The pulse radiolysis of sodium tetraphenylborate and sodium hexachloroiridate in aqueous solution


Material Information

The pulse radiolysis of sodium tetraphenylborate and sodium hexachloroiridate in aqueous solution
Physical Description:
xi, 169 leaves : ill. ; 29 cm.
Crawford, Charles L., 1963-
Publication Date:


Subjects / Keywords:
Pulse radiolysis   ( lcsh )
Sodium compounds   ( lcsh )
Chemical kinetics   ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1991.
Includes bibliographical references (leaves 155-168)
Statement of Responsibility:
by Charles L. Crawford.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001709823
oclc - 25541226
notis - AJC2108
sobekcm - AA00004749_00001
System ID:

Full Text






Dedicated to

my parents, Nathan and Doris


The author expresses his sincere appreciation to his

research director, Prof. Robert J. Hanrahan, and to Prof. M.

Luis Muga for their advice, encouragement, and friendship

throughout this work.

Special thanks are given to Avinash Gupta and Zuoqian

Li for their friendship and assistance.

Appreciation is given to Ravi Bhave for his efforts in

initiating the experimental setup. Mohammad Gholami and

Sandra Roberts are also acknowledged for their contributions

towards much of the data collection.

Thanks to Prof. D. N. Silverman of the J. Hillis Miller

Health Science Center for the use of the stopped-flow appa-

ratus. Experimental assistance with this instrument, pro-

vided by Dr. Chingkuan Tu, is also gratefully acknowledged.

The author's deepest thanks are expressed to Kama

Siegel for her unprecedented companionship and editorial

prowess -- both of which have helped make this work possi-


Thanks to Dave Burnsed for his assistance with the




ACKNOWLEDGEMENTS .................................. iii

LIST OF TABLES ..................................... vi

LIST OF FIGURES .................................... vii

ABSTRACT .... ..................................... X

1. INTRODUCTION .................................. 1

1.1 Foreword .............................. 1
1.2 Review of Previous Work ................. 3


2.1 Pulse Radiolysis Setup ................... 19
2.2 Absorption Spectrophotometry ............. 22
2.3 Light Source Systems ................... 30
2.4 Reaction Cells and Cellholder ............ 37
2.5 Monochromators .......................... 47
2.6 Febetron 706 System ..................... 50
2.7 Transient Recorder and Computer System ... 59
2.8 Computer Program ....................... 63
2.9 Reagents and Their Purification .......... 65
2.10 Sample Irradiation ....................... 67

IN AQUEOUS SOLUTION .................... 69

3.1 Experimental Results ..................... 69
3.2 Discussion .................... ........ 83
3.3 Conclusions .............................. 91

IN AQUEOUS SOLUTION ..................... 93

4.1 Experimental Results ................... 93
4.2 Discussion ............................ 101
4.3 Conclusions ............................. 114





RADIATION SOURCE ...................




REFERENCES .......................... ..............

BIOGRAPHICAL SKETCH .............................








Table Page

1 Radiation Chemical Yields for Water ........... 9

2 Transient Decay Rate Constants ............... 78

3 Physical Constants for Chemical Dosimetry
Systems .............. .. .................. 120

4 Absorbed Dose for Chemical Dosimetry Systems .. 127

5 Chemical Reaction Systems for Computer
Simulation Studies ....... .................. 129


Figure Eae

1 Scheme for the radiolysis of water ............ 8

2 University of Florida pulse radiolysis system 21

3 Experimental arrangement using the McPherson
218 monochromator ............................. 26

4 Experimental arrangement using the Jarrell-Ash
monochromator ................................. 27

5 Photomultiplier tube-base schematic used with
the EMI 9250 QB PMT in the model 3262 housing 29

6 Lead shields for PMT photocathodes in the
Pacific Precision Instruments PMT housings .... 31

7 Photomultiplier tube-base schematic used with
the Hamamatsu R928 side window PMT and Model
3150 housing ................................. 32

8 Xenon-arc lamp spectrum ...................... 34

9 Xenon-arc lamp schematic diagram .............. 35

10 Flow-reaction cell with modified UV-VIS
cuvette ........................ ................ 39

11 Non-flow-reaction cell from UV-VIS cuvette .... 41

12 Reaction cellholder with adjustable razor
blades mounts and wire-mesh screens ........... 42

13 Depth-dose curves for microscope coverslide
experiment ................................... 45

14 Absorbance change with number of pulses
delivered for microscope coverslide experiment 46

15 Entrance and exit slit system flange adaption
to the Jarrell-Ash monochromator .............. 51


Figure Page

16 Febetron 706 system schematic ................. 53

17 Refurbished electron beam tube with ion pump .. 56

18 Total beam calorimetric dosimetry arrangement 58

19 Timing sequence for pulse radiolysis experiment 62

20 Graphical display of optical signal versus time
following pulse radiolysis of a N20-saturated
solution containing 10 mM NaTPB .............. 71

21 Transient absorption spectrum ................. 72

22 Graphical display of optical signal versus time
following pulse radiolysis of a N20-saturated
NaTPB and NaN3 solution ...................... 74

23 Curve fit of data from Fig. 20 (a) ............ 76

24 Curve fit of data from Fig. 20 (b) ............ 77

25 Carbonate competition experiment for determining
the OH + TPB- rate constant ................. 80

26 Graphical display of optical signal versus time
following pulse radiolysis of a N2-saturated
solution contain NaTPB and t-butanol ......... 82

27 Comparison of experimental and computer
simulated concentration versus time plots ..... 90

28 UV-VIS spectrum ......... ............ .......... 94

29 Carbonate competition experiment for determining
the OH + Ir(III) rate constant .............. 96

30 Transient digitizer traces of Ir(IV) absorption
recorded at 488 nm with N20-saturation ........ 98

31 Transient digitizer traces of Ir(IV) absorption
recorded at 488 nm with N2 and air saturation 100

32 Change in first order decay of hydrated electron
with Ir(III) concentration .................... 102

33 Comparison of experimental and computer
simulated Ir(IV) concentration verses
time plots; N20, alkaline ..................... 109



34 Comparison of experimental and computer
simulated Ir(IV) concentration verses
time plots; nitrogen-saturated ............... 111

35 Comparison of experimental and computer
simulated Ir(IV) concentration verses
time plots; aerated ........................... 112

36 Computer simulations of the transient
Ir(IV) trace ................................. 113

37 Change in absorbance with accumulated
irradiation pulses for the modified
Fricke system ................................... 124

38 Thiocyanate dosimetry ........................ 125

39 Hydrated electron dosimetry .................. 126

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Charles L. Crawford

December, 1991

Chairman: Dr. R. J. Hanrahan
Major Department: Chemistry

A pulse radiolysis facility for the study of fast

chemical kinetics in aqueous solution, based on a Febetron

706 electron beam accelerator, has been established at the

University of Florida Radiation Chemistry Laboratory.

In previous work on the radiolysis of tetraphenylborate

(TPB)'solutions carried out in this laboratory, it was found

that several organic products, including benzene, phenol,

and biphenyl, are produced with substantial yield. However,

the reaction mechanism has not been determined. Since the

tetraphenylborate anion, TPB-, is a reducing species, it

should be more readily subject to attack by OH- than by eaq

or by H.. The lack of reactivity between TPB and eaq- has

been confirmed by directly monitoring the transient signal

due to eaq-. Concerning the reaction with OH., two schemes

were investigated: (1) a rapid electron transfer from

B(C6H5)4~ to OH.; (2) OH. addition to B(C6H5)4~. Comparison

of transient absorption spectra resulting from the two

different schemes above suggests that the OH. addition is

the dominant reaction under conditions of N20 saturation,

with an experimentally determined second-order rate constant

of 6.2 x 109 M-is-1. A mechanism based on an initial first-

order self-decomposition of the OH* adduct, (C6H5)3BC6H50H,

is proposed.

Reactions initiated by OH. radicals or eaq- in aqueous

IrCl63- solutions were studied. The rate constant for the

respective reactions were found to be 4.9 x 109 M-1s-1 and

6.1 x 109 M-1s-1. The oxidation product, IrC162- disappears

rapidly in N20-saturated basic solution or in either neutral

N2-saturated or aerated solution, but is nearly inert in

neutral solution with N20 present. The complex IrCl62-

reacts rapidly with hydrogen peroxide in basic media, as

confirmed on the benchtop and by stopped-flow kinetics. It

is therefore inferred that reaction with HO2- may account

for the loss of IrCl62- under basic conditions. Loss of

Ir(IV) in neutral N2-saturated solution without added N20

may involve electron transfer from Ir(II), and loss of

Ir(IV) in aerated solution is attributed to reduction by

superoxide ion, 02-.

Kinetic modeling on the respective mechanistic schemes

of the above systems gives good agreement with our experi-

mental results.


1.1 Foreword

The pulse radiolysis technique offers a convenient way

of generating a variety of unstable or transient species

under well-defined conditions. This technique involves the

initiation of a chemical reaction process via a short pulse

of ionizing radiation. Of major interest is the kinetics

of, as well as both the qualitative and quantitative meas-

urement of these transient species involved in the fast

elementary steps of the reaction process.

The modification of a gas-phase pulse radiolysis system

to a liquid-phase pulse radiolysis system was undertaken to

expand the chemical kinetics facility of the University of

Florida Radiation Chemistry Laboratory. Efforts to obtain

kinetic information related to various aqueous solution

systems were made in order to supplement existing aqueous-

solution-based studies carried out in this laboratory in-

volving techniques such as steady-state Co-60 --radiolysis

and solar-assisted photoelectrochemistry.

To study fast chemical reaction systems, two necessary

requirements are the initiation of the process on a time-

scale that is shorter than or comparable to the process

being observed, and the detection of a physicochemical

change associated with the process. Our liquid-phase pulse

radiolysis system was developed around a Febetron 706 elec-

tron beam (e-beam) accelerator capable of delivering a

short, 3 nanosecond full width at half-max (FWHM) pulse of

600 KeV electrons with a peak current of 8 x 103 amps. The

monitoring technique employed was time-resolved absorption


A microcomputer-based data acquisition system was

integrated into the pulse radiolysis system to acquire,

analyze, and store data obtained from the study of fast

chemical processes.

The kinetic studies involving aqueous solutions of

sodium tetraphenylborate (NaTPB) were undertaken to aid in

developing a plausible mechanism concerning the radiolytic

degradation of the TPB- ion. Examination of the iridium

chloride system was carried out to expand the solution

conditions under which this system had been previously

studied by pulse radiolysis. Due to the similarities of

flash photolysis and pulse radiolysis, specifically in the

case of radiation-induced redox reactions, previous photoly-

sis work on the iridium chlorides served as a reference to

the pulse radiolysis data collected in this study.

Various chemical dosimeter methods applicable to pulsed

electron sources were studied to establish the proper dose

delivered to the liquid sample in the Febetron 706 system,

as modified for liquid-phase work.

1.2 Review of Previous Work

1.2.1 Pulse Radiolysis

At the University of Florida Radiation Chemistry Labo-

ratory, Febetron 706 studies have been completed on the gas

phase reaction kinetics and mechanisms of various chemical

systems such as the alkyl and perfluoroalkyl iodides,1

oxygen/ozone,2 and OH reaction dynamics with small mole-

cules.3-5 The identity and reactivity of important tran-

sient species generated in the pulse radiolysis of gaseous

systems, at atmospheric pressures or less, was of previous

interest in these studies.

Historically, pulse radiolysis was developed to eluci-

date postulated transient behavior in radiation-induced

chemical reaction systems, and apparatus was first brought

into operation in about 1960.6,7 Early pulse radiolysis

systems used visible absorption spectroscopy to monitor

intermediates produced from intense single electron acceler-

ator pulses lasting typically around one microsecond or

less, and much of the developmental work on pulse radiolysis

involved liquid phase systems.8-10 Development of the pulse

radiolysis technique has since brought about variations of

the method to include stroboscopic techniques capable of

picosecond resolution11 and multiple time-resolved observa-

tion techniques such as electron spin resonance (ESR),12-14

conductivity,15-17 emission spectroscopy,18 and resonance

Raman spectroscopy,19-21 to monitor reaction intermediates.

The pulse radiolysis method is the radiation chemistry

analog of flash photolysis, in that the photoflash and

photodissociation of molecules is replaced by a short, high-

energy pulse of electrons, causing excitation and ionization

of molecules by electron impact.22 The technique of pulse

radiolysis, used to study fast reactions, has been applied

in areas of aqueous inorganic, organic, and biochemistry.

Other fast reaction techniques such as flash photolysis,

stopped-flow techniques, and temperature jump or pressure

jump relaxation methods have been designed to monitor con-

centrations and to measure rate coefficients.23 However,

the pulse radiolysis technique was the first method capable

of studying processes occurring at or below 10-9 seconds.24

The characteristics of the electron pulse used in

electron beam pulse radiolysis are its time profile, peak

current and energy, cross-sectional homogeneity, and diver-

gence. For adequate time resolution of the observed tran-

sient signal, the time profile of the pulse should be very

short compared to the relaxation of the intermediates. The

pulse current determines the number of interactions between

the chemical system and the e-beam which, in turn, deter-

mines the concentration or amount of species formed. The

maximum electron energy determines the penetration depth of

the e-beam into the reaction cell and should be high enough

to adequately irradiate the volume of the chemical system

being monitored. The cross-sectional homogeneity and diver-

gence of the beam determine how uniformly irradiated the

optically-sampled reaction region is.

Transformations over the years of the standard pulse

radiolysis installation have led to increasingly sophisti-

cated experimental setups.25-28 Spectrographic methods

involving oscilloscopes and Polaroid cameras have been

largely replaced by digital signal processors, or transient

digitizers that have an interface for computer control.11

Signal processing has been aided by the use of backing-off

circuits which can measure small transients on top of the

analyzing light.29 Lasers and the capability of commercial

xenon-arc lamps to supply pulse current on top of DC current

have provided intense analyzing light sources.30-32 Lastly,

on-line computer systems have been developed which not only

greatly reduce the time required for data analysis but also

improve both quantity and quality of the data


1.2.2 Radiolysis of Water and Aqueous Solutions

Early investigations on the radiolysis of water have

played a significant role in the development of radiation

chemistry.36-39 In addition, sophisticated pulse radiolysis

techniques achieving the time resolution of radiation in-

duced chemical events have been applied more often in the

radiolysis of water than in any other liquid.40 As a

result, most of the aspects of the radiation chemistry of

water and aqueous solutions are now reasonably well under-


A few practical purposes for water irradiation studies

arise from 1) the importance of understanding the effect of

radiation on biological systems in which water is the main

substrate, and 2) various aspects of nuclear reactor tech-

nology concerning water's use as both a coolant and modera-

tor.41,42 Water and aqueous solutions have been studied, in

particular, for three reasons: (1) their role as solvents is

central to chemistry in general and to radiochemistry in

particular; (2) they are readily available and easy to work

with; (3) water is a polar liquid that responds in charac-

teristic ways to radiation.

Important features of the radiation chemistry of water

are: (1) a description of the primary radical and molecular

products and how they evolve in both time and space; (2) the

effect of pH on the radiation chemical yields of primary

products; (3) the track structure and related phenomena

associated with the passage of a charged particle through a

liquid; and (4) a detailed account of all the radical reac-

tions involving the primary radicals and radiolysis


Upon initial ionization of a water molecule by a fast

primary charged particle such as an electron, the liberated

electron from the ionization event will typically have

sufficient energy to further ionize several adjacent water

molecules, which leads to the creation of clusters of ions

along the track of the ionizing particle. These clusters

are called spurs; detailed descriptions of their structure

and distribution along the particle tracks can be found in

the literature.44 Once the dry electrons have been slowed

to thermal energy, they undergo a hydration (or solvation)

process which involves the water molecules becoming oriented

about the charged species. This process occurs on the order

of 10-11 seconds, which is the relaxation time for dipoles

in water. Alternatively, some water molecules are excited

to upper electronic states from which they can autoionize,

dissociate, or fall back to the ground electronic state.

Figure 1 from Buxton43 shows the time scale of events initi-

ated by the absorption of energy by water from an incident

ionizing radiation, such as an energetic photon or charged

particle. Also shown are the various routes to formation of

the primary radical and molecular products.

At approximately 10-10 seconds after the ionizing

event, all the physicochemical processes are complete and

the chemical stage begins. Initially, this process involves

diffusion of the radiolysis products out of the spurs where

they either react together to form molecular and secondary

radical products, or they escape to the bulk solution,

becoming homogeneously distributed. Appendix B lists the

elementary reaction steps in the radiolysis of liquid water,

along with their appropriate rate constants.


- \




OH + H3 +



* molecular products form in the spurs
* diffusion of radicals out of the spurs

eaq ,H, OH,H21 H 202, H3O

Note: A//-



, upon interaction with ionizing radiation

,the electron in a presolvated, or "dry" state

, hydrated electron

Fig. 1. Scheme for the radiolysis of water.

STime scale (sec)







At approximately 10-7 seconds, the spur expansion is

complete, resulting in the so-called "primary yield" of the

radiolysis products eaq-, H30+, OH, H, H2, H202, and HO2.

Radiation chemical yields are expressed as G-values, as

shown in Table 1,38 where G is the number of species (repre-

sented by X) created or destroyed per 100 eV of absorbed

dose, and are written as G(X).

Table 1

Radical and Molecular Product Yields in Irradiated Water

Primary yield eaq- H+ OH H H2 H202 HO2

G( molecules ) 2.63 2.63 2.72 .55 .45 .68 0.026
100 eV

The yields are essentially independent of pH except for

extreme acidic conditions (pH < 3). The G-values are also

dependent on the quality of the ionizing radiation, or the

mean linear energy transfer (LET). Under extremely intense

irradiations, at dose rates so high that the spurs overlap,

yields may also depend on the dose rate.40 In the case of

high-LET radiation45 such as a particles, heavy ions or

fission fragments, or very high dose rates ( > 109

rad/sec),46-49 lower radical and larger molecular yields


In the radiolysis of dilute solutions, most of the

energy is absorbed by the solvent, and effects due to energy

deposited directly in the solute are unimportant for solute

concentrations less than about 0.01 M. Thus all the ob-

served chemical changes in irradiated dilute aqueous solu-

tions are brought about indirectly by the molecular and

radical products of the water radiolysis; this phenomenon is

referred to as the "indirect effect."40 In the presence of

a solute, transient species and final products result from

attack of H, eaq-, OH, and H202 on the added substrate.

Radicals derived from the substrate may also react to form

final products. As evinced by the G-values of the primary

radicals, the radiolysis of water produces approximately

equal numbers of powerful oxidizing and reducing radicals.

The standard reduction potentials for the hydroxyl radical,

OH, and the hydrated electron, eaq-, are 2.72 V and -2.9 V,


For pulse radiolysis chemical applications, well-de-

fined conditions, i.e. either totally oxidizing or totally

reducing, are desirable, and these conditions can be

achieved by interconversion of the primary radicals.50

Briefly, for a convenient way of obtaining almost totally

oxidizing conditions, N20-saturated (N20 = 2.5 x 10-2 M)

solutions are used. N20 converts eaq- to OH, or

eaq + N20 ----> N2 + 0 ----> OH + OH" 129

Acidic conditions of pH s 3 should be avoided due to the

interconversion of eaq- to the hydrogen atom, which is a

major reducing species in acidic media,

eaq- + H+ ----> H 103

For the creation of almost totally reducing conditions,

addition of an organic solute to an inert gas-saturated

solution leaves eaq- as the predominant reducing primary

species. Reactions involved are

OH + ROH ----> R(CH.)OH + H20 I-1

H + ROH ----> R(CH.)OH + H2 I-2

In general, the organic radicals formed, R(CH-)OH are less

powerful reductants than eaq- and are rather unreactive. In

particular, N2-saturated solutions with added tertiary

butanol were used in our work.

1.2.3 Aqueous Solution Studies of Aromatic Systems

In aqueous systems containing organic solutes, reac-

tions occurring between solute and the primary products of

water radiolysis are predominately abstraction, addition,

and one-electron reduction. Organic radicals produced by

these processes can further react before being converted to

stable products, thus leading to complex radiolysis mecha-

nisms, which have been reviewed.51-53

Mechanisms can be somewhat generalized within families

of organic compounds containing the same functional group

due to the tendency of the primary products, eaq-, H and OH,

to react with these functional groups as opposed to the

molecule as a whole.40 Particularly, benzene can be uti-

lized as an analogous system for many aromatic compounds,

with addition to the unsaturated bonds being the character-

istic reaction path. Reaction of the primary radicals from

water radiolysis with aromatic systems exhibit rate con-

stants in the range of 105 to 107 M-s -1 for eaq-, and > 109

M-1s-1 for H and OH.40

C6H6 + eaq ----> C6H6- (+ H20) ---> C6H7. + OH- I-3

C6H6 + H ----> C6H7. I-4

C6H6 + OH ----> C6H60H. I-5

Products of these addition reactions are cyclohexadiene

systems containing an unpaired electron delocalized about

the benzene ring. The dienyl radicals, namely cyclohexadi-

ene, C6H7., and hydroxycyclohexadiene, C6H6OH., have been

investigated by electron spin resonance techniques that

indicate a positive electron density assignment to the

ortho- and para-carbon atoms.54

Extensive pulse radiolysis work has shown these dienyls

exhibit a reasonably strong absorption (e = 5 x 103 Mlom-I)

in the 310-315 nm range.55,56 Rate constants involving the

primary radicals with benzene as well as the dimerization

and disproportionation of the resulting cyclohexa- and

hydroxycyclohexadienyls have also been reported.55-61

Steady-state y-radiolysis studies involving liquid

chromatography,62 radio-liquid chromatography,63 and gas

chromatography,64,65 as analyzing methods, indicate a com-

plex mixture of products formed from the combination and

disproportionation of the cyclohexa- and hydroxycyclohexa-

dienyl radicals. Phenol and biphenyl are the major products

reported, in addition to mixed dienyls and regenerated

benzene. All of the previous studies were carried out in

N20-saturated or deoxygenated aqueous benzene solutions. In

aerated aqueous benzene solutions, eaq- and H are scavenged

by 02. The hydroxycyclohexadiene radical reacts with 02

forming phenol and phenol-like products.66'67

1.2.4 Sodium Tetraphenylborate System

Pulse radiolysis studies of the aqueous NaTPB solutions

have been carried out in this laboratory to help establish a

mechanism consistent with the end product yields obtained

from Co-60 7-radiolysis on aqueous NaTPB solutions.68 These

steady-state radiolysis studies were done in conjunction

with the United States Department of Energy operation at the

Savannah River Plant in South Carolina and concerned the

problem of cesium-137 separation and removal from nuclear

waste obtained from spent fuel reprocessing.69 Precipita-

tion of the radiocesium as its insoluble tetraphenylborate

salt is a method of separating it from the nonradioactive

components and concentrating it to reduce the volume of the

liquid radioactive waste. Other decontamination methods

include ion exchange and evaporation techniques.70'71 The

radioactive precipitate will be stored in large underground

steel tanks until it is sent to the Defense Waste Processing

Facility (DWPF) for immobilization in glass.72'73

The TPB precipitate will undergo intense B- and 7-

radiation from the cesium-137 during processing and storage

that will partially decompose the tetraphenylborate, produc-

ing several organic compounds. Understanding the reaction

scheme which leads to these various organic products as well

as their impact on subsequent processing of the precipitate

are important aspects concerning the design and operation of

the DWPF.68

Since the tetraphenylborate anion, TPB-, is rather

different structurally from benzene or simple aromatics, it

is not obvious that previous work on the mechanism of ben-

zene radiolysis is relevant. However, several lines of

evidence will be presented in this dissertation which indi-

cate that there is a substantial connection. The prepara-

tion and properties of aqueous metal tetraphenylborate salts

have been described by Flaschka and Barnard.74 LiTPB and

NaTPB are found to be the only TPB salts that are apprecia-

bly soluble in water; all TPB salts tend to be soluble in

polar organic solvents.

Only a few mechanistic studies involving primary radi-

olysis radical reactions with the TPB- anion have been found

in the literature. Horii and Taniguchi, using pulse radiol-

ysis, studied the oxidation intermediates of the TPB- anion

in aqueous solutions using the oxidizing radicals N3', Br27,

and SCN2 .75 Kinetic as well as spectral information was

reported for the resulting intermediate, B(C6H5)4', which

was produced in all cases by an electron transfer from the

TPB- anion to the oxidizing radical. Liu et al. reported an

absorbing species at A = 300 nm for the expected radical

derived by OH" addition to the phenyl group of the TPB~

anion.76 Another transient, derived from H' addition to the

phenyl group was also discussed; however, no kinetic infor-

mation was supplied regarding the fate of these resulting

transient species.

Other pulse radiolysis investigations involving the

NaTPB salt include ion association studies, in which the

spectra of the diamagnetic species, Na-, the solvated elec-

tron, eaq-, and the ion pair species (Na+,eaq-) are reported

for alkali metal solutions in various non-aqueous


The photochemistry literature of the metal tetraphenyl-

borates was surveyed to find possible mechanisms concerning

the reactions of the TPB- anion and subsequent intermediates

under UV-irradiation. Photolysis of aqueous solutions of

NaTPB using UV light with a wavelength of 2537 A generates

biphenyl as the major product in oxygenated solutions and 1-

phenyl-l,4-cyclohexadiene as the major product in the ab-

sence of 02.83,84 Other photochemical studies involve the

photolysis of tetraphenylborates in non-aqueous organic


Additional photochemical studies involving TPB- report

on photochemically-induced electron transfer from the TPB

anion to oxygen,88 and various intermolecular charge-trans-

fer transitions of donor-acceptor ion pairs involving the

TPB- anion.89-91

1.2.5 Aqueous Solution Studies of Transition-metal Complexes

In aqueous systems containing transition metal complex-

es, the predominant reaction occurring between the solute

and the primary products of water radiolysis is a one-elec-

tron transfer, either oxidation or reduction. This electron

transfer can occur by an inner- or outer-sphere

mechanism,92,93 often leaving the complex ion in an unstable

oxidation state. Aqueous solutions of both transition

metals and their complexes, most notably bio-inorganic

molecules, have been irradiated and the radiation chemical

processes which occur have been extensively reviewed.94'95

In aqueous systems, equation of the metal complex from

ligand exchange with water molecules can occur, and has been

reported in both pulse radiolysis studies96-99 and photo-

chemical studies.100,101 Changes in the stereochemistry

and coordination number of the transition metal complex upon

pulse irradiation have also been investigated.102,103

1.2.6 Hexachloroiridate System

The redox chemistry of the IrCl62- and IrCl63- complex-

es has been of recent interest due to the prospect of using

the system to produce molecular hydrogen by the solar photo-

chemical reduction of protons in acidic aqueous solu-

tions.104 The solar chemistry of metal complexes has been

reported by Gray and Maverick.105 Solar chemistry is de-

fined as that area of photochemistry in which the excitation

energies fall within the spectrum of solar irradiation at

the earth's surface. The proposed scheme involving aqueous

HC1 solutions of IrC162-/3- is

IrC163- + H ----> IrC162- + 1/2H2 I-6

IrC162- + Cl- ----> IrC163- + 1/2C12 I-7

The net result is the conversion of HC1 to hydrogen and

chlorine using UV and visible light as the only energy

sources. While Gray reports that the second reaction occurs

with low efficiency at wavelengths higher than 430 nm, the

first unfortunately occurs only with UV light, e.g. at 254


In this study we attempt to further elucidate the

radiation chemistry of the chloroiridates under various

aqueous solution conditions, namely in alkaline, N20-satu-

rated solutions and in neutral, aerated solutions.

Dynamic processes occurring in the IrCl62- / IrCl63-

system have been previously investigated by photochemical

techniques.100,101,106,107 Dainton and Rumfeldt made perti-

nent observations concerning the chemistry of the system,

although their experimental procedure involved Co-60 y-

radiolysis rather than direct studies of the transients.108

Mills and Henglein used steady-state gamma radiolysis

of aqueous IrCl63- to explore the formation of colloidal

iridium and the catalytic properties of the colloid with

respect to H2 formation.109'110 Pulse radiolysis of NaIrClg

aqueous solutions, carried out by Mills and Henglein, exam-

ined the reduction of IrCl63- by eaq- and the (CH3)2COH'

radical. A fundamental pulse radiolysis study on the aque-

ous solution redox reactions of hexachloroiridate complexes

by Broszkiewicz addressed both the reduction and oxidation

of aqueous IrC163".111 From his work, the author concluded

that octahedral IrCl63- is oxidized by a single electron

transfer to octahedral IrCl62- without changing its struc-

ture or ligand composition, i.e. equation. Edwards also

commented that the iridium chlorides are typical examples of

complex ions that are relatively inert toward substitution,

but which undergo rapid electron transfer reactions.112

Further pulse radiolysis studies on transition metal

hexachloride complexes have been reported by

Broszkiewicz.113-115 Other pulse radiolysis studies involv-

ing aqueous IrCl63- have examined its oxidation by the

azidyl radical, N3*,116 and the hydroxyl radical, OH*.93


2.1 Pulse Radiolysis Instrumentation

Reactive intermediates generated in radiation-induced

chemical processes are short-lived; thus the pulse radioly-

sis instrumentation must be able to monitor the transient

behavior with a high degree of resolution over extremely

short intervals. Since the concentration of such intermedi-

ates can influence both the mechanistic pathway and the

final product yields, detection techniques that are sensi-

tive to very low concentrations of radicals are necessary

for a complete characterization of the transient behavior.

The University of Florida pulse radiolysis system

configuration incorporating computer control is shown in

Figure 2. The accelerator used is a Febetron (brand name

for the impulse generator produced by Field Emission Corp.)

706, which operates on the principle of the Marx-bank cir-

cuit. Other accelerators used in time-resolved studies

include the linear accelerator linacc), and the Van de

Graaff generator. One may see that the other components

shown comprise an absorption spectrophotometer. Absorption

spectroscopy is the most widely used technique among others

mentioned previously, for monitoring transient behavior in

pulse radiolysis.

>, < O
10 0
>1 r- >1
a 4 k
,-4 s4i (1)
S4 3 : 0 (U N 4 -N

O O 0 -- 0 *4 0
aO O 6 Oi a) E

O 4 O ,- 0 4 4
00 n E o rA 0 > 0
S0 c N O .M 0 > (0

O 2. -.r-4 r m m
4 toCd -4 0 O
(E C --ri O ) 4 ) C ) U U

-C 0 4 t3) -H r-l 0 4 -
o 0 3-4 3-i : -4 o 3- r 0 0

U* E ** **4 *-4
S> p, 4 cn E & > > > o

E3 3 C C*4*

O 10 O 5M3 O





*-I O >4 4 ** f *
U0 H : p

0 0 )0f 04 0 P

04 rd o 0 (0 a)
z a) 04 ,- .44J4.J >4
C 0 0 E 0 (
0 0 P 0 m -40 U) 0 r Cd
003 4- 04 (4 1 4 0 5-
-P r-A 0) Z 04 010 P- O't
C Eo Ei r c o_ e 0 P- 4-) 1D
S 0) z 0 C d (a (o -C 0. 0)
0 'O h P4 Wd 0 r-4 r-4 0 4 1- r4
U) X a04

LL ----


I z


- To



I. ~it



J~~Q-, 0




i Li






I cl i
i i


2.2 Absorption Spectrophotometry

Components involved in an absorption spectrophotometer

are the light source, an optical train including all lenses,

mirrors and absorption cell, the monochromator, the detector

and a recorder. In flash photolysis and pulse radiolysis,

the absorption cell also serves as the reaction region.

2.2.1 Light Source

The light source used should provide high monitoring

light output to optimize the signal-to-noise ratio, S/N,

which is directly related to the intensity of light striking

the photocathode of a photomultiplier by the equation:

S/N = (Is/ A f)1/2 II-1

where Is is the light intensity at the photocathode and

delta f is the bandwidth of the recorder system.117,118

This relationship is applicable when signal shot noise is

the limiting S/N determining factor as is the case in fast

( > 1 kHz) kinetic spectrophotometry. Lasers provide high

monitoring light intensities but are limited as to their

wavelength of operation. Thus in order to utilize an ab-

sorption technique that is applicable to all transients or

combination of transients throughout the ultraviolet and

visible spectrum, a high power spectral lamp with a continu-

ous distribution is preferred. Xenon-arc lamps filled with

xenon gas at above atmospheric pressures provide the follow-

ing features: high arc stability and luminous flux, high

reliability and long operating lifetimes up to 1000 hours.

In addition, high-current pulses can be delivered to the arc

lamp, which enable very high monitoring lamp outputs during

transient measurement times of microseconds to milliseconds.

2.2.2 Absorption Path and Monochromators

The flow reaction cell previously mentioned allows for

the production of transient species in the irradiated volume

which can be directly monitored by absorption spectrophotom-

etry. For example, using the Beer-Lambert relationship,

A = e bc II-2

where e = molar extinction coefficient (M-1cm-1)

b = pathlength (cm)

c = concentration (M)

and assuming 1) a transient molar absorption coefficient, e,

of approximately 1000 mol-ldm3cm-1, 2) a transient popula-

tion which produces a 10% deflection in the initial light

intensity, and 3) a 1 cm absorption pathlength, one calcu-

lates the initial transient concentration formed in the

irradiated volume immediately following the pulse to be

approximately 4.5 x 10-5 M.

Monochromators function to isolate a small wavelength

band from a polychromatic source and consist of a dispersive

element, either a prism or grating, and an image transfer

system of slits and mirrors. By using a monochromator with

an adequate light flux throughput and a varying wavelength

range, transient signals can be recorded as a function of

wavelength to build a time-resolved transient spectrum.

The throughput factors which determine the radiant

power passing out of the exit slit of a monochromator are

the source spectral radiance, Bx in units of Wmm-2 sr-1 nm1,

and the monochromator throughput factor, Y in units of mm2

sr nm. The product of these two factors gives the output

spectral radiant power, Q, in watts.118 For operating a

monochromator with a broadband or continuum source the

monochromator throughput factor, Y, is defined as

Y =w2 h TopRd II-3

where w = slit width (mm)

h = slit height (mm)

F/n = f-number or effective aperture

Top = optical transmission factor (unitless)

Rd = reciprocal linear dispersion (nm/mm)

Since the light image is usually focused down on the en-

trance slits by a condensing lens, the slit width is the

only experimental variable parameter in determining the

light throughput at a particular wavelength for a monochro-

mator with a fixed optical transmission factor, effective

aperture and reciprocal linear dispersion.117

2.2.3 Photomultipliers

The advantages of using photomultiplier detectors to

display transient absorption signals on transient recorders

are high sensitivity due to secondary emission multiplica-

tion, fast response times relative to the half-life of the

relaxation processes being monitored, and the ability to

handle high light flux levels while maintaining a linear

signal at high output photocurrents.119 Two types of photo-

multipliers were used in this work which have different

spectral response ranges. An EMI Model 9250 QB end-on 50 mm

window PMT fitted in a Pacific Precision Instruments Model

3262 housing was used initially with the geometry shown in

Figure 3. Because of its extended spectral response range

above 600 nm, a Hamamatsu Model R928 side-on 30 mm window

PMT fit into a Pacific Precision Instruments Model 3150

housing was eventually incorporated into the geometry shown

in Figure 4.

The photomultipliers were powered by a John Fluke, Inc.

Model 412A High Voltage DC power supply, which proved to be

a well regulated and stable dc source. The overall gain on

the PMT could be altered by changing the applied voltage,

typically from 1000 Vdc to 1250 Vdc in increments of 10 or

100 volts, while maintaining maximum gain on the operational


The EMI end-on tube has a bialkali photocathode, a

Spectrosil quartz window, and an optical range of 165-650 nm

Cell holder


IP 706





L Xenon arc lamp
LE Condensing lense
IP Ion pump
PB Lead cave / shield
PH PMT housing



PMT- ::


PM Parabolic mirror
PMT Photomultiplier tube
RB Adjustable razor blades
SC Solution cell

Fig. 3. Experimental arrangement using the McPherson 218





Febetron 706

/ ;





L Xenon arc lamp or laser
LE Condensing lense
IP Ion pump
PB Lead cave / shield

Fig. 4. Experimental arrangement

-- Jarrell-Ash
4 'Monochromator

- LE


i__ P


PH PMT housing
PM Parabolic mirror
PMT Photomultiplier
RB Adjustable razor blades
SC Solution cell

using the Jarrell-Ash

,.----'iT_ F

with a peak wavelength, or maximum radiant sensitivity

(mA/W), at 420 nm. Figure 5 shows a diagram of the base

components which were customized to achieve optimum rise

times, gains and linearities for use in the pulse radiolysis

system. The EMI Model GB 1001A fast electronic switching

circuit connected through an optoisolator and switching

transistor gates off the first dynode voltage of the PMT to

eliminate overloading problems caused by short burst expo-

sure to high light intensities. The gating circuit offers a

precise way to regulate the PMT from a + 5 volt gate signal

and has the specifications of an on-to-off fall-time of 0.5

microseconds and an off-to-on rise-time of 1.5 microseconds.

This circuit which is normally applied in flashlamp or laser

systems to protect from overloading conditions was original-

ly employed to eliminate the overloading of the PMT by

fluorescence of OH radicals which occurred in previous gas

phase studies on the argon sensitized pulse radiolysis of

water vapor.3 In the present liquid phase work, such severe

fluorescence problems derived from radiation interactions

with either the aqueous chemical system or the quartz cell

have not been encountered.

Gating of the PMT was applied to reduce the effect of

high levels of electromagnetic interference (EMI) associat-

ed with the high beam current Febetron 706 machine. Ac-

counts of these electromagnetic fields and their effects on

detection electronics are discussed in the literature.26'120



0 I

|,, -

0o 0
o 0

0 0

0 0

I *


In o









S> a

0 0 Sn 0)
o r -

0- 0-- o)

c CO



-. m "3






















By gating off the initial dynode of the PMT during the

radiation pulse, overloading of the photocathode by the X-

ray (X-radiation) and the radio-frequency (RF) noise coinci-

dent with the electron pulse was significantly attenuated.

Further attempts to reduce the EMI effects included the use

of lead shields inserted into the entrance of the PMT hous-

ings, Figure 6, and extensive lead shielding around the

outside of the PMT housing.

The Hamamatsu side-on PMT was obtained to expand the

wavelength range to allow transient absorption measurements

above 600 nm, and was used in all subsequent experiments

after installation. This PMT has a multialkali photocath-

ode, a UV-glass window, and an optical range of 195-900 nm

with a peak wavelength at 400 nm. Figure 7 shows a diagram

of the customized base components used with this PMT along

with a gating circuit that is analogous to the one previous-

ly described.

2.3 Light Source Systems

For a versatile pulse radiolysis system with capabili-

ties of nanosecond sampling times in the UV-VIS region of

the spectrum, stable and high analyzing light intensities

are needed in order to permit the observation of weak or

long-lived absorption signals, to reduce shot noise, and to

increase the signal to noise ratio.119

a) 2 1/16 inch diameter x 1/2 inch thick lead shield
for end-on PMT housing, Model 3262.

b) 1 1/4 inch diameter x 1/2 inch thick lead shield
for side window PMT housing, Model 3150.

Fig. 6. Lead shields for PMT photocathodes in the Pacific
Precision Instruments PMT housings.

0* : ---
S00 C-

0o ___

7 oo I
S-T -i i



0 0





j!o al.
.- r

d 0- E

II c

-^ --.^---- _() ^
0 >0 a

0 0

O a
I--- --------C C ) 7CO
0 0a

-o o
00 ;

n -

D o


= 0








o .

r- I

2.3.1 Xenon-Arc Lamp

The lamp system used in this work consisted of a xenon-

arc high pressure discharge lamp contained in a water-cooled

housing and powered by a water-cooled ignition device / DC

power supply. Several different bulbs including an Advance

Radiation Corporation XSA 350 and 500 watt and a Ushio UXL

302-0 300 watt were used throughout the work. The ARC bulbs

were of the deep UV design, in which a metal vapor, such as

mercury, is present in addition to the xenon gas providing a

strong distribution in the deep UV region (190 to 250 nm).

The Ushio bulb was not seeded with such a metal and its

continuous spectrum is shown in Figure 8. This spectrum was

generated by feeding the PMT output across a load resistor

and into a Linear Instruments Corporation Model 591 chart


The bulb was contained in a PRA-Canada ALH 220 housing

with dual water cooling loops and powered by a PRA-Canada

Model M301 ignition-power supply. The power supply uses a

continuous feedback, constant current mode of operation to

give excellent current stability. Current programming with

square, triangular, or sine wave modulation can be used in

this unit to vary the optical output of the lamp. As shown

in Figure 9, a 2-3 L/minute cooling line was in series with

the power supply and top face plate portion of the housing

which was oriented vertically. A separate 0.25 L/minute





Febetron 706



----------- -- ---------------
i- i !

to drain
i ---- ___- -

Y- (Lam& water in
Lamp& 2-3 I/mi
Housing 2-3
-- 1--

to drain

M 301 Power

DC volts DC amps

LE Condensing lense

EC Electrical connectors
I(o) Initial light intensity
I(t) Transmitted light intensity

M Mirror
SC Solution cell
SH -Shutter

Fig. 9. Xenon-arc lamp schematic diagram.





water in
0.25 1/min


cooling line was used for cooling of the lamp electrodes.

The approximate dimensions of the different optical arrange-

ments are also shown for the lamp housing which contains an

f/2.4 exit lens.

For conditions of 1) low absorbance monitoring on the

order of a 1% change in light intensity or less, or 2) lower

relative light output (i.e. the lower and upper regions of

the lamp output spectrum shown in Figure 8) a short dura-

tion, very high monitoring light output is desired. This

was accomplished by using a conventional signal generator

feeding a BNC connector in the rear panel of the M301 power

supply. The timing and magnitude of the positive going

square pulse was determined by the specifications given by

the manufacturer.121 The pulse width is required to be < 1

second to eliminate damage to the bulb. The pulse magnitude

can be varied between 0 and 1 volt according to the follow-

ing equation

out = idle + [Vmod/4) x 50] II-4
where Iout = M301 output current

Iidle = idle current, typically 3-5 amps
Vmod = output voltage of the signal generator

The upper limit of Vmod, 1 volt, is determined so that Iout

does not exceed the listed nominal operating current of the

lamp which for the Ushio UXL 300 W bulb is approximately 15

amps. To attenuate and fine control the 5-volt TTL type

signal from the pulse generator, a potentiometer voltage

divider circuit was placed in series with the pulse genera-

tor and M301 power supply. Preliminary monitoring of the

pulse parameters was done through a Tektronix Model 35A

oscilloscope prior to actually pulsing the lamp.

2.3.2 He-Ne Laser

A Spectra Physics Model 155 helium-neon laser was used

for alignment of the optical systems, calibration of the

monochromator gratings, and as a light source for absorption

spectrophotometry involving the kinetics of the hydrated

electron, eaq-. Characteristics of the monochromatic 632.8

laser light such as high collimation, low divergence, and

high intensity make this source ideally suited for eaq

studies.122 Figure 4 shows the optical arrangement used for

the He-Ne laser, the Jarrell-Ash monochromator and side-on

PMT. Attempts to monitor the 632.8 nm light with the end-on

EMI PMT were unsuccessful due to this tube's lack of sensi-

tivity above 600 nm.

2.4 Reaction Cells and Cellholder

Two types of reaction cells, both consisting of modi-

fied quartz UV-VIS spectrophotometric cuvettes, were con-

structed for this study. Previous accounts of similar

reaction cells for use in low energy electron beam radioly-

sis studies involving liquids indicated the importance of

using very thin electron entrance windows that would mini-

mize the attenuation of the low energy electron beam in the

liquid.123 A cuvette fitted with a reservoir and flow exit

for flushing purposes was used in all the on-line absorption

detection work and a non-flow cell was used exclusively for

the Fricke dosimetry experiments.40 Both cells were fitted

into an aluminum cell holder to provide reproducible posi-

tioning and to assure the delivery of equal doses.

2.4.1 Flow Reaction Cell

The flow cell, Figure 10, was constructed from a 60 ml

separatory funnel, a 9 2 stopcock, a $ 5/20 glass joint, a

glass/quartz connector seal and a modified quartz UV-VIS

cuvette. A thin mylar sheet = 200 microns thick was held by

epoxy glue to the quartz cell and served as the electron

entrance window. In the initial construction of the cell,

FISHER brand microscopic optical glass cover slides with a

thickness of 0.13 to 0.17 mm were used.124 However, the

mylar sheet material was chosen as the electron entrance

window due to its lesser thickness and the ease by which it

could be mounted and trimmed on the cuvette. The optical

pathlength was 1 cm, and with an electron penetration depth

of ca. 1.5 mm for 600 KeV electrons in water,122 the irradi-

ated volume was 1 x 1 x 0.15 cm. This electron penetration

depth, which is less than the extrapolated range of 1.9 mm,

is taken as the depth over which a reasonably uniform dose

N20 gas

Stopcock, E 2

Separatory funnel,
60 ml

Glass joint,

Modified quartz
cuvette, mylar window

Fig. 10. Flow-reaction cell with modified UV-VIS cuvette.

deposition can be assumed.125 For studies using purging

gases, the gas was continuously bubbled into the upper

reservoir for the duration of the experiment.

2.4.2 Non-Flow Reaction Cell

This cell, Figure 11, was constructed from a UV-VIS

cuvette to provide detection of 304 nm light associated with

the maximum absorption of the ferric ion, Fe3+, the oxida-

tion product derived from the radiolysis of an acid solution

of ferrous ion, Fe2+, in the Fricke dosimetry system as

discussed in Appendix B.40 The electron entrance window was

identical to that of the flow cell and a total volume of 1.4

ml was delivered by syringe to the cell prior to radiolysis.

This cell could be readily removed from the aluminum cell

holder in the pulse radiolysis system and transferred to a

Beckman DU model 2400 spectrophotometer with Gilford solid-

state electronics, allowing absorption measurements to be

made as a function of the number of radiation pulses re-

ceived. A dilution factor involving the actual irradiated

volume of 0.15 ml and the total volume of 1.4 ml was used in

all dose calculations.

2.4.3 Cell Holder

An aluminum cell holder, Figure 12, was used to enable

reproducible placement of the solution cell with respect to

both the electron beam and analyzing light. The holder


1 cm

1 cm

mylar sheet e-beam
entrance window

solution injection
by syringe


total volume =

1.40 +/- 0.02 ml

Fig. 11.

Non-flow-reaction cell formed from UV-VIS cuvette.
Reaction volume consists of 1 x 1 x 0.15 cm region
immediately behind mylar entrance window.

6.0 "

4 thru holes

8.0 "

wire mesh screens,
60 80 %
beam reduction



adjustable razor blades,
entrance and exit


Reaction cellholder with adjustable razor blade
mounts and wire-mesh screens. Electron beam
enters through rear flange and traverses 8.0
inches in air. Radiation-induced reaction region
occurs in cell (not shown) located at intersection
of e-beam and analyzing light.

Fig. 12.

consists of a 1 x 1 inch open rectangular aluminum column

eight inches in length fitted to a six-inch diameter alumi-

num flange that is attached by four threaded screws to the

outer edge of the short-pulse adapter chamber of the Febe-

tron 706. Adjustable razor blades mounted at the entrance

and exit apertures at right angles to the electron beam axis

defined the width of the analyzing light beam. The use of a

very narrow analyzing light-beam allowed for the study of

the solution kinetics by monitoring only a narrow cross-

section for which the dose could be reasonably defined.

Using this arrangement, the electron beam traversed ca.

8 inches of air between the field emission tube exit window

and the thin mylar entrance window of the reaction cell.

This distance creates a slight attenuation of the electron

beam intensity which is primarily associated with a spread-

ing of the effective beam diameter away from the tube face.

For a distance of eight inches the beam will have an effec-

tive diameter of three centimeters as compared to a diameter

of 1.5 centimeters as it exits the tube face. The loss of

beam intensity due to absorption in air should be negligible

due to the maximum penetration range of 80 inches for 600

KeV electrons in air.125,126

Experiments with FISHER brand optical glass microscopic

coverslides placed in the cell holder at the solution cell

position were performed to determine the depth distribution

of the electron beam. Typical depth-dose curves are shown

in Figure 13 which includes data for two experimental condi-

tions. Initially a stack of approximately 20 slides was

exposed to a single Febetron shot with maximum charging of

30KeV. Absorption measurements were then made on each

individual slide to detect the amount of discoloration in

the glass due to the e-beam penetration. Secondly, a simi-

lar stack of coverslides was exposed to 10 cumulative shots

at full charge, followed by individual absorption measure-

ments. Both data sets indicate an extrapolated range for

the 600 keV electrons in the glass to be approximately 7

slides. Using a slide thickness of 0.15 mm, the extrapolat-

ed depth in glass is found to be 1.0 mm. Assuming that

glass is approximately twice as dense as water, these data

indicate a maximum range for 600 Key electrons in water to

be approximately 2.0 mm, which agrees with the earlier

reported value of 1.9 mm.125

Figure 14 shows the absorption due to the discoloration

in a single slide as a function of the number of pulses

delivered at maximum charging. The accumulated discolora-

tion resulting from the e-beam passage through the glass

slide was found to be non-linear, with the absorption fall-

ing off as a function of pulse-number.

Wire mesh screens capable of 60% to 80% electron beam

reductions could be positioned in the cell holder two inches

from the field emission tube window (Figure 12) to aid in

Ten pulses delivered

One pulse delivered

i I L ------ .....
2 4 6 8
Coverslide #

Depth-dose curves for glass coverslide experiment.

-- -.






5 -

Fig. 13.








Absorbance change with number of pulses delivered
for glass coverslide experiment.

5 10 15 20
Pulse #

Fig. 14.

acquiring transient formation signals occurring during the

first few microseconds immediately following the radiolysis

pulse. These screens restrict the flux of the electron

beam which was maintained at a maximum charging voltage of

30 KeV. With the wire mesh screens inserted, attenuation of

the beam intensity decreases the total average population of

initial radicals formed. This beam flux attenuation helps

to prevent radical-radical interactions and provide favora-

ble radical-solute interaction conditions. However, due to

both a malfunction in the fastest data acquisition mode of

the transient digitizer, and electromagnetic interference

associated with the Febetron 706 system that were discussed

earlier, attempts to capture transient formation signals

with enough resolution for kinetic analysis were not suc-


2.5 Monochromators

In the early stages of this work a 0.3 meter McPherson

218 monochromator was used. Later, efforts were made to

reduce electronic interference associated with the electron

beam pulse by removing the monochromator and PMT couple away

from and to the rear of the field emission tube. Due to the

rather permanent fixture of the Mcpherson 218 and its asso-

ciated vacuum pump assembly relative to the initial experi-

mental geometry, a second 0.25 meter Jarrell-Ash 82-41550

monochromator was employed.

2.5.1 0.3 Meter Monochromator

Initial studies were carried out using a McPherson 218

vacuum/atmospheric monochromator which has an f/5.3, a 2400

G/mm plane grating and a 1.33 nm/mm linear dispersion. The

vacuum capabilities were not employed for any of this work

and the 2400 G/mm grating having a 0 5,000 angstrom wave-

length range was eventually replaced with a 1200 G/mm grat-

ing capable of operating in the wavelength range from 0 -

10,000 angstroms. A synchronous motor-driven gear box

supplies 12 scanning speeds from 0.5 angstroms/minute to 200

angstroms/minute and slit widths were adjustable from 10

microns to 2 millimeters. Slit widths of 1.5 millimeters

and a fixed slit height of 8 millimeters were used. Figure

3 shows the optical configuration used with the McPherson

monochromator. Light exiting the xenon arc lamp was passed

through a condensing lens onto a first surface mirror and

reflected through the solution cell. The beam was then

focused on the entrance slit of the monochromator and after

wavelength dispersion the exiting beam was reflected onto

the photocathode of the PMT by a 2.5 cm focal length para-

bolic mirror.

2.5.2 0.25 Meter Monochromator

Initial testing of the Jarrell-Ash monochromator cou-

pled to the end-on PMT with both components located approxi-

mately 2 meters away from the field emission tube showed a

marked improvement in reducing the electromagnetic noise

interference. Using the McPherson monochromator in the

previously shown geometry in conjunction with the gating

circuit produced a net dead-time, caused by gross overload-

ing of the PMT, of > 5 microseconds in the detection system.

When the detection system was relocated, as shown in Figure

4, net dead-times of 2 to 3 microseconds could be attained

without the use of the gating circuit. All subsequent work

was completed with this arrangement and eventually the end-

on PMT was replaced by the side-on PMT. To attain further

reduction of the EMI effects, the entire detection system

and signal lines would have to be either completely removed

from the room containing the Febetron 706 or extensively


In this modified configuration the analyzing light,

after passing through the solution, was reflected by a first

surface mirror back to a plano-convex f/2.5 condensing lens

and focused onto the entrance slit of the Jarrell-Ash mono-

chromator. This monochromator has an f/3.5, a 1180 G/mm

grating and a 3.3 nm/mm linear dispersion. A wavelength

motor drive of 100 nm/minute was coupled to a wavelength

drive knob to enable scanning of the light source spectrum.

The slit arrangement provided with this monochromator

was altered to accommodate the direct coupling of the Jar-

rell-Ash monochromator to both a previous gas phase aluminum

reaction cell, and the Pacific Precision Instruments PMT

housing.128 In this work only, the direct couple of the

monochromator and PMT housing was utilized. This was accom-

plished by screwing the PMT housing containing the side-on

Hamamatsu tube directly to the exit slit flange adapter.

Figure 15 shows the entrance and exit system flange adapt-

ers. Brass slit bodies of fixed width and height, 0.71 mm

and 15 mm respectively, were screwed into the aluminum

flange by turning the two small holes in the slit body with

a sharp tool.

2.6 Febetron 706 System

The principle of pulse radiolysis involves the delivery

of ionizing radiation to a chemical system in the form of a

short pulse that creates a non-equilibrium system with an

adequate amount of transient species that are monitored in

time. An accelerator is used as the source of the high

kinetic energy ions and several related parameters are

important in pulse radiolysis measurements. Primarily, the

radiolytic pulse should be very short relative to relaxation

of the transient system, and the pulse current should be

large enough to produce an observable concentration of

transients. Energy deposition, which depends on 1) the

particle's mass, charge, and velocity, and 2) the medium

density, is critical in determining the distribution (either

homogeneous or non-homogeneous) of radical concentrations


slit area = 0.71 mm x 15 mm

/ / /

front view


side view

_---_ "4 bolt holes

slit holder





4 thru holes

Fig. 15.

Entrance and exit slit system flange adaption to
the Jarrell-Ash monochromator.


The Febetron 706 electron accelerator used in this work

is fitted with a Model 5515 high vacuum field emission

electron tube and has the following specifications: 12

joules maximum pulse energy, 8,000 amperes of peak current,

3 nanosecond pulse duration (FWHM), 5% peak to peak repeat-

ability, and a maximum electron energy of 600 KeV. The

extremely high currents make this model particularly ap-

plicable to gas phase studies.

However, it was suspected that the high vacuum electron

tube had become gaseous due to intermittent use in previous

years and the extended lifetime of the electron tube beyond

its suggested 3500 pulses. The e-beam tube had failed after

ca. 8000 pulses; the beam had burned microscopic holes in

the window which allowed the pressure in the tube to rise to

atmospheric. It was expected that partial operation could

be restored by replacing the window and reevacuating the

tube. This attenuated beam feature would be beneficial to

condensed phase studies by eliminating radical-radical

interactions caused by the absorbance of enormous radiation

doses. Encouraged by other somewhat analogous and confirm-

ing reports,122,129-132 we pursued the application of the

Febetron 706 accelerator to aqueous solution work.

Details of the Febetron 706 system can be found else-

where.125,126,133 Figure 16 shows a block diagram of the

Model 706/2677 system. The Febetron 706 system was main-

tained by keeping the various chambers with their associated

on c ; E
"-Ca 00

SO' -

c a
(10)0 -


0) -

2 E.

I- L



O" .


a o

0 C

>* C








a) 0
--- -C

I 0--
L on




pressure regulators and gauges properly pressurized. A

compressed air tank is used to pressurize the module chamber

at a maximum of 80 psi and the short pulse adapter at a

maximum of 240 psi. Freon-12 was maintained at 13 psi in

the pulser case. The transformer oil in the short pulse

adapter was circulated daily by an electric pump connected

to a large oil reservoir.

The accelerator operates on the principle of the Marx-

bank circuit which consists of a number of connected high

voltage capacitors that are charged in parallel and subse-

quently rapidly switched to a series arrangement through

spark gaps. This circuit, located in the Marx-surge stage

of the pulser, produces a high voltage, extremely high peak

power pulse that is used to pulse charge through a series

spark gap in the central electrode of the short pulse adapt-

er stage of the pulser. This short pulse adapter then

discharges an extremely short 3 ns high current pulse to the

field emission electron tube. This pulse generates electron

emission from the cathode and the electrons are then accel-

erated toward an exit window that functions as the anode.

As mentioned above, due to the lack of an inward de-

flection in the thin exit window of the field emission tube,

it was decided that gaseous conditions existed inside the

tube. Under normal conditions, the thin metal window would

be noticeably drawn in by the high vacuum. Experiments

using the technique of calorimetric dosimetry carried out

directly in front of the tube exit window showed negligible

energy outputs. Therefore a modified electron beam tube

exit system was fabricated.

2.6.1 Refurbished Beam Tube

A modified electron beam tube flange and exit window

system132 was constructed to renew the output of the Febe-

tron 706 system. This modification, Figure 17, was formed

from a stainless steel chamber 4.75 inches in diameter and

1.25 inches thick, and fit with a Model 202-100 Perkin-Elmer

2 liter/second differential ion pump powered by a Model 222-

0200 Perkin-Elmer IONPAK 200 power supply. The high vacuum

current limiting power supply operated at a voltage between

+4000 volts DC and +6000 volts DC and the system was capable

of attaining pressures below 1 x 10-6 Torr.

The former anodic window of the field emission tube was

cut away and a thin, approximately 0.001 inch stainless

steel window and o-ring were set into the compartment. The

compartment was attached by three threaded screws to the

field emission tube window face plate flange of the Febetron

706 with a second o-ring seal. Initially, the thin window

was placed 0.8 inches away from the original anode window

location. Although an adequate vacuum was attained, further

dosimetry experiments showed no improvement with this geome-

try so the window was set further into the compartment 0.4

i! C

0n --

S------- > 0L


>0 0
",' 6o
--L uj w

=- o
o N


--_--- ( .













jCi &


inches away from the original window location. After rough

pumping the chamber down to a few microns of pressure and

attaining the minimum +4000 volts DC required for proper

operation of the ion pump, calorimetric dosimetry measure-

ments were again performed.

2.6.2 Calorimetric Dosimetry

The total-beam calorimetric technique, as illustrated

in Figure 18, involves the measurement of the heat deposi-

tion in an aluminum disk upon firing the Febetron 706 at

maximum charge. The thickness and diameter of the disk,

0.033 inches and 1.385 inches respectively, were chosen to

assure complete e-beam deposition. Aluminum is suitable due

to its high thermal conductivity, specific heat, and low

backscatter energy coefficient. The temperature of the

calorimeter was conveniently measured by attaching a thermo-

couple and measuring the voltage output with a recording

millivoltmeter. The thermocouple was attached to the alumi-

num disc in this particular calorimeter by drilling a small

hole in the center, inserting the thermocouple wires, and

swaging the aluminum around them with a punch. The calibra-

tion of the thermocouple/chart recorder was accomplished

using room temperature of = 25 C as the base line and

normal body temperature of = 37 C as the maximum scale

value. For the maximum temperature reading, the disk was

held firmly in the palm of the hand until a steady equili-

brated signal was obtained.

Febetron 706

ion pump



chart recording

37.0 C

T cooling
S25.0 C

S aluminum disc

(4 ft.)

Fig. 18. Total beam calorimetric dosimetry arrangement.

Using the equation

Q = m Cp A T II-5

with Q = heat required to raise the specific heat of the

calorimeter by A J, Joules

m = 2.3 +/- 0.05 grams (mass of the aluminum disk)

Cp = 0.22 cal/ C gram (specific heat of aluminum)

A T = 0.53 C (difference between pre- and post- pulse

read from chart recorder)

a value of 1.12 +/- 0.06 Joules was determined as the beam

energy absorbed per pulse. These positive results indicated

that the placement of the new anode window was a critical

factor with possible arcing complications arising when the

new window was not located close enough in to the original

window position.

2.7 Transient Recorder and Computer System

While early pulse radiolysis studies used the oscillo-

scope-Polaroid photograph technique for recording transient

signals, development of sophisticated transient recorders

followed that now have replaced this manual data analysis

method.117 Digital recorders now digitize the input signal

in real time and can resolve the signal into separate chan-

nels thus providing a stored digitized record of the tran-

sient trace as a function of time. In addition, interfacing

of the recorder with a computer allows for integrated con-

trol of the operational parameters such as time base, sensi-

tivity, delays, signal processing, and data transfer.

The data acquisition and analysis system developed for

this work is a second generation system having been preceed-

ed by a Biomation Model 610B transient recorder supported by

an IMSAI 8080 S-100 bus microcomputer.128 Current instru-

ments include an Analogic Data Precision DATA 6000 waveform

analyzer fitted with a Model 620 2-channel plug-in digitizer

interfaced through a serial port at 9600 baud to a Tandy

1200 IBM-PC/compatible laboratory microcomputer.

The DATA 6000 has two signal-input ranges of +/- 1.2

and +/- 3.6 volts full scale, and a maximum sampling rate of

100 MHz. This sampling rate gives a period (time between

sampled points) of 10 nanoseconds for collection of 512

points using an 8-bit recorder. Although the maximum period

attainable is 300 seconds, a maximum period of 1 microsecond

was used in this work giving a sampling rate of 1 MHz. The

DATA 6000 is noted for its versatile capabilities of signal

capture, processing and data transfer, although the lack of

any fine control on the input voltage parameter was found to

be a rather severe experimental limitation in this work.

The DATA 6000 is controlled by a single internal 68000

microprocessor and a CRT display screen enables visual

inspection of signals before data transfer.134

Data recording parameters were selected either manually

through keypads on the front panel or remotely via computer

control. Remote control of the DATA 6000 via an RS-232

serial interface port utilized three basic functions: con-

trol of data recording parameters, use of processing and

display, and transfer of data to and from the digitizer. To

use these functions, a computer program written in BASIC

language provided either keystroke emulation by key code

number or direct control using English commands. These

operations were sent to the DATA 6000 as string commands and

data transfers to the computer were sent in ASCII number

format. Various timing loops set up in the BASIC program

assured proper hand-shaking. During operation several

parameters such as trigger source, timebase, screen display

and marking were initially set manually and subsequently

controlled remotely through the computer program. For

short-lived transient signals, the multiple trace delayed

mode with signal averaging was used which provides an en-

hancement to the S/N ratio in proportion to the square root

of the number, n, of signals averaged.118 Otherwise, the

single trace delay mode was used with no signal averaging.

As shown in Figure 2, the triggering and delays were

originated from a single initial manual trigger pendant

delivering a 90 volt pulse derived from three 30 volt bat-

teries in series with a standard RC trigger circuit. A

typical timing sequence is presented in Figure 19. Delay of

the Febetron firing was accomplished using an Evans Associ-

ates programmable time delay (PTD) module, Model 4145-2.

Three such modules are present which could be used in series




Z U-

c z

+ I



! O











;- I c i


-- -




' i
' i

for a maximum delay of 3 milliseconds. The voltage divider

and amplifier in the circuit are required because the incom-

ing trigger for the Febetron must be 90 volts in magnitude

while the PTD uses a TTL type input and output. The lamp

pulsing and shutter control was provided by a Vincent Asso-

ciates shutter driver and timing unit, Model SD 10, which

controlled a Model 225L Vincent Associates UNIBLITZ shutter.

The shutter was not incorporated into the timing sequence

but was manually switched on and off during a calibration

procedure to set the zero and full light levels, Vo and Vm

respectively. The lamp pulse duration was controlled by the

shutter driver and timing unit which provides a positive

going 5 volt square pulse with a variable 8 to 8,000 milli-

second duration.

2.8 Computer Program

Full interaction between the operator and the

computer/transient recorder was provided by a single BASIC

program. This program, developed by J. Brogdon and R. J.

Hanrahan with assistance from B. Yarborough and Brent Jones,

was locally written in accordance with suggested BASIC rou-

tines provided by the DATA 6000 manufacturer.135 A complete

listing is provided in Appendix C. The BASIC source code

was compiled from the RUN menu of Microsoft Quick BASIC

version 4.0, and was executed directly from memory or from a

separate executable program written to disk.

Initially, the opening menu would offer all options

available, and upon full completion of any commands the main

menu would again be presented. Prior to data collection the

CALIBRATION command was used to set all data recording

parameters and to provide a dark or no light trace, VO, and

a full light trace, VM. The DATA command was entered during

collection, storage, and averaging of actual traces by the

DATA 6000. Upon completion of this loop three files were

stored into computer memory: all digitized voltage values

(QY(I)), a time dependent concentration file [QX(I),

CON(I)], and a time dependent optical density file [QX(I),

OD(I)]. After transfer of the data from the digitizer to

the computer, the SAVE command could be executed to save a

disk copy of the digitized voltage values. Using the PLOT-

TER command provided a graphical presentation of the digi-

tized voltage optical trace on the computer CRT screen.

Often hard copy of the graphical displays would be obtained

to catalog the data.

The subroutine at line 1570 (LEFT & RIGHT BORDERS)

allowed for selection of any portion of the trace followed

by storage of either a concentration or optical density file

corresponding to the section of the trace within the bor-

ders. The subroutine at line 3980 (LINEAR REGRESSIONS)

provided linear least squares analysis of the concentration

data to aid in determining the kinetics involved in the

transient decays. Operation of the MANUAL command enables

adjustment for the true full light signal, VM. Even though

the 100% light signal was set during the calibration steps,

any fluctuation in the light during the actual data collec-

tion mode could be accounted for by taking an average of the

light signal immediately preceding the pulse as an accurate

measure of the true 100% light intensity. Selection of the

molar absorption coefficient, E, which was dependent on the

specific transient being observed was also done through the

MANUAL command. The DIRECTORY command provided directory

inspection to insure the storage of all files to disk and to

access available memory.

2.9 Reagents and Sample Preparation

2.9.1 Sodium Tetraphenylborate and Sodium Azide

Kodak reagent-grade sodium tetraphenylborate and sodium

azide were used as received.

2.9.2 IrC163-(hydrate) and IrC162-

The sodium hexachloroiridate(III) hydrate and sodium

hexachloridate(IV) complexes, which were obtained from Al-

drich Chemical Company, were stored in a dessicator and used

without purification.

2.9.3 Sodium Carbonate

Fisher Scientific Company-supplied Na2CO3 was used in

all the carbonate competition experiments.

2.9.4 Sodium Hydroxide and Ethanol

Fisher Scientific Company-supplied NaOH and USP reagent

quality 200 proof absolute CH3CH2OH from Florida Distillers

Company were used in the hydrated electron dosimetry stud-


2.9.5 Sulfuric Acid and Ferrous Ammonium Sulfate

Fisher Scientific Company H2SO4 and Fe(NH4)2(SO4)2'6H20

were used as received for the Fricke dosimetry studies.

2.9.6 Potassium Thiocyanate

Fisher Scientific Company KSCN was used as received for

the thiocyanate dosimetry studies.

2.9.7 Tert-Butyl Alcohol

Fisher Scientific Company reagent-grade (CH3)3COH was

used as received for the hydrated electron rate-constant


2.9.8 Potassium Permanganate

Fisher Scientific Company reagent-grade KMNO4 and NaOH

were used as received for the water distillation procedure.

KMNO4 solid was dissolved in laboratory-distilled water that

was made alkaline by addition of NaOH. Upon distillation,

any organic impurities were removed by an oxidation reaction

involving the MnO4- ion.

2.9.9 Oxygen and Nitrogen

Liquid Air Corporation-supplied 02 and N2, both of

which were > 98% purity, were used as received for purging

gases in the 'Super' Fricke studies and the hydrated elec-

tron studies, respectively.

2.9.10 Nitrous Oxide

Matheson CP-grade N20 with > 99.0% purity was used as

received for generating oxidizing conditions for all studies

involving the OH' radical reactions.

2.9.11 Sample Preparation

Aqueous solutions were prepared by dissolving known

amounts of solid reagents in doubly distilled water. A

Model B5 Mettler analytical balance with 200 gram capacity

and +/- 0.05 milligram accuracy was used in all mass meas-

urements. Reference spectra of the Iridium(III) chloride

complex, Iridium(IV) chloride complex, and the Sodium Tetra-

phenylborate solution were measured using a Hewlett-Packard

model 2400 UV/VIS spectrophotometer.

2.10 Sample Irradiation

Prior to radiolysis, solutions which were to be irradi-

ated under oxidizing conditions were purged for approximate-

ly thirty minutes with nitrous oxide, N20. For work in-

volving reducing conditions, nitrogen, N2, was used as the

purge gas. Purging of the solutions was done by introducing

a disposable pipette directly into the 60 milliliter separa-

tory funnel serving as the upper reservoir of the solution

flow cell. Gas flow was maintained throughout the pulse

experiments. Prior to calibrating the DATA 6000, the solu-

tion was introduced by gravity flow into the modified cu-

vette, which was then inserted into the aluminum cell hold-

er. The cell holder was designed to accommodate the solu-

tion cell to give reproducible positioning and uniform

irradiation in the region that was sampled by the analyzing

light beam. After completing the calibration procedure, the

data collection routine was executed and the Febetron was

charged and triggered. Typically, due to product buildup or

gas bubble evolution on the inner surface of the electron

beam entrance window, a fresh aliquot was flushed into the

cell and the system was recalibrated after each pulse.


3.1 Experimental Results

3.1.1 Oxidation Intermediates of Tetraphenylborate Ions

The pulse radiolysis of aqueous sodium tetraphenylbo-

rate (TBP) solutions under oxidizing conditions was studied

with two different methods of generating the oxidizing

species. As mentioned previously, using N20-saturated

solutions the predominant oxidant produced is the OH- radi-

cal generated from the direct radiolysis of water and the

interconversion of eaq to OH. by N20. Alternatively, a

secondary oxidant, N3., is prepared under conditions of N20-

saturation and the addition of excess sodium azide, NaN3.

The azidyl radical, N3*, is formed from the oxidation of the

azide anion by the OH. radical,

OH. + N3 --> OH- + N3. III-1

Both the N3" anion and N3. radical have absorption maxima at

wavelengths < 280 nm.136 Experiments with N3. radical were

carried out to utilize the selective electron transfer path-

way of oxidation as compared to other possible reactions of

the OH- radical with aromatic systems, namely abstraction of

an H atom and addition of the OH. radical to the aromatic


The pulse radiolysis of a nitrous oxide-saturated (N20

S2.4 x 10-2 M) solution of 10 mM NaTPB was carried out at

room temperature. Using the Xenon arc lamp in the pulsing

mode, the transient absorption signals were recorded at

various wavelengths from 300 to 800 nm. Figure 20 shows the

absorption signal recorded at 330 nm for a single pulse

experiment. The appearance of an incipient secondary ab-

sorption occurring approximately 20 to 30 microseconds after

the initiating e-beam pulse was characteristic in all the

time-concentration traces recorded in the 300-400 nm range.

A transient absorption spectrum, Figure 21, was gener-

ated by plotting the maximum absorption or "end of pulse"

values as a function of the wavelength at which they were

recorded. This spectrum shows an absorption maximum at 330

nm; the only other transient spectral feature is a small

absorption (ca. 10% of the main peak) centered at about 750

nm. This spectrum is to be compared with that reported by

Liu et al. for the radical derived by OH- addition to the

phenyl groups of TPB- -- a spectrum with a large absorption

at 300 nm that was reported to change only slightly over the

first few microseconds.76

Assuming the absorption at 330 nm is caused by the

radical derived by the addition of OH* to the phenyl groups

of TPB-, its extinction coefficient at the absorption maxi-

mum of 330 nm can be calculated from the transient spectra

shown in Figure 21 (a) by making use of the value for G(OH)

in N20-saturated solutions equal to 5.4 (= G(OH) + G(eaq-).

MULT, X-AXIS BY 8.18E-83 Sec. Y-AXIS BY 8,19E+81 Units. SOURCE:TPB338.1


0.0 0.2 8.4 0.6 0.8 1.8 1.2

MULT. X-AXIS BY 8.18E-83 Sec. Y-AXIS BY 8.1IE+81 Units, SOURCE:TPB345.3


8.8 8.5 1.9 1,5 2.8 2,5 3.8

Fig. 20.

Graphical displays of optical signal versus time
following pulse radiolysis of N20-saturated 10 mM
NaTPB solution. a) recorded at 330 nm, 200 nsec
period, 100 usec full scale; b) recorded at 345
nm, 500 nsec period, 250 usec full scale.


0.8 *






I co


F- F-
o n
E-1-- h

o o

* r


oo uj os d d



I ....A









In N20-saturated TPB- solutions, this total OH* yield of 5.4

will have been completely converted to the TPBOH-" radical

by the time optical measurements can be made. Thus the end

of pulse optical density can be used to measure the extinc-

tion coefficient of the TPBOH-" adduct. The value obtained

from Figure 21 is approximately (10,000 +/- 10%) M-icm-1 for

the maximum peak at 330 nm.

The optical densities were calculated from the initial

pre-pulse light intensity determined during calibration, Vf,

and from the transmitted optical signal, Vt, using the


OD = -log (Vt / Vf) = -log[ (QY(I) Vo) / Vf) ] III-2

Vf is obtained from subtracting the dark signal, Vo, from

the initial 100% light intensity, Vm, determined under the

experimental conditions prior to irradiation and Vt is ob-

tained from subtracting Vo from the digitized optical sig-

nal, QY(I).

The pulse radiolysis of a N20-saturated solution of 10

mM NaTPB with excess NaN3 z 0.2 M, was carried out at

various wavelengths yielding typical absorption traces, as

shown in Figure 22. The accompanying transient absorption

spectrum is plotted in Figure 21. The transient absorption

traces resulting from the NaTPB with added NaN3 conditions

also illustrate a secondary transient production. The

MOLT, X-AXIS BY 8.18E-83 Sec, Y-AXIS BY 8.16E+01 Units, SOURCE:TPBN3.13



8.8 8.2 6.4 8.6 8.8 1.8 1.2

MULT. X-AXIS BY 8.18E-83 Sec, Y-AXIS BY 8.10E+01 Units. SOURCE:TPBN3.12



8.0 0.5 1.8 1.5 2.0 2.5 3.0
? I

Fig. 22.

Graphical display of optical signal versus time
following pulse radiolysis of N20-saturated 10 mM
NaTPB, 0.2 M NaN3 solution. a) recorded at 330
nm, 200 nsec period, 100 usec full scale; b)
recorded at 330 nm, 500 nsec period, 250 usec full











transient absorption spectrum contains two absorption bands,

one at 330 nm and a broad band above 700 nm. This transient

absorption spectrum as well as the associated kinetics of

the 330 nm decay are in good agreement with published re-

sults obtained under similar solution conditions.75'76

3.1.2 Transient Kinetics

Kinetic analysis of the optical traces was carried out

using a curve fitting routine from the BASIC program, and

the best fit was judged from the correlation coefficient and

visual inspection. Figures 23 and 24 show typical curve

fits for both the initial and secondary transients produced

under N20-saturated millimolar NaTPB conditions. Analogous

curve fitting was carried out on the experimental absorption

traces obtained from N20-saturated solutions of NaTPB con-

taining excess NaN3. The resulting decay rate constants are

presented in Table 2.

3.1.3 Carbonate Competition Method

The reaction of OH. radicals with the TPB- anion occurs

too rapidly to be observed directly using the experimental

conditions previously described. However, the formation

rate of the OH. radical reaction with TPB~ can be obtained

by the carbonate competition method137 using a secondary

scavenger, CO32-, which competes with the TPB~ anion for the

OH. radical. The two competing reactions are

MULT. X-AXIS BY 8.18E-83 Sec. Y-AXIS BY 8.19E+81 Units. SOURCE:TPB338.1

\ (a)

? I

0.2 8.4

[ K= 69382.1

BY 0.10E-84


Sec. Y-AXIS BY 8.18E+82 Units. SOURCE:TPB338.1

2.4 1

2.9 2.5 3.0 3.5 4.0
? I

INTCPT: 2.16E+01 (STD DEV= 1.27E-02) SLOPE: 6.94E+04 (STD
R= .9983794 E= 18800 STD DEU PTS= 2.820695E-92

DEU= 4.26E+02)

Fig. 23. Curve fit of data; a) transient trace presented
from Figure 20 (a); b) First-order fitting routine
applied to data between x = 20 and x = 40 micro-





I I I '

Y-AXIS BY 8.18E+81 Units. SOURCE:TPB345.3



E= 3.6322

G IL +




o. (b)

0.5 1.8 1.5 2.0 2.5
INTCPT=-5.18E+03 (STD DEV= 4.49E+82) SLOPE= 3.63E+88 (STD DEV= 2,66E+06)
R= .9912943 E= 3800 STD DEV PTS: 2388.339

Fig. 24.

Curve fit of data; a) transient trace presented
from Figure 20 (b); b) Second-order fitting rou-
tine applied to data between x = 80 and x = 250












MULT, X-AXIS BY 0.18E-03 Sec.

Table 2

Transient Decay Rate Constants

N20-saturated Initial Transient Secondary Transient

Solutions Decay Rate Constant Decay Rate Constant

( x 104 sec-1 ) ( x 108 Msec )

10 mM NaTPB 5.88 0.98 3.76 0.42

10 mM NaTPB 8.25 0.59 3.10 0.91
0.2 M NaN3
a) 2 a) 4

b) 3.9 0.2 b) -

a) reported by Horii and Taniguchi

b) reported by Liu et al.

OH- + TPB ---> products k III1-3

OH- + C032- --> OH + CO3 (detected) k2 130

Using the He-Ne laser experimental setup, the CO3- radical

anion was monitored as a function of added TPB- anion. The

CO3- radical anion has an extinction coefficient on the

order of 1000 M-1 cm-1, a known formation rate constant of

4.1 x 108 M-1 -1, and a broad absorption band centered on

600-610 nm.138 The equation representing the competition is

[C03 ]inf. = x [OH.]o III-4
k2[CO32-] + kl[TPB-]

with [C03-]inf. = end of pulse concentration of C03- ion

[OH.]o = initial concentration of OH- (includes

primary OH' from water radiolysis and

from interconversion of eaq- by N20)

Taking [OH.]o as equivalent to the [C03-]inf. generated with

no TPB- present, or [C03-]o, equation III-4 can be rear-

ranged to

[C03-]o kI[TPB-]
= 1 + III-5
[C03 ]inf. k2[CO32-]

From a plot of [CO03-o / [CO3-]inf. versus [TPB-] / [C032-],

as shown in Figure 25, kl/k2 is found to be 15. Substitu-

tion of k2, the C03- reference formation rate constant gives

a rate constant for the formation (reaction III-3 above) of

the initial transient = (6.2 0.6) x 109 M'-sec-1.







(TPB ) / (CO3 )

Carbonate competition experiment for determining
the OH + TPB- rate constant.











Fig. 25.

i I i i I i

Radical formation rate constants for TPB' equal to 1.4 x 109

M-1s-1 and 7.1 x 108 M-1 s-1 have been reported for the

electron transfer reaction involving TPB- and N3..75,76

3.1.4 Reduction Intermediates of Tetraphenylborate Ions

Pulse radiolysis experiments were performed using

reducing conditions to examine the possibility of a rapid

reaction between the TPB~ ion and eaq. Because the TPB~

anion is a strong reducing agent and negatively charged,

this reaction was presumed unlikely. However, by monitoring

the eaq- transient signal as a function of added TPB~ any

such reaction would be evinced by either a decrease in the

magnitude of the end of pulse eaq- transient absorption or a

change in the eaq- transient decay kinetics. A typical

transient trace of the pulse experiments conducted on ap-

proximately 10 mM NaTPB solutions under reducing conditions

is shown in Figure 26. The life-time of the eaq transient

was found to be approximately 8-10 microseconds with a best

fit analysis giving a second order decay rate constant of

5.7 x 1010 M-s-1.

3.1.5 Computer Simulation

Modeling calculations were utilized to help establish

the detailed chemical mechanism of multi-step reactions

found in aqueous pulse radiolysis systems. Differential

rate equations for the suspected reaction sequence are

-9.5 MULT. X-AXIS BY 8.18E-04 Sec. Y-AXIS BY 9.18E+91 Units. SOURCE:TPBEAQ38

-1.5 .





8.0 8.2 9.4 8.6 0.8 1.8 1.2
? I

MULT. X-AXIS BY 8.19E-85 Sec. Y-AXIS BY 8.19E+06 Units. SOURCE:TPBEAQ38
3.5 (1/C):(1/Co)+K*t
-K 5.736563E+10

2.5 ... --


1.5 .(b)


2.8 3.8 4.8 5.8 6.0 7.8
INTCPT=-8.54E+04 (STD DEV= 2.85E+03) SLOPE= 5.74E+10 (STD DEV: 5,63E+08)
R= .9901459 E- 14860 STD DEU PTS: 9860.368

Fig. 26. a) Graphical display of optical signal versus time
following pulse radiolysis of an N2-saturated 10
mM NaTPB, 0.04 M t-butanol solution; 20 nsec
period, 10 usec full scale; b) Curve fit of data
between x = 3 and x = 7 microseconds.

integrated to give computer-simulated time-dependent concen-

tration profiles that are compared with those measured

experimentally. These comparisons help to establish, or

validate, proposed mechanisms. The computer modeling was

done using a Gear integrator program.139 This program,

written by R. L. Brown140 of NIST, was modified in this

laboratory to run under Microsoft FORTRAN 77 on an MS-DOS-

based microcomputer.

The input parameters to the program consist of a list-

ing of all mechanistic steps and their rate constants (see

Appendix B), the identification of all reactants with their

initial concentrations, and the time-step which controls the

50 computation intervals.

3.2 Discussion

3.2.1 Electron Transfer

The oxidation intermediates of TPB- ions in aqueous

solutions have been examined by Horii and Taniguchi,75 and

Liu et al..76 From their results with N20-saturated solu-

tions of NaTPB and NaN3, the proposed route to the formation

of several major products is a rapid electron transfer from

B(C6H5)4 to N3 yielding the B(C6H5)4 radical and N3", or

B(C6H5)4- + N3' ----> B(C6H5)4' + N3 III1-6

As shown earlier the azidyl radical, N3', is produced ini-

tially from primary OH radical reaction with N3-, reaction

III-1. The transient absorbing at 330 nm and 800 nm (Figure

21) is attributed to the B(C6H5)4' radical. Both peaks

decay with the same first-order rate constant, indicating

that they represent different absorbances of the same tran-

sient species.76 Loss of the 330 nm peak is attributed to a

rapid self-decomposition first-order decay, producing a

second transient that undergoes a relatively slower decay.

Horii and Taniguchi proposed the reaction,75

B(C6H5)4' ----> B(C6H5)2' + (C6H5)2 III-7

The secondarily-produced transient B(C6H5)2* was suggested

to disappear following second-order kinetics, although no

product species were discussed.75

The pulse results from our oxidation experiments on TPB

solutions containing no added N3- anion could be interpreted

as above, i.e. a rapid electron transfer from B(C6H5)4- to

OH', instead of N3*, or

B(C6H5)4- + OH' ----> B(C6H5)4' + OH- III-8

yielding the B(C6H5)4' radical. As shown in Table 2, the

decay kinetics of the transients produced with and without

added N3- are similar, that is a rapid, first-order decay

followed by a secondary transient decay. However, compari-

son of the transient spectra, Figure 21, suggests that two

different transients are formed. With added azide, the

transient spectral feature at 750 800 nm is about 70% of

the magnitude of the main peak centered at 330 nm. With no

added azide, the transient spectral feature at 750 800 nm

is < 10% of the magnitude of the main peak centered at 330


3.2.2 OH Addition

An alternate mechanism to the electron transfer from

TPB~ to OH', as described above, is the addition of the OH*

radical to the aromatic ring, which would lead to an OH*

adduct transient species different from the TPB- radical, or

B(C6H5)4" + OH' ----> (C6H5)3BC6H50H" 301

The following self-decomposition first-order decay

scheme has been considered for the observed loss of this OH'

addition adduct.

(C6H5)3BC6H50H- ----> B(C6H5)3 + C6H50H'- 302

C6H50H'- + H20 ----> C6H60H' + OH- 303

Triphenylborate and its hydrolysis products, (C6H5)B(OH)2

and (C6H5)2BOH, all have absorptions at wavelengths below

300 nm.74 The secondarily-produced transient C6H50H'-, is

suggested to undergo a rapid hydrolysis, yielding the hy-

droxycyclohexadienyl radical C6H60H.. The hydrolysis of

C6H5OH-" has been reported in a previous aqueous phenol

pulse radiolysis study.141 In neutral solutions a slow

reaction between the solvated electron, eaq-, and phenol

that forms C6H5OH-' was reported. The radical anion,

C6H5OH-', subsequently undergoes rapid protonation by H20 to

yield (C6H5)(OH)H'. This species is a structural isomer of

the hydroxycyclohexadienyl radical, C6H6OH*. A molar ex-

tinction coefficient of (3800 +/- 800) M-1 cm-1 at a wave-

length maximum of 330 nm was assigned to the (C6H5)(OH)H'


Another possible competing reaction pathway for the

C6H50H'- radical anion would be a bimolecular radical anion

reaction. The bimolecular disappearance of negatively

charged radical species has been previously suggested in

pulse radiolysis studies involving the reaction of the

hydrated electron, eaq-, with benzene.58 In this aqueous

benzene study, C6H6-, the addition product of eaq- and

benzene, was reported to decay via a rapid bimolecular proc-

ess before protonation in solutions of pH neutral to 13.

C6H6 + eaq ----> C6H6 III-9

C6H6g + C6H6" ----> products III-10

No specific products of the radical anion reaction, III-10,

were discussed.

In the present study, concerning the combination reac-

tion of C6H50H'-, a possible reaction forming a charged

dimer species, with subsequent formation of biphenyl and

hydroxide ion, is proposed below

C6H50H'- + C6H5OH'- ----> [HO(C6H5)-(C6H5)OH]2- 304

[HO(C6H5)-(C6H5)OH]2- ----> C6H5-C6H5 + 2 OH- 305

The scheme involving radical combination leading to an

intermediate dimer is prominent in the mechanism involving

the radiolysis of aqueous benzene, discussed below.

The hydroxycyclohexadienyl radical, C6H60H', mentioned

above (reaction 303), is an important intermediate in the

formation of major products in the radiolysis of aqueous

benzene solutions; the results of which are in good agree-

ment with the end products derived from Co-60 steady-state

r-radiolysis of NaTPB solutions carried out in this lab.86

The proposed mechanism concerning the benzene

oxidation55'64'65 is described as

C6H6 + OH' ----> C6H6OH' III-9

2 C6H60H' ----> [HO(C6H6)-(C6H6)OH] 306

[HO(C6H6)-(C6H6)OH] ----> C6H5-C6H5 + 2H20 307

2 C6H60H' ----> C6H6 + C6H5OH + H20 308

The hydroxycyclohexadienyl radical, C6H6gH*, has a reported

absorption maxima at 313 nm.58 The combined bimolecular

rate constant for its disappearance via reactions 306 and

308 has been determined to be (7.05 0.35) x 108 M-is-1.58

The disproportionation to combination ratio for the disap-

pearance of C6H6OH" was found to be < 0.4, indicating that

the combination reaction 306 is predominant.63 The major

products produced are biphenyl, benzene and phenol. In the

Co-60 r-radiolysis of aerated 0.05 M sodium tetraphenylbo-

rate solutions the products and their G-values were found to

be biphenyl, 1.9; benzene, 1.3; and phenol, 0.41.68

It is also reasonable to expect the H. atom to add to TPB-

ion, giving a transient species that decomposes via a two-

step process giving the cyclohexadienyl radical

B(C6H5)4- + H. ----> (C6H5)3BC6H6* 309

(C6H5)3BC6H6 ----> B(C6H5)3 + C6H67 310

C6H67 + H20 ----> C6H7' + OH- 311

The (C6H5)3BC6H67 transient has been proposed by Liu et al.

as the weakly absorbing species at higher wavelengths

(Figure 21) in the pulse radiolysis of TPB- solutions with-

out added azide.76 The cyclohexadienyl radical could also

react with the hydroxycyclohexadienyl radical in a bimolecu-

lar process producing benzene56

C6H7* + C6H60H' ----> 2 C6H6 + H20 312

Concerning the reduction of TPB-, a rapid reaction

between eaq- and TPB~ appears to be negligible from the

results of the NaTPB pulse radiolysis under reducing condi-

tions. Comparison of Figure 26 with eaq- traces found in

the hydrated electron dosimetry experiments, Appendix A,

shows little difference in either the magnitude or kinetic

behavior of the eaq- transient signal. The eaq- transient

signal decayed with a second order rate constant of 5.7 x

1010 M-1s-1 with added TPB~; for the eaq- dosimetry experi-

ments, the eaq- signal exhibited a second-order decay = 4.4

x 1010 M-s1-1.

3.2.3 Computer Simulations

Computer simulations were carried out with a Gear

integrator utilizing the data set consisting of the elemen-

tary reaction steps describing the transients of irradiated

pure water (Appendix B). The mechanistic scheme also in-

cludes reactions of eaq- with N20, and OH* with TPB-.

Figure 27 illustrates typical results of the simulated

calculations for conditions involving N20-saturated solu-

tions containing 10 mM TPB-. Time-concentration profiles

for all species in the system are predicted over the ca. 200

microsecond data collection time period. Experimental data

points indicate the observed decay of the transient trace

produced in a typical experimental run, recorded at 330 nm

(Figure 20 (b)). These experimental data points represent

the combined absorbance of all the components in the system

contributing to the absorption trace at 330 nm. Dotted

lines show predicted yields of the various individual tran-

sients formed. The solid line represents the combined

concentrations of the following intermediates; the initial

OH* adduct (C6H5)3BC6H5OH' with a molar extinction coefi-

cient of 10,000 M-1 cm-1, the radical anion C6H5OH" and the

hydroxycyclohexadienyl radical C6H60H' with molar extinction

coefficients of 3800 M-1cm-1, and the intermediate dimers,

[HO(C6H6)-(C6H6)OH] and [HO(C6H5)-(C6H5)OH]2- with molar

extinction coeffiecients of 7,500 M-cm-1.