation and autoxidation of CoOEP is not accompanied by core expansion, as
there is no decrease in the vibrational frequencies of Com(L)20EPX" compared
to CoOEP. The slight increases in v 4 frequencies reflect a decrease in the
eg( 1r*) orbital occupancy upon oxidation of CoII to CoIII and are not related
to core size. (The Cu and NiOEP systems show the opposite behavior and exhibit
decreases in the v4 frequencies). RR wavenumbers reflect the structure of
the ground state of the scattering species. Thus, red-shifts in the absorption
spectra without concomitant decreases in RR frequencies suggest stabilization
of the excited state, i.e. a decrease in the cg” *) rather than an increase in
the a2u(1r) orbital energies upon axial ligation and metal oxidation of these
species. These considerations indicate that the spectral red-shifts observed
upon metal oxidation and ligation of CoOEP are not simply comparable to those
caused by ligation alone of Cu and NiOEP. Whether these conclusions apply
to ligation alone of cobaltous porphyrins under anaerobic conditions cannot

be determined from this analysis.

88

bad
-v
mew
amm
awe
Ham
wfiw

man

.55

«Bow

.ozoaoaeifiosl

n :: uca 652..an u Ea 6523.53 n :3 “com: mco33>oenn<

.33. do: $53330 ..8 3:259: m9 0

mmm
mmm
cmm

Ham

Hmm
cum

«mm

.55

coma
wwwfi
new“
bvwa
mHmH
comm
mama

bmmfi

So

mmmw

mama

mam“

mam“

puma

“cm“

mono—oEoo anon—co v5 =5 .==o 8:35 382 ..8 2.68:
5598.2 3830 v.3 memo—.2595 .353— anaconda

mumfi
wmmfi mama
pmmfi mung
mama mum“
$3
Nam"
cmmfi mama
Nana mom~
an: fig:
Ami—=3 Egg

«ivm:ban.

dc 98 can moocoeouom a

wfimfi
mama
mama
«flag
~m¢~
mHmH
oncmfi

mama

a:

awn“
Nana
awn“
mung
cpmfi
mama
m>m~

m>m~

v:

. mam. cocoaomom m

txmmONAECEoO
Liam—02:32 :aoo
-eoaamomEOoz :zoo
amozoo

oomofioazz
ofiommoamez
ammotzavzo

wmcaxofimmoqfiv

89

COmOEP+'2ClO4‘. Having established the combined effects of metal
oxidation and weak axial ligation on the electronic transitions and RR vibrations
of ConOEP, we can consider like effects in the analogous 1r cation radical
system, CoIIOEP+'. Comparison of Soret maxima and RR frequencies (Fig.
4-7c and d) of CoIIOEP+'ClO4' to those of ComOEP+'ZClO4‘ reveals Optical
red-shifts without porphyrin cation radical core expansion in the latter compound.
The spectral differences between these two species are analogous to those
observed when comparing CoHOEP to Com(MeOH)20EPC104'. In the former
( 1r cation) case the Soret maximum shifts 1150 cm‘l, or from 376 to 393 nm;
while in the latter (ring-neutral) case, the shift is from 391 to 409 nm or 1180
cm‘l. In both cases, RR results imply no change in the porphyrin core size.
The similarity of these examples suggests that in both instances we are comparing
a four-coordinate cobaltous species to a six-coordinate cobaltic one and that
ring planarity and D4}, symmetry are maintained throughout. Furthermore,
the absorption spectra imply that the behavior of the Soret absorption in these
TI’ cations is little changed from that of the ring-neutral analogues.

As discussed above, a key assumption in our analysis concerns the effects
of ring oxidation & s3 on macrocycle geometry. Based on X-ray studies of
crystaline compounds,13 we assume that in general the core geometries of
porphyrin 1r cations, MPJ", are not significantly different from those of the
parent compound MP. By "parent" we specifically refer to a complex as nearly
identical as possible to the MP‘“ in all aspects other than macrocyclic oxidation
state. That is, the oxidation state and coordination geometry of the central
metal must be the same. For example, the "parent" compound for
ConOEP+'C104" is CoHOEP, and the "parent" for CoIH(CIO4-)20EP+° is
Com(MeOH)20EPClO4'. (The cobaltic Ti cation radical complex is now written

in a manner as to imply axial ligation by the perchlorate anions. Attempts

90

to exchange the perchlorate ligands for methanol result in reduction of the
macrocycle and formation of the cobaltic dimethanol ring-neutral complex.)
Thus, our analysis suggests that the core sizes of all four of these complexes
are the same, approximately 1.97 A.

COmOEP“ZBr'. To date, EPR and ENDOR measurements are the most
reliable criteria to establish the MP“ electronic ground state (2A2u or 2A1“).8
EPR measurements of ComOEP“2ClO4' and ComOEP“ZBr' are consistent
with the 2A2u and 2A1u states, respectively.“ Perhaps the most interesting
aspect of the correlations presented in Figure 4-11 and 4-12 is the apparent
insensitivity of the RR frequencies and Soret transition energies of the
MIIOBP“CIO4' complexes to radical electronic state. This implies that we
can use these correlations for structural analysis of OEP“ compounds of either
radical type, and we can extend our discussion to include ComOEP“ZBr".71

Aside from differences in relative intensities, the RR spectra of
ComOEP“2ClO4‘ and ComOEP“2Br' are similar (see Fig. 4-7d and g; Fig.
4-9). Table 4-4 compares the RR frequencies of these two complexes. The
vibrational frequencies of the latter (presumably a 2A1u complex), however,
are 6-7 cm‘1 lower than those of the former (presumably a 2A2u complex),
with the as yet unexplained exception of v10. Because the core size correlations
seem independent of radical electronic state, and because the frequency
differences between these two complexes are small, we speculate that they
can be attributed to differences in porphyrin core geometry caused by the
different axial ligation of these two species. Decreases in frequency of the
macrocyclic vibrations (above 1400 cm‘l) may result from either expansion
or buckling of the porphrin core. It is reasonable to assume that ComOEP“ZBr’
is a six-coordinate complex with a planar porphyrin core. The v3, v11, v2

and v10 frequencies (Table 4-4) can then be used, along with R and A values

91

Table H

Comparison of Resonance Raman Frequencies (cm-1) and of
Depolarization Ratios for Cobaltic OEP“ Complexes

COmOEP+'ZClO4" ComOEP+'2Br‘
v4 1361 (0.2) ?
03 1505 (0.5) 1498 (0.3)
v11 1605 (0.7) 1598 (0.5)
92 1617 (0.5) 1611 (0.4)

910 1642 (0.7) 1649 (0.7)

92

from Table 4-1, to predict a core size of 1.99 :i: 0.02 A for this complex. This
is to be compared to 1.974 :t 0.008 A predicted similarly for the core size
ComOEP“ZClO4". However, the possibility that the vibrational frequencies
of CoInOEP+’ZBr' are lowered from those of ComOEP“2ClO4‘ as a result
of porphyrin core distortions, rather than core expansion, cannot be ruled out

by this analysis.72

CONCLUSIONS

While the interpretation of the spectra described in this paper essentially
ignores radical designation as 2A2u or 2A1“, other researchers have emphasized
the species symmetry of MP“ compounds. Characterization of the MP“ as
2A2u or 2A1”, as revealed by spin density profiles, has been based primarily
on MO calculations,47a:73'75 as well as EPR and ENDOR
studies.498’249473973t75'77 These studies have established many useful
generalizations concerning radical type. For instance, the porphyrin ring
substituents have a profound influence, and structures of the type MTAP“ (where
A is an alkyl or aryl group in the meso position) exhibit 2A2u characteristics,8
while MHP“ (where HP=hydroporphyrin) tend to be classified as 2A1u
radicals.73,76t77 Structures of the type MOAP“ (octaalkylporphyrin where
the Cb substituents are not necessarily identical) may exhibit either state
depending primarily on the metal center and the axial ligands.8’24t73 Examples
of the latter type include the high valent catalytic intermediates of heme
peroxidases and catalases. The compound I structures of these enzymes are
usually decribed as oxoferryl protOporphyrin IX 1; cation radicals,
O=Fe1V(X‘)PP“.2 The electronic state is considered to be mediated by the
identity of the sixth ligand X', and MO calculations predict that ligation by

tyrosine or thiolates promotes the 2A1u configuration (e.g. CAT-I and CPO-I,

93

respectively)73’76 while imidazole ligation generally favors a predominantly
2A2" state (e.g. HRP-0.73,” EPR and ENDOR studies, however, are interpreted
to suggest a 2A2u configuration for both HRP—I and CPO-~I.78 The EPR
measurement of CPO-I (chloroperoxidase compound I) was judged inconsistent
with a 2A1u configuration owing to the relatively strong antiferromagnetic
coupling (between the 8:1 oxoferryl structure and the S=l/2 1f cation radical)
suggested by the spectra. The authors reason that such coupling is more likely
in the 2A2u configuration where relatively large spin densities are placed on
the pyrrole nitrogens.73

Distortions of the porphyrin ring may provide an orbital overlap pathway
for the strong I J |=250 cm'l, antiferromagnetic coupling displayed by
FeHI(Cl’)TPP“SbC16'.55’56 While such magnetic coupling was previously thought
to occur only in 2A2u radicals,8 more recent interpretations76 suggest magnetic
coupling between the paramagnetic metal center and porphyrin 1r cation radical
can occur in either radical type, with the radical type influencing the magnitude
of the interaction.553 While the weak coupling, |J|=2 cm‘l, displayed by
HRP-I78 has been attributed to possibly a planar configuration of the porphyrin
ring,56 the stronger, IJ [=37 cm'l, coupling for CPO-I78 may reflect a more
buckled configuration (other explanations are also possible11176’78979). Recent
work clearly indicates that evaluation of porphyrin core geometry must be
addressed in order to interpret results from magnetic based techniques.

The similarity of the vibrational frequencies in the 1400-1700 cm"1 range
of the pr0posed 2A2u and 2A1u 1r cation radicals ComOEP“2ClO4' and
ComOEP“2Br', respectively, suggest that these frequencies are insensitive
to radical type. There are, however, small wavenumber differences which can
be attributed to relatively minor differences in porphyrin core geometry of

these two compounds. Because these geometric differences are as yet

94

speculative, other interpretations are plausible. Although these vibrations of
the porphyrin macrocyclic may well be insensitive to radical electronic state,
the abnormally high values of the v (Cbe) modes provide a clear basis to identify
ring centered oxidations, particularly in MOEP“ species.22 Both the frequency
and relative intensity of the v4 mode (1340-1370 cm'l) may reflect 2A2u y_s_.
2A1u character. More information concerning electronic states will be provided
by RR excitation profiles. Our intent in this work is to establish an approach
to evaluate core geometry of oxoferryl porphyrin 1r cation radicals in enzymes

in order to complement the NMR, EPR, and ENDOR results for these sytems.

ACKNOWLEDGMENTS
T.O. thanks Dwight Lillie. and Bob Kean for their computer programming
and Harold Fonda for discussion. This work was supported by NIH grants

GM-25480 (G.T.B.) and GM-36520 (C.K.C.), and NSF grant CHE-86-10421 (G.E.L.).

10.

ll.

12.

REFERENCES AND NOTES

(a) Frew, J.E., Jones, P. in "Advances in Inorganic and Bioinorganic
Mechanism"; Academic Press: New York, 1984; Vol. 3, pp. 175-215;
(b) Hewson, W.D.; Hager, L.P. in "The Porphyrins"; Dolphin D., Ed.;
Academic Press: New York, 1979; Vol. 4, pp. 295-332.

(a) Norris, J.R., Sheer, R.; Katz, J.J. in "The Porphyrins"; Dolphin, D.,
Ed.; Academic Press; New York, 1979; Vol. 4, pp. 159-195; (b) Lubitz,
W.; Lendziar, F.; Plota, M.; Mobins, K.; Trinkle, E., Springer Series in
Chem. Phys. 1985, 42, 164-173.

Fuhrhop, J.H., Struct. Bonding 1974, _1_8_, 1-67.

Dolphin, D.; Mulijinani, Z.; Rousseau, K.; Borg, D.C.; Fajer, J.; Felton,
R.H., Ann. N.Y. Acad. Sci. 1973, 206, 177-200.

Wolberg, A.; Manassen, J., J. Am. Chem. Soc. 1970, 9_2_, 2982-2991.
Phillippi, M.A.; Goff, H.M., J. Am. Chem. Soc. 1982, 104, 6026-6034.
Carnieri, N.; Harriman, A., Inorg. Chim. Acta 1982, §_2_, 103-107.

(a) Fajer, J.; Davis, M.S. in "The Porphyrins"; Dolphin D., Ed.; Academic
Press: New York, 1979, Vol. 4, pp. 197-256, (b) Fajer, J.; Borg, D.C.;
Forman, A.; Felton, R.H.; Vegh, L.; Dolphin, D., Ann. N.Y. Acad. Sci.
1973, 399, 349-364.

(a) Browett, W.R.; Stillman, M.J., Inorg. Chim. Acta 1981, 49, 69-77;
(b) Browett, W.R.; Stillman, M.J., Biochim. Biophys. Acta 1981, 660,
1-7.

(a) Godziela, C.M.; Goff, H.M., J. Am. Chem. Soc. 1986, _1_0_8_, 2237-2243;
(b) Goff, B.M.; Phillippi, M.A., J. Am. Chem. Soc. 1933, 13, 7567-7571.

Morishima, 1.; Takamuki, Y.; Shiro, Y., J. Am. Chem. Soc. 1984, 106,
7666-7672.

Shimomura, E.T.; Phillippi, M.A.; Goff, H.M., J. Am. Chem. Soc. 1983,
103, 6778-6780.

95

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

96

(a) Barkigia, K.M.; Spaulding, L.D.; Fajer, J., Inorg. Chem. 1983, 22,
349-351; (b)Spaulding, L.D.; Eller, P.G.; Bertrand, J.A.; Felton, R.H.,
J. Am. Chem. Soc. 1974, g, 982-987.

Nadezhdin, A.D.; Dunford, H.B., Photochem. Photobiol. 1979, _2_9_, 899-903.

(a) Teraoka, J.; Ogura, T.; Kitagawa, T., J. Am. Chem. Soc. 1982, 104,
7354-7356; (b) Van Wort, H.B.; Zimmer, J., J. Am. Chem. Soc. 1985,
107, 3379-3381.

(a) Oertling, W.A.; Babcock, G.T., J. Am. Chem. Soc. 1985, 107, 6406-6407;
(b) Ogura, T.; Kitagawa, T., submitted.

(a) Cotton, T.M.; Parks, K.D.; Van Duyne, R.P., J. Am. Chem. Soc. 1980,
_1_(2, 6399-6407; (b)Lutz, M. in "Advances in Infrared and Raman
Spectroscopy"; Clark, R.J.H., Ed.; Wiley, New York 1984, Vol. XI, pp.
211-300.

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

(a) Chottard, G.; Battioni, P.; Battioni, J.-P.; Lange, M.; Mansuy, D., Inorg.
Chem. 1981, _2_0, 1718-1722; (b) Aramaki, S.; Hamaguchi, H.; Tasumi,
M., Chem. Phys. Lett. 1983, g, 555-559.

Kim, D.; Miller, L.A.; Rakhit, G.; Sprio, T.G., J. Phys. Chem. 1986, 20,
3320-3325.

Oertling, W.A.; Salehi, A.; Chang, C.K.; Babcock, G.T., J. Phys. Chem.,
submitted.

Salehi, A.; Oertling, W.A.; Babcock, G.T.; Chang, C.K., J. Am. Chem.
Soc. 1986, 108, 5630-5631.

Edwards, W.D.; Zerner, M.C., Can. J. Chem. 1985, 6_3_, 1763-1772.

Dolphin, D.; Forman, D.C.; Borg, J.; Felton, R.H., Proc. Natl. Acad. Sci.
USA 1971,68, 614-618.

Wang, C.B.; Chang, C.K., Synthesis 1979, 548-549.

Falk, J.E. in "Porphyrins and Metalloporphyrins"; Elsevier: New York,
1964, 798. .

Alternative structural assignments for these compounds, particularly
Com(MeOH)20EPBr‘, are possible. A mixed diaxial species of the type
(Br‘)Com(MeOH)OEP was prOposed by Johnson, A.W.; Kay, I.T., J. Chem.
Soc. 1960, 2979-2983 and may better explain the similar, yet distinct
absorption maxima of Com(MeOH)20EPClO4' and Com(MeOH)20EPBr‘
(see Table 4-1 of the Discussion section). Further addition of MeOH to
the latter comfiound produces a sample with identical absorption maxima
to that of CoI (MeOH)20EPClO4'. For this work, however, a distinction
between dimethanol ligation YE.- mixed axial ligation of these complexes
is not necessary. See also: Sugimoto, H.; Ueda, N.; Mori, M., Bull. Chem.
Soc. Jpn. 1981, 5_4, 3425-3432.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

97

Under certain conditions, however, molecular oxygen (02) was observed
to promote oxidation of metalloporphyrins. For instance, in the presence
of excess (“01 mM) tertbutyl ammonium perchlorate, oxidation of CoOEP
to ComOEP“2ClO4' was observed, presumably via the formation of HClO4
in CH2012 solution, as discussed by Fukuzumi, S.; Ishikawa, K.; Tanaka,
T., Chem. Lett. 1986, 1-4. Upon laser irradiation in Soret resonance
at 406.7 nm or 390 nm, CHzClg solutions of CuOEP, ZnOEP, CoOEP and
ConOEP“ClO4' in the presence of 02 were observed to undergo
one-electron oxidation requiring various exposure times for each species.
These effects were not observed with 363.8 nm excitation. Similar
photochemical oxidation of CoTPP has been discussed by Gasyna, Z.;
Browett, W.R.; Stillman, M.J., Inorg. Chem. 1984, 2_3_, 382-384; and by
Yamamoto, K.; Hoshino, M.; Kohno, M.; Ohya-Nishiguchi, H., Bull. Chem.
Soc. Jpn. 1986, _5_9, 351-354. Both MgOEP and MgOEP“Br' were difficult
to work with for RR experiments; many spectra of these- compounds were
collected at the outset of our study, but they are not heavily relied on
in this work.

Fuhrhop, J.H.; Mauzerall, D., J. Am. Chem. Soc. 1969, 91, 4174-4181.

Konishi, S.; Hoshino, M.; Imamura, M., J. Am. Chem. Soc., 1982, 104,
2057-2059.

Fuhrhop, J.H.; Wasser, P.; Riesner, D.; Mauzerall, D., J. Am. Chem. Soc.
1972, 9_4_, 7996-8001.

Spiro, T.G. in "Iron Porphyrins'" Lever, A.B.P.; Gray, H.B., Eds.,
Addison-Wesley: Reading, M.A. 1983; Part 2, pp. 89-159.

(a) Abe, M.; Kitagawa, T.; Kyogoku, Y., J. Chem. Phys. 1973, 99, 4526-4534;
(b)Gladkov, L.L.; Solovyov, K.N., Spectrochim. Acta 1986, Vol. 42A,
1-100

’Boldt, N.J.; Donohoe, R.J.; Birge, R.R. and Bocian, D.F., J. Am. Chem.
Soc. 1987, in press.

(a) Brunner, H.; Mayer, A.; Sussner, H., J. Mol. Biol. 1972, 10, 153;

(b) Yamamoto, T.; Palmer, G.; Gill, D.; Salmeen, I.T.; Rimai, L., J. Biol.
Chem. 1973,2481 5211-5213.

Corwin, A.H.; Chivvis, A.B.; Poor, R.W.; Whitten, D.G.; Baker, W.E., J.
Am. Chem. Soc. 1968, _9_0_, 6577-6583.

(a) Whitten, D.G.; Baker, B.W.; Corwin, A.H., J. Org. Chem. 1963, 3,
2363-2368; (b) Tsutsui, M.; Valapoldi, R.A.; Hoffman, L.; Suzuki, K.; Ferrari,
A., J. Am. Chem. Soc. 1969, 9_1, 3337-3341.

We emphasize that the presence of trace amounts of diacid salts in
preparations of MP“ in CH2C12 is not always evident from absorption
spectra. For example no diacid contributions to the absorption spectrum
in Figure 4-5c of CohIOEP“2ClO4' are apparent, yet RR excitation at
406.7 nm reveals their presence as shown by Figure 4-3b. Likewise, RR
measurements at 406.7 nm of ComOEP“2Br' preparations invariably
reflect the presence of porphyrin acid impurities. Earlier, the blue-shifted

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

98

Soret absorption of meta110porphyrin dications allowed the detection of
porphyrin diacid impurities from uv-visible absorption spectra of these
samples (see ref. 47a). It was not until the application of the RR technique,
however, that the presence of the diacid was recognized in metalloporphyrin
1r cation radicals (ref. 21).

Spiro, T.G.; Strekas, T.C., J. Am. Chem. Soc. 1974, 96, 338-345.

Choi, S.; Sprio, T.G.; Langry, K.C.; Smith, K.M.; Budd, D.L.; LeMar, G.N.,
J. Am. Chem. Soc. 1982, 104, 4345-4351.

(a) Stewart, B.; Clark, R.J. H., Struc. Bonding (Berlin) 1979, 3__6, 1- 80;
(b) Rousseau, D. L.; Friedman, J. M.; Williams, P. F, Topics Curr. Phys.
1978, _1_1_, 202- 252.

Shelnutt, J.A., J. Chem. Phys. 1981, 74, 6644-6657.
Spiro, T.G.; Strekas, T.C., Proc. Natl. Acad. Sci. USA 1972, _6_9_, 2622-2626.

Verma, A.L.; Mendelsohn, R.; Bernstein, H.J., J. Chem. Phys. 1974, 61,
383-390.

Spaulding, L.D.; Chang, C.C.; Yu, N.-T.; Felton, R.H., J. Am. Chem. Soc.
1975, _9_7_, 2517-2524.

Johnson, E.C.; Niem, T.; Dolphin, D., Can. J. Chem. 1978, 56, 1381-1388.

(a) Fajer, J.; Borg, D. C.; Forman, A.; Dolphin, D.; Felton, R. H., J. Am.
Chem. Soc. 1970, _9__2, 3451- 3459; (b)Mengersen, C.; Subrammanian, J.;
Fuhrhop, J. H, Mol. Phys. 1976, 3_2, 893- 897.

(a) Gouterman, M., J. Chem. Phys. 1959, 30, 1139-1161; (b) Gouterman,
M., J. Mol. Spec. 1961, 6, 138-163.

Huong, P.V.; Pommier, J.C., C.R. Acad. Sci., Ser. C. 1977, 285, 519-522.

(a) Warshel, A., Rev. Biophys. Bioeng. 1977, 6, 273; (b) Callahan, P.M.;
Babcock, G.T., Biochemistry 1981, _2_0, 952-958.

(a) Fu31wara, M.; Tasumi, M. J. Phys. Chem. 1986, _9_0, 5646- 5650; (b) Fonda,
H. N.; Babcock, G. T., Proc. 7th Int]. Cong. Photosyn. 1986, in press.

Kincaid, J.R.; Urban, M.W.; Watanabe, T.; Nakamoto, K., J. Phys. Chem.
1986, 90, 5646-5650.

Ozaki, Y.; Iriyama, K.; Ogoshi, H.; Ochiai, T.; Kitagawa, T., J. Phys. Chem.
1986, 10, 6105-6112.

The decrease in K value upon oxidation for v10 is not considered +significant
owing to the uncertainty in the v10 measurement of ZnOEP+ "C104 by
Soret Excitation RR. Visible excitation of this compound at 647.1 nm,
which should more clearly show the vm band, was not possible owing
to intense fluorescence emission centered around 675 nm. Thus it is possible
that the relatively high 910 value measured for ZnOEP“ CIO4‘ anomalously

55.

56.

57.

58.

59.

60.
61.

62.

63.

64.

65.

66.

67.
68.

69.

99

lowers the K value (for 910) of the MOEP“ series.

(a) Scholz, W.F.; Reed, C.A.; Lee, Y.J.; Scheidt, W.R.; Lang, G., J. Am.
Chem. Soc. 1982, iii, 6791-6793; (b) Buisson, G.; Deronzier, A.; Duée,
E.; Gans, P.; Marchon, J.-C.; Regnard, J.-R., J. Am. Chem. Soc. 1982,
10_4_, 6793-6796.

Gans, P.; Buisson, G.; Duee, E.; Marchon, J.-C.; Erler, B.S.; Scholz, W.F.;
Reed, C.A., J. Am. Chem. Soc. 1986, 108, 1223-1234.

(a) Kitagawa, T.; Ogoshi, H.; Watanabe, E. Yoshida, Z., J. Phys. Chem.
1975, _7_9_, 2629-2635; (b) Shelnutt, J.A. ; Ondrias, M.R., Inorg. Chem.
1994, _23, 1175-1177.

(a) Shelnutt, J.A.; Straub, K.D.; Rentzepis, P.M.; Gouterman, M.; Davidson,
E.R., Biochemistry 1984, _2_3, 3946-3954; (b) Shelnutt, J.A.; Alston, K.;
Ho, J.-Y.; Yu, N.-T.; Yamamoto, T.; Rifkind, J.M., Biochemistry 1986,
_2_5, 620-628.

the model presented here is an oversimplification. Similar trends in the
eg(1r *) orbital energy levels (LUMO) are revealed by reduction potentials
for MOEP complexes. Comparative measurements of the first oxidation
potential reflect trends in the HOMO energies. The results show that
both the HOMO and the' LUMO rise in energy in the series Ni Cu, Zn,
MgOEP; however, the HOMO levels increase more, thus producing the
spectral red-shift. See ref. 29 and Fuhrhop, J.-H.; Kadish, K.M.; Davis,
D.G, J. Am. Chem. Soc. 1973, §_5_, 5140-5147.

Timkovich, R.; Tulinsky, A., J. Am. Chem. Soc. 1969, 11, 4430-4432.
Collins, D.M.; Hoard, J.L., J. Am. Chem. Soc. 1968, 92, 3761-3771.

Mitra, S. in "Iron Porphyrins"; Lever, A.B.P.; Gray, H.B., Eds.;
Addison-Wesley: Reading, Mass, 1982, Part II, pp. 3-42.

Scheidt, W.R.; Cohen, I.A.; Kastner, M.R., Biochemistry 1979, 18,
3546-3552.

Fleischer, E.B.; Miller, C.; Webb, L., J. Am. Chem. Soc. 1964, _8_6_,
2342-2343.

Masuda, H.; Taga, T.; Osaki, K.; Sugimoto, H.; Yoshida, A.-I.; Ogoshi,
H., Inorg. Chem. 1980, 950-955.

Reed, C.A.; Mashiko, T.; Bentley, S.P.; Kastner, M.E.; Sheidt, W.R.;
Spartalian, K.; Lang, G., J. Am. Chem. Soc. 1979, 101, 2948-2958.

Spiro, T.G.; Strong, J.D.; Stein, P., J. Am. Chem. Soc. 1979, 101, 2648-2655.
Meyer, E.F., Acta Cryst. 1972, 828, 2162-2167.

Setsume, J.; Ikeda, M.; Kishimoto, Y.; Kitao, T., J. Am. Chem. Soc. 1986,
108, 1309-1311.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

100

Kim, D.; Su, Y.O.; Spiro, T.G., Inorg. Chem. 1986, 2_5, 3988-3993.

Actually the question of radical ground state for these compounds is not
completely clear-cut. Of the C104' complexes used in the correlations,
only MgOEP“ and ZnOEP“ have been firmly assigned a ground electronic
state (2A1u) based on EPR studies (ref. 8). Coupling between the
paramagnetic metal centers and the porphyrin radical of CoIIOEP+'ClO4"
and CuOEP“ render these com lexes EPR silent. Although the uv-visible
absorption spectra of Cu, Co and NiOEP“ imply 2A2u character for
their ClO4‘ complexes, the electronic state of CuOEP“ClO4‘ is assigned
as 2A1u based on NMR spectra (ref. 10a). To our knowledge, the EPR
spectrum of NiOEP“ClO4‘ (ref. 29) has never been discussed in terms
of 2A2u or 2A1u ground state symmetry.

Based on RR frequencies, the precurser compound, ComOEPBr', is predicted
to have a contracted core, 911.957 :t 0.009 A. We also note that this
one-electron oxidized compound displays certain Raman absorption
spectroscopic features analogous to a structure known to possess a buckled
porphyrin core, namely the u-nitrido dimer, (FeOEP)2N. See: (a) Hoffman,
J.A.; Bocian, D.F., J. Phys. Chem. 1984, _8_8_, 1472-1479; (b) Scheidt, W.R.;
Summerville, D.A.; Cohen, I.A., J. Am. Chem. Soc. 1976, fl, 6623-6628;
for spectra and crystal structure of u-nitrido dimers, respectively.
However, it is not yet clear if the geometry of this precurser (not "parent"
as defined above) is relevant to the structure of the two-electron oxidized

product, CoIIIOEP+‘ZBr‘.

Hanson, L.K.; Chang, C.K.; Davis, M.S.; Fajer, J., J. Am. Chem. Soc.
1981, 103, 663-670.

Loew, G.H.; Herman, 2.8., J. Am. Chem. Soc. 1980, 102, 6174-6175.

Fujita, 1.; Hanson, L.K.; Walker, F.A.; Fajer, J., J. Am. Chem. Soc. 1983,
105, 3296-3300.

Fujita, E.; Chang, C.K.; Fajer, J., J. Am. Chem. Soc. 1985, 107, 7665-7669.

Chang, C.K., Hanson, L.K.; Richardson, P.F.; Young, R.; Fajer, J., Proc.
Natl. Acad. Sci. USA 1981,_'_7_8_, 2652-2656.

(a) Roberts, J.E.; Hoffman, B.M.; Rutter, R.; Hager, L.P., J. Biol. Chem.
1981, _2_5_6_, 2118-2121; (b) Rutter, R.; Hager, L.P., J. Biol. Chem. 1982,
_2_5_7_. 7958-7961; (c) Schultz, C.E.; Rutter, R.; Sage, S.T.; Debrunner, P.G.;
Hager, L.P., Biochemistry 1964, 2_3_, 4743-4754.

Sontum, S.P.; Case, D.A., J. Am. Chem. Soc. 1985, 107, 4013-4015.

CHAPTER 5

DETAILS OF PEROXIDASE CATALYSIS SUGGESTED BY RESONANCE RAMAN
MEASUREMENTS OF IRON-OXYGEN STRETCHING FREQUENCIES

OF HORSERADISH PEROXIDASE INTERMEDIATES

INTRODUCTION

Horseradish Peroxidase (HRP) uses hydrogen peroxide to catalyze the
oxidation of indole acetic acid (a growth hormone) in plant roots. The generally
accepted reaction pathway, as pr0posed independently by Chance1 and George,2
occurs in four steps:
1. HRP+ H202 + HRP-I + H20
2. HRP-I + AH + HRP-II + A'
3. HRP-II + AH + HRP + A‘ + H20
4. 2A' + products
Other pathways, however, may have physiological significance.3 The prosthetic
group of the enzyme in the native state (HRP) consists of a ferric protOporphyrin
IX ligated on the proximal side of the heme plane by histidine nitrogen. In the
first intermediate state, compound I (HRP-I), the heme most likely retains the

histidine imidazole (Im) ligand and undergoes two-electron oxidation, forming

101

102

an oxoferryl protoporphyrin IX 17 cation radical, O=FeIV(Im)PP“. The second
intermediate, compound II (HRP-II), is formed via the oxidation of the substrate
(AH) which leaves the low-spin (S=l) oxoferryl-imidazole linkage intact but
reduces the porphyrin macrocyclic back to the neutral state.4'6 The
intermediates of the HRP reaction are of interest because of the possible general
application by nature of similar structures in the catalysis of not only other
peroxidases, but also catalases, oxygenases and oxidases.7

The oxoferryl structure of HRP-I and HRP-II is well characterized by a
variety of physical techniques. Three unpaired electrons (S = 3/2) are indicated
by magnetic susceptibility measurements of HRP-I,8 consistent with a low-spin
(S=1) FeIV ferromagnetically coupled to a S=1/2 radical center. Massbauer
measurements of HRP-I and HRP-II are compatible with similar FeIv
configurations for both intermediates.9 170 electron nuclear double resonance
(ENDOR) Spectroscopy of HRP-I prepared with H21702 demonstrates the
incorporation of a single 170 atom in the first intermediate.10 Proton nuclear
magnetic resonance (NMR) studies of the second intermediate and its synthetic
models provide further evidence for the oxoferryl structure in HRP-II.11
Although somewhat controversial,12:13 recent evidence from extended X-ray
absorption fine structure (EXAFS) spectroscopy presents detailed structures
for HRP-I and HRP-II with relatively short (1.65 A) FeO bond lengths and
contracted (1.99 A) heme center-to-pyrrole nitrogen distances.14115 Recent
resonance Raman (RR) measurements of the FeO stretching frequency, v(FeO),
of HRP-1116917 and model oxoferryl porphyrinsls“21 confirm the double-bonded
character and short bond length of the oxoferryl structure. Furthermore, the
frequencies of the porphyrin core vibrational modes are consistent with the
1.99 A core size for HRP-II“:23 and for the synthetic oxoferryl complexes“):24

We present here RR evidence that these structural features are maintained

103

in HRP-I as well. RR studies of HRP-II also reveal a pH dependency for v(FeO)
which suggests H-bonding of the oxo ligand to a protonated distal histidine with
a pKa of 8.6.25126 We report a v (FeO) for the first intermediate at pH 7 which
is very similar to that of HRP-II at pH 11 in the absence of H-bonding. This
suggests that the oxoferryl of HRP-I is not H-bonded at pH 7. The significance
of these structural features to the kinetics and mechanism of peroxidase catalysis
is discussed.

RR studies of the first intermediate are complicated by its reactivity and
photolability.27 Past studies”,29 were interpreted to suggest that cryogenic
techniques may be insufficient to stabilize HRP—I to laser irradiation in the
region of the Soret absorption (300-450 nm). Because of this we have adopted
an alternative approach, using pulsed, near-uv RR excitation of flowing samples
of HRP-I generated by rapid mixing.30 Similar approaches have been applied
to reaction intermediates of cytochrome c oxidase, and both pulsed31132 and
continuous wave (CW) laser excitation33 have been used. The high frequency
(1200-1700 cm‘l) spectrum of HRP-I we obtained earlier with pulsed RR
excitation30 has been independently confirmed by Ogura and Kitagawa34 with
CW excitation. We have also carried out an extensive RR study of porphyrin
11 cation radicals (P“) and other oxidized forms of Co, Cu and Zn
octaethylporphyrin (OEP).35937 We present here preliminary results from an
additional RR study of synthetic ferric porphyrin cation radicals.38 Based on
these recent results from the P“ model compounds and further characterization
of the RR scattering from HRP-II presented here, we have reanalyzed the high
frequency RR spectrum obtained for HRP-l. We find no vibrational frequencies
characteristic of a metalloporphyrin 11 cation radical (MP“). In view of the
evidence in favor of the P“ formulation for HRP-I from uv-visible absorption,39
electron paramagnetic resonance (EPR),40941 ENDOR,42 and NMR43144

Spectroscopies, we offer possible explanations for the RR results.

104

EXPERIMENTAL

Materials. Horseradish peroxidase was purchased from Sigma (Type VI)
and used without further purification. Sodium phosphate and sodium carbonate
buffers (50 mM) were used at pH 7 and 10.8, respectively, for all enzyme and
hydrogen peroxide solutions. 30% H21602 was purchased from Mallinkrodt.
H21302 was made from 1302 (98%, Cambridge Isotope Laboratories) using
glucose oxidase (SIGMA) according to the procedure of Asada and Badger.45
HRP and peroxide solutions were quantified photometrically by using extinction
coefficients at 403 7117146 and 240 nm47 of 103 mM'lcm‘1 and 43.6 M‘lcm"1,
respectively. —

HRP-I. Flowing HRP-I samples (pH 7) for RR measurements were prepared
by rapid mixing of equal volumes of cooled solutions of 0.1 mM HRP and 1 mM
hydrogen peroxide. These were loaded in 10.0 ml syringes and driven through
two eight-jet tangential mixers in series“8 with a Sage 355 syringe pump. After
mixing, the HRP-I sample passed through a 0.5 mm i.d. capillary. Raman
scattering from HRP-I was measured both with pulsed excitation at 390 nm
and 420 nm, and with CW excitation at 406.7 nm. Enzyme solutions were
collected during the measurement and subsequently allowed to relax back to
the native state overnight, whereupon they were pre-filtered (0.45 11111 pore
size, Gelman) and concentrated (the enzyme concentration is halved during
the mixing experiment) in a Diaflow ultrafiltration cell (Amicon, PM 10) and
used again. HRP-I samples prepared from the recycled solutions had a much
longer lifetime owing to the removal of potential substrates present in the
preparations.49 Initial RR spectra of HRP-I were obtained using 0.65 - 0.85
ml/min flow rates which produced dead times of 3.1 - 2.4 s between mixing
and laser excitation. As a spatial beam width of 1.5 mm incident on the 0.5

mm i.d. capillary was used, these flow rates insured that 3.6 - 4.7 scattering

105

volumes (one scattering volume is 0.2 111) passed through the capillary between
laser pulses. After preliminary characterization of the spectrum under these
conditions, we determined that a 0.22 ml/min flow rate (corresponding to a
9.3 3 dead time) produced identical results for the recycled preparations. At
this slower flow rate, 1.2 scattering volumes passed between pulses, corresponding
to the minimum necessary to- insure that each pulse was incident on a fresh
sample aliquot. This was used to increase accumulation time for the low
frequency spectra which suffered from intense background scattering from
the capillary walls. Under the conditions described above for our system we
estimate that 1.0 mJ pulses at 390 nm provide 3» 35 photons absorbed/molecule,
for those molecules from which Raman scattering is collected, regardless of
flow rate. For the CW experiment at 406.7 nm, we estimate 22 and 69
photons/molecule absorbed by the sample at flow rates of 0.70 and 0.22 ml/min,
respectively, for a 15 mW laser power, and 0.5 mm beam waist. Thus, for pulsed
excitation, photon flux is independent of flow rate above a threshold which
insures that each pulse is incident on a fresh sample aliquot. Photon flux then
depends directly (and exclusively) on pulse energy if spatial beam width and
capillary i.d. (which determine scattering volume) remain constant. On the
other hand, for CW excitation, photon flux is inversely proportional to flow
rate and directly proportional to laser power for a system in which the laser
beam waist and capillary i.d. are equal (i.e. all photons are incident on the sample
and all sample molecules are illuminated).

The formation of HRP-I was confirmed for each rapid mixing experiment
by visual detection of the flowing green solution in the capillary prior to the
onset of the RR measurement. After laser irradiation a sample aliquot was
collected and checked by a uv-visible absorption measurement during each run.

For the recycled samples these spectra were essentially that of HRP-I. For

106

the less stable fresh samples these spectra reflected a mixture of HRP-I and
HRP-II, confirmed by monitoring the return to the native state. The total time
from mixing to the absorption spectrum was typically '1. 3 minutes or longer.
The formation of HRP-I prior to the RR measurement was further confirmed
by an independent optical absorption experiment in which the rapid mixer with
a section of rectangular tubing (0.3 mm path, Vitrodynamics) in place of the
capillary was used.

HRP-II. HRP-II samples were prepared in the cooled cylindrical quartz
cell which was spun during RR measurements. Addition of stochiometric amounts
of p-cresol (Aldrich) and 1.5 x stochiometric quantities of hydrogen peroxide
to solutions of native HRP at pH 7 resulted in formation HRP-II. RR spectra
of these samples were collected with OMA detection for 2.5 min after which
absorption spectra were monitored and the experiment was repeated with a
fresh sample. HRP-II samples at pH 10.8 were prepared by adding 1.5 equivalents
of peroxide to solutions of the native enzyme. RR spectra of these more stable
solutions were collected for 10 min, and absorption spectra were monitored
before and after laser irradiation.

Model compounds. Cl'FemOEP was prepared from OEPHZ (Porphyrin
Products) by standard methods.50 Cl‘FemOEP“CIO4‘ was prepared by A.
Salehi by stirring Fe(ClO4)3 in a CH2C12 solution of Cl'FemOEP. Sample
integrity was monitored by optical absorption before and after RR measurements.
CH2C12 was freshly distilled from CaHz.

Instrumentation. Laser pulses (5 ns duration, 10 Hz) from 390-435 nm were
provided by a DCR-IA based Quanta Ray System by using three laser dyes
(Exciton): LD390 (in methanol), DPS (in dimethylformamide) and Stilbene 420
(in methanol). CW emission at 363.8 nm came from a Coherent Innova 90-5

Argon ion laser. Raman scattering from these sources was measured with a

107

 

 

A
n
V

  
 

 

 

 

   
  
 

Lama.)
P.C. diam. .

8 mm. diam.

1*:

 

 

. 05 mm.
Tr" View in direction of arrow X

.A ll.

8 Section B-B

 

1_ 4
1 cm.

Figure 5-1 . End view (a), longitudal section (b) and cross-section (c) of
the eight-jet, tangential mixers. Ref. 48.

108

8:08:80 :a :o CESEEC was 3a.. :9. mm 3

9E. .:o:3:oE=.=m:_ £5 38: 8.5908 cm? was :8 wE:
15% 5.3.6 Bot—05:8 A0 mtmv @280 a :m 33:28 :42:
Eon.“ E: 33723 a: 8:88 @2338 mm .moEobooE

= <20 88888 use .6698 8...; oboe S.,: 5.:
080m of wcmm: an 033.3: “3:: 030320 28 ...oquoEo
..o:oE 23.5 a :_ vomhoammc £333.35 3: xoam a :33
@3828 $3 Cm 5333 :3 29:8 9: E0...“ mieguaom
:aEwm dong <Tmon waruuz 9: Eob AS: 38 2593:
EEC. 2: 2:3 2149: nomad 25 033m mam 3:25 95 E
can .82: c254 :ozoxm wEaEa an 825.8 me? An: 3
63.5%.: H 883: m: 8 E: can “a 53:62: .893 .m5me
Raw: .3 ©2395 3353 7mm: wfizoc .«o mECEoSmon
:wEam 85:89. conga .8.“ Com: 5:2sz:90 329::me

.NA 95»:

1109

00¢

Nmmm

 

U

 

 

.3. D

Nuvucougsa_mug

mega .ugggzm
«an .uum

<Humua
uennma

 

 

an n
“ «.mgm _ m ea c

 

 

A

unuam

 

 

 

 

.«ususmmss
«kmN aqua

 

VA

.2383:
auva

aoao.ugn
‘ cam”

 

 

 

us-osu:gu
Q~Nn

 

 

 

 

 

 

 

HquG
mtm Guam

 

hoaxmsnu
w-~

 

 

110

commercial Spex/EGécG PAR diode array spectrometer described earlier.30
Instrumentation used to measure Raman scattering from the 406.7 nm Krypton
ion laser (spectra Physics) emission included a Spex 1401 double-monochromator
and is described elsewhere.36 Uv-visible‘absorption spectra were measured
with a Perkin Elmer Lambda 5. Figure 5-1 and 5-2 diagram the experimental

apparatus.

RESULTS

High frequency RR scattering from HRP intermediates. Figure 5-3 shows
RR spectra obtained with 390 nm excitation of horseradish peroxidase at pH
7 . The spectrum of the native enzyme, Figure 5-3a, is typical of a
five-coordinate, ferric protoheme complex with a high spin (S=5/2) ion. The
high frequency vibrations from 1450 to 1700 cm'1 are inversely proportional
to the center-to—pyrrole nitrogen distance, or core size of the porphyrin
macrocycle.51’52 Thus, these modes are sensitive to the spin state of the central
iron as controlled by the ligands normal to the heme plane. The bands in Figure
5-3a at 1498, 1548 and 1573 cm"1 are assigned to the normal modes v3, v11
and v 2, respectively, as described by Abe _e_t_ _a_l_.53 The shoulder at 1548 cm"1
possibly contains contributions from V38 as well.54 The band at 1630 cm"1
may also correspond to a superposition of two vibrations: one is a stretching
mode of the peripheral vinyl, v (C=C), and the other is the v10 mode of the
porphyrin macrocycle.21’55 The intense band at 1373 cm‘1 of the native enzyme
is assigned to v 4, the oxidation state marker. Its frequency is inversely related
to the occupancy of the porphyrin eg( 7' *) orbitals. Although these are the lowest
unoccupied molecular orbitals (LUMO) of the porphyrin, they may be populated
by overlap with the iron dxz and dyz orbitals. This is reflected by lowered

vibrational frequencies of modes with CaN stretching character such as v 4.56

Figure 5-3 .

111

High frequency RR spectra of HRP samples at 2-10 C,

l mJ/pulse. Sample conditions and total accumulation
times: (a) spinning cell, 20 min; (b) flowing in the capillary
of the rapid mixer at 0.65-0.85 ml/min, 40 min;

(c) Spinning cell, 22 min.

112

HRP Xex = 390 nm

7
H
p

 

a) native

0mm. l

mmm. I
won. I

“W

mhfl |

b) compound I

>tmzm._.ZH Z<_>_<m

OnQ l
an l
mmm. I

mom. |

wN! I.

mNN_ I

c) compound II

I300 I 500 [700

RAMAN SHIFT (cm")

IIOO

 

113

Figure 5-3b and c show the RR spectra of HRP-l and HRP-II, respectively.
The similarity of the 390 nm RR spectra of these two intermediates is surprising
considering the significant differences displayed in the vibrational frequencies
and RR intensities of metalloporphyrin 1r cation radical (MP+') and their neutral
meta110porphyrin (MP) parent compounds.35’36’57 This strong similarity was
obscured in our previous analysis of the high frequency RR Spectrum of HRP-I30
because we compared it to RR spectra of HRP-II produced with excitation at
514.5 or 457.9 nm from early work.58v59 Our spectra of HRP-II with excitation
from 390—435 nm match the more recent work with 406.7 nm excitation of Terner
and coworkers22 as well as Kitagawa and coworkers,17 both in vibrational
frequencies and relative intensities. Accordingly, a reanalysis of our high
frequency RR spectrum of HRP-I is in order. From the RR spectra in Figure
5-3b and c we assign the bands at 1378, 1508, 1563 and 1630 cm‘1 to v4, v3,
0 11 overlapped with v33 and v(C=C), respectively, for both HRP-I and HRP-II.
(The resonance enhancement of v10 at this excitation wavelength is negligible,
see below.) Thus, the spectra differ only in the relative intensity of the v2
mode, which appears at 1583 cm‘1 in the spectrum of HRP-II (Fig. 5-3c) and
is all but absent in the spectrum of HRP-I (Fig. 5-3b).

Model compounds. Figure 5-4 depicts the changes in vibrational frequencies
and relative RR intensitites between a ferric octaethylporphyrin 1r cation radical,
Cl‘FeIIIOEP+’CIO4' and its parent compound, Cl‘FemOEP. Studies of analogous
tetraphenylporphyrin60 complexes suggest that both of these compounds exhibit
a five-coordinate, high-spin (S=5/2) configuration for the heme iron and are
similar in all aspects except the oxidation state of the macrocycle. In the
spectrum of the cation radical (Fig. 5-4a) we recognize the v4, v29, v3, v11,
v 2 and v10 modes at 1368, 1388, 1491, 1584, 1600 and 1625 cm‘l, respectively.

Although the high frequency envelope from 1560-1660 cm‘1 is poorly resolved

RAMAN INTENSITY

114

 

 

 

  
 

 

 

II:
<3 8
Xe, = 363.8 nm 1:” 3.2
‘ m '2'
(D I
'-‘-’ 8
8' a 2
a) Cl’Fe"'OEP+'CIOZ I2 3_- I
h ' '
9!
__ T“)
a ‘2 a
.... I O
I {j 8 ?
I 9.! (D
' o _
a 2‘ a
9. I* g, 7
I :0 V
:0 T h
In T V
9 s ..
- - a)
' ' 3 6‘
m I T
b) Cl'FemOEP 2
I
l r r I I I I I 1
900 ”00 I300 l500 I700
RAMAN SHIFT (cm")
Figure 5—4. RR Spectra of model compounds in CHZClz. Conditions:

40-50 mW incident on sample in spinning cell at 25 C.
Accumulation times: (a) 5.3 min; (b) 10 min.

115

in Figure 5-4a, three bands with distinct depolarization ratios are evident from
polarized scattering (not shown). These vibrational assingments are made by
analogy to our previous study of Co, Cu and ZnOEP+' complexes.36 The normal
mode assignments in the 1450-1700 cm"1 range from that work are based on
core size frequency dependencies and depolarization ratios of the Raman bands
of these MOEP“ species. The assignments of v4 and v29, however, to the
features at 1368 and 1388 cm“1 in Figure 5-4a are tentative because these modes
were not included in the previous structural analysis of MOEP spectra.36
Frequency shifts. The vibrational modes above 1350 cm"1 of the neutral
species, Cl'FemOEP are straight-forward: v4, v 29, v 3, v 11, v2 and v10
occur at 1375, 1406, 1494, 1557, 1581 and 1629 cm‘l, respectively (Fig. 5-4b).
The RR enhancement of v33 observed for protoheme complexes does not occur
here owing to the higher symmetry the OEP macrocycle, and of course there
are no vinyl substituent vibrations. The changes in the vibrational frequencies
which accompany ring oxidation of these ferric compounds agree qualitatively
with those observed with other MOEP“ species,36 but are smaller in magnitude
for v4, v3, and um. We find that stretching modes with predominantly CaN
character (v4) or CaCm character (v3 and v10) decrease in frequency, while
those with Cbe character (v11 and v2) increase in frequency when the
porphyrin ring is oxidized. We see these frequency shifts regardless of whether
the electron is abstracted from the a1u(") or a2u(1r) molecular orbital. Thus,
the observed changes in vibrational frequencies are not completely consistent
with predictions made based on either the spin density profiles of
metalloporphyrin 1r cation radicals (MP*")28’30’61 or the nodal structure of
aluh) and 82u(1r) orbitals.56962:63 These two approaches (which yield
predictions inconsistent with one another) fail possibly because they consider

only changes in the Ir bonding of the macrocycle upon oxidation and ignore

116

possible changes in the a framework.64 Consideration of both a and IT effects
to the force constants may better explain the changes in vibrational frequencies
of the porphyrin core upon oxidation.

Low frequency RR scattering from HRP intermediates. Figure 5-5 shows
low frequency RR scattering from pulsed, 390 nm excitation of HRP, HRP-I
and HRP-II. The spectrum of the native enzyme, Figure 5-5a, is similar to
that measured with 406.7 nm excitation reported earlier;16 however, there
are some differences in relative intensities produced by the different excitation
frequencies. Spectra of HRP-I at pH 7 prepared with both H21602 and H21802
appear in Figure 5-5b. The feature at 791 cm"1 in the spectrum of the 160
sample is replaced by a band at 754 cm“1 in the 180 sample. This is clearly
a mode involving an oxygen atom incorporated from the peroxide. A 35 cm‘1
decrease is expected for a Fe-O harmonic oscillator upon substitution of 18O
for 160. Thus, this band is assigned to v(FeO), the symmetric stretch of the
oxoferryl of HRP-I.

Figure 5-5c and d show spectra of HRP-II at pH 7 and 10.8, respectively.
As shown by recent work,16’25 at pH 7 the oxoferryl stretch of HRP-II appears
as a weak feature at 774-779 cm‘1 in RR spectra produced with 406.7 nm
excitation. This band shifts to 740-743 cm"1 upon substitution of 18O for 160.
Our RR spectrum obtained at A ex = 390 nm of the second intermediate at pH
7 is consistent with these findings. At our signal/noise level, the v(FeO) band
is not clearly visible in Figure 5-5c, although its contribution appears at
approximately 775 cm‘l. On the other hand, for HRP-II at pH 10.8, RR excitation
at 390 nm results in strong enchancement of v(Fe0) as shown in Figure 5-5d.
Our results at pH 10.8 match those of Hashimoto gt 11:17 and Sitter gt 31.25
We find the v(FeO) at 787 cm‘1 for HRP-II at high pH, shifting to 753 cm'1

upon 180 substitution. Thus, the v (FeO) measured for HRP-I at pH 7 appears

117

Figure 5-5 . Low frequency RR spectra of HRP samples obtained under
390 nm excitation. Total accumulation times: (a) 85 min;
(b) 160 sample: 144 min, 180 sample: 78 min; (c) 52 min;
(d) 160 sample: 48 min, 180 sample: 48 min. The
feature at 981 cm"1 in (b) is due to sulfate ion in the
H21302 preparation.

RAMAN INTENSITY

118

-H28

HRP x9, = 390 nm

a) native

pH?

 

  
  
 

 

g _
‘f a) o 2
#2 I
g HZIGO2
I
I N
— In
(D 2
0‘ I
I
b) compound I \
pH 7 \
B
H2 02
In
(D
O)
m l

g N

£0 0)

l I

HZIGOZ
c) compound 11
pH 7
l L L J 1 l 1 J 1
400 600 800 IOOO IZOO

RAMAN SHIFT (cm")

RAMAN INTENSITY

119

-679
-II3I

 

3
(D IN
Eve ID 8
I IN _ O) _
I I0 I l
O'I
I
I6
__ H2 02
'2
I
I 33
8 IO 0"
m :9 I
(:0 compound 11' j 9 I
[s
I
pH l0.8 '
I8
H2 02
l J l 1 l l I l l
400 600 800 I000 |200

RAMAN SHIFT (cm")

120

to be almost identical to that of HRP-II at pH 10.8. By analogy to the work
of Sitter £11.25 and Hashimoto gt a_l.26 this suggests that the oxo ligand of
HRP-l is not H-bonded at pH 7 as is the case for HRP-II. Experiments to confirm
this are in progress.

Aside from the similarity of the v(FeO) of HRP-l at pH 7 to HRP-II at
pH 10.8, other low frequency features require comment. The spectra of both
of these species (Fig. 5-5b and (3) exhibit slight frequency decreases in v 7 (N 679
cm'l) compared to HRP-II at pH 7 (682 cm‘l, Fig. 5-5c). This is consistent
with observations by Sitter fit 11:25 concerning the pH dependence of the RR
spectrum of HRP-II. The feature at 641-644 cm'1 in the spectra of HRP-I (Fig.
5-5b), however, does not appear in the high pH HRP-II spectra (Fig. 5-5d). We
offer no assignment for this mode at present but we note that an out-of-plane
asymmetric vinyl wag is expected in this region.65 Although Terner gt a_l_.16
assigned this mode to the poorly resolved band at 'b 691 cm“1 labelled in Figure
5-5a and c, the 641-644 cm"1 value is closer to the expected frequency (630
cm‘l). The assignment by Terner and co-workers represents the first possible
observation of this mode in a heme complex. Its activation could be caused
by the co-planar configuration of the pyrrole 1 vinyl group with the heme in
HRP, which may serve to stabilize the porphyrin cation radical of HRP-I.66
There has been no vinyl—protein interaction unique to HRP-I proposed. Should
the 641-644 cm'1 feature arise from the vinyl groups, its unique activation
in HRP-I might be an indication of such or else it could be due to electronic
factors.

Excitation profiles. The similarity between our RR results for HRP-l at
pH 7 and HRP-II at pH 10.8 prompted us to examine the spectrum of both the
- second intermediate at high pH and the first intermediate at neutral pH as

a function of excitation frequency. Figure 5-6 shows the spectrum of HRP-ll

Figure 5—6 .

121

High frequency RR scattering from HRP-II, pH 10.8.
Accumulation time for each spectrum was 10 min using
1 mJ/pulse. Excitation wavelengths were as indicated.
The 1049 cm‘1 peak is from 0.2 M NO3" used as an
internal standard to measure relative intensities. The
984 cm‘1 feature is from 804’.

122

mnm...

 

 

HRP II pH I08

0) 435 nm

 

 

 

 

b) 425.5 nm

>._._mzw._.ZH Z<§<m

   

    

d) 406.7nm

' I700

I500

I300

IIOO

 

e) 390 nm

900

RAMAN SHIFT (cm-W

123

at pH 10.8 at five excitation frequencies between 435 and 390 nm. Because
the differences in the high frequency spectra of HRP-I and HRP-II both at pH
7 (Fig. 5-3b and c, respectively) is limited to the relative intensities of the
features at 1563 and 1584 cm’l, the behavior of these bands is of principal
interest. With 435 nm excitation (Fig. 5-6e) high frequency bands appear at
1509 (v3), 1584 (v2), 1603 (v37) and 1635 cm‘1 (v10). Figure 5-6b shows
that the v10 intensity falls off quickly as A ex moves to the blue and that with
425.5 nm excitation the v(C=C) mode at 1630 cm‘1 begins to dominate this
region. Also, at this )‘ex9 a shoulder is apparent at 1563 cm'l. In Figure 5-6c-e
this feature becomes more obvious as RR excitation is varied toward higher
energy, while the intensity of v2 (1584 cm‘l) decreases. Finally, with ,390
nm excitation (Fig. 5-6a), the feature at 1563 cm‘1 dominates this region of
the spectrum and the 1584 cm'1 mode now appears as a shoulder. Figure 5-7
shows the RR scattering from our flowing HRP-I sample at pH 7 at three
excitation frequencies in the Soret region. Although the spectrum excited at
420 nm (Figure 5-7a) is very weak owing to the short accumulation time, the
dominant features can be recognized at 1378, 1562, 1581, and 1632 cm"1 and
can be assigned to v4, v 11 + v 33, v 2 and v (C=C), respectively. The intensity
of 92 (1581 cm‘l) is greater relative to the intensity of the band at 1562 cm‘1
at this excitation wavelength. The spectrum in Figure 5—7b was obtained under
CW excitation at 406.7 nm. Vibrational bands are apparent at 137 8 cm'1 (V 4),
1399 cm‘1 (v29), 1428 cm'1 (<5S (=CH2)), 1477 cm’1 (v23), 1508 cm'1 (v3),
1562 cm'1 (v11 + v33), 1583 cm'1 (V 2), 1603 cm‘1 (v 37), 1618, 1631 cm"1
(V(C=C)), and 1641 cm‘1 (V10). The intensities of the features at 1583 and
1562 cm‘1 are roughly equal in Figure 5-7b. Finally, Figure 5-7c juxtaposes
the scattering excited at 390 nm with the other HRP-I spectra. Comparison

of Figure 5-6c-e with Figure 5—7a-c suggests that the RR excitation profiles

124

Figure 5-7- High frequency RR scattering from HRP-I, pH 7.
(a) 1.9 mJ/pulse, 4 min; (b) 15 mW, 0.5 s/cm‘l; sample
flow rate, 0.22 ml/ min; similar results were obtained
flowing at 0.7 ml/min; (c) 1 mJ/pulse, 40 min.

125

pH?

HRP I

 

ohm. ..

 

 

a) 420nm
b) 406.7 nm
(cw)

>._._mzm._.ZH Z<_>_<m

c) 390nm

 

 

RAMAN SHIFT (cm-'I

126

of the vibrational modes of the scattering species in the HRP-II pH 10.8 and
HRP-I pH 7 samples are similar. In both cases the v 2 mode at'b1583 cm‘1
dominates the region from 1550-1600 cm“1 with 420 nm excitation, while a
feature at «.1563 cm'1 dominates under 390 nm excitation. The intensities of
these features are roughly equal with Aex = 406.7 nm. The similarity of these
excitation profiles suggests that the heme complexes of HRP-II (at pH 10.8)
and HRP-I (at pH 7) have similar electronic states.

The large relative intensity of the oxoferryl stretch in the RR spectra
excited at 406.7 nm of HRP-II, oxoferryl myglobin23 and model compounds18
has been noted to suggest the presence of a charge transfer (CT) electronic
transition along the oxoferryl axis.16:19 Figure 5-8 shows RR spectra of HRP-II
(pH 10.8) obtained with seven excitation wavelengths from 432-390 nm. The
V(Fe0) is obvious only in spectra produced with excitation to the blue of 410
nm, significantly higher energy than the Soret absorption maximum at 419 nm.
A crude excitation profile (EP) of the v(FeO) = 787 cm‘1 vibration is shown
in Figure 5-9. Along with this, the excitation profiles (a plot of relative Raman
intensity is. excitation frequency, 1 ex) of the porphyrin core vibrations v 7
= 680 cm"1 and v4 = 1379 cm'1 are also displayed. The latter two agree well
with the excitation profiles for the v 7 and v4 modes of deoxy myoglobin
observed by Bangcharoenpaurpong _e_t_ _a_l_.67 On the other hand, the
Fen-N(histidine) stretching frequency, v (FeNIm), of deoxy myoglobin displays
a different excitation profile than the v(FeO) of HRP-ll. While the EP of the
former vibration of deoxy myoglobin matches the Soret absorption and was
judged inconsistent with CT enhancement,67 the EP of the V(Fe0) vibration
displayed in Figure 5—9 appears to have a maximum around 400 nm, well to
the blue of the Soret absorption maximum. This is consistent with but not proof
of the occurrence of a CT transition on the high energy side of the Soret (0-0)

transition of HRP-II.

Figure 5-8 .

127

Low frequency RR scattering from HRP-II, pH 10.8.
Accumulation time was 10 min for (aHd), 20 min for
(er), using 1 ml pulses. The 983 cm‘1 peak is from
0.2 M 80 ‘ used as an internal intensity standard. The
1049 cm‘ feature is from NO3'.

128

HRP II pH IO.8

 

b) 425.5 nm

IIIIIIIII

  

c) 4|6 nm

>tmszZH 232$;

 

e) 406.7 nm

 

f) 404 nm

soo‘aoo

 

 

l200

I000

400

RAMAN SHIFT (cm")

129

 

 

  
 

 

 

e E
C
E 3 E {3-
n o
s! g g g HRP-I pI'I I08
I I 5.1".
V4 ‘ I379 CITITI
E
C
>_ o
.___ a
(I) I
Z
LIJ
I—
Z
Z
4
2 VIFeO) - 787 cm"
<1
0:
LIJ
Z
’—
<1
_I
LIJ
0:
117-680 cm”l
23 2'4' ' ' '2'5' ' ' T2'6

EXCITATION FREQUENCY (cm"/|OOO)

Figure 5-9 - Excitation profiles of W; = 680 cm'l, v (FeO) = 787 cm"1
and V4 = 1379 cm‘l.

130

DISCUflON

Prediction of HRP-l vibrations from structural correlations. The similarity
between the high frequency RR spectra of HRP-l and HRP-II in Figure 5-3b
and c is unexpected. The vibrational frequencies between 1450 and 1700 cm"1
are a sensitive probe of the structure of both the neutral metalloporphyrin52
and the metalloporphyrin 1r cation radical.36 Studies of model compounds
suggest that oxidation of the macrocycle does not significantly influence
porphyrin core geometry, particularly if the oxidation state and axial ligation
of the metal center remain unchanged.36168 This is in agreement with EXAFS
measurements of HRP-I and HRP-II by Penner-Hahn 91 31.15 Thus, we expect
the porphyrin core sizes of HRP-I and HRP-II to be similar. Despite the structural
similarities between MP and MP‘“ the electronic states of the cation are clearly
distinct from the netural form and our recent studies of
metallooctaethylporphyrin 1r cation radicals (MOEP+' where M = Zn, Cu, Co,
and Ni) demonstrate that vibrational frequencies and relative intensities in
the RR spectra of the MOEP“ species are substantially changed from those
of the parent MOEP compounds.36 Thus we expect significant differences in
the RR spectra of HRP-I y_s_. HRP-II, analogous to those we observed in the
MOEP“ 1g. MOEP spectra. Based on our recently presented36 structural
correlations and assuming a core size of 1.99 A, we expect v3, 911 and v 2
to occur at approximately 1501, 1590 and 1606 cm'l, respectively, for HRP-I.
Here we have assumed that the frequencies of the v(Cbe) modes, v11 and
92, are 10 cm“1 less for PP“ compared to OEP‘" compounds as they are in
the ring-neutral complexes.69 Thus, the V3 value of 1508 cm"1 for HRP-I
(Fig. 5-3b), unchanged from that of HRP-ll (Fig. 5-3c), and the absence of features

in the 1590 to 1610 cm'1 range, suggest that the species producing the spectrum

131

in Figure 5-3b is not a porphyrin 1r cation radical. Since the evidence for a
porphyrin radical in HRP-l, particularly from ENDOR results,42 is strong, we
consider here possible explanations for the RR result.

HRP—II. The trivial explanation is that our HRP-I spectrum is produced
by bona-fide HRP-II contaminants in the sample, produced either by
photo-reduction or reduction by impurities acting as substrates. However,
although similar, the RR spectra of HRP-I and HRP-II at pH 7 are clearly distinct,
most notably in terms of the behavior of the oxoferryl stretching mode (Fig.
5—5). Furthermore, optical absorption spectra following each RR measurement
of HRP-l solutions (particularly for the recycled samples) were like that of
HRP-I. Thus, reduction to HRP-II (which would be irreversible) is not the cause
of the similar RR spectra of the two intermediates (Fig. 5-3b and c).

Photochemistry. Both frozen and solution samples of HRP-I are known
to undergo photoreaction when exposed to light in the Soret band region.27
The photoproduct, called Y, has an absorption spectrum like that of HRP-II
(i.e. characteristic of an oxoferryl, ring-neutral heme), and gives an EPR signal
of a free radical located on a protein residue.70 Although a different
interpretation has been offered,71 photo-induced electron transfer from a nearby
amino acid to the porphyrin cation radical could leave a structure analogous
to cytochrome c peroxidase compound 1.4453 This species would have an
oxoferryl, ring-neutral heme with a protein free radical and would account
to the EPR and absorption spectrum of the photoprod'uct.71 If this were the
case, the RR spectrum of Y should resemble that of HRP-II. Thus, our spectrum
(Fig. 5—3b) might represent scattering from Y despite the measures taken to
prevent the photoreaction. Under our typical experimental conditions (1 mJ/laser
pulse) we estimate a maximum of 35 photons/molecule was absorbed by the

scattering sample. With a quantum yield of 0.003,'71 this would correspond

132

to less than 1196 photoconversion possible per 5 nanosecond laser pulse. Thus,
a maximum of 1196 impurity could be reflected in the RR spectrum only if Y
is formed in less than 5 ns. We estimate that a 17% impurity with an extinction
coefficient at 420 nm of 100 mM‘lcm'1 would be required to produce scattering
”ex = 390 nm) at intensitites comparable to that of the green cation radical,
HRP-I, with an extinction coefficient at 400 nm of 50 mM‘lcm'l. Thus, an
1196 photoconversion could feasibly cause artifacts in the HRP-I RR spectrum.
The close similarity of the RR excitation profiles of HRP-l at pH 7 and HRP-ll
at pH 10.8 are consistent with possible photoconversion as they suggest scattering
species with similar electronic absorption spectra. On the other hand, similar
experiments with CW excitation at 406.7 nm at a wide range of flow rates (from
0.2-20 ml/min) carried out both in our lab (Fig. 5-7b) and elsewhere,34 as well
as pulsed measurements at laser energies ranging from 0.5-2.0 mJ/pulse, all
produced similar results. These facts argue against not only artifacts due to
photoconversion to Y but also possible scattering from excited states of HRP-I.
Spin delocalization. Using an approach similar to ours, but with CW
excitation at 406.7 nm and high flow rates (5-20 ml/min.) Ogura and Kitagawa34
recently obtained a high frequency RR spectrum of HRP-I, although with a
very low signal-to-noise ratio. Their result is similar to our high frequency
spectrum30 (see also Fig. 5-3b) and is also atypical of a MP‘“. As these authors
point out, extensive spin delocalization of the porphyrin radical onto the axial
oxoferryl-imidazole system might account for the absence of vibrations
characteristic of the oxidized porphyrin in the RR spectrum of HRP-I.
Delocalization of the porphyrin cation radical spin onto an axial pyridine ligand
has been described for PyZnTPP‘” (TPP = tetraphenylporphyrin).72 Perhaps
more significant, the spin systems of the oxoferryl (8:1) and porphyrin cation

radical (S=1/2) may interact strongly in compound I structures.73174 While

133

these factors may well influence vibrational frequency shifts between oxidized
and neutral porphyrin complexes, this may not be sufficient to explain the
essentially identical RR spectra of HRP-II (pH 10.8) and HRP-I (pH 7) presented
in Figures 5-6c-e and 5-7, respectively.

ENDOR measurements clearly show that the unparied electron in the radical
site of HRP-I is hyperfine coupled to both 1H and 14N nuclei.42 These could
come from either the methine and b-carbon substituents and the pyrrole nitrogen
of the porphyrin or from the a- and B -protons and imidazole nitrogen of the
axial histidine. Although deuterium substitution ENDOR experiments which
would demonstrate that the radical site is indeed the porphyrin rather than
the histidine are mentioned by Roberts e_t_ £1.42, they are not presented.

HRP-1". Given the possibility of a product of photo-induced electron transfer
within the heme pocket, we must address the relationship of the photoproduct
to the green form of HRP-I and to the catalytic sequence. We will call this
putative photoproduct HRP-1* to acknowledge the possibility that it may be
distinct from the product Y proposed earlier,70,71 and to emphasize that it
is formally equivalent to HRP-I in oxidation state. Figure 5-10 depicts HRP-1*
in rapid equilibrium with HRP-l. In the dark, formation of HRP-1* is negligible,
but under high illumination HRP-1* is favored, being reached presumably through
an excited state of HRP—l. HRP-1‘I lacks both the prophyrin radical and the
H-bonded oxoferryl. Devoid of these two features, HRP-1* is an unreactive
structure and must return to the green form, HRP-l, in order for peroxidase
catalysis to continue. The heme of HRP-1* is thus identical to that of HRP-II
at high pH. This accounts for the similarity of the spectra in Figures 5-6 and
5-7. In this interpretation, HRP-I and HRP-1* have metal centers with identical
valences and axial ligands and thus differ only in the porphyrin oxidation state.

Our RR studies of (CH3OH)2ComOEPClO4‘ and (C104‘)ZComOEP+' suggest

134

I»...

HRP-I

Figure 5-10.

 

 

HRP-II

Schematic diagram of the active site of HRP catalytic
intermediates, showing HRP-I, the postulated photo-
product HRP-I“, the reaction of HRP-l with a phenolic
substrate, and HRP-II.

135

similar six-coordinate cobaltic solution structures with 1.97 A core sizes for
both complexes.36 By analogy to these and other MOEP“, MOEP pairs, we
conclude that HRP-l and HRP-1* possess similar geometries. Thus, should our
HRP-I RR result actually be due to an HRP-1* photoproduct, we prOpose that
the structural information contained in the RR spectrum is still valid to the
green P+' form of HRP-l. Therefore, given either possible explanation (i.e.
photochemistry or spin delocalization) for the similarity of the high frequency
RR scattering of HRP-l and HRP-ll, the results presented in Figures 5-3 and
5—5 suggest that the core sizes of these two species are equal and that, at pH
7, the oxoferryl of HRP-I is not H-bonded as is the oxoferryl of HRP-II. In
the absence of the H-bond at high pH,25,26 HRP-II is unreactive. The association
between this H-bond and the reactivity of HRP-II suggests that the oxoferryl
unit requires the H-bond for activation, presumably in order to facilitate the
formation of leaving H20 as depicted in reaction 3, above. Thus, the absence
of the H-bond in HRP-l helps explain the initial reduction of the porphyrin radical
cation, rather than the ferryl structure, to form the second intermediates.
Kinetics and mechanism of HRP-l reduction. A host of kinetic studies
of the reaction of HRP-l with a variety of substrates suggest the involvement
of a distal amino acid residue exhibiting a pKa of 5.1,5 identified possibly as
histidine 42 (His-42).6b For the substrates p—cresol and L-tyrosine, the reaction
proceeds by a base catalyzed mechanism.6 Figure 5-10 proposes a mechanism
for the reaction of p—cresol with HRP-l at pH 7. The substrate is oriented by
H-bonding to distal His-42. This is consistent with studies which show that
the acid form of p—cresol will bind to the cyanide complex of HRP at a distinct
but nearby binding site at pH 7.75 As an electron is transfered from the substrate
to the porphyrin cation radical, a proton is transferred along the H-bond to

His-42. This homolytic cleavage of the H-A bond (or H-O bond of p-cresol)

136

results in the formation of the free radical of the substrate76 along with HRP-ll.
The oxoferryl of HRP-ll is H-bonded to the protonated His-42 residue.25v26
The reaction of HRP-II with the substrate is influenced by an acid group with
pKa 8.6.5 Thus, the participating acid group is most likely His-42 in both cases

as was speculated earlier.77 The pKa is shifted by the positive charge of the
I porphyrin cation in HRP-l. This mechanism is consistent with Hammett and
Okamoto-Brown plots constructed from rate constants for the reduction of
HRP-l by a series of substituted phenols. These were taken to suggest a
mechanism wherein the neutral substrate donates an electron to HRP-l and

simultaneously loses a proton.78

CONCLUSIONS.

Using RR spectroscopy applied to a flowing sample of HRP-I (pH 7) prepared
by rapid mixing, we have characterized a distinct intermediate formed by the
reaction of HRP with hydrogen peroxide. This compound is two (oxidation
equivalents above the native state and decays through a species with an absorption
spectrum characteristic of HRP-II. The RR scattering in the 1300-1700 cm'1
range obtained from this intermediate is not characteristic of a metalloporphyrin

17 cation radical and is suggestive of an oxoferryl heme with the same geometric
and electronic structure as that of HRP-II (particularly _the high pH form). We
have considered two possible explanations for our result: (i) photo-induced
electron transfer to the porphyrin radical cation of HRP-l occuring in less than
5 ns and resulting in a new species, HRP-1*, which produces the RR spectrum,
or (ii) extensive delocalization of the porphyrin radical of HRP-I onto the axial
ligands.

The low frequency RR scattering in the 600-1000 cm'1 range from this

species reveals a v(FeO) at 791 cm‘l, typical of the non-hydrogen bonded form

137

of HRP-ll. This suggests that the oxo ligand of HRP-I is not hydrogen bonded
at pH 7. This proposal is consistent with the kinetics of the reaction of HRP-I
with p-cresol. Because the oxoferryl structure requires the hydrogen bond for
activation,25v26 this result explains why the porphyrin radical is the site of
reduction of HRP-I. We represent a mechanism for HRP-I reduction by p-cresol
in which a proton from the substrate binds to His 42 and forms a hydrogen bond

to the oxo ligand of HRP-ll.

FUTURE WORK

At the present time it is not possible to address conclusively the question
of photoconversion v_s. radical delocalization to explain the high frequency RR
result for HRP-l. In addition to the work presented here, we have utilized
cryogenic techniques, both in aqueous (unpublished results) and antifreeze
solvents,30 to obtain RR spectra of HRP-I similar to those we obtained with
both pulsed, 390 nm and CW, 406.7 nm excitation. Variations of the rapid mixing
approach have included the following: (i) off-resonance CW excitation at 363.8
nm which failed to produce detectable Raman scattering, and (ii) comparative
CW measurements at 406.7 nm at different flow rates (we used 0.22 and 0.7
ml/min; Ogura and Kitagawa34 report flow rates of 5-20 ml/min), all of which
produced similar results. Felton e_t_ 53.58 also describe results similar to these
obtained under 457.9 nm excitation for HRP-I samples at temperatures of 0
to -30 C. In conclusion we must assume that if these results do not represent
Raman scattering from the green MP‘“ form of HRP-I, such scattering is
prohibitively difficult to obtain under CW or nanosecond pulsed laser excitation.
On the other hand, in collaboration with A. Salehi, RR studies of the model
compound O=FeIVOEP+’79 are in progress and will greatly facilitate further

analysis. In particular, methods to produce 180=FeIVOEP+' are being developed.

138

Time-resolved Optical absorption measurements of flowing HRP-I samples
illuminated by pulsed laser irradiation are planned. These experiments will
be carried out in collaboration with Dr. Dewey Holten and coworkers and will
conclusively address the question of the involvement of HRP-I photochemistry
in these RR results.

Regardless of these complications, the pH dependence of the v(FeO) observed
for our putative HRP-I species should be determined. Low frequency data should
be obtained from samples at pH 4, 5 and 6 in order to estimate the pKa of the
proposed histidine group involved in the reaction. Also, the experiment should
be repeated in D20 buffer at pH 7 to establish more firmly the absence of
hydrogen bonding of the oxo ligand. Finally, RR studies of intermediates involved
in the catalysis of other peroxidases (notably mye10peroxidase, lactoperoxidase
and heme bromoperoxidase) have already begun. This work will be carried out

in the MSU Shared Laser Laboratory with Drs. R. Wever and J. Manthey.

ACKNOWLEDGMENTS
We thank Dr. T. Kitagawa for manuscripts prior to publication and Drs.
R. Wever and J. Manthey for discussion and critical reading. T.O. thanks A.

Salehi for technical assistance. This work was supported by NIH grant GM-25480.

10.

11.

12.

13.

14.

REFERENCES

Chance, B., Arch. Biochem. Biophys. 1952, 4_1, 416-424.
George, P., Nature 1952, 169, 612-613.

Smith, A.M.; Morrison, W.L.; Milham, P.J., Biochemistry 1982, _2_l_,
4414-4419.

Frew, J.E.; Jones, P. ' in "Advances in Inorganic and Bioinorganic
Mechanism"; Academic Press: New York, 1984; Vol. 3, pp. 175—213.

Hewson, W.D.; Hager, L.P. in "The Porphyrins", Dolphin, D., Ed.; Academic
Press: New York, 1979, Vol. 7, pp. 295-332.

(a) Dunford, H.B.; Stillman, J.S., Coord. Chem. Rev. 1976, _1_9, 187-251;
(D) Dunford, H.B., Adv. Inorg. Biochem. 1982, _4, 41-68.

Naqui, A.; Chance, B., Ann. Rev. Biochem. 1986, _5_5_, 137-166.

Theorell, H.; Ehrenberg, A., Arch. Biochem. Biophys. 1952, 4_1, 442-461.
Moss, T.H.; Ehrenberg, A.; Bearden, A.J., Biochemistry 1969, 8, 4159-4162.
Roberts, J.E.; Hoffman, B.M., J. Am. Chem. Soc. 1981, 103, 7654-7656.
LaMar, G.N.; de Ropp, J.S.; Latos-Grazynski, L.; Balch, A.L.; Johnson,
R.B.; Smith, K.M.; Parish, D.W.; Cheng, R.-J., J. Am. Chem. Soc. 1983,
105, 782-787.

Chance, B.; Powers, L.; Ching, Y.; Poulos, T.; Schonbaum, G.R.; Yamazaki,
1.; Paul, K.G., Arch. Biochem. Biophys. 1984, 235, 596-611.

(a) Chance, M.; Powers, L.; Kumar, C.; Chance, B., Biochemistry 1986,
23, 1259—1265; (b) Chance, M.; Powers, L.; Poulos, T.; Chance, B.,
Biochemistry 1986, .2_5_, 1266-1270.

Penner-Hahn, J.E.; McMurry, T.J.; Davis, I.M.; Balch, A.L.; Groves, J.T.;
Dawson, J.H.; Hodgson, K.O., J. Biol. Chem. 1983, 258, 12761-12764.

139

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

140

Penner-Hahn, J.E.; Smith-Eble, K.; McMurry, T.J.; Groves, J.T.; Dawson,
J.H.; Hodgson, K.O., J. Am. Chem. Soc. 1987, in press.

Terner, J.; Sitter, A.J.; Reczek, C.M., Biochim. Biophys. Acta 1985, 828,
73-70.

Hashimoto, S.; Tatsuno, Y.; Kitagawa, T., Proc. Japan Acad. 1984, 6_0,
Ser. B.; 345—348.

Bajdor, K.; Nakamoto, K., J. Am. Chem. Soc. 1984, 1_(_)§, 3045-3046.

Proniewicz, L.M.; Bajdor, K.; Nakamoto, K., J. Phys. Chem. 1986, _9_0_,
1760-1766.

Chappacher, M.; Chottard, G.; Weiss, R., J. Chem. Soc., Chem. Commun.
1988, _2, 93-94.

Kean, R.T.; Oertling, W.A.; Babcock, G.T., J. Am. Chem. Soc. 1987, in
press.

Terner, J.; Reed, D.E., Biochim. Biophys. Acta 1984, 789, 80-86.

Sitter, A.J.; Reczek, C.M.; Terner, J., Biochim. Biophys. Acta 1985, 828,
229-235.

Kean, R.T.; Babcock, G.T., manuscript in preparation.
Sitter, A.J.; Reczek, C.M.; Terner, J., J. Biol. Chem. 1985, 260, 7515-7522.

Hashimoto, S.; Tatsuno, Y.; Kitagawa, T., Proc. Natl. Acad. Sci. USA
1986,83, 2417-2421.

Stillman, J.S.; Stillman, M.J.; Dunford, H.B., Biochemistry 1975, _1__4,
3183-3188.

Teraoka, J.; Ogura, T.; Kitagawa, T., J. Am. Chem. Soc. 1982, 104,
7354-7356.

Van Wart, H.B.; Zimmer, J., J. Am. Chem. Soc. 1985, 107, 3379-3381.
Oertling, W.A.; Babcock, G.T., J. Am. chem. Soc. 1985, 107, 6406-6407.

Babcock, G.T.; Jean, J.M.; Johnston, L.H.; Palmer, G.; Woodruff, W.H.,
J. Am. Chem. Soc. 1984, 106, 8305-8306.

Babcock, G.T.; Jean, J.M.; Johnston, L.H.; Woodruff, W.H.; Palmer, G.,
J. Inorg. Biochem. 1985, _2_3_, 243-251.

Ogura, T.; Yoshikawa, S.; Kitagawa, T., Biochim. Biophys. Acta 1985,
_8_3_2_, 220-223.

Ogura, T.; Kitagawa, T., submitted for publication.

Salehi, A.; Oertling, W.A.; Babcock, G.T.; Chang, C.K., J. Am. Chem.
Soc. 1986, _1_Q_8_, 5630-5631.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.
46.

47.

48.
49.
50.

51.

52.

53.

54.

55.

56.

141

Oertling, W. A.; Salehi, A.; Chung, Y.; Leroi, G. E.; Chang, C. K; Babcock,
G..,T J. Am. Chem. Soc. 1987, submitted for publication.

Oertling, W.A., Salehi, A.; Chang, C.K.; Babcock, G.T., J. Phys. Chem.
1987, in press.

Salehi, A.; Oertling, W.A., unpublished results.

Dolphin, D.; Forman, A.; Borg, D.C.; Fajer, J.; Felton, R.H., Proc. Natl.
Acad. Sci. USA 1971, 6_8, 614-618.

Aasa, R.; Vanngard, T.; Dunford, H.B., Biochim. Biophys. Acta 1975,

391, 259-264.

Schulz, C.E.; Devaney, P.W.; Winkler, H.; Debrunner, P.G.; Doan, N.;
Chiang, R.; Rutter, R.; Hager, L.P., FEBS Lett. 1979, 103, 102-105.

Roberts, J.E.; Hoffman, B.M.; Rutter, R.; Hager, L.P., J. Biol. Chem.
1981, 256, 2118-2121.

La Mar, G.N.; de ROpp, J.S., J. Am. Chem. Soc. 1980, 102, 395-397.

La Mar, G.N.; de Ropp, J.S.; Smith, K.; Langry, K.C., J. Biol. Chem. 1981,

_2_5_8, 237-243.

Asada, K.; Badger, M.R., Plant 6c Cell Physiol. 1984, 25, 1169-1179.
Schonbaum, G.R.; Lo, S., J. Biol. Chem. 1972, 247, 3353-3360.

Worthington Biochemical Corporation, "Worthington Enzyme Manual",
1972, p. 41.

Gibson, Q.H.; Milnes, L., Biochem. J. 1964, _9_1_, 161-171.
Chance, B., Arch. Biochem. 1949, _2_2_, 224-252.
Adler, A.; Kampas, F.; Kim, J., J. Inorg. Nucl. Chem. 1970, 12, 3443-3445.

Spaulding, L.D.; Chang, C.C.; Yu, N.-T.; Felton, R.H., J. Am. Chem. Soc.
1975, _9_7, 2517-2525.

Spiro, T.G. in "Iron Porphyrins", Lever, A.P.B.; Gray, H.B., Eds.;
Addison-Wesley: Reading, MA, 1983; Part 2, pp. 89-159.

Abe, M.; Kitagawa, T.; Kyogoku, Y., J. Chem. Phys. 1978, _6_9_, 4526-4534.

Choi, S.; Spiro, T.G.; Langry, K.C.; Smith, K.M.; Budd, D.L.; La Mar, G.N.,

Desbois, A.; Mazza, G.; Stetzkowski, F.; Lutz, M., Biochim. Biophys. Acta
1984, 785, 161-176.

Spiro, T.G.; Strekas, T.C., J. Am. Chem. Soc. 1974, g, 338-345.

57.

58.

59.
60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.
76.

77.

142

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

Felton, R.H.; Romans, A.Y.; Yu., N.-T.; Schonbaum, C.R., Biochim. Biophys.
Acta 1976, 434, 82-89.

Rakhit, G.; Spiro, T.G., Biochem. Biophys. Res. Comm. 1976, _7_1_, 803-808.
Boersma, A.D.; Goff, H.M., Inorg. Chem. 1984, _2_3_, 1671-1676.

Hanson, L.K.; Chang, C.K.; Davis, M.S.; Fauer, J., J. Am. Chem. Soc.
1981, 103, 663-670.

Kim, D.; Miller, L.A.; Rakhit, G.; Spiro, T.G., J. Phys. Chem. 1986, fl,
3320-3325.

Maggiora, G.M., J. Am. Chem. Soc. 1973, 95, 6555—6559.

Kashigawa, H. in "Biomolecules: Electronic Aspects", Nagata gt _a_l_., Eds.;
Elsevier: New York, 1985; pp. 31-50.

Choi, S.; Spiro, T.G.; Langry, K.C.; Smith, K.M., J. Am. Chem. Soc. 1982,

194, 4337-4344.

La Mar, G.N.; de Ropp, J.S.; Smith, K.; Langry, K.C., J. Am. Chem. Soc.
1983, 105, 4576-4580.

Bangcharoenpaurpong, K.T.; Schomacher, K.T., Champion, P.M., J. Am.
Chem. Soc. 1984, 106, 5688-5698.

(a) Spaulding, L. D., Eller, P. G.; Bertrand, J. A; Felton, R. H, J. Am. Chem.
Soc. 1974, _9_6, 982- 987; (b) Gans, P.; Buisson, G.; Duee, E.; Marchon, J. -.;C
Erler, B..;S Scholz, W.F.; Reed, C.A., J. Am. Chem. Soc. 1986,108,
1223-1234.

Callahan, P.M.; Babcock, G.T., Biochemistry 1981, _2_0_, 952-958.

Chu, M.; Dunford, H.B.; Job, D., Biochem. BiOphys. Res. Commun. 1977,

E, 159-164.

Nadezhdin, A.D.; Dunford, H.B., Photochem. Photobiol., _2_9_, 899-903.

Fujita, 1.; Hanson, L.K.; Walker, F.A.; Fajer, J., J. Am. Chem. Soc. 1983,

LIE, 3296-3300.

Sontum, S.P.; Case, D.A., J. Am. Chem. Soc. 1985, M, 4013-4015.

Schultz, C.E.; Rutter, R.; Sage, J.T.; Debrunner, P.G.; Hager, L.P.,
Biochemistry 1984, 23, 4734-4754.

Critchlow, J.E.; Dunford, H.B., J. Biol. Chem. 1972, 247, 3714-3725.
Shiga, T.; Imaizumi, K., Arch. Biochem. Biophys. 1975, 167, 469-479.

Dunford, H.B.; Cotton, M.L., J. Biol. Chem. 1975, 250, 2920-2932.

143

78. Job, D.; Dunford, H.B., Eur. J. Biochem. 1976, 66, 607-614.

79. Chang, C.K.; Kou, M.-S., J. Am. Chem. Soc. 1979, 101, 3414-3415.