Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Flash photolysis of intermediates in hydrated electron reactions Vidyarthi, Suinil Kumar 1973

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1973_A1 V53_9.pdf [ 6.85MB ]
Metadata
JSON: 831-1.0060230.json
JSON-LD: 831-1.0060230-ld.json
RDF/XML (Pretty): 831-1.0060230-rdf.xml
RDF/JSON: 831-1.0060230-rdf.json
Turtle: 831-1.0060230-turtle.txt
N-Triples: 831-1.0060230-rdf-ntriples.txt
Original Record: 831-1.0060230-source.json
Full Text
831-1.0060230-fulltext.txt
Citation
831-1.0060230.ris

Full Text

)7-z  c I  FLASH PHOTOLYSIS OF INTERMEDIATES IN HYDRATED ELECTRON REACTIONS  by  SUNIL KUMAR VIDYARTHI  B.Sc. Hons., Patna University, India, 1966 M.Sc, Patna University, India, 1968  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY  in the Department of Chemistry  We accept this thesis as conforming to the required standard  UNIVERSITY OF BRITISH COLUMBIA JUNE, 1973  In p r e s e n t i n g  t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r  an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e  and  study.  I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying o f t h i s t h e s i s f o r s c h o l a r l y purposes may by h i s r e p r e s e n t a t i v e s .  be granted by  the Head of my  I t i s understood t h a t copying or  of t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be written  Department or  permission.  Department of The U n i v e r s i t y o f B r i t i s h Columbia Vancouver 8, Canada  publication  allowed without  my  -i-  ABSTRACT A t r i p l e f l a s h p h o t o l y s i s apparatus  i s described.  Hydrated  e l e c t r o n s ( e aq ~ ) p r o d u c e d by U.V.  f l a s h p h o t o l y s i s o f some common i n o r g a n i c  anions  a t 633 nm.  (OH",  SO^)  were monitored  T h i s s p e c i e s was  identified  by i t s a b s o r p t i o n s p e c t r u m and by t h e e f f e c t o f w e l l known e l e c t r o n scavengers. In t h e a b s e n c e o f an e l e c t r o n s c a v e n g e r , by a s e c o n d o r d e r p r o c e s s t o f o r m a c o m p a r a t i v e l y i n t e r m e d i a t e , X.  e" d e c a y e d l a r g e l y aq l o n g l i v e d (-r^ms)  Results i n d i c a t e that X i s probably a dimeric  hydrated  2  e l e c t r o n , (e£ ) q  o r  a  a  p r o d u c t t h e r e f r o m and can be p h o t o d i s s o c i a t e d i n  t h e 300 - 400 nm r e g i o n t o r e g e n e r a t e Metal ions A g  +  r e a c t w i t h e~ t o p r o d u c e t h e r e s p e c t i v e aq s h o r t l i v e d m e t a l atoms ( h y p e r - r e d u c e d s t a t e s ) w h i c h a l s o r e g e n e r a t e e~ aq by a s e c o n d f l a s h . M° + hv  300 nm)  (A^  and T l  hydrated e l e c t r o n s .  +  * M  Ag° has an a b s o r p t i o n s p e c t r u m w i t h x s t r e n g t h o f 0.63.  +  m a x  +  e^  q  a t 315 nm and an o s c i l l a t o r  T h i s band i s a s s i g n e d as c h a r g e - t r a n s f e r - t o - s o l v e n t  (CTTS). A c e t o n e , w h i c h r e a c t s r a p i d l y w i t h e" , p r o d u c e s a t r a n s i e n t , aq p r e s u m a b l y ( C H ) C 0 ~ , a t pH 11. The a b s o r p t i o n s p e c t r u m o f t h i s t r a n s i e n t was r e c o r d e d , (x = 255 nm, e ^ 1.3 x 10 * M" cm" ) and i t s p h o t o max max ^ 3  2  1  l y s i s i n t h i s r e g i o n was a g a i n shown t o r e g e n e r a t e  1  1  e~ . q  H e n c e , i t seems t h a t t h e a b s o r p t i o n s p e c t r a o f many s u c h t r a n s i e n t s may  be CTTS bands i n t h a t t h e y p r o d u c e e"  upon p h o t o l y s i s .  -il The decay of e  was followed in the presence of some inorganic  gases, e.g., SF , N 0 and Xe. 6  2  -  The possibility of intermediate formation  in those reactions are discussed. The carbonate radical ion, CO3, was produced during the flash photolysis of an N 0 saturated alkaline (pH 11) 2  solution containing C 0 and was studied by its absorption spectrum and 2  decay kinetics. It was demonstrated that HCOO" ions, methanol or 2-propanol, commonly used OH scavengers, yield e" upon flash photolysis.  -i i i CONTENTS CHAPTER  PAGE  I.  INTRODUCTION  1  1. Production of e~ aq (a) Pulse Radiolysis (b) Flash Photolysis  3 3 4  2. Intermediates in Hydrated Electron Reactions (a) Hydrated Electrons In 0H"/H Solution (b) Unusual Oxidation States of Metal Ions (c) Inorganic Gases (d) Reaction of e~ with Acetone aq 2  II. EXPERIMENTAL  12 13 18 20 24 27  1. The Flash Photolysis Apparatus (a) Flash lamps (b) Electronic Circuit (c) Spectroflash lamp (d) Reaction Vessel and the Flow System (e) Transmittance of Lamps and Filters  27 27 28 29 30 30  2. Detection System (a) Laser-Photomultiplier Arrangement (b) Spectrographs  32 32 33  3. Special Methods (a) Removal of C0 from 3D Water (b) Preparation of Aqueous Xenon Solution  35 35 35  4. Chemical Compounds and Gases  36  2  -IV-  CHAPTER  PAGE  III. EVIDENCE OF DIELECTRONS IN AQUEOUS SOLUTIONS  39  Results  39  1. Hydrated Electrons from the Main Flash (a) Decay of (b) Identification of the Absorption at 633 nm (i) Hydrated Electron Scavengers (ii) OH Scavengers (iii) Absorption Spectrum  39 41 45 45 46 48  2. Hydrated Electrons from the Auxiliary Flash (a) Filter Experiments (b) Absorption Spectra (c) Triple Flash Experiments (d) Absorption Coefficient (e) Effect of Added Solutes  50 52 56 57 6.1 62  Discussion  68  IV. FLASH PHOTOLYSIS OF Ag° and Tl°  74  1. Results from Ag(I) Solutions  74  Hydrated Electrons from the Auxiliary Flash (a) 0H"/Ag Solution (b) Ag S0i Solution (c) Na S0i and Ag S0i Solutions (d) OH Scavengers +  2  t  2  t  2  (  75 75 75 76 81  2. Results from Tl(I) Solutions  86  Discussion  89  -V-  CHAPTER  PAGE  V. TRANSIENT PRODUCED IN FLASH PHOTOLYSIS OF AQUEOUS ACETONE SOLUTIONS AT pH 11  93  Results 1. Decay of e" in the Presence of Acetone aq 2. Hydrated Electrons Produced by the Auxiliary Flash (a) Filter Dependence (b) Delay Dependence (c) Dependence Upon the Flash Energy  93 96 98 98 101  3. Triple Flash Photolysis  103  4. Absorption Spectrum  106  5. Decay of Ac'  109  Discussion  112  J  VI. INORGANIC GASES  116  1. Nitrous Oxide (a) 0H"/N0 Solution (b) S0^/N0 Solution (c) CO3/N2O Solution (d) Absorption Spectra (e) Double Flash Photolysis (f) N 0" ? (g) NO" ? (h) CO3 ? (i) N 0" in C0 Free 3D H 0 ?  116 118 120 120 122 122 126 126 128 130  2. Sulfur Hexafluoride  132  3. Xenon  137  2  2  2  2  2  2  -viCHAPTER  PAGE  VII. FLASH PHOTOLYSIS OF WATER AND AQUEOUS SOLUTIONS OF SOME OH SCAVENGERS  139  1. 3D Water  139  2. Aqueous OH Scavengers (a) Hydrated Electrons from the Main Flash (b) Double Flash Experiments  139 139 143'  Discussion  145  VIII. CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK  148  APPENDICES  152  REFERENCES  161  -vi i LIST OF FIGURES FIGURE  2-1  PAGE  Schematic drawing of the apparatus  27F  2-2(a) Triggering and time delay sequence (b) Capacitor-spark gap assembly circuits  28F  2-3  Transmission curves for some Corning glass filters  30F  2-4  Reproduction of an oscilloscope trace showing the duration of typical Main and Auxiliary flashes  33F  2- 5  Apparatus used for storing small amounts of Xe  35F  First and second order kinetic plots for the decay of aq f°H° i 9 fash  42  3-2  Matheson and Dorfman treatment of a combined first and second order process  44  3-3  First order plot for the decay of e" in the presence of 1.1 x lO" M NaN0  47  Absorption spectrum of the transient produced in the U.V. photolysis of 0H~/H solution  49  Reproduction of oscilloscope traces showing the absorbance at 633 nm as a function of time in an 0H"/H solution in double flash experiments  51  3-6  Absorption spectra of an 0H"/H solution 55 ysec after the Auxiliary flash and 630 usee after the Main flash  58  3-7  Reproduction of oscilloscope traces showing the absorption at 633 nm in the triple flash experiments on an 0H"/H solution  60  Reproduction of an oscilloscope trace for a solution made in C0 - free 3D water  66  Oscilloscope traces showing absorption by e^q at 633 nm during the Main and Auxiliary flashes in a 10" M solution of Ag in 10" M Na S0  77  Plot showing the decay <j>f the transient as a function of time in a 5 x 10~ M Ag solution  79  3- 1  e  w  n  t n e M a i n  5  a q  3  3-4  2  3-5  2  2  2  3- 8  2  4- 1  5  +  3  2  4-2  6  lt  -viiiFIGURE  PAGE  4-3  Absorption spectra in the wavelength region 250 - 425 nm for a 10" M Ag solution  82  4-4  Oscilloscope traces for experiments in a solution containing 3 x 10" M HCOONa and 1.1 x 10" M Ag at pH 7  84  4- 5  Oscilloscope traces from double flash photolysis experiments for a solution containing 1 mM iso-propanol and 10" M T1 S0^  87  First and second order kinetic plots for the decay of e~ following the Main flash in a solution containing " 2.5 x 10" M acetone at pH 11.  95  5-2  Oscilloscope traces showing absorption due to e" at 633 nm during the Main and Auxiliary flashes  97  5-3  Dependence of the percentage of e^n regenerated by the Auxiliary flash upon the flash energy  82  5-4  Absorption spectrum of the transient produced in the flash photolysis of a solution containing 5 x 10" M acetone at pH 11.  5- 5  Second order kinetic plot for the decay of the total amount 110 of egq regenerated by the Auxiliary flash as a function of delay time  6- 1  First order kinetic plot for the decay of a long lived transient at 633 nm in a solution containing 4 x 10" M NaOH saturated with N 0  117  Second order kinetic plot for the decay of a transient produced during the flash photolysis of a deoxygenated 2 x 10" M Na C0 solution  123  First order plot for the decay of the above transient in a N 0 saturated Na C0 solution  124  Absorption spectra of an 0H~ solution saturated with (1) H and (ii) N 0  125  5  +  3  5  +  5  2  5- 1  6  ac)  6  107  3  2  6-2  3  2  6-3  3  2  6-4  2  2  6- 5  3  2  First order plot for the decay of eI in the presence of 2.5 x 10" M SF  134  Q  6  H  6  7- 1  First order plot for the decay of e^g following the Main 143-'' flash in a solution containing 3 mM alcohol and 3.5 x 10" M NaN0 6  3  -ixLIST OF TABLES TABLE  PAGE  1-1  Comparison between the limiting quantum yield, <j> , for the production of hydrated electrons and the concentration of egq produced by flash photolysis  11  1- 2  Spectral properties of some intermediate products in  15  0  the flash photolysis of aqueous anion solutions 2- 1  Source, grade and purity of the chemicals and gases used . 37  3- 1 3-2  Transmission characteristics of some Corning glass filters Ratio of absorbance due to Auxi1iary/Absorbance due to Main flashes with different Corning glass filters  54 55  3- 3  Effect of added solutes on the second peak heights  65  4- 1  Ratio of peak absorbances (due to eaq) for 5 x 10~ M Ag solution with various light filters 6  80  +  4- 2  Ratio of peak absorbances for 2 x 10" M T l solution with various light filters  5- 1  Ratio of peak absorbances with various glass filters in an acetone solution  5-2  Time (trn ) for the second peak height to reach maximum in double flash experiments on solutions containing various concentrations of acetone  100  5-3  Results from triple flash experiments on a solution containing 5 x 10" M acetone  104  7-1  Results from the flash photolysis of aqueous methanol, T^-lf 2-propanol and formate ion solutions with the Main flash energy of 625 joules  7-2  Dependence of the second peak heights upon various glass filters used during the double flash photolysis of alcohols and formate ions  5  +  ax  6  88 99  14-4  -xAC KNOWLE DGEMENTS  I am indebted to Drs. N. Basco and D.C. Walker for their invaluable guidance throughout this work. Through discussions and criticisms they have greatly advanced my understanding of the subject. I am obligated to Dr. Geraldine Kenney-Wal1 ace for her kind and expert assistance during the initial stages of this work. I would like to thank Dr. G.B. Porter for his comments on the rough draft of this thesis. I am grateful to Kiss Gwen Thomas for reading the final copy. I am also grateful to Dr. Roger May for many useful discussions. Many thanks to Mr. Eric Fisher, Mr. Cedric Neale and all other members of the technical staff for their invaluable assistance in the construction and maintenance of the Flash Photolysis apparatus. I am also grateful to the University of British Columbia for financial assistance. Finally, I would like to thank everyone who in some way has been of assistance during this work.  CHAPTER I INTRODUCTION A process in which one reactant loses an electron to another is called an electron-transfer reaction and in principle is the simplest form of an oxidation-reduction reaction. This loss of electron may be caused by the presence of an electron acceptor in the immediate vicinity or by the introduction of excess energy - thermal, photochemical or electrostatic. The released electron may either be incorporated into the orbitals of an atom or molecule having positive electron affinity or if it travels through a medium of condensed matter, may end up in a potential energy "trap" induced by its own polarisation field.  The  electron thus trapped in a solvent differs from a free electron in that it is less mobile, more localised and thermodynamically more stable. This electron localised in a sheath of oriented solvent molecules has all the characteristics of an ordinary chemical reagent and is called a "soivated electron". In water, this species is referred to as a "hydrated electron" and designated by the symbol, e^q. In the photolysis of halide ions, Franck and Scheibe (1) proposed H-atom formation in a reaction between H ions and electrons +  which, according to their spectroscopic theory, were ejected into the solvent during the light absorption.  Stein (2) and Platzman (3) in the  early 1950's postulated the existence of hydrated electrons in explaining the influence of C0 and 0 on the bleaching of aqueous methylene blue 2  2  solutions and presented a detailed treatment of the thermalisation and hydration of secondary electrons. The existence of hydrated electrons  -2was then demonstrated by a number of radiation chemists based on competition kinetic studies. The first report of a transient absorption in radiolysis of water which was attributed to e~ was published by q  Keene (4) in 1960.  Boag and Hart (5,6) in 1962, discovered an optically  absorbing species, established its maximum at 720 nm, and concluded from its chemical and physical properties that the species involved was e" . aq This was followed by the observation of the absorption spectrum of hydrated electrons in many aqueous systems and which confirmed the original postulation of the existence of hydrated electrons. The spectrum presented by Keene (7) in 1964 is used as a reference spectrum during the present studies. Extensive research v/ork on hydrated electrons has been carried out in laboratories across the world in the past decade. Several books (8-11) and numerous review articles have been published in this field. This Chapter of the thesis presents a brief summary of some aspects of the subject that are pertinent to this work.  -31.1 PRODUCTION OF e" §4 Although several methods are available, two of the most useful are the techniques of pulse radiolysis and flash photolysis. (a) Pulse Radiolysis The technique of pulse radiolysis utilizes a short (10~ to 10~ sec) pulse of ionizing radiation, usually electrons, of 5  12  sufficiently high intensity to produce an instantaneous concentration of transient species, large enough to be observed by conventional absorption spectroscopy. These techniques have been well discussed in references 9 and 12.  In the pulse radiolysis of water, when the  ionizing radiation penetrates 1 iquid water, energy is lost in several stages (steps 1-1 and 1-2), HO  i-i ]  "  > HO  2  "sec  2  +  2  E  » se  e  + ;  * "t  e  e  E C  * aq e  where the secondary electron, e~ , a result of the ionisation of water, ec  loses its energy to excite another water molecule. The subexcitation electron, e~ thus formed is believed to become thermalised in 10"  12  e>  sec and finally hydrated in 10" sec. In the meantime H 0 transforms 11  +  2  into an OH radical and H^0 ion and hence the net ionisation reaction +  can be represented by process 1-3. 1-3  2H0 2  > H0  +  3  + OH + e~  q  The ionising radiation also produces radicals following excitation and dissociation of H0*, as in reaction 1-4, 2  1-4  H,0*  > H + OH  -4-  so that the primary reactive species produced are H, OH, and e" . These aq processes occur in minute local volumes called "spurs" along the path of high energy particles.  In the spur, the following reactions determine  the fate of hydrated electrons. e"aq + OH 1-5 •> OH. aq 1-6 1-7  e" + H aq e + e aq aq —  > Ho + OH' aq >H + 20H aq  k = 6  z  2  These are followed by the reaction of  k = 7  3.0 x 10  10  M~ sec" (13)  2.5 x 10  10  M~ sec" (13)  1  1  1  1  6.0 x 10 M~ sec" 9  1  1  (13)  with the molecular products  formed in the recombination of H-atoms and OH-radicals, and the trace impurities present in water. The yield of hydrated electrons by pulse radiolysis is quite high (depending upon the dose and the pH of the solution) but due to the above rapid reactions the lifetime of e~^ is usually short (in the order of about 10" sec). Thus, using high energy particles, it is not 9  possible to exclusively produce hydrated electrons as the reactive species. Other products, such as OH radicals and spur products in comparable amounts, are always formed. (b) Flash Photolysis Flash photolysis, on the other hand,is a technique to produce high concentration of free radicals (or ions) by using an intense pulse of light of any desired wavelength. Selective illumination of a substance may thus be performed using this technique by the insertion of glass filters having specified transmittance in different regions of the spectrum. This technique was employed for the studies reported in this thesis.  -5(i) Of Pure Water Photolysis of liquid water with U.V. light has been reported to follow process 1-8 (14-17). 1-8  H0 — — 2  > H + OH  Sokolov and Stein (18) reinvestigated the system using 185 nm light. They studied the excited water molecules, their reaction with scavengers and the production of e" . Using N 0 as an electron scavenger, which 2  yields easily measurable N gas by reaction 1-9, 2  1-9  N^0 + e"—> N + 0* aq aq z2  2  k = 5 x 10 M"i sec"  1  3 9  they obtained evidence that e~ is produced during the photolysis of aq water but the yield is less than 10% that of H-atoms. Recently Boyle et al (19) have observed a transient absorption in the flash photolysis of pure, oxygen-free, water with 195 nm light and attributed i t to the hydrated electron. Identification of the species was based on the absorption spectrum and kinetic properties of the transient in the presence of specific hydrated electron scavengers. The energetics of the overall reaction 1-10 was considered. hv  1-10  2H0 (aq) 2  » e~ + OH (aq) + H 0 q  3  (aq)  It was estimated to be endothermic by 5.8 eV. The wavelength limit for the formation of eV was found to be >, 192 nm, i.e., 4: 6.5 eV. This then implies that the ionisation potential of liquid water is approximately half of that in the vapor phase (12.6 eV). The quantum yield of e" aq in this process was rather small and estimated to be only 0.004. It  -6may be pointed out here that in the above work, only about 2% absorption measured at 578 nm, corresponding to about 3 x 10~M 8  , was obtained.  As will be seen later in this thesis, such small concentration of may be produced by flash photolysis of any trace anionic and organic impurities. Moreover, it was not possible in the present study to reproduce the results of Boyle et al using the especially purified water even though the intensity of the light flash was about a factor of 3 larger. (ii) Of Aqueous Solutions of Alcohols Matheson et al (20) postulated the formation of e~ by the photolysis of water in an aqueous solution containing 0.2M methanol. Sokolov and Stein (18) pointed out that at the above concentration, methanol would absorb half of the incident light below 200 nm and that the ionisation potential of methanol is lower than that of water. Methanol and isopropanol, with very low reactivity towards e" , have been quite frequently used as specific OH-radical scavengers in Radiation Chemistry. OH-radicals react with the alcohols by abstraction of a hydrogen atom. 1-11  CH0H + OH  1-12  (CH)CH0H + OH  3  3  2  > CH-0H + H 0 2  k = 4-6 x l O ^ S e c " (13) 1  2  n  >(CH ) C-0H + H 0 k = 1-4 x 10 M Sec (13) 9  3  2  2  -1  _1  12  Based on some results during this project it can be said that in the above work, e" was actually produced by the photolysis of aq methanol rather than water itself (Chapter VII).  -7-  (iii) Of Aqueous Solutions of Inorganic Anions; CTTS Spectra Flash photolysis of aqueous inorganic anions with ultraviolet light of sufficent energy has been explained as actual photodetachment of an electron which then becomes hydrated. 1-13  Y" — — — > Y + e" aq aq aq an  The photochemistry of such processes involving in particular the halide ions has been reviewed by Stein (21). The absorption spectra of Cl", Br", and I" ions in solution consist of bands with high oscillator strength, which are not observed in the gas phase. These spectra were first interpreted as due to the Charge Transfer to the Solvent (CTTS) by Franck and Scheibe (1). It has been suggested that an excited state is first obtained (22), 1-14  Y"  + hv  >  aq  '  Y"  *  aq  which may return to its initial state by energy loss through nonradiative processes. However, as an alternative, the water dipoles may begin to take up a new equilibrium position which can lead to a state where the electron is ejected and hydrated. 1-15  Y" * aq  » Y + e" aq aq an  Weiss (22) regards this model to be similar to that of an F-center and not a free hydrated electron,mainly on the basis of small variations in absorption spectra for e~ , obtained from different anions in aqueous q  solution. Several theories have been put forward regarding the mechanism  -8-  of reaction (1-13) since the original Franck and Scheibe's CTTS model. For instance, Franck and Platzman proposed a theory in 1954 which has been modified by Smith and Symons (23) to explain some environmental effects, e.g., temperature, solvents and added solutes, on the absorption bands. However, the theory failed to predict the exact location of the CTTS bands of the halide ions (21). Taking into account the discrepancies in the above theories, Stein and Treinin (24) developed a new theory which retained the physical picture of the bound excited electron but introduced a parameter, r , the effective radius of the solvent cavity occupied by the original ion. CTTS "y• "  hv  E  H +  ( i o> * K  /r  + K  In this expression Ki and K are constants for the series of monoatomic 2  halides, depending upon the optical dielectric constant, D p and the Q  dielectric constant of the solvent, D . E is the electron affinity 5 y of the halogen atom Y, hv^j-r^ is the energy of the transition at band maximum and H is the heat of solvation of Y. The spectroscopic values of r calculated from the above equation were in fair agreement with Q  ionic radii calculated from the partial molar volumes of the respective ions. Use of the above mentioned environmental effects have been made for the case of certain aqueous anion solutions to identify the CTTS nature of their absorption bands. Shirom and Treinin (25) studied the effects of temperature, solvent and added solutes on the absorption spectrum of H~ ions, which consists of three bands. The results were 3  compared with similar effects in case of halide ions. The band at 200 nm  - 9 -  was assigned to be due to CTTS on the basis of the evident similarity. A thorough study of the solvent effects on halide ions was done by Burak and Treinin and a solvent scale for CTTS spectra of anions was reported (26). Although the studies were restricted to monovalent anions only, it was suggested that this scale could be used to identify a CTTS band for most anions. Whether an electron is actually transferred from an ion to the solvent during the transition designated by a CTTS band can be ascertained by comparing the x of the bands and the free energy change for the max reaction ( 1 - 1 6 ) . mav  1-16  X  n+  +H  +  >X  (n+1)+  + %H  2  Such a comparison for several ions was made by Marcus (27). It was observed that of those studied the most electronegative ion, F" ( A G for reaction 1 - 1 6 = 9 4 kcal/mole) has its absorption maximum at the shortest wavelength  (150  nm), while Cr  +2  ions, for which process  (1-16)  is exothermic and A G = - 9 kcal, absorbs at the longest X ( 3 8 0 nm). In addition to the characteristics of the CTTS spectra described above, these bands share a common feature. Photolysis of ions with definite CTTS bands (e.g., halide ions and OH" ions) with the corresponding wavelengths of light produce hydrated electrons, which can be easily identified by their strong absorption in the visible. This property of the CTTS bands can therefore be employed to identify such absorption spectra. Inability to use different solvents, experimental limitations to study temperature effects and a restricted range of solute concentration has forced the author to base his conclusions regarding CTTS bands mainly on  -10the above type of experimental observations. Calculation of the oscillator strength for the transition has been used to support such conclusions in some cases. Quantum yields of hydrated electrons in the photolysis process 1-13, have been reported for some ions. These data have been compared with the concentration of hydrated electrons produced during the flash photolysis of the corresponding ions in Table 1-1.  It may be pointed  out here that information regarding the wavelength limit of such photolysis processes may be used to estimate the ionisation potential of the respective ion.  -11-  Table 1-1 Comparison between the limiting quantum yields, <|> , for the production of hydrated electron quoted in the literature (28) and the concentration of hydrated electron produced by flash photolysis in this work.  Literature data Ions  r  o  y  Yields from this work  Wavelength (nm)  [e*aq ] x 10?M  Wavelength (nm)  0.230.29  253.7  50.0  OH"  0.110.50  184.9  20.0  <220  SO "  0.71  184.9  14.0  <220  -  -  10.0  <220  -  -  4.0  <220  -  -  0.9  <220  2  u  CO" 2  3  *  HCOO" *  CIO^  280 - 240  * These ions have been shown to yield hydrated electrons by process 1-13, for the first time during this work.  1.2  INTERMEDIATES IN HYDRATED ELECTRON REACTIONS This section gives a brief account of some of the existing  problems in Radiation Chemistry that have been dealt with in this thesis. During the past decade research on the reactions of the hydrated electron with various molecules has shown that short lived intermediates are often formed in such reactions. The reactants involved include inorganic, organic or biological molecules, neutral atoms and positively or negatively charged species. The nature of the intermediates are diversified in an analogous way.  In many cases an adduct type of  structure for such intermediates have been suggested. results in a loosely bound complex of type, S  This reaction  e" , or species of type aq  S~, where S represents the reactant of any of the types mentioned above. These intermediates often may have narrow absorption bands with high oscillator strength which, when photolysed, may lead to charge-transferto-solvent. The experimental set-up (described in Chapter II) assembled in this laboratory during this work is designed for investigating such intermediates. Using any of the aqueous anion solutions in Table 1 - 1 , significant concentrations of hydrated electrons can be produced by a U.V. flash. A second light flash can be used to photolyse the intermediates formed following the decay of the hydrated electrons in the presence or absence of the reactant, S, at a known concentration. This involves the technique of double and triple flash photolysis, which is described in Chapter II.  -13Due to the diversified nature of the reactants and the intermediates, the results are presented in separate chapters (III to VI) in this thesis. The following parts of the present section offer a brief introduction to the various systems involved in this study; (a) The formation of an intermediate has been reported in a solution free of any such reactant (36). The subject is briefly discussed in this part; (b) Several metal ions are known to form short-lived hyper-reduced states by reaction with e!j . A good deal of work has been done on these states, particularly in the case of transition metal ions. The topic is discussed + + with respect of Ag and Tl ions; (c) Among other inorganic solutes, gases like N 0 and SF have been 2  6  suggested to form negatively charged species. Traces of C0 present 2  in aqueous solutions may yield an interesting species, CO3 radical anion. The possibility of xenon being used as a reactant is discussed; (d) Acetone is known to react rapidly with e~ and was studied as an example of an organic scavenger. The last part of this section deals with this reaction and a possible reaction intermediate. (a) Hydrated Electrons in 0H~/H? Solution Flash photolysis of any of the aqueous anion solutions listed in Table 1-1 may be employed to produce hydrated electrons in detectable concentration. However, the fate of the radical, Y, and its products (reaction 1-13) must be considered. Table 1-2 presents the available spectral data for such intermediate products from the ions used to produce e" during this work. The spectral regions of absorption by  -14-  the photolysis products of I", HCOO" and SO^ complicate the present studies. The C O 3 radical anion, on the other hand, absorbs in the same region as the hydrated electron itself (x^ /! 720 nm). 3  =  Also, due to  the extreme reactivity of e"^, the life-time of e~ is severely curtailed by the presence of the above radicals or their products. OH-radicals, however, have only a weak absorption in the U.V. Thus, along with the fact that minimal amounts of complicating products are formed, OH" ions were found the most suitable source to produce long lived e"^. The OH-radicals are produced in reaction 1-17. 1-17  OH" — — > OH + e" aq aq x< 220 nm  The solution is saturated with hydrogen gas, which reacts with the OH-radicals and in turn produces additional  by reactions 1-18 and  1-19. 1-18  H + OH  » H + H0  k  18  = 5.0 x 10 M" sec" (13)  1-19  H + 0H~  » e"  k  19  = 2.0 x 10 M" sec" (13)  2  2  7  9  1  1  1  1  At this concentration (7 x 10~ M), H also minimises the extent of k  2  reaction 1-20. 1-20  OH + OH  > H0 2  k  = 5.0 x 10 M" sec" (13) 9  20  1  1  This system has been successfully employed by previous workers (36,37) to produce long-lived e" . In the absence of any specific scavenger, e~ decay by a second order process. Basco et al (36) in a double flash photolysis experiment indicated the presence of a long-lived (t^ up to about 0.1 sec)  -15Table 1-2 Spectral properties of some intermediate products in the flash photolysis of aqueous anion solutions.  Ions  Radicals or products  x ,„ m  ( n m )  e „ m  v  ( -i -i) M  Reference  cm  h  370  1.56xl0  30  SO^  455  4.6x10  31  C0  CO3  600  1.86xl0  32  HCOO"  HC00(?)  250  2.25xl0  33  OH"  OH  _*  3.7xl0  34,35  I" 2_ SO4  2_ 3  5  2  3  3  2  ( 260 nm) e  * The absorption spectrum of OH-radicals in water has been observed in the region 200-300 nm (35) and it appears that X  < 200 nm.  -16-  intermediate - postulated to be a hydrated electron dimer - in reaction 1-21. 1-21  e" + e^ q  q  > [^"^q  *\  k  The hydrated electrons were monitored at 633 nm  = 5 x  (G633  l° M Sec"i 9  _1  = 1 .2  X  lO^fOcnf  1  (7)). About 10~M e" were produced by a U.V. flash using the aboveaq mentioned H/0H" solution. Long lived e" (t>0.2 ms) decayed mainly 6  2  *i  aq  by a second order process and a value for k i comparable to the published data was obtained. After the majority of e" had decayed, an auxiliary 2  aq  flash of restricted wavelength, unable to photolyse OH", generated an additional absorption signal at 633 nm due to e~ . This was explained on the basis of the photolysis step 1-22. 1-22  ^l'\ AQ  Q  —  > 2e"  A>700 nm  a q  Stable dielectron species are known to occur in polar solids (38,39) and in metal-ammonia or -ether solutions (40). This, however, was claimed to be the first demonstration of short lived dielectrons in water. Czapski and Peled (41) questioned the existence of dielectrons. It was demonstrated that no absorption could be obtained in the region 600-1000 nm in the pulse radiolysis of an aqueous alkaline solution containing 10~M alcohol. An upper limit for the extinction coefficient 2  of these species was set at ^15 lOcm" . However, the results presented 1  in Chapter III indicate that the intermediate absorbs near 350 nm rather than in the wavelength region previously suggested (36). Furthermore, in their system, the lifetime of e" , produced in comparable concentration  -17was only about 50 usee and never greater than 100 ysec.  It will be  seen later (Chapter III) that these species could not be formed in significant amount when e"^ decayed with a lifetime of <100 ysec. This also is consistent with Czapski and Peled's findings. For a solution containing 10~ M NaOH saturated with H Gopinathan et al 3  2  (37) suggested the occurrence of an e~^ complex, (^ - ~) q which could a+  e  a  be photolysed to produce hydrated electrons by photolysis with light centered around 300 nm. . (Na .e~)  1-23  The complex (^ - ~) q a+  e  a  hv a q  w a s  > Na + e"  X^300 nm  +  a q  .  considered the intermediate involved mainly  because of the increase in the concentration of e"^ on photoionisation with added Na ions (as NaClOiJ. Since di electrons had been reported +  to be absorbing in the infrared (36), involvement of such species in process 1-23 was not considered. This long lived intermediate (36) is of considerable interest from the point of view of (i) spur reactions in Radiation Chemistry, (ii)  the reaction of alkali metals with water yielding H gas and  (iii)  . the cathodic electrolytic decomposition at high pH in which  2  elr. aq may be involved in comparable reaction schemes (42). Therefore, a detailed investigation of this system was carried out by the author on an improved and more versatile triple flash photolysis apparatus (described in Chapter II), especially built for these and related studies, The wavelength limits for the photoionisation of these species were reinvestigated and effect of added solutes were studied. Chapter III presents results and discussion from these studies.  -18(b) Unusual Oxidation States of Metal Ions Transition metal ions are known to form transient ions (or atoms) by reacting with e" . For divalent ions, reaction 1-24 occurs aq  with a high rate constant (45). 1-24  M  + e" aq  2+  9-  M  +  In liquid solutions, the absorption spectra of these species (in an unusual valence state of the metal) have been obtained in pulse radiolysis experiments (44-47).  X  m a x  for most of these bands are in the  region of about 300 nm, with high extinction coefficients. These unusual oxidation states are also referred to as the hyper-reduced states of the metal ions involved, Mi , Cd , Zn etc. Similarly, +  +  +  hyper-oxidised states of metal ions have been observed when reaction takes place with OH-radicals.  Recently Buxton . et al have reported the  formation of such hyper-oxidised, states in the reaction of the hyperreduced states of the transition metal ions with N 0 (48). The hyper2  reduced states have also been observed for some metal ions of the Lanthanide series. For instance, the absorption spectra of Eu(II), Yb(II) and Sm(II) have been reported (49). It has been suggested that the absorption around 300 nm by the hyper-reduced transition metal ions may lead to charge transfer (46).  It is conceivable that these intermediates are in their ground  electronic states as the excited states will be extremely short lived. In case of Ag(I) solutions, producing Ag° by reaction 1-25, 1-25  Ag + " +  e  > An°  k  = 3.2 x 10*° M" Sec" 1  25  1  the ground state is 5S, (a 5s configuration). Brown and Dainton (50) 2  1  -19have tentatively assigned the absorption spectrum of Ag° in glasses to  h H  3 a 5 PV , , * r 5S, transition. Ag° has also been identified in the 2 or 2  2  3  n B  pulse radiolysis of Ag(I)-doped sodium metaphosphate glasses (51) by photometric methods, and in pyrex and sodium tetraborate glasses by EPR (52). EPR studies have also shown their presence in gammairradiated AgN0 ices (53). There is evidence for the existence of 3  Ag° in liquid ammonia as. well (54). Since Ag° is known to decay predominantly by reaction 1-26, 1-26  Ag° + Ag  » Ag  +  k  2  = 5.9 x 10 M" sec' (47) 9  26  1  1  it can be seen on the time scale of the present experimental set-up if a dilute solution of Ag ions (10" - 10" M) is used. Chapter IV of +  6  5  this thesis presents some results and discussion from flash photolysis experiments using aqueous Ag(I) solution. Experiments were performed mainly to investigate the nature of the Ag° absorption band by using the technique to determine if the band is of the CTTS type as described earlier (1.1, b, i i i ) .  Thallous, a non-transition metal ion, has been  reported to give similar results to Ag(I) solutions. Cercek et al (55) have obtained the absorption spectra of T l  and T l ions in a pulse  2 +  2  radiolysis experiment. T1°, which may be produced in reaction 1-27, 1-27  T l + e~ +  q  *T1°  k  = 1.1 x 10^ M" sec' (45) 1  27  1  reacts with T l ions at a very fast rate by reaction 1-28. +  1-28  Tl° + T l  +  »Tl£  K=kp/kg= 2.3 x 10 M~  Thus, the absorption spectrum of Tl° could these workers.  3  1  (55)  not be recorded by  -20Due to the noted similarity between Ag° and Tl in aqueous 0  solutions, Thallous solutions were also examined in an analogous manner by the author. Results obtained from this system (Chapter IV) were used to add credence to the conclusions drawn in case of Ag°. (c) Inorganic Gases Several inorganic gases have been found to react rapidly with hydrated electrons in aqueous solution (8). The products of these reactions are varied and in few cases negatively charged transient intermediates have been postulated, (i) Nitrous Oxide N 0 is one of the most popular hydrated electron scavengers. 2  The reaction seems to proceed as follows: 1-32  e~ + N 0  1-33  N 0~ + H 0 -—* N0H + OH"  q  »N0  2  2  2  2  k  aq  = 8.7 x 10 l O S"i (63) 9  32  2  N0H  *N  2  + OH  2  Some controversy exists regarding the precise mechanism of this reaction. It appears that the first step is the formation of a negative ion by reaction 1-32. This may dissociate to give N and 0~ by process 1-34, 2  in a time probably comparable to the vibrational period of the molecule, 1-34  N0" 2  a  q  » N  2  +  0" a  q  or it may have a sufficiently long lifetime to enter directly into reaction with solutes. Anbar et al (64) studied the deamination of ethylene diamine by pulse radiolysis in aqueous solutions and support the idea of  -21a long lived N0~ . However, results of Adams (65) are in conflict with 2  this concept of N 0~ with a lifetime greater than 10" sec. 9  2  q  It has been suggested that N0~ is stable in alkaline solutions 2  (Henglein in 65, & 66) and its rapid decay in less basic solution is attributed to the equilibrium 1-33.  To the author's knowledge, there  is no direct evidence for the existence of such species. Based on the suggested stability of N0~ in alkaline solutions, several experiments 2  were carried out to try to detect N0~ in the flash photolysis of 2  aqueous 0H~ solutions containing N 0 and are presented in Chapter VI. 2  (ii) Sulfur Hexafluoride SF is also known to react rapidly with e~ (67) with the 6  q  following proposed stoichiometry. 1-35  e~ + SF + 4H 0 = SO*' + 6F" + 8H  +  q  6  2  The mechanism of the above reaction involves a negative ion, SFg, which dissociates into SF radicals and F~. The rest of the products are 5  formed by step-wise hydrolysis of the SF radicals. The rates of b  hydrolysis for some of these intermediates have been studied by conductometric pulse radiolysis (68). Rumfeldt (69) explained his results on the effect of SF on the photochemistry of aqueous solutions by assuming 6  SF as an intermediate in reaction 1-35. 6  No speculation on its lifetime  or absorption region was made. Recently Crawford and Rumfeldt (70), on the basis of a molecular orbital scheme of SF , suggested that the add6  ition of an electron to SF to yield SFg would involve the use of a high 6  energy antibonding orbital. The dissociation of SFg is suggested to take the following path in the presence of an anion, A".  -221-36  SFg  1-37  SF + A"  1-38  SFg  » SF + F" 5  > SFi + A  5  > SF  h  +  F"  However, the existence of long lived SFg and SF5 is not implied in the above mechanism. On the contrary, it was suggested that these species may be in repulsive electronic states in which case 1-35, 1-36 and 1-37, 1-38 would constitute single concerted reactions. Once again, there is no evidence for the existence of SFg and SF5 ionic species in aqueous solutions. Thus, the decay of e~ in the aq  presence of SF was followed. Several experiments were carried out to 6  investigate these intermediates by flash photolysis. The results are reported in Chapter VI. (iii) Xenon The solubility of noble gases in water increases with increasing atomic no. (71), Xe being the most soluble (^1 M at saturation) among the non-radioactive elements. It has been observed that the ionisation potential of a molecule may be greatly lowered in the condensed phase (19) or in aqueous solution. Bearing this in mind, attempts have been made during this work to photoionize aqueous Xe solutions (Chapter VI) with far-ultraviolet light down to ^190 nm. The ionisation potential of Xenon in gas phase is %12.0 ev (29). Although it has been suggested that inert gases should not display any reactivity towards e~ (8), the author reinvestigated this aq  claim in these particularly favorable conditions. Experiments were  -23carried out to observe the extent, if any, of the reaction 1-39, 1-39  Xe + e" aq  where e  > Xe" aq  was produced by 1-17, and attemnts were made to photoionise  any Xe" ., if formed. (iv) Role of CO2 in Aqueous Alkaline Solution Due to the higher solubility of C0 in water, the presence of 2  some dissolved C0 is inevitable if water is in contact with the atmos2  phere. In alkaline solutions these molecules are conveniently converted into carbonate ions. 1-40  C0 + 20H"  » CO3" + H 0  2  2  When hydrated electrons are produced by process 1-17, comparable concentration of OH-radicals are also formed. These radicals are known to 2_  react with C0  3  ions at a fast rate (32),  C O 3 " + OH  1-41  » CO3 + OH"  kkl = 4.1 x 108 ~ i sec"i (32) M  Carbonate radical anions, produced in reaction 1-41, have been reported to have an absorption spectrum in the visible, with x  max  = 600 nm and  max ^-86 x 10 fOcm" (32). As will be seen later in this thesis,  e  =  3  1  hydrated electrons were monitored at 633 nm ( - = 1.2 x ']Q M " cm") aa k  1  1  e  in this work and at this wavelength, CO3 radical anions have an extinction coefficient of 1.53 x 10 M" cm". 3  1  1  In this work the author has observed the absorption spectrum of C0 (Chapter VI) and compared it with that of e~ in this system. 3  Experiments were performed to demonstrate that the absorption signal  -24obtained at 633 nm due to e" has entirelv different characteristics aq than that due to C0 radical ions. The distinction of the two species 3  was conclusively established by repeating the experiments in C0 -free 2  water. (d) Reaction of Hydrated Electron with Acetone Acetone has been shown to be an excellent scavenger of hydrated electrons based on competition kinetic studies (56) as well as direct determination of the rate constant by pulse radiolysis (57). However, it was observed that in the gamma-radiolysis of DMSO (dimethyl sulfoxide) solutions, high concentrations of acetone do not diminish the N yield 2  when in competition with N 0 for the solvated electron (58). This was 2  explained on the basis of an electron-transfer reaction, 1-29, involving a long-lived intermediate, postulated to be the radical anion, [(CH ) C0]-. 3  2  1-29  [(CH ) C0]" + N 0 s 3  2  >N'+ 0" + (CH ) C0 s  2  2  3  2  Rather similar observations have been made during the gamma-radiolysis of HMPA (hexamethyl phosphoric triamide) solutions, where an increase in the G(N ) was observed, if 0.7 M acetone was added to a solution con2  taining 3 x 10~ M N 0 (59). Once again, electron transfer from the 2  2  [(CH ) C0]~ radical anion to N 0 was considered 3  2  2  to explain the results.  In high energy electron radiolysis of HMPA solutions containing 3 x 10"  2  M acetone, an absorption signal at a considerably lower wavelength than the Amax for e~ s v/as actually observed, with a lifetime that ruled out e~ as the absorbing species (59). This absorption, tentatively assigned to [(CH ) C0]" in HMPA, may be part of a CTTS band. 3  2  -25In the radiolysis of aqueous acetone solution, Reisz (60) suggested that the hydrated electrons react with acetone to form an anion radical, which is the precursor of isopropyl alcohol, the predominant radiolysis product in the system. This is consistent with the previously reported mechanism for the formation of isopropanol radicals in the pulse radiolysis of aqueous acetone solutions (61). Reactions 1-30 and 1-31 were suggested (57). 1-30  (CH ) C = 0 + e~  1-31  (CH ) C - 0' + H 0 ^ z = ± (CH ) C - CH + H 0  3  MCH ) C  2  3  - 0"  2  +  3  2  3  3  2  2  A transient absorption in the 200 - 300 nm region was observed and assigned to the isopropanol radicals. The above results were obtained from neutral or acidic solutions. The equilibrium constant for reaction 1-31 was obtained, pK = 12.2, and it was suggested that in strongly alkaline solutions, [(CH ) C0]~ 3  2  predominates. The absorption spectra  (in the region x>250 nm) in alkaline solutions of acetone indicate that \nax °^  a D S O r D  i n g species is < 250 nm (62).  This thesis reports experiments carried out (Chapter V) to investigate the intermediate in reaction 1-30.  The spectral region from  about 230 nm to 400 nm was scanned. These studies were conducted at high pH (^11) for two reasons, (a) photolysis of OH" with U.V. light aq  provides a source of e~ (Table 1-1) and (b) the reverse of the reaction q  1-31 is favored. Double and triple flash photolysis experiments were performed to establish the nature of the absorption spectrum. Kinetic measurements were made by monitoring e~ at 633 nm. q  -26-  The purpose of this investigation was to produce the above discussed intermediates using a U.V. flash and then study them by the subsequent flash(es). As observed before in this laboratory (36) regeneration of e" aq bv the second flash was recorded for an intermediate described earlier (a). This thesis reports an analysis of this system in a greater detail as well as an investigation as to whether such effects could be observed for other transients described above. In other words, is the production of e~ by flash photolysis of the intermediates in q  hydrated electron reactions a common feature of these species? In addition to the study of these electron - adduct intermediates, some interesting results were obtained during the flash photolysis of aqueous OH scavenger solutions. These results are reported in Chapter VII.  CHAPTER II EXPERIMENTAL II.1  THE FLASH PHOTOLYSIS APPARATUS The apparatus is schematically shown in Figure 2-1. Two  flash lamps, one on either side of the reaction vessel, lay on brass metal supports in an all brass casing. The casing had essentially three parts. The bottom hollow cylindrical  part was connected to  the top two by means of hinges. The top part had supplementary supports to ensure tight fitting of the tubes. The casing v/as 60 cm long and 5.4 cm in diameter, and was horizontally secured on a 1.5 metre long optical bench. For the maximum efficiency of the photolysis the inside of the casing was covered with aluminium foil.  Appropriate  slits were cut to securely fit the aluminium frame for holding glass filters and holes were made for the intake and exit of the flow system for filling up the reaction vessel. (a) Flash lamps The photolysis lamps L and A were 50 cm long and were made from Suprasil tubing (8 mm I.D., 1.5 mm wall thickness) with B-10 quartz sockets at each end.  The electrodes were tungsten rods soldered onto  brass B-10 cones which were sealed into the sockets with black wax. The lamps were evacuated and filled through a hole in the electrically grounded electrodes. The charging electrodes were connected to the capacitor assemblies Cj, C , and C by means of high voltage doubly 2  3  shielded cables. The ground electrodes were linked with the brass container and hence the casing served as the 'ground' of the system. The 11  Figure 2-1  Schematic drawing of the apparatus. V, reaction vessel; L, Main flash lamp; A, Auxiliary flash lamp; C C , and C , Capacitor-spark gap assemblies shown in Fig 2-2; F, light filters; S, Spectroflash lamp; G, spectrograph; H, He-Ne laser; I, interference filter (narrow band pass at 632.8 nm); P, photomultipler detector; 0, oscilloscope; M, plane mirror. l9  2  3  _28_ electrodes were cleaned and noli shed quite frequently for maximum efficiency. The lamps were cleaned with 10% HF solutions. (b) Electronic Circuit The Main flash lamp (L in Fig. 2-1), so called because it was always used as the first flash lamp in a multi-flash experiment, obtained its energy from a 50 yF Dubilier Capacitor. The charge was held off by a spark gap connected to a pulse transformer as shown in Figure 2-2 (b).  Upon triggering the pulse transformer the gap closes and the  capacitor discharges across the lamp filled with a low pressure of Argon gas (^10 torr) producing an intense light flash. The Auxiliary flash lamp circuit was similar except that a 33.5 yF capacitor was employed. An additional 33.5 yF capacitor assembly was connected in parallel to the above, so that the Auxiliary lamp, A, could be discharged twice producing altogether three flashes. By adjusting the distance between the two nodes of the spark gap, the voltage range under which the lamps were to be fired could be altered. Considering the amount of light required and the endurance of the suprasil lamps a safe region of <3 to 8 kv was selected. The triggering circuit is shown in Figure 2-2 (a). Generally, the Main flash was fired manually and the same pulse fired the delay unit, Di.  If there was no delay in the breakdown of the lamp, this was very  efficient. But sometimes a variable delay was observed before the Main flash, the reason for which could be dirty electrodes, old lamps, etc. Hence a phototube that produced a 5-20 volt signal by picking up the scattered light from the Main flash, was used to trigger both the oscilloscope and the electronic delay unit, D . x  The rest of the sequence  -28F-  TO SCOPE  to lamp  •T  (b)  TO SPEC.  Ca)  Figure 2-2  (a) Triggering and time delay sequence for three photoflashes and a spectroflash. Tj is a manually operated trigger pulse generator; T , T , J and T are trigger pulse generators for the Main, Auxiliary (first flash), Auxiliary (second flash) and spectroflash respectively. D , D and D are variable time delays. Any unit may be by-passed. 2  k  3  5  x  2  3  (b) Capacitor-spark gap assembly circuits for the Main and Auxiliary lamps (C C and C in Fig 2-1). T is the trigger pulse input. ls  2  3  -29(Fig 2-2 b) was unaltered. The Auxiliary lamp was fired by the trigger output from the delay unit, D  ls  and could be fired again by a trigger  from D . D and D were used to set delays between any of the three 2  1  2  flashes in the range of ^ 1 ysec to ^100 msec. The electronic delay unit D was used to select a time lapse between the photolysis flash lamp 3  and the spectroflash lamp. In triple flash experiments where both the capacitors connected to the Auxiliary lamp were fired at the same energy, the intensity of the second Auxiliary flash was usually found to be lower than the first. This was probably due to a back discharge from the second capacitor. The reason for this behaviour is not well understood. (c) Spectroflash Lamp The conventional vertical type of spectroflash lamp (S in Fig 2-1) employed in previous similar types of experiments by Kenney (72) was found adequate for absorption spectra recorded in the visible and infra-red regions. It consisted of a fine capilliary with similar electrodes to the photo-flash lamp. The capilliary was filled with Argon at about 30 torr pressure and a 2 yF capacitor charged to 9.0 kv was discharged to produce a narrow and intense light flash. A small hole on the other end of the casing was made for convenience of alignment by using the already coaxial laser beam, H. For studies in the U.V. region of the spectrum, a modified version of the horizontal type of spectroflash lamp was used. The basic design of this lamp has been described by Morse (73). This lamp had considerably higher intensity in the U.V. and had a comparatively shorter  -30duration of the flash. The original design was slightly modified for the present work by replacing the otherwise solid charging electrode with a hollow one provided with a glass window. This served the same purpose as the hole in the vertical type of lamp. (d) Reaction Vessel and the Flow System The reaction vessel, V, in Fig 2-1, was usually a 42.5 cm long, 1.2 cm in diameter and 1 mm walled suprasil tube placed in the centre of the brass cylinder described earlier. Two suprasil windows were fused on either ends of the tube which was provided with two side arms for the inflow and outflow of the solutions. The vessel was cleaned before each set of experiments by flushing it with 2D water, alcohol, dichloromethane, more 2D water and finally with 3D water. High purity water was prepared by redistilling distilled water with acidic dichromate (called 2D at this 60  stage), gamma-irradiating with 2Mrad dose from a Co  Gammacell and then  refluxing over an alkaline potassium permanganate solution for at least overnight. The water thus prepared was termed triply-distilled or 3D and was used for making all experimental solutions. Solutions were deoxygenated by bubbling He or H gas through them in a pyrex reservoir 2  and were pushed into the reaction vessel via an all glass system under a small pressure of the same gas. The used solution could either be drained off or replenished by pushing in fresh solution in a similar way. The flow system was not open to the atmosphere. (e) Transmittance of Lamps and Filters The spectrosil grade of quartz used to make the photolysis and spectro-flash lamps was able to transmit light of wavelength > 190 nm.  100  260  340  420  500  580  660  740  Wavelength (nm)  Figure 2-3  Transmission curves for some Corning Glass filters used at F in Fig 2-1, Curves were obtained from "Corning Glass Filters" catalogue.  -31Various Corning Glass filters of thickness 1.0 and 2.5 mm which were custom-made to fit the aluminium frame F (Fig 2-1) were used to perform photolysis with restricted wavelengths of light. The transmittance of the filters normally used are given in Figure 2-3. The actual % transmission values relative to the maximum of each filter are tabulated in Table 3-1 for every filter used during this work. Both the data in Fig 2-3 and Table 3-1 were obtained from the Catalogue for Glass Filters (Corning Ltd.). In this thesis, these filters will be referred to by their catalogue number.  II.2 DETECTION SYSTEM (a) Laser-Photomulti pii er Arrangement Either a 9 mw, (Optics Technology model 233) or a 2 mw(Spectra Physics model 132) C.W.,  He-Ne laser was used in this arrangement. The  laser beam, after passing through the axis of the reaction vessel, was centered on a 1P28 Photomultipier tube via a 3 mm thick narrow bandpass (1.5 nm) interference filter (632.8 nm).  A 1 K ohm load resistance  was generally used, but this could easily be replaced by loads varying from 50n to 1 Mft. A fluke 412B High Voltage power supply was used for the photomultipier. Various oscilloscopes, e.g., Tektronix models 454, 532 and 545, and a Tele-equipment model D-43 to display the signals for the photomul tipier.  were employed  By plotting the output  vs the high voltage, using different neutral density filters in the path of the laser beam, a linear region for the response of this particular photomultipier tube was obtained. This was 570 - 720 volts and the tube was always operated within this region. The traces were recorded on black and white Polaroid film type 47, speed ASA 3,000 by using a Polaroid camera attached to the oscilloscope. The oscilloscopes were triggered externally be the pickup signal at the beginning of the Main flash. Thus, recorded traces represented the absorption of the 632.8 nm laser line. The decays of these signals were analysed for kinetic measurements. Use of an IBM 360/67 computer was made to convert the directly obtained % absorption into Absorbance(A), 1/A, and log A. Additional Fortran IV programming was 'used to yield the least square slopes for the plots of time vs 1/A and log A with correlation coefficients and standard deviation.  -33The results from the double and triple flash photolysis experiments were analysed as follows. Because of the observed variation in the intensity of the Main flash, the maxima (or peaks) of the traces obtained by the second or the third flash were usually expressed as the ratio of second to first or third to first peak heights. For quantitative and conclusive purposes the individual peak heights were converted into absorbances and the results expressed as ratio of absorbance from Auxiliary to absorbance in Main. A typical analysis is presented in Appendix 1. The relatively long duration of the photolysis flashes, as shown in Fig 2-4, restricted meaningful analyses for decays to delays of > 40 psec. (b) Spectrographs "Medium" and "Small" Hilger and Watts Quartz prism spectrographs were used to record the spectra in the U.V. and Visible regions. These instruments have relatively high dispersion in the lower wavelength region and hence the calibration errors were small. The spectrographs were calibrated by using a low pressure Hg-lamp. A Bass-Kessler (74) type grating spectrograph was employed for the visible to near infra-red region. A laser was used for optical alignment. A quartz convex lens was used between the specflash and the reaction vessel to produce a parallel beam from the former. By selecting the time on the delay unit the spectra could be recorded at any desired lapse of time after the Main or Auxiliary flashes. IIford HP3 plates and films were used to record absorptions in the region 230 - 500 nm and Kodak moderately fast I.P. films (35 mm) were  _33F.  Figure 2-4 Reproduction of an oscilloscope trace showing the duration of typical Main and Auxiliary flashes. Delay time between the flashes = 82 psec. Ordinate measures the scattered light from the lamps as the detector was deliberately exposed to it. The intensity of the lamp output should not be compared.  -34-  employed for the 500 - 800 nm region. Characteristic curves for both types of films were drawn at different wavelengths.  A Joyce Loebel  Mark 2 Microdensitometer (dual beam), capable of up to x 50 magnification and excellent reproducibility was used for the comparison of plate or film densities. Experimental conditions were so adjusted that the data recorded from the microdensitometer traces were always in the linear region of the characteristic curves drawn. The developing and processing of the films and plates were standardised for the best results and the same procedure was followed throughout the entire study.  -35II.3 SPECIAL METHODS (a) Removal of CO2 from 3D Water Freshly collected 3D water was boiled on a heating mantle for about an hour and then cooled down to the room temperature in an atmosphere of Matheson grade He gas. A portion of this water was used to prepare the stock solution of Ba(0H) which was filtered to remove any 2  precipitated BaC0 . The experimental solution was prepared by diluting 3  this stock solution. As will be seen later (Chapter VI), even traces of C0 in 2  water used to prepare the experimental solution yield misleading results. Hence, the relevant experiments were repeated with the solutions made from the above water sample before drawing significant conclusions. (b) Preparation of Aqueous Xenon Solution Due to the much higher cost of pure Xenon gas, a special method was employed to make aqueous Xe solution rather than direct Xepurging through water. Small amounts of Xe were stored over Hg in an apparatus shown in Figure 2-5. Bulbs A and B, containing clean Hg were connected together through taps hi and B and a piece of tygon tube, C. D is .a B-10 socket x  that could be linked to either the vacuum line or the end of the reaction vessel in the flash photolysis set-up (Fig 2-1). The apparatus was evacuated down to a pressure of ^10~ torr and then filled with Xe through 3  tap Ci until the Hg levels in both bulbs were the same. For saturating any solution with Xe, this apparatus was connected to the reaction vessel by a line thoroughly flushed with He or H . 2  The liquid was left in contact  with Xe for 1-2 hours in which time a fair amount of Xe should be dissolved.  -35F-  Figure 2-5  Apparatus used for storing small amounts of Xe. A and B, glass bulbs f i l l e d with Hg; A , Bj and stODCOcks; x  D, and B-10 socket to connect i t with reaction vessel, V in Fig 2-1.  -36-  II.4 CHEMICAL COMPOUNDS AND GASES The source, grade and purity of the different chemicals and gases used during this work are indicated in Table 2 - 1 . The data was obtained from the publications of the manufacturers. The chemicals were used without further purification unless otherwise indicated. The aqueous solutions were prepared by weighing the right amount of the solids and measuring the required amount of liquids with pipettes. Hydrogen gas was passed through a Fischer Scientific Pdcatalyst deoxygenator to remove traces of Oxygen. Nitrous oxide was purified by bubbling it through concentrated solutions of FeSO^ and NaOH, and 3D water.  -37-  Table 2-1 Source, grade and purity of the chemicals and gases used during these experiments  Chemical Formulae NaOH NaoSO^.lO H 0 2  Manufacturer  Grade  Purity(%)  BDH  AR  98.0  BDH  AR.  99.0  NaClOi+.HjO  Fischer  Purified  Nal  Allied  Reagent  NaN0  3  BDH  AR  98.0  NaC0  3  BDH  AR  99.0  BDH  Reagent  98.0  BDH  AR  98.0  Ba(N0 )  BDH  AR  99.0  Ag S0  BDH  AR  99.0  Alfa Inorganic  Purified  98.0  BDH  AR  99.5  Fischer  Certified  99.0  2  HCOONa Ba(0H)  2  3  2  2  4  CuSO^.SH^ HzSOit (CH)CH0H 2  BDH  ARISTAR  . 98.0  CH0H  BDH  AR  98.0  Fischer  Certified  99.0  3  3  (CH ) C0 3  2  -38T a b l e 2-1  (continued)  Chemical Formulae  Manufacturer  Grade  Purity(%)  Gases He  Matheson  Research  99.9999  H  99.9995  Ar  99.999  Xe  99.995  N0  98.0  NO  98.5  SF  99.8  2  2  6  NH  99.999  0  99.99  3  2  CHAPTER III. EVIDENCE FOR DIELECTRONS IN AQUEOUS SOLUTIONS RESULTS III -1 HYDRATED ELECTRONS FROM THE MAIN FLASH Flash Photolysis of aqueous OH" ions by light of A < ^ 220 nm was used to produce e~ by 3-1. q  3-1  0H^ + q  hv (A  < «\, 220 nm) -2  » OH + e~  q  _2  Many other anions, e.g., SO 4 , I , CO etc. are known to absorb light 3  in U.V. to produce hydrated electrons by the general photolysis process 3-2. 3-2  Y^ + hv ( U . V . ) — > Y + e;  q  The nature of the radicals, Y, is varied (see 1.2). For reasons discussed in Chapter I, the relatively pure chemical system involving reaction 3-1 was found most suitable for these studies. The concentration of e" could be varied by changing the Main flash energy. A typical flash of about 500 joules on a solution of NaOH in the concentration range of 10~ to 4 x 10~ M (at higher concentration, most of the 3  3  light was absorbed near the wall of the reaction vessel, and a nonuniform concentration of e~ resulted) created up to 2 x 10~ M hydrated 6  q  electrons. The 633 nm laser line was used to monitor e~ and the beam aq  passed through the centre of the reaction vessel.  Hydrated electrons  are known to absorb strongly at this wave length with e = 1.2 x 10* M" cm" 1  (7).  1  1  _40_  For reasons that have been mentioned earlier (1.2) and which will be further discussed in this chapter, hydrated electrons were required to react predominantly by reaction 3-3, 2H0 2  aq aq  3 - 3  e  + e  >  M "* -  H  2  + 2 H 0  "  k = 5 x 10 M~ Sec" (13) 9  1  1  3  and not with OH-radicals or impurities. OH-radicals are known to react with e~ by reaction 3-4 at a fast q  3-4  OH + e"  »0H"  aq  = 3 x 10 M" Sec" (13) 10  aq  1  1  rate, hence an OH-radical scavenger was required for long lived e" . Formateions or isopropanol, good OH-radical scavengers, were used in 7  some experiments.  However, saturating the solution with H gas was 2  found most efficient and suitable, for i t converts the OH-radicals into e" by reactions 3-5 and 3-6. 3-5  OH + H —» H + H 0  k = 5 x 10 M" Sec' (13)  3-6  H + 0H^  k = 2 x 10 M' Sec" (13)  2  7  2  »e~  q  1  1  5  7  q  1  1  6  The concentration of H gas at atmospheric pressure is 7 x 10"* M. The 1  2  extent to which hydrated electrons react with OH-radicals or impurities was experimentally indicated by the mean lifetime of the e~ observed. aq  It varied between 200 - 100 ysec depending mainly upon the purity of water, NaOH and the deoxygeneation by the H saturation procedure. There 2  was no complication due to the product of the reaction, 3-7, at this concentration of H . 2  3-7  OH + OH  >H 0 2  2  k = 5 x 10 M" Sec" (13) 9  7  1  1  -41The other reaction that occurs in this system is 3-8,  3  -  8  +  H  —  *  H being the product of reaction 3-5, and this reaction may be significant. It can be seen after consideration of the above reactions that an H saturated NaOH is a very pure solution as there are no net chemical 2  products and hence can be used to study transients that are very reactive. Repeated flashing does not alter the chemical system either. (a) Decay of e~  q  Hydrated electrons thus produced were long lived (half life 200 psec) and this decay was followed at 633 nm. decay by combined first and second order rates.  They were found to Figure 3-1 shows  typical first and second order kinetic plots for the decay of the absorbance at 633 nm due to e'~ following the Main flash.  As indicated  q  in Fig 2-4, the first 60 ysec are complicated due to the tail of the Main flash.  The approximate limiting slope of the 1/A vs time plot gives  a value of k = 6 x 10 M" 9  sec" , where k is defined by equation (i).  1  1  3  3  (1)  -de^/dt - k [ e ; F 3  q  Matheson and Dorfman (9) report a procedure for analysing combined first and second order kinetics.  According to this method, i f k and k^. are g  the second and first order rate constants for disappearance of e~ , q  equation ( i i i ) can be deduced by integrating equation ( i i ) . -da^/dt •  k U'r t  +k;[.^][S]  (11)  200  Time  Figure 3-1  300  (/xsec)  First and second order kinetic plots for the decay of e" following the Main flash. aq  -431/A = e / A i + k /k e^[e f - 1] kf0  k  2  s  9  (iii)  f  where S is the reactant (impurities etc.), k^ = k^[s], A and A are x  2  the absorbances due to e" at times ti and t and e = t - t i . Figure 2  ?  aq  A  3-2 shows a plot of 1/A against 1/A for e = 200 ysec. From the line 2  2  drawn, a slope of 1.67 ± 0,02 and an intercept of 2.53 ± 0.1 (Least Squares fit) was obtained. This yielded the following, ks = 4.9 x 10 M" sec" and k = 2.5 x 10 sec" y  1  1  3  1  f  The initial concentration of e~ produced in this experiment q  was about 2 x 10" M and hence the fraction of e~ decaying by reaction 6  q  3-3 can be calculated. Using equation (ii) and assuming [S] » [e" ], one aq  can obtain equation (iv). ]/d[S] = 1 + k [e"  d[e"  s  (iv)  ]/k  f  This, on integration in the time limit 0  to00,  gives equation (v).  that 60% of e~ reacted by the bimolecular reaction 3-3. If only 10" M 6  q  °f I«aq produced initially, this value would be 45%. A more rigorous calculation was done, without the assumption that [s] >> [e~ ]. The aq e  w e r e  value thus obtained was less than 10% higher and hence the above value is acceptable under experimental error limits.  _44_  Figure 3-2  Matheson and Dorfman (9) treatment of a combined first and second order process. Aj and A are absorbances at times ti and t respectively, e = t - ti = 200 ysec. 2  2  2  -45-  III.1  (b) Identification of the Absorption at 633 nm To show that the absorption signal at 633 nm in the flash  photolysis of 3 mM NaOH solution was due to hydrated electrons, the effects of various specific hydrated electron scavengers and OH-radical scavengers were studied. Furthermore the absorption spectrum of the species was taken and compared with the published spectrum of e~ . (i) Hydrated Electron Scavengers Many hydrated electron scavengers were studied during this work and a few transients were characterised. These transients will be dealt with individually in later chapters of this thesis. In the presence of e"aq scavengers, the absorption signal at 633 nm was found to decay at a faster rate depending upon increasing the concentration of the scavengers. For instance, bubbling the experimental solution with 0 , N 0 and SF for 2  2  6  a progressively increasing period of time caused the 633 nm absorption signal to decay by an increasingly faster first order rate until it could no longer be observed initially.  The specific rate constants for the  reaction 3-9, 3-9  S + e^  > Product(s)  q  where S stands for 0 , N 0 and SF , could not be compared with the published 2  2  6  bimolecular rate constants because the concentration of these gases could not be evaluated quantitatively. The concentration at saturation could be calculated from the known solubility of the gases in water but in all the three cases the half life of e^ in the presence of saturation concenq  tration of these gases is in the range of 10" sec. This is too short to be 9  -46observed on the present detection system. These difficulties were avoided by using sodium nitrate. 3-10  NO3 + e  »Product(s)  aq  The published values for the specific rate constant of reaction 3-10 range from 0.92 to 1.1 x 10  10  M" sec" (13). Figure 3-3 shows a first 1  1  order plot for the decay of the absorption signal at 633 nm in a deoxygenated aqueous alkaline solution (pH = 11.5) containing 10~ M sodium k  formate and 1.1 x 10" M sodium nitrate. During the first 40 psec, the 5  tail of the Main flash had significant intensity to produce enough e" aq  to complicate the decay due to reaction 3-10.  But after this time  period, a good first order decay was observed. The value of k from the slope of the straight line in Fig 3-3 is 1.1 x 10  10  10  obtained  M" sec" 1  1  and is in excellent agreement with the published values. (ii) OH-scavengers The decay of the absorption signal at 633 nm after the Main flash was followed in presence of 2.0 and 4.0 x ~\0~ M 2-propanol and k  3.0 and 5.0 x 10~ M sodium formate. The mean lifetime of the signal 5  was about 20% smaller than that in absence of these OH-scavengers and was attributed to the effect of impurities associated with these chemicals. It was concluded, however, that the OH-radical scavengers had essentially no effect upon the absorption signal. Thus the absorption at 633 nm did not arise from a species having OH as a precursor.  -47-  Fiqure 3-3  First order plot for the decay of e^q in a solution containing 1.1 x 10~ M NaNC^, lO M HCOONa at pH 11.5 5  -4  -48(iii) Absorption Spectrum A vertical  type of specflash lamp was discharged at 81 joules  and the spectra were recorded on a Medium Hilger spectrograph for the 400 - 550 nm region and on a Bass-Kessler spectrograph (74) for the 500 750 nm region, using Ilford HP3 plates and Kodak I.R. films respectively. All the recorded data were corrected for the film or plate 'gamma' obtained from the characteristic curves drawn for each wavelength. Finally, the points were normalised by some arbitrary factor to compare them with the published spectrum (solid line in Fig 3-4). Figure 3-4 shows the absorption spectrum of a transient produced in the Main flash from an H saturated 3 mM NaOH solution obtained 70 2  ysec after the flash. The solid line represents the published spectrum °f e;  q  (7).  The error limit in the wavelength of these data is + 5 ran. The absorbance values had higher errors at shorter wavelength because of the lower density and the variation in the specflash light output. This probably is the reason for a relatively poor agreement in this region.  1  1  400 Figure 3-4  1  500 X(nm)600  1  700  1  S00  Absorption spectrum of the transient produced in U.V. photolysis of OH"/Hi solution. O-absorbance observed 70 ysec after the Main flash. Solid line is the published absorption spectrum of e" (7).  -50•III.2 HYDRATED ELECTRONS FROM AUXILIARY FLASH A long lived (observable up to <v 20 msec) transient species was observed in double and triple flash photolysis of H saturated aqueous 2  alkaline solutions, which produced hydrated electrons upon photolysis with light of restricted wavelength. When the Auxiliary lamp was discharged (at ^ 800 joules) after most of the hydrated electrons produced during the Main flash had decayed, a second absorption signal at 633 nm was observed. Pyrex filters (0-53) were used between the Auxiliary lamp and the reaction vessel. The decay of this absorption peak (or the second peak) was similar to that shown in Fig 3-1.  The effects of e~ and OH radical scavengers upon the q  second peak were similar to that on the first peak. An absorption spectrum similar to that of e" was obtained (III.2 (b)). Hence the transient aq  absorbing at 633 nm during the second peak was attributed to e~ . q  Figure 3-5 shows results from a typical set of double flash photolysis experiments, (a), (b) and (c) show the second peaks at different delay times between the two flashes. No such absorption peak was obtained if the Main flash did not precede the Auxiliary (Fig 3-5 d). The height of the second peak decreased with increasing delay times after ^ 500 - 600 ysec. At delays greater than 20 msec, the second peak height was negligible. A maximum auxiliary signal was obtained in the 500 - 600 ysec region, indicating that species responsible for the reproduction of e~aq in the second peak was formed and decayed during the lifetime of e" .  -51-  (O 50  ( d ) 50-  Abs  1 msec  Figure 3-5  200  /xsec  Reproduction of oscilloscope traces showing the absorption at 633 nm as a function of time in an 0H~/H solution. The arrows indicate when the Auxiliary flash occurred. In (d) Main lamp was not fired. 2  -52Light of X > 260 nm only was photolysing the solution during the Auxiliary flash in the experiments of Fig 3-5. The vertical arrows indicate the time at which the Auxiliary lamp was fired.  (In this text the species  responsible for the production of the second peak will be referred to as ' X ' ) .  The largest second peak obtained during these experiments was  20% of the absorbance due to e~ from the Main flash. Production of this aq peak required the formation and decay of e" from the Main flash. In aq the absence of e" from the Main flash, n o second absorption peak was aq observed. Also, the second peak height depended upon the concentration of e ^ produced during the Main flash and on its lifetime. The largest second peaks were obtained from solutions in which the mean lifetime of e^q was > 200 ysec. When the lifetime of e~ with respect to the q  first order component was less than 100 ysec, no auxiliary absorption peak could be produced. Due to the relatively small size of the second peaks, a detailed study of the second peak height dependence upon the delay time was not possible, however, it depended markedly upon the lifetime of e~ in the particular solution being studied. Hence, a very meaningful lifetime for X cannot be quoted. In a solution where great care was taken to avoid any impurities and where e~ decayed with mean q  lifetime > 200 ysec, this was about 5 msec. III.2 (a) Filter Experiments The variation of the second peak height upon changing the filters between the Auxiliary lamp and reaction vessel was studied and the results are reported in Table 3-2. Table 3-1 shows the transmittance at different wavelengths for various Corning glass filters used during these experiments. The  -53numbers have been obtained from the curves in Fig 2-3, assuming the maximum transmittance of each filter as a 100% transmission. Table 3-1 shows the corresponding values for these filters also. In the case of 7-54, the data for only the relevant region of the spectrum are indicated. The numbers in brackets represent the long wavelength limit bf the band. The maximum transmittance at about 320 nm is an absolute value of 89% transmission. Double flash photolysis experiments (at the same delay) were repeated with various filters and the results were expressed as the ratio of absorbance due to Auxiliary to absorbance in Main (^/\ /^ - ). ux  al  n  The brackets after the filter number indicate where 1 or 2 layers of the filter were used (Table 3-2). From the results with 3-74 and 3-70 filters, where negligible Auxiliary peaks were obtained, the long wavelength limit for the photoionisation of X can be put at 400 nm.  There was only about 10% decrease  in the ratio by changing the filters from 9-54 to 7-54, hence light of X < 240 nm was not very significant. A decrease of about 20% by using 0-53 instead of 9-54 indicates that the absorption maximum for the process either must be at X > 300 nm, as 0-53 transmits only about 50% of the light at this wavelength, or there is enough light in the flash to cause this effect. This also indicates that the photodissociation of X to produce e~ is not caused by a narrow band in the region 260 q  320 nm.  It seems that the band responsible for the production of the  second peak is mostly in the region 320 - 400 nm but may extend down to about 260 nm with significantly lower absorption coefficient in the latter region. However, if the intensity of the light flash is so high  Table 3-1 Transmission characteristics (relative to maximum) of Corning glass filters used between Auxiliary lamp and reaction vessel  Filter Number  Wavelength (nm) of Transmission 3%  10%  50%  90%  97%  . 9-54  216  223  240  288  300  7-54  231(411)  236(401)  253(390)  283(367)  298(353)  0-53  279  290  303  334  344  0-52  339  342  350  380  389  3-74  385  390  415  461  497  3-70  486  493  507  548  573  +  * Error limit = ± 3 nm +  7-54 has a transmission band in the relevant region of the spectrum. The wavelengths in brackets indicate the long wavelength limits of this band. The maximum transmission at ^ 320 nm is an absolute transmission of 89%.  Table 3-2 Ratio of Absorbance due to Auxiliary/Absorbance due to Main flashes with different Corning glass filters between the Auxiliary lamp and the Reaction vessel (for the same delay between the flashes)  Filter Number  R a t i 0  Aux Main  A  /A  9-54  o.n  7-54  0.103  0-53(1)*  0.087  0-53(2)  0.070  0-52(1)  0.035  0-52(2)  0.025  3-74  < 0.01  3-70  < o.oi  ;  The error limits for the ratio values are ± 0.01. * Numbers in brackets represent the number of layers of the filter used.  -56that even a small % transmittance of the filters permits sufficient light, the whole region 260 - 400 nm must be taken into account. Basco, Kenney and Walker (36) were misled to conclude that the wavelength region responsible for the photodissociation of X was > 700 nm on the basis of the results obtained with 7-54 and 3-70 filters. The present results can explain the finding in the case of 7-54 which has an effective band in the U.V.  However, the second peak obtained  with 3-70 cannot be explained except by concluding that some stray U.V. light must have by-passed that filter in those experiments. The present apparatus was especially made for.these experiments and no such error is likely. The present filter data are also not inconsistent with Czapski and Peled's (41) finding that no absorption in the I.R. region could be observed in the pulse radiolysis of a similar solution. It will be shown later that even if the probe were carried out in the correct region (320-400 nm) no X should have been observed because of the noted dependence of X upon the longevity of e~ . aq So far as the wavelength region is concerned these results are in reasonable accord with results of,Gopinathan, Hart and Schmidt (37) discussed earlier (1.2). It is quite likely that the same intermediates were involved in both these experiments. III.2 (b) Absorption Spectra The absorption spectrum of the species absorbing at 633 nm in the second absorption peak v/as taken 55 ysec after the Auxiliary flash. Figure 3-6 shows the absorbance in the region 500 to 850 nm, •  -57" (i) 55 ysec after the Auxiliary flash (with 0-53 filters between the Auxiliary lamp and reaction vessel) which was fired 600 ysec after the Main flash, (ii) 630 ysec after the Main flash alone when most of e" aq in the first peak have decayed (no Auxiliary). These results were recorded on a Kodak fast I.R. film. The solid line in Fig 3-6 again is the published spectrum of e~ (7), normalised appropriately. The aq vertical bars on the absorbances represent the estimated error limits for these measurements. Only the absorbance in the 500 - 850 nm region is reported here because in the region 300 - 500 nm, no absorption above the variation in the specflash lamp intensities could be observed. Hence, a lower limit of 0.1 was placed on the sensitivity for detection of absorbance in this region. (This value was obtained from the actual plate densities obtained from the fluctuations in the light output of the spectroflash lamp converted into absorbances in a similar way as above.) An attempt to observe absorbance due to X in the 320 - 400 nm region was also unsuccessful because of the same reasons.  However, it can be stated  that for X, A < 0.1 in this wavelength region. It is clear from Figure 3-6 that the absorbing species at 633 nm which gives the second absorption peak by photolysis of X is the hydrated electron. The experimental data are in reasonable agreement with the published spectrum of e'^. III.2 (c) Triple Flash Experiments The maximum amount of hydrated electrons regenerated by the Auxiliary flash was only 20% of the total e" produced by the Main flash.  -58-  Figure 3-6  Absorption spectra of an 0H"/H solution, (i) 55 sec after the Auxiliary flash (A > 260 nm) which was fired 600 ysec after the Main flash, (ii) 630 ysec after the Main flash (no Auxiliary). 2  v  -59Amongst the possible reasons are the following: (i) a relatively small extinction coefficient of X, (ii) alow flash intensity and/or (iii) formation of X by only 20% of the initially generated e" in the aq Main flash. Triple flash photolysis experiments were carried out by discharging the Auxiliary lamp twice (see Chapter II) with 100 to 400 ysec delay. Figure 3-7 (a) shows the oscilloscope trace for absorption due to e"aq at 633 nm in such an experiment without any filters between the Auxiliary lamp and reaction vessel. This represents a comparison of the intensities in the three flashes. Figure 3-7 (b) shows the results from a similar triple flash experiment with 0-53 filters between the Auxiliary lamp and reaction vessel. The vertical arrows indicate the delay times at which the Auxiliary was flashed. It is evident that < 20% as much e'^ were regenerated by the third flash compared to the second. When this is compared with the relative heights of the second peak obtained by discharging the Auxiliary lamp after 530 and 870 ysec of the Main flash, it is established that the comparatively smaller third peak is not due to the normal decay of X. Further confirmation of this fact is obtained from Figure 3-7 (c), where the Auxiliary was fired at 3.0 kV, 530 ysec after the Main flash and then again fired at 7.0 kV, 110 ysec later. The electrical energy output of a flash is given by 1/2 cV , where c is the capacitance and V is the charging voltage. Hence, 2  the energy of third flash was a factor of 5.5 (obtained from (7/3) ) higher 2  than that of the second flash.  It is observed that the absorption produced  by the first Auxiliary flash is barely detectable in this experiment.  -60-  Figure 3-7  Reproduction of oscilloscope traces showing the absorption at 633 nm in the triple flash photolysis experiments on an OH'/Ho solution:  (a) no filter in F (Fig 2-1); (b) filter 0-53 in F. Arrows indicate the times at which the Auxiliary lamp was fired (fired at 7 kV both times); (c) as in (b) except that the Auxiliary lamp was fired at 3 kV the first time and at 7 kV 110 ysec later.  -61These results indicate that > 80% of the X formed is normally photodissociated by an Auxiliary flash at 7.0 kV. This then implies that only 25% of the initially produced e" have formed X and can be observed aq  in this system after 600 ysec. Hence, in a typical experiment, the maximum concentration of X after 600 ysec must be only about 2 x 10~ M. 7  III.2 (d) Absorption Coefficient If the approximate concentration of X formed is 2 x 10~ M 7  (see preceeding section) and the upper limit for the absorbance has been established to be 0.1 (section 2 b), then an upper limit on the absorption coefficient of X can be placed at 1.2 x }0 M'krrf . k  1  This value is over  the range where X is photodissociated to give e" as well as the visible aq region of the spectrum. A lower limit of the absorption coefficient for X may also be estimated as follows. From the Lambert-Beer's law we have, log  IQ/IJ.  =  2.3 zi [ X ] , where I and I are the incident and transmitted light, e Q  t  the mean extinction coefficient over the 320 - 400 nm region and l the path length. As the absorbance of the solution is very small, log L/I+ o r . = -log(l-I a/I.) o £ IaJ l o, where I a is the intensity of light n  absorbed by X . Hence, the fraction of the incident light absorbed by X following the Auxiliary flash may be approximated by, I /I = 2.3 A[X] a  0  £  (vi)  In this laboratory, on a similar set up used for gas phase flash photolysis, the extent of photolysis of gaseous bromine at 1 torr pressure was measured by Thomas (75). From the known absorption spectrum of bromine,  -62it could be calculated that 2.3 a I U o  x 10~ einstein cm nm". 6  If  1  the same light output is assumed in the 320 - 400 nm region and the efficiency of the present apparatus is considered the same, E for X in this region can be calculated from (vi). The whole range is considered to be 80 nm, hence e = 2 x 10 M"cm" 3  1  1  where it is assumed that I £ [X], a  as ~ 80% of X is photolysed by the Auxiliary flash. The distance between the Auxiliary lamp and the reaction vessel is greater in this set up than the one used to obtain the data on bromine. Hence, this is probably the lower limit for the absorption coefficient. Therefore, in the wavelength region responsible for the photodissociation of X, the mean extinction coefficient is ^ 2000 fOcrrf , but does not exceed 12,000 M" 1  cm" at any X.  1  In a related experiment on a similar solution and using  1  the same technique to produce e" but using a Q-switched Ruby laser aq to monitor both e~aq and X, an absorption of the 347 nm line was observed, Hydrated electrons also absorb at this wavelength, hence the absorbance, experimentally observed, was compared with the calculated absorbance due to e"^ at 347 nm up to about 600 ysec after the Main flash.  Due  to much scatter in the data points, no kinetic data on X could be obtained but one positive conclusion was made. Assuming 20% conversion of e~ to aq  X, the extinction coefficient of X at 347 nm was calculated to be ^ 6000 M" cm". 1  1  This is consistent with the present findings.  III.2 (e) Effect of Added Solutes The double flash experiments were repeated on different chemical systems and the results were compared to that obtained from Na0H/H  2  system. The delay between the two flashes was 800 ysec. Table 3-3  -63-  shows the effect of added solutes on the ratio (absorbance due to Auxiliary^absorbance due to Main flash) at 633 nm. It is evident that virtually no effect was observed on the height of the second peak either (a) by flooding the solution with Na  +  ions (using 1 M NaCl0^ solutions) or (b) by excluding Na ions by the +  use of  NHijOH  or  Ba(0H) 2  solutions. The relatively smaller peaks are  probably due to the shorter half life of e" from the Main flash as aq the ratio A / A j depends strongly upon the lifetime of e'^. These results were used in deciding whether the (Na*e~ ) complex suggested aq aux  ma  n  +  by Gopinathan et al ( 3 7 ) is species X (see discussion of these results later in this Chapter). Most of the experiments reported here involve a solution saturated with H 2 to scavenge OH-radicals. At this concentration of H  2  (7 x 1 0 _ t f M), the mean life times of OH and H according to reactions  3-5 and 3-6 are about 30 ysec each. Some experiments (results in Table 3-3) with formate ions and isopropanol at concentrations used to give the OH-radicals a lifetime of about 1 ysec were performed to indicate the role of H and OH-radicals in the formation and decay of X.  (The  small lifetime of OH precludes the formation of H atoms by reaction 3-5.)  In these systems hydrated electrons decayed at a somewhat faster  rate which is probably due to the impurities associated with these chemicals. The second absorption peak heights, however, were comparable to the results obtained fromanOH"/H system with comparable lifetime of 2  aq*  e  Si il m  a r  filter data and dependence upon delay times were also  observed. In Chapter VII, these solutions have been shown to yield additional effects. Nevertheless, the results reported here clearly \  -64" demonstrate that X is not destroyed nor is its formation inhibited by these solutes, i.e., X is not a strong oxidising agent nor is it formed from OH-radicals. Results with 0 and H 0 strongly confirm the earlier findings 2  2  2  that X would not be formed if the lifetime of e~ were curtailed by adding some e"aq scavengers. This also indicates that the second peak is not obtained by the photolysis of species like H 0 or H 0 itself, 2  2  2  2  which may be formed by reactions 3-7 and 3-12. 3-12  H 0 + e; 2  2  >H 0  q  2  (8)  2a q  In Chapter VI (inorganic gases), a transient species identified as the CO3 radical anion is reported to absorb strongly at 633 nm in the flash photolysis of a N 0 saturated alkaline solution. This has 2  been shown to be due to the presence (non-removable by normal means) of C0 in the 3D water. Simply by using Ba(0H) solution at a concentration 2  2  of 2 x 10" M, saturated with H does not exclude the possibility that 3  2  X is CO3 radical ions. The solubility product of BaC0 is 5 x 10~ , 9  3  _2  hence it can be calculated that the maximum concentration of CO ions, 3  in a solution containing 2 x 10~ M Ba 3  +2  ions, will be 2.5 x 10~ M in 6  the dissolved state. This is a high enough concentration to produce about 2 x 10" M CO3 radical ions (if X is C0 radical ion). For this 7  3  reason, in one set of experiments most of the C0 was removed from the 2  3D water by a method described in Chapter II and this water was used to make up a 2 x 10~ M Ba(0H) solution. After saturating this solution 3  2  with H , similar results were obtained in double flash photolysis experi2  ments. Fig 3-8 shows a typical oscilloscope trace for the absorbance  -65due to e~ produced by both Main and Auxiliary (with 0-53 filter) q  flashes at 633 nm, the latter being fired 650 ysec after the Main. This not only excluded the possibility of X being CO3 radical ions, but also confirms that X is not a product of the reaction 3-11. co 2  3.11  +  e  ;q  —>  [co2];q  The intermediate C0 has been reported to have a prominent absorption 2  band in the relevant region (33). The reported filter data (III.2 (a)) also would be inconsistent with C0 being X. 2  In one set of experiments, 3 x 10  M 2-propanol and 3 x 10"  _lf  3  M of deoxygenated sodium sulfate solution was used to produce e" . The alcohol was added to scavenge the SOij radical anions produced by reaction 3-2.  The experimental half life of hydrated electron in this •  case was about 160 ysec and second absorption peaks were obtained which were comparable to the Ma0H/H solutions. 2  Comparable results were also obtained when the water, used to prepare these solutions, was obtained from a different source. The ionic impurities in this water supply were very small (specific conductivity of the sample = 1.8 x 10 mhos). This indicates that such impurities, 8  if associated with the 3D water, are not responsible for the effects observed.  -66-  100  -M 14-f 100 JJS  Fiqure 3-8  Reproduction of an oscilloscope trace obtained in an experiment similar to one shown in Fig 3-5 (a) except that the solution was made in C0 -free 3D water. 2  Table 3-3 Effect of Added Solutes on the Second Peak Heights  Chemical  Concentration  Experimental Half life of e* aq  *Ratio A /A . aux mam  NaOH/H only  3 x 10" M  200 ysec  0.10  Excess Na as NaClO^  10" - 1 M  150 ysec  0.08  NH4OH/H2  pH = 10.4  120 ysec  0.06  Ba(0H) /H  2 x 10" M  200 ysec  0.10  H0  10' - 10" M  20-0 ysec  < 0.001  NaOH, HCOONa  3 x 10" M HCOO" pH = 11.2  170 ysec  0.09  NaOH 2-propanol  3 x 10 M pH = 11.4  160 ysec  0.08  NaOH/0  pH = 11.5  20 ysec  < 0.001  6  pH = 11.5  10 ysec  < 0.001  NaOH/NO  pH = 11.5  20 ysec  < 0.001  Na S0 2-propanol  3 x 10" M (both)  160 ysec  0.08  2  +  2  2  2  3  3  5  2  a  2  NaOH/SF  a  b  2  2  3  1+  3  5  -1+  3  * Error limits = ± 0.02 a - 1 minute of purging with the gas b - Only about 15 sec df contact with N 0 2  _68. DISCUSSION X was postulated previously as a hydrated electron dimer formed in the reaction 3-13, 3-13  eaq + eaq  aq  mainly on the basis of the wavelength region responsible for its photolysis to produce hydrated electrons again (36). 3-14  (e ") 2  aq  + hv  » 2e aq  Due to some experimental oversight (see III.2 (a)), the wavelength region was assigned to X > 700 nm which also seemed to support the view that X is a dielectron (will be designated by e ), by comparison with 2  some dielectron species in solids (38,39). The bands of such species are shifted to the red relative to the monomeric species. The present results indicate the near U.V. region to be responsible for the photolysis of X. Does this still support the identification of X as e ? Some recent calculations by Fueki (76) on trapped 2  dielectron species predict only a very small blue-shift for X in water, compared to e~ . q  max  of e  2  However, if one considers the thermodynamics  of reaction 3-13, about 3.5 eV are required for the regeneration of e~q from X, which is about twice the solvation energy of e~ (1.7 eV) q  (77). This value of 1.7 eV also corresponds to the X , (720 nm) for max r  the hydrated electron. These values are consistent with reaction 3-13 which will be governed predominantly by the difference in the heat of solvation of the two species involved. Therefore, one may consider 3.5 eV reasonable  -69as the solvation energy of e formed by reaction 3-13 and for its photo2  dissociation. There is no evidence to suggest that the 320 - 400 nm region is the only range where X absorbs as e may be expected to have a much 2  broader band involving transitions at longer v/avelengths.  The near U.V.  region only represents the wavelength region responsible for the photoionisation of X. The results from 9-54 and 7-54 filters indicate that the high energy tail may extend to about 260 nm. On the basis of the above argument X may be e but there is 2  no direct proof of this assignment. The kinetic data are consistent with e as well as products formed by reaction of e with the cations 2  2  present in the solution.  3-15  M" (.f-J +  +  » M<5->  2 +  The fact that X can be observed in systems containing Na , NH^ or Ba  2  ions suggests that if X is the product of reaction 3-15, where one of the above cations are involved, the absorption spectra of all these products have to be similar. This is conceivable if the intermediates are a complex of type (e~ "M-e" ). Absorption spectra of similar interag aq mediates, e.g., Na", K~ and Cs~ have been reported (78) in THF, formed by the association of  with two solvated electrons (in two steps) or  by the disproportionation of the neutral metal atoms, M°. The x S  m3V  ilia X  values depend upon the nature of the cation to a great extent, being the largest for Na . s  This is inconsistent with X being a similar species.  However, it is conceivable that approximately 10~M Na ions were present 5  +  in all solutions, dissolved in the 3D water from the pyrex vessels used  -70during distillations.  If such is true, X may be attributed to Na" but aq  the results in the presence of excess Na ions imply that they must be +  formed by reaction 3-16, 3-16  -) > ' (e "), 'aq  Na > Na ' ""aq  2  e'aq + e~„ "aq  4 - 2  an  2  +  0  with the first step being rate-determining.  It is evident from the results  using different cations and excess Na ions that X cannot be (Na ' q) +  +  e  a  complex suggested by Gopinathan et al (37). (From the filter experiments it seems very possible that the same species is involved in their system.) Moreover, using equation (v) (Section 1 (a)), where A y is the fraction of e~ reacted with Na ions, it can be calculated that A y < 0.1 at the +  q  concentration of Na normally used. The rate constant of the reaction +  3-17, 3-17  Na + e"  > Product  +  aq  has been reported (13), as k  < 10 M" sec" . Hence, 20% regeneration 5  17  1  1  of e~ by the Auxiliary flash (under the most favorable conditions) q  cannot be explained. k  It may be emphasised here that the quoted value of  is only an upper limit.  17  X may also be attributed to H~ since it is the conjugate acid aq of e as given by equilibrium 2  3  "  2_,  1 8  (ef)  aq  + H * — = ± H  aq«—  If such is the case, H~ 3-19  (eh  a  q  aq  has to be formed by reaction 3-19,  + H0 2  > H;  q  +  0H;  q  -71-  because the pH of the experimental solutions was only about 11.  H~  q  could also be formed by reaction 3-8, H-atoms being produced by reaction 3-5.  It can be calculated that 10-20% of e~ react with H-atoms. The aq  comparable amount of e~ regenerated by the Auxiliary flash in presence q  of OH-scavengers like 2-propanol and formate ions which exclude reaction 3-5 and hence 3-8, indicate that if H~ is formed in both reactions 3-8 q  and 3-19, then X cannot be H~ . Unless reactions 3-5 and 3-6 occur in aq one step represented by 3-20. 3-20 H + 0" > e" aq 2  Results with H atom scavengers (e.g., OH", 2-propanol and HCOO" ions) also exclude the possibility of X being a product of reaction 3-21, if the H atom is a product of this reaction as suggested (79). 3-21  e~  + H0 2  > ?  k  = 16 IM" sec" 1  21  1  (79)  The kinetic data are consistent with the above reaction and 10-20% of e"  initially produced may undergo this reaction if the value of k  21  is correct. Long lived triplet states of H0 or OH" which may be produced 2  during the Main flash are not responsible for the production of e~ by aq the Auxiliary flash. This can be established on the basis of the following: (i)  The maximum second peak is not obtained at the earliest delay time  at which meaningful measurements could be made. (ii) There are no first or second absorption peaks observed if pure 3D water is flashed.  -722_  (iii) S0  solutions show comparable results.  4  (iv) Small concentrations of specific e~ scavengers eliminate the formation of a second absorption peak, indicating e~ is a necessary aq precursor. The possibility that X is a transient formed in the reaction of e~aq with a trace impurity cannot, however, be totally excluded.  But  for this to be the case, such impurity has to be present in the system at the same concentration throughout this entire study, during which a variety of modifications have been made to the system. One or more of the following reasons may be given for the low maximum second absorption peak: (i)  This constitutes the yield of e (or Na~ or H~ ). 2  q  (ii) Only a fraction of e survive during reaction of e~ following the 2  q  Main flash. (iii) e may have alternative decomposition paths and an absorption spec2  trum spread over the whole visible region, hence only part of the light absorbed causes the photodissociation. (Note that all wavelengths greater than 400 nm were also illuminating the solution except when 9-54 filters were used.) (iv) e is formed both in singlet and triplet states and only one of 2  these is photodissociated by the light used. The effect of e~ scavengers upon X is not clear. It may be q  an effect on the formation of X but at the same time be quite reactive towards this species. The reactivity of the transient with di-or trivalent metal ions which are capable of accomodating two electrons could  -73not be tested due to the slight solubility of the metal hydroxides. Also due to the formation of the hyper-reduced states of these ions (see Chapter IV) such studies would not be possible. Finally, the two basic conditions to observe the photogeneration of e"aq from X that must be fulfilled are: (i) a considerable proportion of e~  should decay by the bimolecular  reaction 3-13 and (ii) the mean lifetime of e~ with respect to all other reactions must q  be quite long, say > 100 ysec. This indicates that if X is formed by intraspur reactions in pulse radiolysis experiments, it will soon be destroyed by products such as OH-radicals, H 0 ions and H 0 and by +  3  2  2  reducible solutes. Hence the lifetime of X would be quite short in a radiolysis system and it would be difficult to detect it.  It seems  that flash photolysis of 0H"/H solutions is most favorable for observing 2  such a species. At this point it is understandable why Czapski and Peled (41) were unable to observe dielectrons in a Pulse Radiolysis. System. The reasons being twofold, the misconception about the wavelength region and the fact that in 200 ysec hydrated electrons had decayed through about 14 half lives in their system. Considering the long life of X in the absence of e" scavengers, it may be said that X plays aq an important role in any physical or chemical process in which water is reduced to hydrogen gas.  CHAPTER IV FLASH PHOTOLYSIS OF AQUEOUS Ag° AND Tl° Various hyper-reduced states of metal ions produced in the pulse radiolysis of aqueous solutions containing these ions have been identified and shov/n to have strong absorption bands in the U.V.' region (see Chapter 1.3).  These bands are usually fairly narrow and have high extinction  coefficients, indicating that they may be CTTS bands (46). The following results were obtained during the study of aqueous Ag(I) and T1(I) solutions by flash photolysis. The experimental set up was essentially the same as described in Chapter II. IV.1  RESULTS FROM Aq(I) SOLUTIONS <  Preliminary results were obtained in the flash photolysis of 3 x 10" M sodium hydroxide solution deoxygenated with He in the 3  presence of 5 x 10~ M Ag ions. A transient species was observed 6  +  which produced hydrated electrons in a double flash photolysis experiments with restricted wavelength of light. As discussed in Chapter III, hydrated electrons were produced in the photolysis step 4-1. 4-1  OH" + hv (x < % 220 nm) aq  *> OH + e" an  v  In the presence of Ag ions these decayed mainly by reaction 4-2. +  4-2  Ag + e^ +  q  * Ag°  k = 3.2 x 10 M" sec" (52) 10  1  1  2  Due to the relatively high value of k and the long duration of the Main 2  flash, kinetic analysis for the decay of e~ by reaction 4-2 was complicated.  -75" Figure 2-4 (Chapter II) shows the relatively high intensity of the flash at 40 ysec. By this time the absorption due to e~ (at 633 nm) in the aq above experiments had decayed to less than 5% at 5 x 10" M Ag . Hence G  +  these studies were also carried out in various other systems. HYDRATED ELECTRONS FROM THE AUXILIARY FLASH (a) 0H"/Aq Solutions +  Auxiliary absorption peaks at 633 nm due to hydrated electrons could be obtained by firing the Auxiliary flash lamp after a delay of 50 to 250 ysec following the Main flash.  In these experiments, various  Corning glass filters eliminated the possibility of e" being produced aq in reaction 4-1 by the Auxiliary flash. By changing the filters from 0-53 to 0-52 the second peak was reduced by a factor of 2. 3-74 reduced it down to < 10% and 3-70 practically eliminated it. The second peak heights were negligible after delay times > 200 ysec and increased progressively with increasing energy of the Auxiliary flash. The solubility product of AgOH is 2 x 10' , thus at pH 11, the 8  maximum concentration of Ag ions in solution was only 6.7 x 10~ M. +  6  Therefore, these experiments were repeated and studied in detail in solutions at natural pH, using S0^ as the source of e~ from the Main q  flash. (b) Ag2S0it Solutions Deoxygenated lO'^Msilver sulfate solution produced less than 10" M hydrated electrons by a Main flash 8  energy up to 700 joules.  .76. (It should be noted here that meaningful observations cannot be made for concentration of e~ < 10" M using the author's detection system.) This aq 8  is in accord with expectations based on calculation of the half life of aq (c)  e  i n  ttlls  s  Na2SQtt  y (\ ^ ^ )* and Ag2S(h Solutions s t e m  0 , 2  s  In Ag S0t /Na S0 solutions the short wavelength light 2  f  2  tt  (x < ^ 220 nm) of the Main flash (no filter between it and the reaction vessel) was absorbed by SOl^ according to reaction 4-3,  4-3  SO^ + hv  (X < * 220 nm)  »  + S0^  to generate hydrated electrons (80,81). Strong absorption of the 633 nm laser light by e" occurs during the Main flash, as shown in Fig 4-1, aq e"aq being the only species in these systems absorbing significantly at this wavelength. This reacts rapidly with Ag . +  With a 0-53 filter  between the Auxiliary lamp and reaction vessel an Auxiliary absorption peak such as that shown in Fig 4-1 was obtained. This Figure shows traces obtained for 35 and 95 ysec delays between Main and Auxiliary flashes. At very short delay times the second absorption peak was comparable in size to the first peak showing that most of the hydrated electrons originally formed were regenerated by the second flash containing only X > 270 nm. The magnitude of the second absorption peak decreased with increasing delay time between the firing of the Main and Auxiliary lamps. Fig 4-2 shows these data for a solution containing Ag at 5 x 10" M, +  6  the ordinate being proportional to the concentration of the intermediate  (a)  (b)  H ht  H K  IO/xsec  20/xsec  Time  Figure 4-1  t  Time  Oscilloscope traces showing absorption by e" at 633 nm during Main and Auxiliary flashes +  a  "  in a 10 M solution of Ag in 10 M Na S0i,. A 0 - 5 3 filter was included between the Auxiliary lamp and the reaction vessel. The arrow indicates the triggering pulse for the Auxiliary lamp, (a) shows a time delay of 35 ysec and (b) 95 ysec between the Main and Auxiliary flashes. 5  3  2  .78.. at various times compensated for small variations in the Main flash intensity as registered by the hydrated electron concentration in the Main absorption peak. The decay of the transient species was approximately first order, as indicated by the inset of Fig 4-2, with a half life of ^ 42 ysec. Ag° produced in reaction 4-2 has been reported to decay largely by reaction 4-4. 4-4  Ag° + Ag  +  »Ag  k = 5.9 x 10 M" sec" (47) 9  2  1  1  k  On adding an additional layer of the 0-53 filter the absorption in the Auxiliary peak was reduced significantly for a given delay time. On replacing the 0-53 filter by a 0-52 filter the Auxiliary absorption peak was reduced to about half that found for the same delay with 0-53. On further exchanging the filter to 3-74 the absorption was reduced to ^ 25% that found with 0-53 whereas 3-70 eliminated the Auxiliary absorption peak. These data are given in Table 4-1 for a 5 x 10" M Ag solution. 6  +  These filter experiments indicate that: (a) almost all e~ may be regenerated by an Auxiliary flash containing q  X > 270 nm at short times, so the absorption is substantial at X > 270; (b) a considerable fraction of the absorption band leading to photoregeneration of e" aq lies in the range 270 to 320 nm; (c) the band responsible is fairly broad and certainly extends to ^ 400 nm. Absorption spectra were taken at various times after either flash using light from an 80 J spectroflash directed down the length of the reaction vessel and focussed onto the slit of a small Hilger and  0.300.2540 0.20-  80 t((JS)  120  O.I5-  O.IO-  0.05-  i  20 Figure 4-2  i  i  40  1  60  80  1  IOO  D E L A Y TIME (JJ sec)  1  '  120  Plot showing the decay of the transient as a function of time in a 5 x 10~ M Ag solution. The ordinate is the ratio of absorbances due to e~ during the peak of the Auxiliary flash relative to the Main flash. The abcissor represents the time delay between the flashes. Inset shows the same information plotted according to first order rate low. (The particularly large uncertainty in the first data point arises mainly from the existence of significant absorption still be e" from the e  q  aq  Main flash.)  -80Table 4-1 Ratio of peak absorbances (due to e" ) in Auxiliary/Main flashes for 5 x 10"M Ag solution with various light filters between the Auxiliary 6  +  lamp and the reaction vessel for the dame delay time between flashes (60 ysec).  Filter #  Aaux/Amain  9-54  0.19*  7-54  0.15  0-53  0.11  0-52  0.06  3-74  0.03  3-70  < 0.01  * Error limit = ± 0.02  -81and Watts quartz prism spectrograph containing 35 mm HP4 photographic film.  Immediately after the Main flash the absorption spectrum  corresponded quite closely to that of e" , as shown previously (Chapter aq  III). (i)  Fig 4-3 shows the spectrum obtained for three conditions: 35 ysec after the Main flash, when ^ 90% of e~ had reacted with q  Ag present at ^ 10" M; +  5  (ii) 13 ysec after the Auxiliary lamp, which was fired 38 ysec after the Main. In this experiment a 0-53 filter was present between the Auxiliary lamp and the reaction vessel; and (iii) 30 ysec later, when e~ had again reacted with Ag . aq (d) OH-radical Scavengers +  In the above solutions, several species including Ag°, Ag  2  and Ag  2+  are thought to arise from the following reactions  4-2  Ag + e"  > Ag°  +  aq  4-3  Ag° + Ag  * Ag  4-5  Ag + OH  » Ag  +  2  +  2+  + OH"  Any of these could be responsible for the second absorption peak. (Even in a solution of S0^ alone, OH-radicals may arise from S0^ radical anions.) However, since Ag  2+  arises from reaction of OH-radicals it may  be eliminated by the use of OH scavengers. (i) Formate ions Two mixtures were used, containing 3.0 x 10" M NaOH, 5.1 x 3  10" M Ag S0 and 10' M HCOONa in one and 1.1 x 10" M Ag S0 and 3.0 x 6  3  2  4  10' M HCOONa in the other. 3  5  2  l+  -82-  250  300  350  400  X(nm) Figure 4-3  Absorption spectra in the wavelength range 250 - 425 nm for a 10" M Ag solution: (i) 35 usee after the Main flash (0); (ii) 13 usee after firing the Auxiliary flash (i.e., during the flash) which was fired 38 ysec after the Main flash («); (iii) 30 usee after the Auxiliary flash which was fired 38 usee after the Main flash (A). A 0-53 filter was placed between the Auxiliary lamp and the reaction vessel. The dashed line is the published spectrum of Ag° (46). 5  +  450  -83Comparable traces to those shown in Fig 4-1, were obtained from the first solution, indicating that OH radicals were not involved in the production of the species yielding e~ during the Auxiliary flash. Therefore, Ag  2+  q  was excluded. Some additional investigations carried out during this work indicate that flash photolysis of HCOO" ions with light of X < ^ 220 nm produce hydrated electrons (Chapter VII). HCOO" + hv aq  > HCOO + e" aq  The yield is smaller than in the case of the usual 0H~/H system and 2  the lifetime shorter. Use of this finding was made in the experiments involving the second mixture. Fig 4-4 shows the oscilloscope traces from a double flash experiment performed on these solutions with delay times (a) 28 ysec and (b) lOO^tsec using 0-53 filters. The second peak height was negligible at delays > 200 ysec. It should be mentioned here that the second peaks observed in the double flash photolysis of pure formate solutions (Chapter VII) do not affect the above results because of their negligible size (relative to the peaks obtained in the experiments, Fig 4-4) and entirely different dependence upon delay time. (ii) Isopropanol In addition to scavenging OH radicals 2-propanol has been reported to react rapidly with S0j| radical anions, by reaction 4-6. 4-6  (CH)CH0H + SO; 3  2  » Products  k = 8.8 x 10 Nf sec" (84) 7  8  1  1  -84-  (b) c o 0  (/>  A  1  CL  5%  <  V  /  t k-  t  ZOfjLS  gure 4-4  Oscilloscope traces for experiments similar to Fig 4-1. Solution consisted of 3 x 10" M HCOONa and 1.1 x 10" M Ag at pH 7. Auxiliary was fired (a) 28 ysec and (b) 100 ysec after the Main flash. A 0-53 filter v/as used between the Auxiliary lamp and the reaction vessel. 3  +  5  -85-  It was shown by the following experiments that the sulfate radical anions, SO^, were not responsible for the regeneration of e'~ by the Auxiliary flash during these experiments. first consisted of 3 . 0 x  Two mixtures were again employed. The  1 0 " 3 M NaoSO^,  2-propanol and  Ag 2 S0i + .  In the second case, the findings of Chapter VII were again  M  1.05  x  x  and the second,  10"3  isopropanol and  5.2  M Ag2S0it  1.5  x  10"3 M  10"6  TO"5M  utilized in producing e~ by the U.V. photolysis of 2-propanol. Comparable second absorption peaks were obtained from these solutions, also indicating no detectable effect of the presence of 2-propanol in the solutions.  -86" IV.2  RESULTS FROM THALLIUM(I) SOLUTIONS Thallous ion solutions were studied in a similar way. Fig  4-5 shows the oscilloscope traces obtained from double flash photolysis of a deoxygenated solution of 10~ M 2-propanol and 10~ M Tl^O^. 3  5  Filters 0-53 were used between the Auxiliary lamp and the reaction vessel. The delay times between the flashes were (a) 25 ysec and (b) 175 ysec. The results in this case were quite similar to those with Ag ion solu+  tions as can be seen by comparing the above mentioned figure with Fig 4-1 and the data in Table 4-2 with Table 4-1. The dependence of the ratio of absorbance during Auxiliary to absorbance during Main flashes upon the filters used, is given in Table 4-2.  When these data were compared with the transmission charac-  teristics of the filters, it is evident that the intermediate was photodissociated with light of wavelength around 300 nm. In this case, reactions 4-7 and 4-8 are known to occur (45,55) and are expected to be involved here as in the Ag(I) solutions, but k > k^ (55). 8  4-7  Tl + e  4-8  Tl° + Tl  *T1°  +  aq  +  »T1  k = 1.1 x 10 M' sec" (55) 10  1  1  7  2  One again, since the presence of 2-propanol did not affect the results, the possibility that T l  2 +  is the species that can be photodissociated to  produce e~aq , can be discarded because reaction 4-9 produces these species, 4-9  T l + OH +  »T1  2+  + OH'  -87-  (a)  Time  Figure 4-5  Oscilloscope traces from double flash photolysis experiments for a solution containing 1 mM isopropanol and 10~ M TloSOi^. 0-53 filters were used. Auxiliary flash occurred (a) 25 ysec and (b) 175 ysec after the Main flash. 5  -88-  Table 4-2 Ratio of peak absorbances in Auxiliary/Main flashes for T l solution +  (2 x 10~ M) with various light filters between the Auxiliary lamp and 5  the reaction vessel for a 30 ysec constant delay time between the flashes.  Filter #  * Error limit = ± 0.04  Aaux' /Amain  9-54  0.35*  7-54  0.31  0-53  0.27  0-52  0.20  3-74  < 0.04  3-70  < 0.04  DISCUSSION The laser photometry data indicate that hydrated electrons, produced during the Main flash, react rapidly with Ag according to +  reaction 4-2, thereby creating a transient species (half life <\> 42 ysec at 5 x 10~ M Ag ) which can be photodissociated by the Auxiliary flash 6  +  to yield e" again. Light filter experiments show that the transient is photolysed by wavelengths centred around 300 nm but extending to 400 nm. This is entirely consistent with the spectra shown in Fig 4-3. Curve (i) shows the spectrum of the transient 35 ysec after the Main flash, curve (ii) its loss during the Auxiliary flash and (iii) its formation again 30 ysec after the Auxiliary flash. Absorption at long wavelengths in curve (ii) also corroborate the formation of e~ . q  The transient  evidently has an absorption band centre ^ 315 ± 5 nm and this is the band which leads to photodissociation to e~ . Also sketched in Fig 4-3 is the spectrum reported by Baxendale et al (46) with X  max  at 310 nm and assigned by them to Ag° which compares  very closely to our transient. Consequently, it is concluded that photolysis during the Auxiliary flash can be represented by photoionisation process 4-10. 4-10  Ag° — ^  • A g + e~ +  aq  This conclusion is corroborated by a comparison of the estimated extinction coefficient, and by the decay kinetics displayed in Fig 4-2. The extinction coefficient of Ag° at * following procedure.  m a x  (315 nm) was estimated by the  The flash intensity absorbed is given by the  -90analytical form, I = ate"* * (82) where b is l / t . The value of 'a' a max was obtained experimentally from a solution which was the same as that of Fig 4-3 except that it contained no Ag (i.e., e" very long-lived) ag because in this solution the maximum absorbance could be related directly 3  m  +  pOO  to the total number of electrons formed, i.e.,  \ I dt = a/b . It is  •J- a  +  now assumed that all e  aq  %  0  react with Ag to give Ag° and that the latter  decay by reaction 4-4 with k = 6 x 10 M" sec" . Thus [Ag°] at 35 ysec 9  1  1  6  was calculated by integration of the equations (see Appendix 2) d[e" ]/dt = a te" - k [Ag ][ ; ] bt  +  q  2  e  q  d[Ag°]/dt = k [Ag ][e" ] - k^[Ag][Ag°] aq +  +  2  This value of [Ag°] at 35 ysec was used to evaluate the extinction coefficient of Ag° from the experimental absorbance shown in Fig 4-3, curve (i).  The value obtained was about 2.5 x lO Nf cm", which com4  1  1  pares remarkably well with the published value of 2.3 x 10* M" cm" at 1  313 nm (46). From the above value for ^  m x  1  1  and the absorption spectra  if Fig 4-3, the oscillator strength for the transition resulting in the photoionisation of Ag° was calculated to be 0.63 (see Appendix 3). The rate of decay of the transient is consistent with it being Ag° but not with Ag or Ag . Ag is formed in reaction 4-4 and may decay 2  2  2  by reaction 4-11 to yield Ag 4-11  Ag + Ag  +  2  Both Ag and Ag 2  2+  2+  > Ag  2+  (47,83). + Ag°  k  = 3.8 x 10 M* sec" (47) 9  n  1  1  are believed to absorb at X < 400 nm and the 310 nm  band has recently been tentatively assigned to Ag rather than Ag° (83). 2  However, when the time at which Ag°, Ag and Ag 2  2+  reach their maximum  -91concentrations are calculated - by the procedure indicated above, using the published values of k , kk and k 2  and [Ag ] - i t is found that the +  n  maxima occur essentially during the Main flash for Ag°, 50 - 85 ysec afterwards for Ag and at even later times for Ag . (See Appendix 2) 2  2  Thus, laser photometry data can only be explained i f Ag° was the photodissociated species. The decay shown in Fig 4-2 yields a value of  kk = 3.3 x 10 M " sec" for reaction 4-4, which is somewhat smaller 9  1  1  than the published values, averaging 6 x 10 M " 1 sec" (47,83). 9  1  Cercek et al have obtained the absorption spectra of T l \nax  (  = 4 0 0 nm  max  e  =  > max e  ^  x  tions (55).  = K  3  2  x 1 0  c m _ 1  )  "  cm") and T l 1  1 0 n s i n  *  n e  P ^ u  se  + 2  (X  max  2  = 260 nm,  radiolysis of Thallous solu-  Consequently, the filter experiments strongly imply that  none of the above two species were involved in the present system, but that Tl° was the short lived transient which may be photodissociated by light between 250 to 350 nm yielding e" according to process 4-12 as aq follows 4-12  Tl° + hv  » T l + e~ aq +  The similarity between the results from silver (I) and thallium (I) solutions adds credence to the assignment of the 315 nm band to Ag° in the case of A g . +  The absorption band of Ag° in Fig 4-3 may be assigned to a CTTS transition on the basis of the following: (i)  The occurrence of process 4-10.  (ii)  The high oscillator strength for this transition.  ( i i i ) The fact that x electron transfer for  m a x  (90.8 kcal) from Fig 4-3 and the free energy of  -92Ag° + H  +  > Ag + ^H +  2  of 18.5 kcal fits agreeably with the comparison table for several ions with CTTS bands, reported by Marcus (27). This may be a common feature of the unusual valence states of the transition metal ions, most of which absorb in this region. When Ni  2+  solutions were examined in a similar way, analogous effects were  found, implying that the transient ion Ni may be photodissociated back +  to N i  +2  and e~ by light around 300 nm. q  However, Cu  +2  ions showed  negative results but then one might argue that Cu ions are far more +  stable than any hyper-reduced states and do not really come in the same category. Further reduction of cuprous ions- by hydrated electrons has  not been observed, either in flash photolysis or pulse radiolysis  experiments. (It should be noted here that "these transients were very short lived and could not be involved in the experiments on the bimolecul combination of e"aq (Chapter III) even if transition metal ions were present as trace impurities.) This conclusion is contrary to that of Brown and Dainton (50) who ascribed the Ag° band to  5 Py  ^-5 S^. 2  2  2  Q  r  The over-riding factor seems to be that the latter would not be expected to yield e" upon photolysis. Attempts were made to produce e~ from metallic silver in this apparatus, either in the form of a deposited Ag-mirror on the reaction vessel wall (partial coating only) or as colloidal sol (a resultant Ag  +  solution after multiflashing). No absorption due to e~ above the experi mental limits were observed.  CHAPTER V TRANSIENT PRODUCED IN FLASH PHOTOLYSIS OF AQUEOUS ACETONE SOLUTIONS AT pH 11 RESULTS Hydrated electrons were produced by U.V. photolysis of OH" ion solution saturated with hydrogen as before (Chapter III).  They  were long lived (t, ^ 0.2 msec) and decayed by mixed first and second order reactions. A 633 nm laser line was used to monitor e~ and the aq species was identified as e" by spectrographs studies. The pH of aq the solutions used throughout these experiments, with added acetone, was about 11. Flash photolysis of aqueous acetone solutions at natural pH, with no other additions, did not produce hydrated electrons in Main or Auxiliary flashes. However, when e" were generated as above, at pH aq 11 using NaOH, they could react with acetone to produce a transient and this is shown to produce e" again during the Auxiliary flash. V.l  aq  DECAY OF e" IM THE PRESENCE OF ACETONE ag Decay of the absorption due to e~ at 633 nm produced by the q  Main flash was followed in presence of 2.5 to 30.0 x 10" M acetone. 6  The plots of log A as a function of time were always reasonably straight lines, indicating that e~ decayed by a pseudo first order rate. At pH q  11 the rate of reaction 5-1 has been followed, 5-1  (CH ) C = 0 + e" 3  2  > (Ac")  Products  -94and the rate constant ki has been reported to be 5.6 x 10 M" sec" 9  (34).  1  1  Only at lov/ concentrations, when one could be sure that acetone  was not lost by evaporation, could the value of kj obtained be compared with the published data. Fig 5-1 shows the first order plot of absorbance (at 633 nm) due to hydrated electrons in the presence of 2.5 x 10~ M acetone. 6  From the slope of the straight line in the Tog A vs  time plot, one obtains a bimolecular rate constant of 5.2 x 10 M" sec" , 9  1  1  which is in good agreement with the published data. At higher acetone concentrations, acetone is probably lost by H bubbling because the value of kj was found to be much less than the 2  above value for the bimolecular rate constant. Consequently, the decay of e" was actually used to estimate the concentration of acetone in aq these solutions. Also, when compared with the decay of e~ in the aq  absence of acetone, these plots (as in Fig 5-1) were used to obtain the approximate amount of e~ which reacted with acetone only.  -95-  L°9 A 1-5 1  1.3 1.1 2-9 2.7 40  Figure 5-1  80  120 Time (/AS)  160  First order kinetic plot for the decay of e~ following the ag  Main flash in a solution containing 2.5 x 10 M acetone at pH 11. 6  - 96 V.2  HYDRATED ELECTRONS PRODUCED BY AUXILIARY FLASH A transient (which will be designated by Ac") was produced  following the Main flash. This regenerated e~ during the Auxiliary aq flash when Corning filters were restricting the wavelength of the light illuminating the solution in the reaction vessel. The lifetime (1/e) of this species was about 2 msec. Fig 5-2 shows oscilloscope traces of 633 nm absorption signals in the double flash photolysis of a deoxygenated solution containing 7.0 x 10" M acetone at pH 11. 6  [The concentrations of acetone quoted do not represent the actual amount of acetone added to the solution, but calculated from kj/ki, where k  x  is the first order rate constant (sec" ) for the decay of e~ in that 1  q  particular solution and k is the published bimolecular rate constant L  (M  _1  sec~Vj. Fig 5-2 (a) shows the second absorption peak due to e" , aq  where the Auxiliary flash (with 9-54 filters) followed the Main flash after 92 usee. No such peak was obtained when the Main flash did not precede the Auxiliary (Fig 5-2 (c)).  Furthermore, as the delay time  was increased there was a regular decrease in the height of the second peak. A delay time of 1.2 msec is shown in Fig 5-2 (b). The height of the second peak compared to the first was found to depend upon (a) the filters used between the Auxiliary lamp and the reaction vessel, (b) the delay between the two flashes and (c) the charging voltage, i.e., the energy of the flashes. By optimising these conditions the maximum yield of e~ regenerated by the Auxiliary flash was about 60% of e~ produced q  in the Main flash.  q  100  o  (b)  100-  a L.  o  <  200 50-  (c)  0 2 0  Figure 5-2  Time (LIS)  Oscilloscope traces showing absorption due to e" at 633 nm aq during Main and Auxiliary (x only > 220 nm, 9-54 filters) flashes. Auxiliary was fired at delay times (a) 92 ysec and (b) 1.2 msec, (c) represents the case when the Main flash was not fired. Solution contained 7 x 10" M acetone at pH 11. 6  - 98 (a) Filter Dependence The transmission characteristics of the glass filters used during these experiments are given in Table 3-1.  Results in Table 5-1  show the dependence of the second absorption peak heights (expressed as the ration ^ / ^ - j ) P u  aux  ma  n  on  these filters. The solution contained 5 x  10~ M acetone and the delay between the two flashes was 208 ysec. 6  Comparing the transmission of 9-54 and 0-53 filters with the results in Table 5-1, i t is clear that the band responsible for the regeneration of e~ aq by the Auxiliary flash must be centered around the 230 - 280 nm region, and apparently extends to about 400 nm. (b) Delay Dependence The dependence of the second peak heights upon the delay time between the Main and Auxiliary flashes was studied for several concentrations of acetone.  The concentration of acetone was estimated as  described before. The % absorption due to e~ during the Auxiliary q  flash corrected for that in the Main (by taking the ratio) was plotted as a function of delay time between the flashes. The delays at which the maximum second peaks in solutions containing various concentrations of acetone were obtained and are noted in Table 5-2. These values of tmax „ for each concentration of acetone were compared with the lapse of time by which the absorption signal due to e" produced by the Main flash had decayed to < 5%. The comparison indicated that t ,„ occurs at a time when only about 5% of the total e~ max a< produced are still present. J  Table 5-1 Ratio of Absorbance due to Auxiliary/Absorbance due to Main flashes with various glass filters between the Auxiliary lamp and the reaction vessel at a constant delay time of 208 ysec.  Filter #  * Error limit = ± 0.03  Aaux/Amain  9-54  0.22*  7-54  0.19  0-53  0.09  0-52  0.06  3-74  0.03  3-70  < 0.02  -100Table 5-2 Time U  m a x  ) for the second peak height to reach maximum in double flash  photolysis experiments on aqueous alkaline (pH ^ 11) solution containing various concentrations of acetone  Concentration of acetone (x 10 M) 6  t  m a x  ( sec)  2.5  230 ± 30  5.0  170 ± 30  8.0  90 ± 20  15.0  40 ± 15  -101(c) Dependence Upon the Flash Energy Several experiments were performed to test the effect of flash energy on the yield of e~ and Ac". These involved two series, one in which the Auxiliary lamp was fired at a fixed energy  800 joules) at  the same delay time, while the energy of Main flash was varied..  The  ratio of Auxiliary to Main peak heights steadily diminished with increasing Main flash energy in the range 500 - 800 joules.  This was so  because at the fixed acetone concentration only a certain amount of e~ aq reacted with acetone, In the second series, the Main was fired at a fixed energy 625 joules) while the energy of the Auxiliary flash was varied. In this case, results in Fig 5-3 were obtained for the Auxiliary flash energy range of 500 - 1000 joules.  (A fixed time delay of 90 psec was  set and 9-54 filters were restricting the light from the Auxiliary flash.) These results indicate that the Auxiliary flash reaches its maximum efficiency in regenerating e~ in the energy range of 900 - 1000 q  joules.  Also evident is the fact that only about 60% of the maximum  obtainable Auxiliary absorption peak is produced in a typical experiment (Main at 625 joules and Auxiliary at 700 joules).  -102-  55 504540 o  ts  a:  35-^  30 i 25 • 7  8  9  10  2 F l a s h E n e r g y( j o u l e s , ) x 10 Figure 5-3  Dependence of the % e~ regenerated by the Auxiliary flash ag compared to the Main upon the flash energy of the Auxiliary lamp (with 9-54 filters). Main lamp was fired at a fixed energy (625 joules) and the Auxiliary followed after a constant delay of 90 ysec.  -103V.3 TRIPLE FLASH PHOTOLYSIS To determine the percentage photolysis of Ac" to yield e~ in aq the Auxiliary flash, the Auxiliary lamp was discharged twice (separated by a delay of 60 ysec), 50 ysec after the Main flash. 9-54 filters were used between the Auxiliary lamp and the reaction vessel. The H -saturated 2  alkaline solution (pH ^ 11) contained about 5 x 10~ M acetone. Three 6  absorption peaks were obtained and are compared with the peaks obtained in theabsence of the filters, in Table 5-3.  [Relatively smaller absorbance  obtained in the third peak is due to a back discharge from Auxiliary 2 capacitor (II. 1 From the triple flash results in the absence of any filters between the Auxiliary lamp and the reaction vessel, it is clear that the third flash puts out only about 60% of the light emitted by the second flash. In the presence of 5 x 10~ M acetone, the half life of e'^ 6  is about 25 ysec. Hence 50 ysec after the Main flash about 75% of e~  q  have decayed. This, when compared with the amount of e~ regenerated q  by the first Auxiliary flash with 9-54 filters, indicates that at least 50% of Ac" was photolysed. Due to the relatively longer lifetime of Ac", the amount of Ac" decayed in 60 ysec (the delay between the two Auxiliary flashes) would not be significant. Hence it would be expected that with 9-54 filters Auxiliary 2 should regenerate about the same amount of e~ as q  Auxiliary 1. This is consistent with the data in Table 5-3. However, if the intensity of Auxiliary 2 was 60% less than that of Auxiliary 1 (due to back discharge) in the experiment with 9-54 filters also, Auxiliary 2 regenerated more  than Auxiliary 1. This may be so because  of the additional amount of Ac" formed by the reaction of e~ regenerated q  Table 5-3 Results from triple flash photolysis experiment on a solution containing 5 x 10" M aceton 6  expressed as actual absorbance due to e~  q  Absorbance at Maximum Peak Heights First Peak  Second Peak  Third Peak  Energy of Flashes  Fi1ter  Main Auxiliary 1 Auxiliary 2 (Joules)  0.26  0.72  0.40  425  700  700  None  0.29  0.11  0.10  425  700  700  9-54  -105by Auxiliary 1 with acetone. In any case, it is evident that an Auxiliary flash at 700 joules photolyses about 50% or more of the total Ac".  These  arguments are, however, based upon the assumption that the concentration of e" aq produced by the Auxiliary flash is proportional to percentage photolysis of Ac" which is justifiable on the basis of the data in Fig 5-3.  -106V.4  ABSORPTION SPECTRUM By the experimental arrangement described earlier (Chapter II.),  the absorption spectrum in the region 230 - 300 nm of the species present after the decay of the majority of e" , was recorded on a small Hilger aq and Watts spectrograph.  At shorter delay times after the Main flash  comparable absorption spectrum of e]j in the visible region was observed. Fig 5-4 shows the absorbance recorded in a solution containing 5 x 10~ M s  acetone (pH ^ 11) saturated with H , 50 ysec after the Main flash. 2  Hydrated electrons remaining after 50 ysec in this solution amount to less than 10% of the total amount produced. (This was based on the measurements from the 633 nm laser photometry experiments.) The curve (with open circles) shows the absorbance due to the transient Ac" and the darkened circles represent the variation in the spectroflash lamp intensity. 255 ± 5 nm.  It is clear that the absorption spectrum of Ac" has  XMAX  =  The absorbance values have been corrected for the variations  in the film (HP3 Ilford) "Gamma" by obtaining characteristic curves for each wavelength. The absorption spectrum in Fig 5-4 is quite consistent with the filter data on the second absorption peaks and hence is assigned to the species responsible for regeneration of e~ by the Auxiliary flash. aq The concentration of Ac" after 50 ysec of the Main flash can be approximately evaluated a£ follows. With 9-54 filters present an Auxiliary flash fired 50 ysec after the Main produced 1.2 x 10~ M e"^ 7  in an acetone solution. If only about 60% of Ac" can be photolysed (at 700 joules on Auxiliary) then at 50 ysec, there must have been 2 x 10~ M 7  of Ac".  Alternatively, knowing the first order rate constants for the  decay of e~ in the presence and absence of 5 x 10" M acetone, one can 6  -•107-  0.20 l 0.16 )  0.12. 0.08  i  0.04 ^ 230  Figure 5-4  250  270  290  A (nm)  Absorption spectrum of the transient produced 1n the flash photolysis of a solution containing 5 x 10" M acetone at pH 11. (o), 50 usee after the Main flash, (•), variation in the speclamp output. 6  -108estimate the concentration of Ac" from the hydrated electron yield. In this way it was calculated that about 70% of e~ react with acetone. q  For this solution the Main flash at 600 joules produced 5 x 10~ M, 7  aq*  e  ^ence, ^ acetone and e~ exclusively yield Ac", the concentration q  of Ac" is about 3.5 x 10" M after most of the e~ have decayed. 7  q  Using 3 x 10" M as the approximate concentration of the trans7  ient giving rise to the absorption at 255 nm after 50 ysec of the Main flash, one can estimate the molar extinction coefficient of the species at x „ to be 1.3 x lO M" 4  1  cm". 1  -109-  V.5  DECAY OF Ac" If the fraction of Ac" photoionised to give hydrated electrons,  a, is assumed to be constant during the time in which the second peak (due to e~ ) has become minimal, a plot of the concentration of q  produced in the second flash as a function of the delay between the flashes should yield the kinetic data on the decay of Ac". Logarithm and inverse of absorbances due to e~ at the maxima of the Auxiliary peaks were plotted vs time delay. (The concentrations of e" from the first aq peak height varied by 2Q%.) Reasonably straight lines for second order plots were obtained. k , defined by equation, 2  -dCAO = k [Ac-]* 2  dt can be obtained from the slope of such straight lines. Consequently, the slope of such plots can be related to k through the equation, 2  k = slope x 2  cxel  where e is the extinction coefficent of e~ at 633 nm, a is the path q  length and a is the fraction of Ac" photolysed by the Auxiliary flash and depends upon the filters used to produce the Auxiliary peaks. Fig 5-5 shows a second order plot; from its slope one calculates k = 3 . 5 x 2  1 0 M " sec" , assuming an approximate value of a of 0 . 3 for 0-53 filters, 9  •e = 1.2  1  1  x lO  4  M"  1  sec" and a = 4 2 . 5 cm. 1  Large errors were associated with this procedure and wide variations in k for reaction 5-2 were obtained. 2  -110-  56 i  o  48 -  4032 < 24 16 8-  0-4  0-8  1.2  1.6  20  24  2-8  T i m e (ms) Figure 5-5  Second order kinetic plot for the decay of total amount of e" regenerated by the Auxiliary flash (with 0-54 filters) as a function of delay time between the Main and the Auxiliary flashes.  -1115-2  Ac" + Ac" aq aq  » Products  The values of k obtained from these experiments were in the range 2  10 to 10 9  10  M" sec" , and did not depend upon the estimated acetone 1  1  concentration in those solutions. These variations were attributed mos to the large uncertainties in the second peak heights. It is also conceivable that the decay of e~ regenerated in the Auxiliary flash by a second order rate law was merely coincidental.  -112DISCUSSION These results indicate that hydrated electrons produced by the photolysis of OH" during the Main flash decay predominantly by reaction 5-1, 5-1  (CH ) C = 0 + e" aq 3  * (Ac")„ 'aq  2  a  J , c  v  with a rate constant close to the published value. A transient intermediate of half life in the range 0.2 to 0.6 msec is produced and it has a strong absorption band centered at 255 nm.  Filter experiments show  that by photolysis of these species with light around 250 nm, hydrated electrons are again produced. Experiments with varying lamp energies indicate that the amount of e" regenerated by Auxiliary flash reaches a maximum at > 900 joules. Triple flash results are consistent with this and they show that > 50% of Ac" are photolysed in a typical flash using a 9-54 filter. The absorption spectrum with x  max  at 255 nm is consistent  with the previously reported transients observed in the pulse radiolysis of aqueous acetone solution (57,85) so far as the region of absorption is considered. The detailed absorption spectrum has not been obtained before. However, Hart et al (57) have observed the absorption spectrum of iso-propanol radicals, (CH ) C-0H, in neutral solutions. At high pH, 3  2  the equilibrium 5-3, 5-3  (CH ) C-0H + OH" 3  2  > (CH ) C-0~ + H 0 3  2  2  would shift to right and the radical anion (CH ) C-0~ should predominate. 3  2  -113Acetone may react with OH radicals produced in the photoionisation of OH" and create acetonyl radicals by reaction 5-4, but at the used concentration (based on the solubility of H in water at atmospheric 2  pressure) of H (7 x 10 M), OH radicals would preferentially react _tt  2  with H by 5-5. 2  5-4  (CH ) C0 + OH  5-5  H + OH  3  2  2  > CHC0CH 3  2  + H0 2  » H + H0 2  Comparison of the filter data with the absorption spectrum indicates that the band with maximum at 255 nm is responsible for the regeneration of hydrated electrons by the Auxiliary flash. This band cannot be assigned to any of the direct photolysis products of acetone as indicated by the experiments with neutral acetone solutions. Results obtained during these experiments can be explained very well by assigning this absorption band to the radical anion (CH ) C' - 0" or (acetone)". The production of e~ in the photolysis aq step 5-6, 3  5-6  2  (CH ) C - 0" + hv (x = 255 nm) 3  > (CH ) C0 + e~  2  3  2  q  indicates that the above band may be regarded as a CTTS band. This also indicates that the ionisation potential of these radical anions in aqueous solutions is about 5.0 eV.  CTTS bands are usually strong and hence the  estimated high extinction coefficient of 1.3 x 10 M" cm" is consistent h  1  1  with the assignment. This conclusion is also consistent with the electron transfer characteristics of similar radical anions in DMSO (58) and HMPA (59). Unfortunately, the suggested electron transfer to N 0 in both 2  -114those systems could not be investigated in the present aqueous solutions due to the unpredictable concentration of acetone and the inability to measure or control the concentration of N 0 in water. 2  Knowing the concentration of acetone in the solution (by back calculation), the rate constant for reaction 5-1 and the rate constant for reaction 5-7 of e~ with possible impurities, q  5-7  Imp + e^q  * Products  one can calculate the fraction of e~ decaying by reaction 5-1 with q  acetone. (k = 6 x 10 sec" obtained by following e~ in the absence 3  1  7  q  of acetone.) The fraction varies between 60 to 80% in the concentration range of acetone during these experiments. This is consistent with up to about 60% regeneration of e~ by the Auxiliary flash with an energy q  in the 900 - 1000 joules region.  (In this particular case the fraction  of e~ aq decaying predominantly with acetone is calculated to be 65%.) It may also be mentioned at this point that almost all of the transient can be photolysed by a flash of > 900 joules. At about 700 joules (a typical Auxiliary flash energy in an experiment), the light output is not sufficient for a 100% photolysis. It has been suggested that radical anions of type (RjCOR^) would decay by a disproportionate reaction (8, p. 137). 5-8  2(R!C0Ri)  » R C0R + RiCH0HR + 20H~ 1  2  2  The results obtained are consistent with a second order decay for (acetone)", probably by the following reaction. 5-2  2(CH C0CH )' 3  3  5 CHC0CH + CHCH0HCH + 20H" 3  3  3  3  -115Th e rate constant k cannot be assigned a definite value but according 2  to the present results it is in the range 10 to 10 9  10  M"  1  sec" . This 1  large uncertainty in the value of k is probably because of the high 2  uncertainties in the measurements of the second peak heights and in assuming the constancy of a . This is, of course, rather an indirect way for obtaining such results. Finally, the fact that these effects are due to the presence of some impurities cannot be exluded. However, the results on the dependence of t  for second peak heights upon the concentration of  m a x  acetone are strongly against this, unless the impurity is introduced as a large impurity in the acetone used - which itself is at very low concentration. (CHC0CH)~ has been suggested to be a good electron carrier 3  3  in electron transport reactions (86). The information obtained from the present studies may be of great assistance in studying these effects by directly following (acetone) at high pH. -  If the formation of such  species is a general phenomenon for ketones, this may be employed to selectively reduce such ketones. Unfortunately, this can be performed only at very small concentrations.  CHAPTER VI INORGANIC GASES VI.1  NITROUS OXIDE A long lived absorption signal at 633 nm detectable up to  ^ 20 msec was obtained in the flash photolysis of 4 x 10~ M NaOH solu3  tion which was deoxygenated with He and then saturated with N 0. 2  The  inset of Fig 6-1 shows a typical oscilloscope tracing amounting to ^ 20% absorption at 633 nm produced by the Main flash fired at 5.0 kV (625 joules). The plot of log absorbance vs time suggests that the transient decays by a first order rate law with a rate constant of 2.8 x 10 sec" calculated from the slope of Fig 6-1. 2  1  The half life  for formation of the species was ^ 50 ysec and did not follow either first or second order rate law.  Matheson and Dorfman's method of analy  sing mixed order reactions, described in Chapter III (Section 1 (a)), could not be used here because of the relatively short time range involved. Purging the N 0 from the solution by slowly bubbling with 2  He eventually reproduced the very strong (^ 90%) absorption signal at 633 nm due to e~ , which decayed with a half life of ^ 150 ysec. Multi q  flashing such solutions progressively produced less and less long lived absorption signals at 633 nm.  However, N 0 solutions in oxygenated 2  pure 3D water alone at natural pH, produced only a very small (< 5%) absorption peak. The following reactions may be involved in the flash photolysis of the above solution.  Figure 6-1  First order kinetic plot for the decay of a long lived transient at 633 nm in a solution containing 4 x 10~ M NaOH and saturated with N 0. Inset shows the oscilloscope trace of the actual absorption signal. 3  2  -1186-2  N 0 + e" > N 0" aq aq N 0" » N + 0" aq aq 2  2  z  6-3  z  2  2  z  z  6-4  N0~ + N 0  > N +N0  6-5  N 0 + 0~  * [N 0i]  6-6  [Mi]  2  2  2  2  2  2  2  * NO + NO"  It is conceivable that there were some carbonate ions present in the system which would produce the C0 radical anions as follows: 3  6-7  CO3  + OH  >  CO3  + OH'  Taking into account the above set of reactions, the experiments described in this Chapter were essentially designed to establish the following: (i)  if the absorption signal is due to N0~; 2  (ii) whether e~ and/or OH radicals are precursors of the species absorbing at 633 nm; (iii) whether N 0 or NO" radical anions produced in reactions 6-4, 6-5, 2  2  and 6-6 were responsible for the observed effects; (iv) whether CO3 radical anions, known to absorb strongly in this region (32),are produced due to the presence of C0 or CO3 in the 3D water. 2  VI. 1 (a) 0H7N20 Solution Experiments were performed with the NaOH/N0 solution described 2  above in the presence of various added solutes. (i) Nitrate ions In the presence of 0.1 M sodium nitrate, the absorbance at 633 nm  -119was reduced by 50%, indicating that e'^ are involved in the formation of the species but are not the only precursor. e"aq at a rate given by 1.1 x 10 constant (13).  10  M  _1  NO3 ions react with  sec' as the bimolecular rate 1  [N 0] in water at saturation is -v 0.05 M, hence e'^ 2  should preferentially react with N O 3 under these conditions. (ii) OH Scavengers The absorption signal was reduced to a barely detectable level (A < 0.005) in the presence of 2 x 10" M formate ions or 2-propanol. 3  This indicates that OH radicals were a necessary precursor of the species which gives rise to the 633 nm absorption. ( i i i ) NO At different concentrations of NO (adjusted by varying the time of bubbling this gas through the experimental solution), with the [N 0] in the range 2  of 0 to 0.05 M and Main flash energies up to 800  joules, no absorption signals at 633 nm were produced. (iv) C0  2  By dissolving some C0 gas into the solution the signal was 2  increased by a factor of % 5. A fast decaying component presumably due to e" aq , produced by reaction 6-8, 6-8  CO3  » C O 3 + e"  X<^220 nm  a q  was followed by a slow decaying signal at 633 nm with lifetime comparable to the species observed in the Na0H/N0 system. 2  -120(v) Ba  +2  Ions In the presence of 5 x 10" M Ba 2  ions, the 0H~/N0 solution  +2  2  produced a signal of A < 0.01 during the Main flash. that even at such high concentration of Ba be present in the solution since K  It should be noted  ions, ^ 10" M CO3 ions can  +2  7  of BaC0 = 5 x 10~ . 9  gp  3  VI.1 (b) S0"/N0 Solution ?  When reaction 6-9 was used to produce e~ instead of process q  6-1, hv  6-9  SQ[\  >  X<^ 220 nm  S O 4 + e~ a q  the following results were obtained. In the absence of N 0, 4 x 10~ M 3  2  sodium sulphate solutions (deoxygenated) produced approximately the same concentration of  by a flash of similar energy, as did the  sodium hyroxide solutions. When this solution was saturated with N 0 2  and the Main lamp discharged at 625 joules, the same longlived transient was observed. The absorbance in this case was, however, a factor of 2 smaller than in the NaOH/N0 system. 2  If 2 x 10" M sulphuric acid was added to the above solution 3  the signal was not detectable. VI.1 (c) C0j/N0 Solution 2  Carbonate radical anions, C O 3 , can be produced by reactions 6-7 and 6-8.  The absorption spectrum of this radical ion has been studied  in the visible region (32) and the extinction coefficient of CO3 at 633  -121nm was obtained from Behar et al's spectrum, indicating e = 1.53 x 10  3  M" cm" (87). These radical anions were produced by flashing a 2 x 10~ M 1  1  3  deoxygenated sodium carbonate solution and the decay was followed by laser photometry. C O 3 radical ions are reputed to decay by second order processes (reaction 6-10 or 6-11). 6-10  CO3 + CO3  » C0 +  6-11  CO3 + CO3 2 ° > 2C0 + HO; + OH"  2  CO4  H  2  The second order rate constant of 1.57 x 10 M" sec" has been obtained 7  1  1  and this depends upon both the pH and the ionic strength of the solution (32). When a solution containing 2 x 10" M Na C0 was flashed, the 3  2  3  absorption decayed according to second order kinetics as demonstrated by the data shown in Fig 6-2. the quoted  £  633  The slope of this plot in combination with  yields a second order rate constant of 1.5 x 10  7  which is in good agreement with the published value.  sec" , 1  M"  1  The deviation  from a true second order rate as observed by Weeks and Rabani was observed after ^ 70% of CO3 radical ions had decayed. A different kinetic behaviour was encountered if the above solution was saturated with N 0. An absorbance of ^ 0.3 was obtained 2  by a Main flash at 625 joules due to the formation of C0 radical anions. 3  In this solution the decay was first order, as shown in Fig 6-3, with a first order rate constant 2.63 x 10 sec" . This value is very similar 2  1  to the first order rate constant obtained from Fig 6-1 in a NaOH/N0 2  solution, indicating that the same species is involved in the absorption  -122at 633 nm in both cases. The inability to measure the exact concentration of N 0 gas in these solutions restricted the possibility of studying the 2  above decay at different known concentrations of N 0. 2  VI. 1 (d) Absorption Spectra Fig 6-4 shows the absorbance due to transients produced by the Main flash on (i) H saturated and (ii) N 0 saturated 5 mM sodium hydroxide 2  2  solutions in the 500 - 750 nm region, followed 200 ysec after the flash. A Bass - Kessler type of grating spectrograph with a 35 mm camera and high speed IR film was used to record these spectra. The experimental points were obtained by converting the true film (or plate) densities into optical density units by using the film "gamma" values obtained at the different wavelengths. The solid lines (i) and (ii) represent the published absorption spectra of hydrated electrons (7) and carbonate radical anions (87) respectively. No strong absorption (A > 0.1) was observed in the 300 - 500 nm region which was studied using a small Hilger and Watts Quartz Prism type of spectrograph and Ilford HP3 films for both of the above solutions. In this region, absorbance values of < 0.1 are not very meaningful due to the variations in the output of the spectroflash lamp. VI.1 (e) Double Flash Photolysis Deoxygenated sodium hydroxide solution (4 x 10" M) was saturated 3  with N 0 and then flashed with both Main and Auxiliary lamps charged to 2  5.0 and 7.0 kV respectively, and separated by a delay of 0.03 to 25 msec.  •123-  10  Figure 6-2  20 30 40 T i m e (ms)  50  Second order kinetic plot for the decay of a transient absorbing at 633 nm and produced during the flash photolys of a deoxygenated 2 x 10 M Na C0 solution. 3  2  3  -124-  L o g A  i  2  Figure 6-3  p  1  4  6  1  8  i  10  1  1  12  14  First order plot for the decay of the transient (same as in Fig 6-2) in a N 0 saturated solution of 2 mM Na C0 . 2  2  3  -125-  500 Figure 6-4  700  600 A(nm)  800  Absorption spectra of an OH" solution, 200 ysec after the Main flash, (i) saturated with H and (ii) saturated with N 0. The solid lines in (i) and (ii) represent the published absorption spectra of e" (7) and CO3 radical anions (87). 2  2  ag  -126For these experiments 0-53 filters were used between the Auxiliary lamp and the reaction vessel. No second absorption peak, due to either the longlived species or to hydrated electrons (fast decaying), could be observed at al1. VI.1  (f)  N2O"?  In view of the above results one can disregard N0~, the electron 2  adduct product of reaction 6-2 6-2  N 0 + e~ 2  > N0~  q  2  as being the longlived species absorbing at 633 nm because this does not involve OH radicals as a precursor.  Furthermore, in the presence of 0.1 M  N0~ ions, N0~ produced by reaction 6-2 should be negligible and a 3  2  reduction of the absorption peak by only 50% cannot be explained by N0". Also, using process 6-9 to produce e" should not affect the yield aq 2  of N0~. Evidently N 0" is quite short lived and this conclusion is not 2  2  in disaccord with a previous suggestion in which it was estimated to have a life of  10" sec (66,85). 8  VI.1 (g) NO"? The absorption signal could possibly be due to NO" or N 0 2  2  produced by the following mechanism, unless the lifetime of N 0" is too 2  short for reaction 6-4 to occur. 6-4  N 0" + N 0  => N + N 0  6-5  N 0 + 0'  * N0  6-6  N 0;  2  2  2  2  2  2  >  2  2  NO + NO"  2  -127Seddon and Young (88) have obtained the absorption spectrum of N 0^ 2  and shown i t to be a strong band with a X  max  at 380 nm.  If the assign-  ment of these authors of the observed absorbance is definitely then the present results could be due to NO".  N 0^ 2  The effect of e~  q  scavengers can be explained by the above mechanism. In the presence of NO3 ions, where reaction 6-2 is negligible, only reactions 6-5 and 6-6 produce NO".  Since 0" required in 6-5 is produced by reactions  6-1 and 6-12, the yield of NO" would be half that in the absence of NOi. hv  6-1 6-12  OH"  * OH + e" aq OH + OH" » 0" + H 0 2  The effect of OH radical scavengers can be conveniently explained as well because reactions 6-12, 6-5 and hence 6-6 would be stopped.  NO"  also gains support from the effect of NO gas upon the absorbance. Reaction 6-13 is a well known (89) fast reaction, 6-13  NO + e" aq  > NO"  but i t has been suggested (88) that NO" may decay by reaction 6-14, 6-14  NO" + NO  at a rate given by km  » (N 0 )~ 2  2  = 3.3 x 10 M" 9  1  sec" . 1  Hence, i f the solution  is saturated with nitric oxide (concentration = 3 x 10~ M) the half 3  life of NO" would be % 0.1 ysec, which cannot be observed by the present detection techniques.  All of the above considerations, however, cannot  explain why the absorption was so low in the presence of Ba  +2  ions.  -128VI.1 (h) COl_ ? The experiments with sodium carbonate solution clearly show that the species involved there could easily be CO3 radical anions produced by the following reactions 6-8  C O3 3  6-15  N 0 + e"  6-7  CO3  h  ^> C O 3 3 +  v  2  H 2  q  " aq e  ° •> N + OH" + OH 2  + OH  »  CO3  + OH"  CO may decay by one or both of the following reactions, 3  6-10,6-11  CO3 + CO3  » Products  and 6-16  CO3 + Y  > Products  where Y is an unknown impurity and reaction 6-16 probably follows a first order rate law causing the observed departure from a true second order rate law expected for reaction 6-10 or 6-11. One explanation has been put forward with H 0 being Y. This autocatalytic mechanism assumes 2  2  that if the bimolecular reactions of CO3 generate H 0 , then the rapid 2  2  reaction of CO3 with H 0 could produce this behaviour as the H 0 builds 2  2  2  2  up (87). However, the comparison of Fig 6-3 and 6-1 indicates that the same species may be involved in both cases. The absorption spectrum in Fig 6-4 confirms that the absorbance at 633 nm in these solutions was due to the carbonate radical anions C O 3 .  The data points are quite con-  sistent with the published spectrum and are definitely very different from the spectrum of e~ .  -129Source of carbonate radical anions could be the dissolved C0 , which yield C0 ions 2  3  C0 + 20H"  *  2  CO3  + H0 2  in the solutions of NaOH used during the experiments, since C0 probably 2  cannot be totally removed by purging with Helium. C O 3 radical anions are produced by reaction 6-7 6-7  CO3  7  + OH  CO3  + OH"  and not by 6-8, the photolysis process, because of the low concentration. In the presence of N 0, when all e~ have been scavenged and converted 2  q  into 0" radical ions in a short time, the [C0 ] increases by a factor of 3  2. One can explain why, in absence of N 0, the absorption due to 003 2  radical ions is insignificant on the basis of the above fact; also that the extinction coefficient of e" at 633 nm (1.2 x 10 M" cm") is about aq 4  1  1  a factor of 8 greater than that of CO3 radical ions (1.53 x 10 M" cm" ). 3  1  1  This assignment not only explains all the results obtained from these experiments, that were considered in favor of NO", but also (i) the effect of Ba  +2  ions, (ii) the absorption spectrum with a maximum at 600  nm and (iii) the effect of nitric oxide gas. NO is known to react with OH radicals at a very fast rate by reaction 6-17 6-17 OH + NO * HN0 2  The rate constant k  = 4.7 x 10 M" sec" has been reported (13) which, 9  17  1  1  at saturation concentation of NO (c = 3 x 10~ M), gives OH radicals a 3  half life of < 0.1 ysec.  -130Even though the above discussion strongly suggests CO3 radical ion to be the absorbing species at 633 nm in the above systems, the possibility of MO" being produced and having an absorption band in the same region as C O 3 , (hence being masked because of a greater concentration of the latter) cannot be excluded.  However, the following  experiments established the need for C0 in the system in order to 2  observe this long lived species absorbing at 633 nm.  Double flash  experiments show that CO3 does not have a CTTS band. (i) N0~ in C0 -free 3D H 0 ? 2  2  2  As discussed above the absorption signal at 633 nm that lived for milliseconds ( C O 3 radical ions) was apparently due to traces of dissolved C0 in the 3D H 0 supply and could not be removed unless a 2  2  procedure described in Chapter II (Section 3 (a)) was employed. A 4 x 10" M sodium hydroxide solution prepared in this water 3  in an atmosphere of helium was degassed and then saturated with N 0. 2  A Main flash at 625 J. did not produce an absorption signal at 633 nm in this solution on a fast (10 ysec/div) or slow (2 msec/div) time scale. This clearly indicates that the source of the long lived species observed in the experiments with undecarbonated water required the presence of C0 and hence that C0 radical anions were responsible. This also 2  3  excludes the possibility of NO" being the absorbing species. In a saturated solution (5 x 10" M N 0), the half life of e~ will be less 2  2  than 2 nsec. (k  , -) = 8.7 x 10 (63).) However, by purging the 2° aq above solutions with Helium for a few minutes much of the N 0 was driven w n  N  9  +  e  2  -131off.  The Main flash now produced a 70% absorption signal at 633 nm  from the resulting solution, which decayed predominantly by a first order rate law with a half life of approximately 20 psec. The pseudo first order rate constant value could not be compared with the published value due to the inability of measuring the [N 0], 2  Double flash photo-  lysis experiments using 9-54 filters were also carried out on this solution but no second absorption signal due to e~ was obtained for any q  time delay between the two flashes. This indicates that either (a) reaction 6-18 6-18  N0" — — * N 0 + e" X<* 220 nm 2  2  a q  does not occur, i.e., if there is an absorption band due to N0~ in the 2  220 - 1000 nm region, it is not a CTTS band, or (b) N0~, if formed, 2  decays by reaction 6-3 6-3  N 0" aq 2  z  > N + 0" aq 2  z  in a time that is too short for the present technique, or (c) the light of wavelength >220 nm causes reaction 6-19.  -132-  VI.2  SULPHUR HEXAFLUORIDE Sulphur hexafluoride is known to react with hydrated electrons  (67) by reaction 6-20 6-20  SF + e"  (SF; ?)  6  k  20  = 1 .65 x 101o M"i sec'i  but the occurrence of species SFg has not been established. From the known solubility of SF gas in water, it can be calculated that at 6  saturation, [SF ] = 2 x 10  _lt  6  M, so that the half life of e" in an SF  6  saturated solution at atmospheric pressure is 0.7 t, = = 2.1 x 10" sec ' * 2xl0 x 1.65xl0 > 7  _lt  1(  Hydrated electrons were produced by a Main flash at 5.0 kV in a 10~ M 3  deoxygenated sodium hydroxide solution by reaction 6-21 6-21  >  OH" — -  OH +  e~  X< 220 nm  a q  and were followed at 633 nm. The half life of the signal (about 75% absorption) was ^ 150 ysec. The kinetics were complicated, as indicated earlier inthis thesis (Chapter III Section 1 (a)). This solution was saturated with SF and then flashed again with the Main at 5.0 kV. No 6  absorption signal at 633 nm was observable, which is consistent with the t^ calculation above. A signal decaying with a half life of 0.2 ysec cannot be observed in this system. Similar to the N 0 case, the procedure of bubbling off an 2  unknown amount of SF was employed to obtain an absorption signal due to 6  e~ . By varying the bubbling time and rate (which could not be standar-  -133dised to use as a measure of relative [SF ]) different concentrations 6  of e~ could be obtained. During these experiments, 2 to 6 x 10~ M 7  q  aq  e  w e r e  P ' roa  ucea  ' d their decay analysed for both first and second ar|  order rate laws. Fig 6-5 shows the log A vs time plot for the decay of ^ 3 x 10~ M e~ . 7  q  The plots always indicated a first order decay of  absorption due to e~ , suggesting that hydrated electrons decayed preaq dominantly by reaction 6-20.  The slopes of these plots were different  in each case, depending upon the time of purging the 0H~/SF solution 6  with Helium, hence upon [SF ]. 6  From the slope of the first order  plot in Fig 6-5 one obtains a pseudo first order rate constant of. 4 x 10 sec" for the decay of e~ . From this and the published 4  1  values of k o» it was calculated that ^ 2.5 x 10~ M SF was present in 6  2  6  the above solution. Similarly, from the different experiments it was calculated that the concentration of SF in these solutions were in the 6  10" to 10" M region. Thus, the [SF ] was about a factor of 10 higher than the [e~ aq ] produced in these solutions, 6  5  6  Double flash photolysis experiments on these solutions containing 10" to 10" M SF yielded no second absorption peak due to e~ 6  5  6  at any delay between the Main and Auxiliary flashes. During these experiments Main was fired at 5 kV and Auxiliary at 7. Vycor filters (9-54) were used between the Auxiliary lamp and the reaction vessel. Attempts to observe an absorption spectrum in the 230 - 400 nm region following the Main flash in the above solution were unsuccessful for any delay time. In the 500 - 800 nm region, the strong absorption due to e~q was observed at times less than 50 ysec after the Main flash.  1.1  -134-  L o g A  1.0 2.9 H 2.8 2-7 2-6  1  2.5 ^ 2.4  J  23 J  2-2 J 20 Figure 6-5  30  40 50 60 T i m e (ps)  70  First order plot for the decay of e~ in the presence of 2.5 x 10- M SF . 6  a q  fi  -135The existence of SFg as an intermediate product in the pulse radiolysis of aqueous SF solutions was suggested by Asmus and Fendler 6  (67) to explain their kinetic data. This species was to dissociate by reaction 6-22. 6-22  SFg  > SFj + F"  In the photochemistry of aqueous SF solutions, Rumfeldt (69) suggested 6  the following mechanism in the presence of anions, A", 6-23  SF' + A"  6-24  SFj  >  5  >  SF4  SF5  +A  + F"  to account for the yield of F" and SG\ ions. These are the final products, the sulphate ions being produced by the hydrolysis of SF^. More recently Crawford and Rumfeldt (70), by carrying out some M.O. calculations, suggest that addition of an electron to SF would involve the use of a 6  high energy antibonding orbital, thus destabilising the molecule. For maintaining maximum ^-bonding step 6-22 was suggested as the dissociative process. The fate of SF radicals by 6-23 and 6-24 was explained in an 5  analogous way.  However, these considerations are not intended to imply  the existence of SFg or S F 5 . The present results are consistent with the above arguments, suggesting that intermediates like SFg and SF5 may not exist in these solutions or that they have a short life time. However, they suggest that these intermediates, if present, do not have a CTTS absorption band in the 220-1000 nm region leading to their photoionisation by process 6-25, 6-25  R" + hv  »R* + e"  -136where R" represents the intermediate.  It is conceivable, however, that  these species may have an absorption band in the above region which leads to excitation followed by dissociation by process 6-26 (i.e., e~ is q  not generated). 6-26  R" + hv  > (R"")*  Rj  +  R  2  For instance, if SFg were the intermediate, SFg + hv  SFj + F"  This argument is only valid if none of the above fragments produced in reaction 6-26 absorb strongly (e > 10 M" cm") at 633 nm. 3  1  1  1  -137VI.3 XENON The literature does not show reports of any experiments to determine if Xenon reacts with hydrated electrons. This section presents some experiments attempted to observe such a reaction. It was estimated that using the procedure described in Chapter II (Section 3 (b)) the concentration of Xenon in the aqueous solution was > 0.1 M. Such solutions in 3D water when flashed with the Main lamp alone at up to 1000 joules did not produce any absorption signal due to e~ . This implies that the photoionisation of aqueous q  Xenon molecules by 6-27, 6-27  Xe + hv — - » Xe + e" +  does not occur for the wavelength region of 185 - 1000 nm with a sufficiently large quantum yield. In a different set of experiments hydrated electrons were produced by the photolysis of OH" ions in an aqueous solution containing Xenon at > 0.1 M to investigate the possibility of Xenon reacting with aq *°  e  e^q.  P * ro<  uce  a  transient which could be photodissociated to regenerate  It is conceivable that reaction 6-28 may occur.  6-28  Xe + e~  > Xe".  Firstly, comparable decay characteristics of e" to those reported in aq Chapter III were obtained from an 0H"/H solution. This solution was 2  then left in contact with Xenon for about 2 hours and by measuring the amount of solution present it was estimated that the concentration of Xenon in this solution was again at least 0.1 M. When this solution was  -138f 1 ashed the half life of e~ v/as about 190 ysec and hence was virtually unaffected by the presence of Xenon. An Auxiliary flash, restricted by 0-53 filters produced practically the same amount of e" as before (in the second absorption aq peak due to X);  Various delay times between the flashes were used.  The results from these experiments suggest that if Xenon does react with e" , by reaction 6-28, the rate of this reaction is very slow and less than 10% of e~ reacts with Xenon even at about 0.1 M. Taking aq 10* sec" as the pseudo first order rate constant for the decay of e~ 1  1  in the absence of Xenon, it can be estimated that the bimolecular rate constant for 6-28 is < 1.0 x 10^ M" sec" . It is clear that even if 1  1  Xenon reacts with e~ (albeit slowly) to produce Xenon", the concentration q  of this species would have been too small, so that no conclusions can be drawn about process 6-29. 6-29 Xe" + hv  »  Xe + e~„  aq  CHAPTER VII FLASH PHOTOLYSIS OF WATER AND AQUEOUS SOLUTIONS OF SOME OH SCAVENGERS 1. 3D WATER An absorption in the visible region of the spectrum due to hydrated electrons has been reported for the flash photolysis of deoxygenated pure triply distilled water (19) at X ^ 190 nm. Some experiments were performed to test this claim, using clean suprasil lamps and discharging them at about 1000 joules. The 3D water sample prepared in this laboratory was employed.  However, for pure water at natural  pH no absorption at all at 633 nm was observed under maximum sensitivity, indicating that the yield of e~ was less than 10~ M. 8  q  2. AQUEOUS OH SCAVENGER SOLUTIONS (a) Hydrated Electrons from the-Main Flash When a solution of some commonly used OH radical scavenger methanol, iso-propanol or sodium formate in the concentration range of 10~ - 10" M was flashed with the Main lamp at 625 joules, well defined 3  2  absorption peaks at 633 nm were obtained. These signals were eliminated if the solution had been left in contact with oxygen for a short period of time before the flash. Other hydrated electron scavengers had similar effects to 0 in eliminating the absorption signal, thereby confirming 2  that e" aq was the source of absorption. Fig 7-1 shows a log A plot for the decay of the species absorbing at 633 nm as a function of time in a solution containing 3.5 x 10" M sodium nitrate, in which the absorption 6  was created by flashing a deoxygenated 3 mM 2-propanol solution. From  -140the slope of the best straight line the bimolecular rate constant for the decay was calculated to be 8.8 x 10 M" sec" . This is in good 9  1  1  agreement with the published value of k for the reaction of e~ with x  q  N O 3 ions (13). N O 3 + e"  7-1  * Products  k = (0.8,1.1)xl0 M" sec" 10  1  1  x  Table 7-1 shows the concentration of e~ produced in a typical flash with or without 9-54 filters between the flash lamp and the reaction vessel. The % absorption obtained directly from the oscilloscope traces were converted into the concentration units using 42.5 cm as the path length and 1.2 x lO M" cm" as the extinction coefficient of e'^ at 4  1  1  633 nm (7). The chemical compounds used during these experiments evidently contained considerable amounts of impurities (up to 0.01 M hence 10" M 5  in the solution) that react with hydrated electrons with high rate constants (^ 10  10  M" sec" ). Analyses of the decay rates therefore 1  1  probably represent decay with these impurites. The half life of e" aq produced from these solutions in absence of any added electron scavengers was in the neighborhood of 50 usee but in some cases were as low as 25 usee. These results indicate that hydrated electrons are produced by the photolysis of the OH scavengers, however, due to their relatively fast rate of decay in absence of any added electron scavengers suggests that they react mainly with the impurities in the system. 7-2 e~ + Imp ^ Products  -141Table 7-1 Results from the flash photolysis of aqueous solutions of methanol, 2-propanol and formate ions with Main flash energy of 625 joules.  Chemical  Concentration  [e~ ] produced x 10 M aq (No filters) (9-54) 7  r  x 10 M 3  CH0H  10.5  2.0 ± 0.3  < 0.08*  (CH)CH0H  11.2  4.3 ± 0.3  < 0.08*  1.2  4.0 ± 0.3  < 0.08*  3  3  2  HCOONa  * Essentially no absorption peaks at 633 nm were observed with 9-54 filters but since the sensitivity of the present system allows meaningful observation of only about 10" Me" 8  the upper limits have been placed.  Log A  1  30  Figure 7-1  40  50 60 70 T i m e ( j j s )  First order kinetic plot for the decay of e~ following the Main flash in a solution of 3 mM alcohol and 3.5 x 10~ M NaN0. q  6  3  -143VII.2 (b) Double Flash Experiments As reported in previous Chapters, many chemical entities in the concentration range of 10~ M react with hydrated electrons to produce 5  transient intermediates. These can be photolysed to produce e~ by the q  Auxiliary flash with light of restricted wavelengths. Hence, if the impurities associated with the present solutions were of similar nature the results from these double flash experiments would be misleading, and no definite conclusions can be drawn. Results in Table 7-2 show the preliminary data obtained during such experiments on the OH scavenger solutions. Main lamp was fired at 5 kV and Auxiliary at 7. The delay between the flashes was about 40 ysec and the second peak heights are expressed as the ratio A  /A  ..  An interesting observation was made by increasing the pH of the 2-propanol solution from 7 to about 11. The ratio A  aux  /A  main  with  9-54 filters between the Auxiliary lamp and the reaction vessel was increased by > 50% over this pH range. The second peak heights in both 2-propanol and formate ion solutions were maximum at the earliest delays between the flashes, hence probably reach maxima during the flash. No second peaks were observed in the case of methanol at any delay time.  -144Table 7-2 Dependence of the second peak heights upon the filters used between the Auxiliary lamp and the reaction vessel during the flash photolysis of alcohols and formate ions  Source of e~  q  * aux^ main A  A  W 1 t h C o r n i n  9 filter #  9-54  0-53  < 0.01  < 0.01  < 0.01  3 mM (CH)CH0H  0.50  0.10  < 0.05  3 mM HCOONa  0.75  0.20  < 0.05  2 mM CH0H 3  3  2  * Error limit = ± 0.05  3-74  -145DISCUSSION These results demonstrate for the first time that aqueous methanol, iso-propanol and formate ion solutions, upon flash photolysis with light of X < ^ 220 nm, produce hydrated electrons. This may be represented by the photoionisation processes 7-3, 7-4 and 7-5. 7-3  CHoOH + hv  7-4  > CH 0H* + e" aq aq (CH ) CH0H + hv : > (CH ) CH0H* + e~  7-5  HCOO" + hv  3  d  i  3  2  3  2  q  » HCOO + e~ aq  Comparison of the yield of e~ produced from the same concentrations aq of methanol and 2-propanol (Table 7-1) and the same flash energy reveals that the quantum yield for the production of e~ is about twice as high aq •for process 7-4 than 7-3; It may be speculated that the positive fragments produced in processes 7-3 and 7-4 lose a proton in the following steps. 7-6 7-7  CH0H* ^ aq (CH ) CH0H 3  3  2  aq  > CH0H + H* aq » (CH ) C-0H + H+ 2  z  3  2  q  In the case of methanol, double flash results definitely indicate that CH0H radicals, produced by reactions 7-3 and 7-6, or their 2  reaction products cannot be photolysed to produce e~ (unless they have aq a very short lifetime). The absorption spectrum of these radicals are unknown in aqueous solutions.  -146The regeneration of e _ by the Auxiliary flash in 2-propanol aq solutions may be produced from impurities. However, i t could possibly be consistent with the acetone results whereby the Auxiliary flash produces e"  from (CH ) C-0H radicals (or their reaction products). 3  2  They have  been reported to absorb in the region 200 - 300 nm (57), consistent with the filter data.  As discussed in Chapter V, the equilibrium 7-8 has  been well established. , 7-8  (CH ) C0H + OH" 3  ^ [(CH ) C0]" + H 0  2  3  2  2  The results from 2-propanol at pH 11 suggest that the same species as in the case of alkaline acetone solutions (Chapter V) are involved.  At  high pH, equilibrium 7-8 is favoured to the right and hence more of the (acetone)" are formed, hence there are more e~ ' produced in process 7-9. [(CH ) C0]" + hv  7-9  3  2  » (CH ) C0 + e ^  q  3  A  2  > 220 nm  The filter data are also consistent with the absorption spectrum of these species (Fig 5-3).  The involvement of impurities to give such effects  can, however, not be entirely neglected. The product of the photolysis step 7-5 has been studied in some detail.  It has been reported that the conjugate base of the HC00  radicals, C0  radical ions, absorb strongly in the U.V. with A  2  250 ± 5 nm, e  = 2250 M"  1  m a x  cm'  1  (33).  m a x  =  Hence, the regeneration of  hydrated electrons during the Auxiliary flash may be explained on the basis of the process 7-10. 7-10  C0  2  + hv  > C0 + e" X > 220 nm 2  a q  -147-  The filter data and the observation of maximum peak heights during the flash is consistent with this hypothesis. Once again, the results could be due to the presence of impurities associated with the formate sample.  CHAPTER VIII CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK Several intermediates in hydrated electron reactions have been produced by flash photolysis in this study. In this course a few novel sources for the production of e~ have also been discovered. Aqueous q  solutions of methanol iso-propanol and formate ions (popular OH scavengers) produce e" upon U.V. flash photolysis. However, in contrast to aq Boyle et al's findings (19), pure water does not yield observable concentrations (10~ M) of e~ even with a flash energy of 1000 joules. 8  q  The intermediates investigated in this work whose precursors are hydrated electrons may be divided into two categories; (a) intermediate X, which is formed in the absence of any specific e~ scavenger and (b) the intermediates that are formed in the presence of various scavengers. The present results indicate that X is a dimeric hydrated electron or a product therefrom, which is formed when the majority of e^ decay by the bimolecular reaction 8-1. 8  -  ]  aq aq  e  + e  ^"^aq  X absorbs light in the wavelength region 300 - 400 nm to undergo the photoionisation process 8-2. 8-2  (e^")  aq  • hv  » 2e;  q  It has not been possible to directly observe the dielectrons but the product of reaction 8-2 could easily be monitored.  -149In the presence of an electron scavenger, hydrated electrons no longer decay by 8-1, but by a pseudo-first order rate lav/ with rate constants depending upon the nature and concentration of the scavenger. The,studies reported here involve Ag and T l , representing the metal +  +  cations, N 0 and SF , representing the reactive inorganic gases, Xenon as 2  6  an unreactive gas and acetone as an organic molecule.  Intermediates Ag°  and Tl° are produced from the above metal ions and absorb light around 300 nm. They are short lived and yield e~ when photolysed with light aq corresponding to their absorption spectra. For instance, Ag° U 315 nm) undergoes reaction 8-3. 8-3 Ag° + hv (X = 315 nm)  = m a x  » Ag + e~ +  q  On the basis of process 8-3 and the high oscillator strength for this transition, this band is assigned to be a CTTS band. SFg and N 0~ are probably too short lived to be examined in an analogous 2  way. In a solution which contains C0 and where all e~ are scavenged 2  q  by N 0, OH radicals form an interesting species CO3 radical anion by 2  reaction 8-4. 8-4  CO3  + OH  » C O 3 + OH"  CO3 apparently does not have a CTTS band as i t does not produce e~ upon photolysis.  However, e~ is not a precursor of these ions. Perhaps i t aq is not surprising to find that Xenon does not react with hydrated electrons  < W > < """'>10  M  1c e s  Q  An intermediate is formed when acetone reacts with hydrated electrons. This relatively long lived species is probably the C H 3 C O C H 3  -150radical anion. 8-5  C H 3 C O C H 3 + e"  > CH3COCH3  It has an absorption spectrum in the U.V. with maximum at 255 nm and estimated extinction coefficient of 1.3 x lO M" cm". 4  1  This is a CTTS  1  band because e~ can be produced by photolysis step 8-6. 8-6  CH3COCH3 + hv  (\\  250  > CH C0CH  nm)  3  3  +  e~  q  These radical anions probably decay by the second order process 8-7 with a rate constant in the neighborhood of 1 0 - 1 0 9  8-7  2CH3COCH3  M" sec" .  1 0  1  > CH3COCH3 + CH3CHOHCH3 +  1  20H"  In addition to these, intermediates that do not require hydrated electrons as a precursor may also produce e" in a double flash experiment* a  CH3C-O1M  q  and HCOO radicals which are produced during the photolysis of  iso-propanol and formate ions yield such results. However, this could be due to the transients formed by the reaction of e~ with impurities q  present in these solutions, because the electron decay was unexpectedly fast in those systems. Thus the intermediates in the hydrated electron reactions studied during this project have all yielded e~ when photolysed with the a  q  light of the same wavelength as their absorption spectra. This may be a common phenomenon among all the intermediates with hydrated electrons as precursors. However, the present study is certainly not sufficient to make such a broad generalisation. Nevertheless, if it were proved to be a common feature of all such transient intermediates it would imply that upon introduction of an electron into the reactant molecule's orbital species  -151-  with CTTS bands result. It is now evident that this technique can be employed to study a number of such intermediates with hydrated electrons as cursors to test the above hypothesis.  pre-  It would be of particular interest  to carry out investigations involving transition metal ions. Curiously enough, most of the hyper-reduced metal ions absorb around 300 nm (46). Similar studies could be performed on the cations of the Lanthanide series (48). Also, there are numerous organic molecules of much biological interest where such short lived transients could play an important role in the mechanism of corresponding biochemical reactions.  152  -153-  1.  SAMPLE ANALYSIS OF AN OSCILLOSCOPE TRACE FROM A DOUBLE FLASH PHOTOLYSIS EXPERIMENT The r e s u l t s f r o m d o u b l e f l a s h p h o t o l y s i s e x p e r i m e n t s  p r e s e n t e d as r a t i o s , A  a l  a UX  / A . Mia I  have been  Following i s a description of the n  p r o c e d u r e and c a l c u l a t i o n s i n v o l v e d i n t h e p r e s e n t a t i o n o f t h e s e d a t a . F i g 4-1 ( a ) has been a n a l y s e d . The f i r s t peak h e i g h t i s measured f r o m t h e b a s e l i n e t o t h e maximum and e x p r e s s e d a s I = 21.1 % a thus, A . m a  n  = l o g ( 1 0 0 / 7 8 . 9 ) = 0.104  (1)  The f i r s t peak t r a c e i s t h e n e x t r a p o l a t e d p a s t t h e t i m e where t h e maximum during the A u x i l i a r y flash occurs.  The s e c o n d peak h e i g h t i s m e a s u r e d  f r o m t h i s e x t r a p o l a t e d l i n e t o t h e maximum o f t h e p e a k .  In t h i s  case,  I a = 8.0 % and h e n c e A  a i l w  = 0.037  (2)  aUX  And f i n a l l y t h e r e s u l t i s e x p r e s s e d a s t h e r a t i o o f (2) t o ( 1 ) , w h i c h here would b e , A  /A . = 0.356 aux' main  The e r r o r i n t h i s method o f a n a l y s i s c o u l d be q u i t e h i g h i f t h e s e c o n d peak h e i g h t s a r e v e r y s m a l l o r t h e t r a c e i s s t e e p when t h e s e c o n d flash occurs.  -1542. KINETIC CALCULATIONS INVOLVED IN THE FLASH PHOTOLYSIS OF AQUEOUS SILVER(I) SOLUTION (a) Calculation of the times for the concentrations of Ag° and Ag  2  to reach maximum The mechanism involved in the formation and decay of Ag^ and Ag is as follows, 2  Y" + hv  Y+ e„ aq  I a  a  (1)  Ag + e~  Ag°  k\ = 3 x lO™ M"I sec"i.  (2)  Ag° + Ag  Ag  k = 5.9 x 109 M"i sec"i  (3)  Ag + Ag  Ag  +  q  +  2  +  2  ++  2  k = 3.8 x 109 M'I sec'i  + Ag?  2  For convenience, during this calculation k* 's will designate bimolecular rate constants and k's psuedo first order rate constants. The flash intensity absorbed is given by the analytical form (82), I  a  a  = ate"  (1)  bt  where b is 1/tmax for the flash, and in the present system m tax = 15 ysec. Hence, v  b = 6.7 x 10* sec"  (ii)  1  1  Now, a q  dt  = ate"  bt  - k [Ag ][e; ] +  x  q  (iii)  -155-  °]  =  d[Ag  [Ag ][e^ ] - k [Ag°][Ag ]  k[  +  (iv)  +  q  2  dt and  d[Ag|]  =  -  R  [ A g  o  ] [ A g  +  .'  3  k  [ A g  +  ] [ A g  +]  ( v )  3  dt  Considering the relatively fast rate of reaction (1), it would be a reasonable approximation to write from (iii) and (iv),  _ d[Ag°]  il'lj dt  _  b  t  .  t  r t  _  0 i r <  _  t  = a t e ^ - k [Ag ][Ag ] -  2  dt  (vi)  4.  Since it can be assumed that the concentration of Ag ions remains constant, one can equate the following, k'i = ki[Ag ]; k = k [Ag ] and k = k [Ag ] +  +  2  +  2  3  3  Therefore,  ^°l  - ate'  d  - k [Ag°]  bt  2  dt Upon integration we arrive at the following, n  -  /+cT ht  a  [Ag°] " _ J _ + f (k -b) \ t  e  ~ 2t k  Q  +  Q  '  e  -bt i )  e  (vii)  (k -b)  2  2  Differentiating equation (viii)with respect to time,  d[Ag°2 dt  =  _ a _ / -bt _ - b t bte  (k -b) 2  v  At maximum [Ag°], d[Ag°] = 0 and t = t dt  kze  - k t _ -bt 2  ( k 2  .  be  .  b )  -156Thus, 1 -/bk  - b2\t  2  m a x  =  e  ( b  - 2 k  (viii)  ) t :  For [Ag ] = 5 x 10~ M, k = 3 x lO* sec" . Using this and the value of +  6  1  1  2  'b'from equation (i), equation (viii) can be satisfied for values of t max of about 30 ysec. If the values of k is assumed to be 3 x 10 and 9  2  1 x 10  10  M" sec" , t 1  1  m a x  of 40 and 20 ysec respectively are obtained.  Similarly, equation (v) becomes, d[Ag ] = k [Ag°] - k [Agtl 2  2  3  dt Putting in the value of [Ag°] from (vii) and integrating one obtains, [Ag ] = ak 2  [ { (k -b)(k -b)t - k - k + 2b } e*  2  3  2  3  (k -b)2  (k -b)2  2  +  e"  3  { (k -k )(k -t>) - (k -b)2+ (k -b)(k -k ) }  k3t  3  2  3  3  2  (k -k )(K -b)2 3  or,  2  3  (k -k ) 3  2  [Ag ] = 2  ak  2  [Ae'  bt  + Be"^* + ce" ^] 1  (k -b)2 2  where, A = (k -b)(k -b)t - k - k + 2b 3  bt  2  2  3  (k -b)2 3  2  (ix)  3  2  •157-  B = 1 k -k 3  and  2  C = (k -k )(k -b) - (k -b) + (k -b)(k -k ) 2  3  2  3  3  2  (k -k )(k -b) 3  -(k -b)  2  3  2  2  3  2  2  (k -k )(k -b) 3  2  2  3  Differentiating equation (ix) with respect to t,  d[AgJ2  J dA -bt _ e  bAe  -bt  dB -k t _  +  e  I dt  dt +  2  k 9 B e  -k t 2  dt  dC -k t _ e  3  k 3 C e  -k tA_ak2_ 3  dt  / (k -b)  (  x  )  2  2  Now, dA-  k -b  dt  k -b  2  ,  3  dB  _  dc = 0 _and . dt dt v  n  A  Also, when [Ag ] is maximum, ^ -^ d  92  2  dt  =  = 0.  = 0 and t = t, max  Therefore, k 2 B e  (b-k )t 2  m a x +  k 3 C e  (b-k )t 3  m a x  (k -b) _  =  2  (k -b) 3  By using [Ag ] = 5 x 10~ M, +  6  k = 3 x 10 sec' , 4  2  1  k = 2 x TO* sec" . 1  3  1  b A  ( x i )  -158Knowing b from equation (i) and using the above values, equation (xi) can be satisfied for t  of about 75 ysec. Once again,  mx  altering  the value of k in the range 3 to 10 x 10 M" sec" changes t from max 85 to 50 ysec. 9  1  1  2  v  (b) Calculation of the concentration of Aq° at a particular time One can calculate the constant 'a' as follows. We know,  but also, ate" dt= bt  a[e~  (-l-bt)l  bt  bS-  /  = a  o  b2  Hence,  From an experiment where e~ was produced in the photolysis aq  step,  _2 SO 4  _  + hv  SOi,  +e  and no Ag ions v/ere present, [e~ ] was obtained by extrapolating the +  oC| 0  decay curve to zero time, [e~q] = 9.45 x 10" M. Using the value of b 6  Q  from equation (i), a = 4.25 x 10* 1  (xii)  -159-  Knowing all the constants and the time, t, the concentration of Ag  1  was obtained from equation (xii). [Ag°] = _ a _ A e (ki-bJV,.  b t  +  e'^  - e' (k -b) x  bt  A /  (xii)  -1603. CALCULATION OF THE OSCILLATOR STRENGTH  The relationship between the oscillator strength, f, and the experimentally determined integrated molar extinction coefficient of an absorption band is given by the following equation. (ref.28, p.172)  edto  Where e is the extinction coefficient, to the frequency and F constitutes a correction factor near unity related to the refractive index of the medium. Since there is no apparent agreement on how to evaluate F (90), a value of unity was assumed for the calculations in this thesis. Since the bands observed during this work, where such calculations were to be done, were nearly symmetrical, another approximation that  \  ed<o  £  e  m  a  x  A  u  h  was made. Values of A u ^ were calculated from the absorption spectra using the following equation,  where Xi and x are the wavelengths at the half-width of the absorption 2  bands in cm, € . g . , in the case of Ag°, Fig 4-3, x = 4 x 10~ cm and 5  1  X  2  = 2.84 x 10~ cm. 5  -161REFERENCES (1) J. Franck and G. Scheibe, Z. Phys. Chem., 1928, 139A, 22. (2) G. Stein, Disc. Farad. Soc, 1_2, 227 (1952). (3) R.L. Platzman, Physical and Chemical Aspects of Basic Mechanisms in Radiobiology, N.R.C. publ., 305, 34 (1953). (4) J.P. Keene in "British Empire Cancer Campaign", 38th Annual Report, Part 2, p. 498 (1960). (5) E.J. Hart and J.W. Boag, J. Am. Chem. Soc, 84_, 4090 (1962). (6) J.W. Boag and E.J. Hart, Nature, 197, 45 (1963). (7) J.P. Keene, Rad. Res. 22, 1 (1964). (8) E.J. Hart and M. Anbar, "The Hydrated Electron", Wiley and Sons, N.Y., 1970. (9) M.S. Matheson and L.M. Dorfman, "Pulse Radiolysis", M.I.T. Press, Cambridge, Mass., 1969. 10) J.H. O'Donnel and D.F. Sangster, "Principles of Radiation Chemistry", Elsevier, N.Y., 1970. 11) M. Burton and J.L. Magee (Eds.), Adv. Rad. Chem., 1969 (Vol. 1), 1970 (Vol. 2). 12) J.P. Keene, in"Pulse Radiolysis", M. Ebert et al, Eds., Academic Press, London, 1965, p. 1. 13) M. Anbar and P. Neta, Int. J. Appl. Rad. and Isotopes, 1967, 18, 493. ~~ 14) J. Barrett and J.H. Baxendale, Trans. Farad. Soc, 56_, 37 (1960). 15) J.H. Baxendale and A.L. Mansell, Nature, 190, 622 (1961). 16) J.H. Baxendale, Rad. Res., V7, 312 (1962). 17) J.K. Thomas and E.J. Hart, J. Phys. Chem., 68, 2414 (1964). 18) U. Sokolov and G. Stein, J. Chem. Phys., 44, 3329 (1966). 19) J.W. Boyle, J.A. Ghormley, C.J. Hochandel and J.F. Riley, J. Phys. Chem., 73, 2886 (1969).  -162(20) M.S. Matheson, W.A. Mulac and J. Rabani, J. Phys. Chem., 67, 2613 (1963). ~~ (21) G. Stein, Action. Chimi et B. des Rad., Ed. M. Haissinsky, 13, 130 (1970). ~ (22) J.J. Weiss in "The Chemistry of Ionisation and Excitation", p. 17 (1967) . (23) M. Smith and M.C.R. Symons, J. Chem. Phys., 25, 1074 (1956). (24) M. Shirom and G. Stein, Nature, 204, 778 (1964). (25) I. Burak and A. Treinin, J. Chem. Phys., 39, 189 (1963). (26) I. Burak and A. Treinin, Trans. Farad. Soc, 59, 1490 (1963). (27) R.J. Marcus, Science, 123, 399 (1956). (28) J.G. Calbert and J.N. Pitts, "Photochemistry", John Wiley and Sons, p. 271, 1966. (29) R.E. Huffman, Y. Tanaka and J.C. Larrabee, J. Chem. Phys., 39, 902 (1963). (30) L.I. Grossweiner and M.S. Matheson, J. Phys. Chem., 6J_, 1089 (1957). (31) L. Dogliotti and E. Hayon, J. Phys. Chem., 71_, 2511 (1967). (32) J.L. Weeks and J. Rabani, ibid, 70, 2100 (1966). (33) J.P. Keene, Y. Raef and E.J. Swallow, in "Pulse Radiolysis", Eds. Ebert et al, p. 99, 1965. (34) J.W. Thomas, Trans. Farad. Soc, 6J_, 702 (1965). (35) J.K. Thomas, J. Rabani, M.S. Matheson, E.J. Hart and S. Gordon, J. Phys. Chem., 70, 2409 (1966). (36) N. Basco, G.A. Kenney and D.C. Walker, Chem. Comm., 1969, 917. (37) C. Gopinathan, E.J. Hart and K.H. Schmidt, J. Phys. Chem., 74, 4169 (1970). (38) J.J. Markham, "F-centers in Alkali Halides", Academic Press, 1966. (39) L. Kevan, D.R. Ranneke and R.J. Friauf, Solid State Comm., 6, 469 (1968) . (40) J. Eloranta and H. Linschitz, J. Chem. Phys., 38, 2214 (1963).  -163- • (41 G. Czapski and E. Peled, Chem. Comm., 1303 (1970). (42 (a) D.C. Walker, Adv. Chem. ser., 81_, 49 (1968). (b) G.A. Kenney and D.C. Walker, in Electroanalytical chemistry, Ed.. A.J. Bard, Vol. V, Marcel Dekker, N.Y., 1971, Chapter 1. (43 N. Basco, G.A. Kenney-Wallace, S.K. Vidyarthi and D.C. Walker, Can. J. Chem., 50, 2059 (1972). (44 G.E. Adams, J.H. Baxendale and J.W. Boag, Proc. Chem. Soc, 1963, 241. (45 J.H. Baxendale, E.M. Fielden, C. Capellos, J.M. Francis, J.V. Davies, M. Ebart, C.W. Gilbert, J.P. Keene, E.J. Land, A.J. Swallow and J.M. Nosworthy, Nature, 201, (1964). 4 6 8  (46 (a) J.H. Baxendale, E.M. Fielden and J.P. Keene, Proc. Roy. Soc, A286 , 320 (1965). (b) J.H. Baxendale, E.M. Fielden and J.P. Keene in "Pulse Radiolysis", M. Ebert et al, Eds., Academic Press, London, 1965, p. 207. (47 J. Pukies, W. Roebke and A. Henglein, Ber. Bunsenges. Physik. Chem. 72, 842 (1968). (48 G.V. Buxton, F. Dainton and D.R. McCracken, J. Chem. Soc, Faraday Trans. I, 69, 243 (1973). (49 M. Faraggi and Y. Tendler, J. Chem. Phys., 56_, 3287 (1972). (50 D.M. Brown and F.S. Dainton, Trans. Farad. Soc, 62_, 1139 (1966). (51 A. Barkatt, M. Ottolenghi and J. Rabani, J. Phys. Chem., 76, 203 (1972). (52) L. Shields, J. Chem. Phys., 45, 2332 (1966). (53) L. Bales and L. Kevan, ibid., 55, 1327 (1971). (54) R.R. Dewald, J.M. Brooks and M.A. Trickey, Chem. Comm., 1970, 963. (55) B. Cercek, M. Ebert and A.J. Swallow, J. Chem. Soc (A), 612 (1966). (56) J. Rabani and G. Stein, J.Chem.Phys., 37_, 1865 (1962). (57) E.J. Hart, S. Gordon and J.K. Thomas, J. Phys. Chem., 68, 1271 (1964). (58) (a) T.K. Cooper, Ph.D. Thesis, U.B.C., 1972. (b) T.K. Cooper and D.C. Walker, Can. J. Chem., 49, 2248 (1971). (59) G.J. Flynn, Private communication.  -164(60) P. Reisz, J. Phys. Chem., 69, 1366 (1965). (61) S. Gordon, E.J. Hart and J.K. Thomas, J. Phys. Chem., 68, 1262 (1964). (62) K.D. Asmus, A. Henglein, A. Wigger and G. Beck, Ber. Bunsenges. Phys. Chem., 70, 756 (1966). (63) S. Gordon, E.J. Hart, M.S. Matheson, J. Rabani and J.K. Thomas, Disc. Farad. Soc, 36, 193 (1963). (64) M. Anbar, R.A. Munoz and P. Rona, J. Phys. Chem., 67_, 2708 (1963). (65) G.E. Adams.in "Radiation Research", G. Stein, Ed., North-Holland Publ. Co., Amsterdam, 1967, p. 195. (66) K.D. Asmus, A. Henglein and G. Beck, Ber. Bunsenges. Phys. Chem., 70, 459 (1966). (67) K.D. Asmus and J.H. Fendler, J. Phys. Chem., 72, 4285 (1968). (68) K.D. Asmus, W. Grubein and J.H. Fendler, J. Am. Chem. Soc, 98, 2625 (1970). ~ (69) R.C. Rumfeldt, Can. J. Chem., 49, 3336 (1971). (70) M. Crawford and R. Rumfeldt, Can. J. Chem., 49, 3336 (1971). (71) CRC Handbook of Chemistry and Physics, 51st Edition (1971). (72) G.A. Kenney, Ph.D. Thesis, U.B.C., 1971. (73) R.D. Morse, Ph.D. Thesis, U.B.C, 1972. (74) A.M. Bass and K.G. Kessler, J. Opt. Soc. Am., 49, 1223 (1959). (75) G.H. Thomas, Unpublished results. (76) K. Fueki, J. Chem. Phys., 50, 5381 (1969). (77) J.H. Baxendale, Rad. Res. Suppl., 4, 139 (1964). (78) M.T. Lok, F.J. Tehan, J.L. Dye, J. Phys. Chem., 76, 2975 (1972). (79) E.J. Hart, S. Gordon, E.M. Fielden and A.L. Mansell, J. Phys. Chem., 70, 150 (1966). (80) J. Barrett, M.F. Fox and A.L. Mansell, ibid., 69, 2996 (1965). (81) E. Hayon and T.T. McGarvey, ibid., 71_, 1472 (1967). (82) N. Basco, A.B. Callear and R.G.W. Norrish, Proc Roy. Soc, A260, 459 (1961).  -165(83) P.L.T. Bevan, Private communication. (84) M. Anbar and E.J. Hart, J. Phys. Chem., 69, 973 (1963). (85) G.E. Adams, B.D. Michael and J.T. Richards, Nature, 21_5, 2148 (1967). (86) (a) G.E. Adams, B.D. Michael and R.L. Wilson, ACS, Adv. in Chem., 81_, 289 (1968). (b) M. Anbar and P. Neta, J. Chem. Soc. (A), 1967, 834. (87) D. Behar, G. Czapski and I. Duchovney, J. Phys. Chem., 74, 2206 (1970). (88) W.A. Seddon and M.J. Young, Can. J. Chem., 48, 393 (1970), (89) S. Gordon, E.J. Hart, M.S. Matheson and J.K. Thomas, J. Am. Chem. Soc., 85, 1375 (1963). (90) R. Benasson and E.J. Land, Trans. Farad. Soc, 1904 (1971 ).  LO  (/O  1166  1461  g.y.i- 6>oLt> M*ML  PATHA  ONiuB&^fTY  UAT £  &0Lb Mei>AL  FBLLOWZhttf*  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0060230/manifest

Comment

Related Items