EPA-650/4-74-004
March 1974
Environmental Monitoring Series
-------
EPA-650/4-74-004
MOLECULAR MODULATION
SPECTROMETRY FOR OBSERVATION
OF FREE RADICALS
by
Harold S . Johnston
Department of Chemistry
University of California
Berkeley, California 94720
EPA Grant No. 801120
ROAP No. 26 AAD
TASK No. 12
Program Element No . A11008
EPA Project Officer: Dr. Joseph J. Bufalini
Chemistry and Physics Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
11
-------
ABSTRACT
Free radical intermediates determine the mechanism of most chemical and
photochemical reactions. Although some indication of these intermedi-
ates can be inferred from kinetic studies, direct observation of the
free radicals is required for unambiguous understanding of the mechanism.
This project has developed phase-shift methods that give both kinetic
and spectroscopic data for reaction intermediates in photochemical
processes. The methods have been designed to observe free radicals at
about lO^1 particles/cc, and at these very low concentrations radicals
have lifetimes of about one to ten seconds. New digital phase-sensitive
devices have been built to measure phase shifts at these very low
frequencies. The spectrometric regions are infrared (IR), ultraviolet-
visible (UV), mass (MS). The reactive intermediates observed and the
various regions of spectra studied by these methods are: C100 (IR, UV,
MS), CIO (UV, MS), HOO (UV, IR) , and N2Os in NC>2 photolysis (IR) .
This report was submitted in fulfillment of ROAP No. 26AAD, TASK 12,
and Grant Number 801120 by the University of California under the partial
sponsorship of the Environmental Protection Agency. Work was completed
as of October 31, 1973.
111
-------
CONTENTS
Page
Abstract iii
List of Figures vi
List of Tables ix
Acknowledgments x
Sections
I Conclusions 1
II Recommendations 2
III Introduction 3
IV Apparatus 4
V Nature of Results 11
VI Chlorine Oxide Radicals 33
VII Spectra and Kinetics of the Hydroperoxyl Free Radical 52
VIII Photolysis of Nitrogen Dioxide 74
IX Discussion 86
X References 90
-------
FIGURES
No. Page
1 Schematic Diagram of Experimental Systems with Infrared 5
Detection of Intermediates
2 Schematic Diagram of the System Using the Mass 8
Spectrometer
3 Phase Relations Between Photolysis Light, Reactant, and 14
Product of a Primary Photochemical Reaction
4 Photochemical Reaction with Simple, Single Intermediate 16
5 Phase Relations Between Photolysis Light, Reactants, 18
Intermediates, and Products in Complex Photochemical
Reaction
6 The Fundamental Modulation Frequency of a Primary First- 23
Order Radical
7 The Modulation of a Secondary Radical 25
8 Modulation Amplitude Relative to the Limiting Amplitude 26
9 Concentration Modulation of a Second-Order Radical 29
10 Concentration Modulation of the Product of a Second- 31
Order Radical
11 C1OO Infrared Molecular Modulation Spectrum 35
12 CIO Modulation Spectrum in Ultraviolet 37
jt ' f
13 High Resolution Molecular Modulation Spectrum 38
14 Calculated Relative Modulation Amplitude for the Two 39
CIO Intermediates
J^
15 Spectrum of New CIO Species 41
X
VI
-------
FIGURES
No. Paqe
16 Resolution of Complex Spectrum into Two Components 42
17 Ultraviolet Spectrum of C1OO Radical as Obtained by 43
Two Different Procedures
18 Observed Points and Calculated Curves for the Phase Shift 50
of CIO
19 Observed Points and Calculated Curves for the Phase Shift 51
of CIO
20 The Modulated Infrared Absorption Spectrum Obtained During 55
the Photolysis of Ozone in the Presence of Hydrogen
Peroxide
21 Six Repetitive Scans of the Modulated Absorption Spectrum 56
Obtained During the Photolysis of Ozone in the Presence
of Hydrogen Peroxide
22 Eleven Repetitive Scans of the Modulated Absorption 57
Spectrum Obtained During the Photolysis of Ozone in the
Presence of Hydrogen Peroxide
23 The Average of Sixteen Scans of the Modulated Absorption 58
Spectrum Obtained During the Photolysis of Ozone in the
Presence of Hydrogen Peroxide
24 Phase Shift of the Infrared Absorption Peaks 60
25 The Modulated Ultraviolet Absorption Spectrum Obtained 62
During the Photolysis of Hydrogen Peroxide
26 The Dependence of the Phase Shift of the Modulation at 65
2200 A on Flashing Period.
vu
-------
FIGURES
No. Page
27 Comparison of Radical and Reactant Amplitudes 66
28 The Modulated Absorption Phase Shifts from 2450 to 67
o
2000 A Obtained During the Photolysis of Hydrogen Peroxide
o
29 The Modulated Absorption Amplitudes from 2450 to 2000 A 68
Obtained During the Photolysis of Hydrogen Peroxide
30 The Phase Shift of the Radical vs the Flashing Period 70
31 The Ultraviolet Spectrum of the Hydroperoxyl Radical 72
32 N02 Decay; Comparison of Experimental Points and 77
Calculated Curve
33 N9°5 DecaY As Observed by Infrared Absorption 78
34 Concentration Ratios Where Nitrogen Pentoxide is a 79
Maximum
viii
-------
TABLES
No. Page
1 Absolute Cross Section for Absorption of Ultraviolet 44
Radiation by the Free Radical C1OO and CIO in Units of
...-18 2
10 cm
2 Comparison of Modulated Signal of CIO and Unmodulated 46
Signal of Ar
3 Comparison of Modulated Signal of C1OO and CIO 47
4 Observed Phases As a Function of Chopping Frequency and 49
Concentrations
5 Data for NO_ Modulation At One Atmosphere Total Pressure, 82
24°C
6 Observed Rate Constant and Rate Constant Ratios 83
7 Elementary Rate Constants for Nitrogen Dioxide 84
Photolysis, 297°K, 1 atm N
IX
-------
ACKNOWLEDGMENTS
The work reported here was carried out by G. E. McGraw, E. D. Morris,
T. T. Paukert, L. W. Richards, J. van den Bogaerde, Ching-Hsong Wu,
Leo Zafonte, and Alan Marker.
-------
SECTION I
CONCLUSIONS
The molecular-modulation method is capable of giving quantitative data
concerning free radical intermediates in photochemical reactions driven
by light of moderate intensity, as in sunlight. This report gives
quantitative spectroscopic data for the CIO free radicals, CIO and
X
C100. These substances and their photochemical reactions may not be
important in photochemical smog, but they have unexpectedly turned out
to have possible significant effect in the photochemistry of the strato-
sphere. This report gives quantitative spectral and kinetic data for
the hydroperoxyl free radical, HOO. This radical is undoubtedly very
important in both tropospheric and stratospheric pollution.
Although this method does work and has given some quantitative data on
important and elusive atmospheric species, it requires a large amount of
time and effort. Radical lifetimes under atmospheric conditions may be
as long as several seconds, and signal-to-noise features are extremely
unfavorable at such frequencies. It is so difficult to obtain data by
this method that it cannot be recommended as a general method of kinetics;
it should be used only when one must study photochemical reactions under
realistic atmospheric conditions.
-------
SECTION II
RECOMMENDATIONS
If this method is to be developed further, one should go to higher
intensities of photolyzing light, shorter optical paths, and faster
electronics. A light intensity of about 10 to 100 times that of natural
sunlight should be a fairly good simulation of realistic atmospheric
conditions - it still is far less than the intensity of flash photolysis.
Radical lifetimes would be 3 to 10 times shorter, and radical concentra-
tions would be 3 to 10 times higher. If the optical path is substan-
tially shorter, there should be less accoustical noise associated with
the long lever arm of extensively folded light paths; and thus smaller
modulation amplitudes should be stable.
-------
SECTION III
INTRODUCTION
Most photochemical reactions in the gas phase occur with highly reactive
free radicals as short-lived intermediates. A firm understanding of
these photochemical reactions demands the direct observation of the free
radicals. A considerable body of data of this kind has been built up by
the method of kinetic spectroscopy and flash photolysis, which uses very
high intensities of photolyzing radiation, with free radical concentra-
tions of radicals. In this laboratory we have developed a new method
that is designed to give the spectra and chemical kinetics of radicals
in this range of concentration. This method uses a continuous measur-
ing probe, infrared beam, ultraviolet beam, or the collection grid of a
mass spectrometer; the photolyzing light is turned on and off at a
regular controlled rate; the radicals form in the light and decay in the
dark; and the continuous probe is modulated by the appearance and
disappearance of radicals. This "molecular modulation" provides a
differential signal that is far more sensitively observed than a direct
signal. Also the step-functions of increasing products and of decreasing
reactants modulate the measuring probe by about the same amount as the
formation and decay of radicals. The amplitude of the modulation as a
function of wave length or mass number gives the spectra of the radicals;
the phase shift between photolyzing light and radical gives the lifetime,
and hence the chemical kinetics, of the radical.
The photolysis of C12 in the presence of 0- was studied primarily as a
check on the new apparatus, since the system has been intensively
studied for more than 20 years. The hydroperoxyl radical HOO is
probably the most abundant free radical in photochemical smog. The prin-
cipal results of this report are quantitative infrared and ultraviolet
spectra and kinetic data for the HOO free radical. This report also in-
cludes a chapter on intermediates in the photolyses of nitrogen dioxide.
-------
SECTION IV
APPARATUS
A. INFRARED SPECTROMETER
The instrument used in this work is a long path molecular modulation
infrared spectrometer. A block diagram of the apparatus is shown in
Figure 1. The reaction cell is a cylindrical quartz tube 91 cm long and
28.7 cm in diameter, with a volume of 67.0 liters. The cell is equipped
with gold coated multiple reflection mirrors which give it a spectro-
scopic path length adjustable from 4 to- 40 meters. The source of the IR
radiation is a Nernst glower which is chopped at 400 cps by an American
Time Products tuning fork. The monochromator is a McPherson model 2051,
one meter grating monochromator equipped with a 150 line per mm grating
and order sorting filters. The IR detector is a liquid helium cooled,
copper doped germanium photo conductor produced by the Santa Barbara
Research Center.
/
The photolytic light for these experiments was supplied by four, 32 inch
G.E. 30 watt, F30T8/BL black lamps. The photolytic photon flux in the
16 2
cell is on the order of 10 photons/cm -sec. The photolysis lamp out-
put for the cell is monitored by a phototransistor, which can be used as
a reference to detect fluctuations in the lamp intensities. The photol-
ysis lamps are driven by a regulated power supply which can electronically
switch them on and off in response to a reference square wave from a
crystal oscillator.
The electronics for the instrument are designed to detect the periodic
concentration fluctuations in the absorbing gases in the cell which are
induced by the flashing of the photolysis lamps. This is done by two
sequential stages of demodulation. The 400 cps AC signal from the
detector carries the modulation information on sidebands at 400±f cps,
where f is the frequency of the flashing lamps. The first demodulation
is carried out by a 400 cps lock-in amplifier which produces a DC signal
directly proportional to the spectroscopic light intensity. Riding on
this DC signal as a low frequency ripple is the modulation information.
-------
Loo Frequency
Modulotion
To Modulation DC
Computer »
90'
DC
Amps
XBL 728-6830
Figure 1 - Schematic diagram of experimental system with infrared
detection of intermediates.
FIGURE 1
-------
This DC signal is split in two with one component being recorded while
the other is AC coupled, filtered, and sent into the second stage of
demodulation. The second stage demodulator is a set of two lock-ins,
operating at frequency f, with one lock-in reference in phase with the
photolysis lamps and one reference exactly 90° out of phase. The out-
puts of these two lock-ins represent the amplitude of the sine and cosine
components of the first fundamental of the modulation signal. From
these values the phase shift and amplitude of the modulation signal can
be calculated. The output of all three lock-in amplifiers are recorded
and time averaged in a signal averaging computer.
B. ULTRAVIOLET APPARATUS
The basic system is analogous to that used in the infrared, although
several new features were developed, especially the digital lock-in
system. The reaction cell is a quartz cylinder 1.8 m long and 15 cm i.d.
with a single mirror at one end. Two photolysis lamps were mounted out-
side the cell and backed by Alzak aluminum reflectors. The spectrometric
light over the region 2000-3500 A is provided by a deuterium lamp
(Bausch and Lomb DE-50A) powered by a well-stabilized power supply. A
tuning fork (American Time Products) operating at 400 Hz chops the
spectrometric beam at the cell entrance. The fork also provides a 400-
Hz reference used in the detection system. A McPherson (Model 218, 0.3 m)
monochromator equipped with a 2400-line/mm grating follows the reaction
cell. An EMI photomultiplier (9526B) is mounted at the exit slit of the
monochromator.
The output of the photomultiplier and the 400-Hz reference from the
tuning fork are fed into a 400-Hz lock-in amplifier which separates the
magnitude of the spectrometric light and the low-frequency modulation
signal. A crystal oscillator is scaled down to control the photolysis
lamps and to provide reference signals to compare with the modulated
signal. The signal and references are sent to a digital lock-in where
long-time ctvej.aging is obtained by means of reversible counters. The
magnitude of the in-phase and 90° lag components is digitalized and
punched into paper tape. This tape is later fed into a computer which
-------
calculates phase shift and modulation amplitude at each wavelength of
the spectrometric light.
The intensity of the photolyzing ultraviolet light was found by following
the rate of decay of NO2. This method requires exact values of the
absorption coefficient of NO, as a function of wavelength (2700-4400A),
quantum yield over the same range of wavelength, the spectral distribu-
tion of the photolyzing fluorescent lamp, and the full set of rate
constants for the elementary reactions in this system (involving NO_,
NO, O, 0^f 0 , NO.,, and ^0-) . To calculate the rate of production of
Cl, it was necessary to combine the above data with the absorption
2
coefficient of C12 over the same range of wavelengths.
In the infrared apparatus the average ultraviolet light intensity was
15 2
8.5x10 photons/cm -sec), and the cross section for light absorption
by chlorine weighted by the spectral distribution of the lamps used was
-19 2
0.935x10 cm . Thus the rate of destruction of chlorine molecules is
d ln[C!2]/dt = a!Q = 0.80xlO~3 sec"1
In the ultraviolet apparatus the photolyzing intensity is larger because
two lamps are used and the cell has a small diameter.
16 2
I = 4.2x10 photons/(cm sec)
al = 3.92xlO~3 sec"1
o
C. MASS SPECTROMETER
A schematic diagram of the apparatus is shown in Figure 2. The photo-
lyzing light was a 500 watt mercury arc lamp, filtered by Corning 9863
and 0160 filters immersed in a heat-removing solution of 0.1 M CuSO.,
o ^
passing a band between 3200 and 3800 A. A rotating disk chopped the
beam to give a square-wave photolyzing beam in the reaction cell. The
chopper blade was driven by a Bodine synchronous motor coupled with a
multiratio gear box to provide six different periods of rotation, 1, 2,
-------
Manifold
Figure 2 - Schematic diagram of the system using the mass spectrometer
-------
4, 8, 16 and 32 sec/cycle. A reference signal was provided by a small
projector light and a photo-transistor placed on the opposite side of
the blade from the photolyzing beam. The reaction cell (98 cc volume)
was constructed of Pyrex with a quartz window fused on the rear through
a graded seal. The front end was hemispherical in shape with a pin hole
33 ym in diameter to allow reaction gases to leak into the mass spectro-
meter. The pin hole was finished by grinding with a tapered copper rod
to give a short, round, flared exit. After the gases left the reaction
cell via the pin hole leak, they spread into an intermediate vacuum
chamber, and a small portion intersected a "skimmer" to enter the high
vacuum chamber of the mass spectrometer as a molecular beam. The size
of the skimmer hole, angles of its opening, and its distance from the
pin hole of the reaction cell were designed to optimize the beam in-
tensity and to avoid standing shock waves from blocking the entrance to
the mass spectrometer. The intermediate chamber was evacuated by a
Welch 1402B mechanical pump, NRC-4-SP type 121 diffusion pump, and a
liquid nitrogen trap; during experiments its steady state pressure was
-4
just below 10 torr. The main chamber of the mass spectrometer was
evacuated by a Marvac IDR2 mechanical pump, a 6 inch ultra-high vacuum
CVC diffusion pump including water cooled chevron baffles and type TSMU
sorbent trap. Also, a type PDV-2 ion pump was installed, but it was not
operated during experiments because it introduced a substantial amount
of noise to the electron multiplier. During experiments the chamber
pressure was below 6xlO~ torr.
The mass spectrometer was of the quadrupole type, Extra-nuclear Labora-
tories, model 324, with a type II high efficiency electron-impact ionizer.
Because of serious attack on the tungsten filament by chlorine and
oxygen, it was replaced by a rhenium wire; 0.018 cm diameter. Six holes,
0.25 cm diameter, were drilled around the filament shield case to
increase pumping speed in the ionization zone. The system was aligned
so that the pin hole, the skimmer, the ionization zone, and the axis of
the mass spectrometer were all in a straight line. Ions were detected
by an electron multiplier, closely coupled to a bakeable preamplifier
with 10 ohm input resistance, which gave a frequency response from DC
to 100 kHz.
-------
In the reaction cell, reactants and products were present typically in
amounts 4 to 6 orders of magnitude higher than free radical inter-
i
mediates. To cover this very wide range, we measured free radicals by
a direct ion counting technique, and reactants by analogue, current
measuring methods Circuit constants were such that the ion counting
method was useful for constituents in the reaction cell between 10 and
10 molecules/cc, and the analogue method was useful from 10 to 10
molecules/cc. An ion event was amplified by the electron multiplier to
give a pulse in the preamplifier. The pulses were passed into a voltage
comparator that rejected as noise pulses below a certain level and sent
i
higher pulses into a "one shot" circuit, which converted them to a uni-
form height of 4 volts and a uniform width of 5 ysec. The complete
period, T, of a cycle of photolysis, light and dark portions, was pre-
cisely divided into 4 equal parts: O to T/4 (light on), T/4 to T/2
(light on), T/2 to 3 T/4 (light off), 3 T/4 to T (light off). Pulses
from the "one shot" circuit were passed to two Hewlett-Packard model
H19-5280 A reversible counters, one adding from 0 to T/2 and subtract-
ing from T/2 to T, called the in-phase counter and the other adding from
T/4 to 3T/4 and subtracting during the other intervals, called the out-
phase counter. To record the mass spectrum at a given number, a large
number of cycles of photolysis was accumulated in the counters. It is
very important to have the time for counting positively to be exactly
equal to the time of negative counting; otherwise unbalanced counts will
accumulate after many cycles. The timing system has an accuracy of one
part per million for 30 seconds of averaging time. After completing a
pre-set number of complete cycles, the counting stopped, and the net
counts from each counter were puched out on paper tape. Later the paper
tape was fed into a computer, which calculated the desired functions.
Analogue signals were processed and recorded by the method of Morris and
g
Johnston .
10
-------
SECTION V
NATURE OF RESULTS
A. SIMPLE GRAPHICAL PRESENTATION
The effect of intermittent ultraviolet illumination upon a photolabile
chemical system is to bring about, through photolytic decomposition of
one of the reactants, a periodic variation in its concentration and
that of the intermediates directly formed. Ensuing reactions bring
about fluctuations in the concentrations of other intermediates and
products, as well. Use of a well-stirred continuous-flow reaction vessel
permits experimental measurements to be made under quasi-steady-state
conditions: the concentrations of the various species fluctuate about
average values which remain constant with time, and spatial concentration
gradients are negligible. Concentrations fluctuations modulate the am-
plitude of the 400 cps IR beam transmitted through the reaction vessel
giving rise to an amplitude-modulated electrical signal from the IR
detector.
Although the frequency of the molecular modulation coincides with the
frequency of pulsed ultraviolet illumination, it is, in general, dis-
placed in phase from the ultraviolet excitation voltage. The extent of
the phase shift depends upon whether the modulation is due to a reactant,
intermediate or product, and also upon the lifetime of the absorber under
the conditions of the reaction, relative to the period of ultraviolet
illumination. Since we are ultimately concerned with determining the
lifetimes of some of the chemical species involved, as well as their
identities, measurement of the phase shift is a means of obtaining the
former information and very often the latter as well.
The output of the dual phase demodulator consists of two D.C. voltages
which permit the evaluation of the phase shift in the following manner.
The input signal to the dual phase demodulator may be represented as:
11
-------
E = E sin (2-rrvt + 6) where E is a measure of amplitude of
molecular modulation
v = frequency of uv excitation
-,aoi"i£>^4- *-<-> 4-Vio n 1 +-Y-snri o 1 o+- 1 i rrVi+- TVi^c is ftC'.C'.nm'n 1 1 Rhp.fi
^W»W V^. »!>» »AM»* _iW^--WM *- »-.» _» - ~J _ - ,_
either by varying reaction conditions, or the rate of pulsing the ultra-
violet light, or both. One can then obtain the lifetime from the
12
-------
measured phase shift and the ultraviolet flashing frequency alone, or
in conjunction with the dependence of the signal amplitude upon flashing
frequency of the ultraviolet lamp.
In general, the identity of an absorber responsible for a given modula-
tion signal can be ascertained from the shape and position of its ab-
sorption spectrum. However, if this is not possible due to either over-
lap of absorption bands of several species in the reaction or to the
fact that the absorption spectrum is not known, then the species can
often be identified as a reactant, intermediate or product by the phase
shift. The principles involved will be illustrated by three
hypothetical reactions.
PRIMARY PHOTOCHEMICAL REACTION
If reactant A absorbs light directly to form product C, without forma-
tion of intermediate species, the mechanism is
A + hv * C + ...
The exciting light is electrically switched on and off to produce a
square wave, Figure 3. When the light is on, reactant A decreases and
product C increases; when the light is off, both A and C remain constant,
Figure 3. If each of the three periodic functions of Figure 3 is re-
solved into its fundamental A.C. component, it is found that the reactant
leads the photochemical light by 90° (phase shift of +90° or of -270°),
and the product lags behind the photolysis light by 90° (phase shift of
-90°). The reactants and products as shown in Figure 3 are as they
would appear in a static gaseous system, with a steady decrease of
reactant and build-up of product. In our actual system, the reactant
and carrier gases flow rapidly through the cell, so that the combination
of stirred-flow and photochemical processes yields reactant and product
concentrations that fluctuate about a steady-state value; the phase
relationships shown in Figure 3 are preserved in this case.
PHOTOCHEMICAL PROCESS WITH FORMATION OF INTERMEDIATE
The mechanism in this case is simply
13
-------
2-m/i
PHOTOCHEMICAL
LIGHT
pr
FAST INTERMEDIATE
TUNED
INTERMEDIATE
.SLOW INTERMEDIATE
I PRODUCT FROM
FAST INTERMEDIATE
^>\PRODUCT FROM
SLOW INTERMEDIATE
PHASE SHIFT
RELATIONS (S)
-90
Figure 3 -
Phase relations between photolysis light, reactant, and
product of a primary photochemical reaction (in static, non-
flowing system) .
14
-------
A + hv + X+**
X+ -v C+"*
Where X is an intermediate (usually, but not necessarily, a free
radical). If the reactant is neither regenerated by the intermediate X
nor attacked by X, then its phase relationship is the same as that of
Figure 3. The phase relationship of intermediate X to photochemical
light varies, depending on the relationship between duration of light
flash and the lifetime of the intermediate in the system, as shown in
Figure 4. Three cases of intermediates are shown: (1) Fast inter-
mediate. The lifetime of the intermediate is very short compared to
duration of light, and in this case the intermediate concentration is
proportional to light intensity at all times, with a zero phase shift.
(2) Slow intermediate. The lifetime of the intermediate is very long
compared to duration of light pulse; the intermediate continues to build
up throughout the light period and continues to decay after the light
turns off. After an induction period the intermediate oscillates with
a triangular wave form about a steady state value, with a low A.C. ampli-
tude and a 90° phase shift. (3) Tuned intermediate. The lifetime of
the intermediate is about the same as the duration of the light; for the
case shown in Figure 4, the half-life of the intermediate is one-fourth
of the duration of the light-on period. The phase-shift of the product
depends on the phase-shift of the intermediate. As can be seen from
Figure 4, a product arising from a "fast intermediate" has a phase-
shift of -90°, and a product arising from a "slow intermediate" has a
phase-shift of -180°. For a "tuned intermediate", the intermediate it-
self will have a phase-shift between 0° and -90°, and the product will
have a phase-shift between -90° and -180°.
COMPLEX PHOTOCHEMICAL REACTIONS
For any mechanism of interest, the expected phase-shifts can be derived
from graphical constructions, such as Figure 3 or Figure 4. However,
for complex systems it becomes more convenient to treat the case
analytically, in terms of a system of simultaneous differential equations,
or numerically on a computer. Results of such an analysis for a simple
15
-------
90 | 9
0
^^
^X
PH
RE
b i
^0 2
>o I
360
+ 10
r\
f
^
+90
- r~
|
UJ
cr
\ «
ASE SHIFT \ o
LATIONS(S) \ %
\. Q-
PHOTOCHEMICAL
LIGHT
PRIMARY
PHOTOCHEMICAL
REACTANT
PRIMARY
PHOTOCHEMICAL
PRODUCT
LIGHT-
-90
Figure 4 - Photochemical reaction with simple, single intermediate
(non-flowing system).
16
-------
general mechanism is given here. The mechanism is:
A + hv * X + Y 1^ = k-j^ IQ[A]
X + Y(M-}A R2 = k2(M) [X] [Y]
X + A -> products R3 = k3[X] [A]
(other reactions not involving A)
Change of species S F = (f/V ([S]-[S] )
s o
by virtue of flow through
reactor
where R's are rates, k's are rate constants, f is volumetric flow rate
and V is volume of reactor, subscript zero refers in concentration of
species as it flows into the reactor. The reactant A is destroyed
photochemically, regenerated by recombination of radicals, and destroyed
by radicals. The total range of phase shifts for this example is given
by Figure 5. Intermediates can have a phase shift between 0 and -90°;
products can have a phase shift between -90° and -180°. So long as
reactants, intermediates, and products have separate wave-length regions
of optical absorption, this method gives kinetic information: life-
times of all intermediates, spectra of intermediates, a direct clue as
to the radical precursors of various products , and a direct clue as to
the interaction of radicals with reactants.
It should be noted that this method gives only the change in reactant
or product concentration per cycle, and thus the signals from reactants
and intermediates are the same order of magnitude, even though reactants
may exceed intermediates by a factor of a million or more.
B. MATHEMATICAL BASIS FOR THE METHOD
The reaction scheme below illustrates the important types of chemical
species in a modulation experiment. The species A through F are stable
molecules; and X, Y, and Z are free radicals. The modulation of each
species is a function of the type of species and of the elementary
17
-------
+90
±180
a^TfJjrj
REACTANT
DESTROYED
BY LIGHT AND
ATTACK
FROM
INTERMEDIATES
REACTANT
DESTROYED
BY LIGHT AND
REGENERATED
FROM
INTERMEDIATES
JTSlHtionS tstV.'SSr* *^1^Q4-Ql *»«! e 1 -i /TV,*- >---,-!,<-.4-= ^4-r-
intermediates, and products in complex photochemical
reaction.
18
-------
A + hv -» 2X (1)
X + B -- Y + C (2)
k2
Y + D -4 Z + E (3)
k3
Z + Z -3- F (4)
Since the concentrations of the species in the chemical system vary
periodically in time, they can be represented by a Fourier expansion of
the form
f(wt) = J [a sin(nt) ] + b
where w is the fundamental angular frequency of the wave obtained from
the period T by the relation w = 2-rr/T. An alternative form of the
Fourier expansion is:
f (wt) = J c sin (nwt + 6 ) + c
u n n
n
The two are related by:
c = (a + b ) ' and 6 = tan (b /a ).
n n n n n n
The coefficients c are amplitudes; the quantities <$ are phase shifts.
In the following analysis, the photolytic lamp is represented by an ex-
pansion involving sine terms only, i.e., all b = 0. Hence all 5 = 0.
For any chemical species, then, 6 is its phase shift relative to the
photolytic lamp.
REACTANT DECOMPOSED BY LIGHT
The differential equation for A is:
f, I 21 , fl
= i- [A]Q - a [A] 2°- + - It odd i sin(nwt) - §-[A] (5)
n
where f is the flow rate into and out of the cell in liters/sec.
V is the volume of the cell in liters
[A] is the concentration of A entering the cell in
molecules/cm
19
-------
2
I is the photon flux in photons/cm .sec
a is the absorption cross-section of. the reactant in
2
cm /molecule
[A] is the concentration of A
u) is the flashing frequency in radians/sec
t is the time in sec
Since to = 27T/T = 2uf where f is the flashing frequency in cycles/sec.,
we can write e = cat = 2irft from which we get de/dt = 2-irf or dt =
d6/27rf, giving
tA] -- T I' odd H sin(ne) }
n
The requirement of a stable D.C. concentration means
ft aI0 fi
( | [A]o - (-£- +L ) [A] = 0
which gives the following simplified differential equation
As long as the concentration modulation, A [A] , is much smaller than the
12
total concentration, [A] , the equation is linear and is easily integrated
giving the concentration modulation
rrT [A]
[A]mod = °2' I, odd ij cos(ne). (7)
IT f n n
20
-------
From the definition of phase shift, we see that the reactant concen-
tration modulation has a phase shift of +90° with respect to the flashing
lamp. The modulation of A is seen to be a triangular wave whose ampli-
tude is inversely proportional to the flashing frequency.
RADICAL FORMED BY THE INITIAL PHOTO-DISSOCIATIVE STEP AND DECAYING BY A
PROCESS FIRST-ORDER IN RADICAL CONCENTRATION
The differential equation describing the radical concentration [X] in
terms of the previously defined quantities a, [A], 01, and I and the
concentration of reactant Bf is:
I 21
o
n
= 2a[A] ( j2 + £ I odd i sin (nut) ) - k^B] [X] . (8)
This can be solved in a straight forward manner like the previous case
to yield:
2al [A] k..[B] k.,
[X] = * T, odd ( ~2 sin(n9) - cos(n0) )/( (=
£ ~ ** -2^1 n 2.
TT f n
al^tA]
When this equation for [X], the radical concentrations, is taken to its
low frequency limit we get:
4crl [A] , aI[A]
' odd sin(ne) +
This is the equation of a square wave with an amplitude of aIQ[A] /k^ [B]
oscillating about a D.C. level of I [A]/k,[B]. Thus the radical concen
tration has a maximum of 2aIQ [A] /k-jjB] the radical concentrations one
obtains from the "steady-state" approximation for [X] . Note also that
the phase shift of the radical concentration is 0°. At high frequencies
we have :
201 [A] , aIo[A]
lim [X] = - 2 - J, odd - ±2 cos(ne) + k°[B] (11)
f-*°° TT f n n 1
21
-------
So [X] becomes a triangular wave with vanishing amplitude oscillating
about a D.C. level equal to one-half the "steady-state" concentration.
The phase shift is -90°.
It is convenient to define the "life-time" of the radical, X, to be
1
Tl kxB
which is the time required for the concentration to drop by a factor
of e. The behavior of the fundamental of [X] at intermediate flashing
frequencies is plotted in Figure 6 as a function of the ratio of the
flashing period (T=l/f) to the radical life-time.
The phase shift of the fundamental of [X] is given by:
6 = tan'1 (b^a.^ = tan"1 (-l/(k;L[B]/2Trf ) ).
So
Thus the radical life-time can be found from just one phase shift
measurement at one frequency if the radical species is known to be
formed in the initial step and to decay by a process first-order in radi-
cal concentration.
RADICAL FORMED BY A RADICAL-MOLECULE REACTION AND DECAYING BY A PROCESS
FIRST-ORDER . IN RADICAL CONCENTRATION
When a radical is formed by the reaction of a preceding radical, X, with
a molecule, B, and is destroyed by reaction with another molecule, D,
the differential equation describing the concentration of the new
radical, Y, is:
dJjYJ-=k1[B] [X] - k2[D] [Y]. (12)
Integration of this equation after changing the variable from t to 6
22
-------
10
T/r = T(k,B)
100
1000
XBL 69'C -59 J3
Figure 6 - Dependence of the fundamental modulation frequency of a
primary first-order radical on the ratio of flashing
period to radical life-time. The amplitude is relative to
the limiting amplitude as T/T approaches infinity.
23
-------
and substituting Eq. (9) for [X] gives
a I [A] 2al [A] k, [B] k [B]k,[D] - (2Trfn)2
° 2' odd ( - ~
TT f n (2irfr n ( (^ [B]/2Trf ) +n)
k?[D] + k, [B] - 2
± - - cos(n9) )/( (k-[D]/2irfr + n ) . (13)
27rf( (k
Since the coefficients b are always negative and the coefficients a
may be either positive or negative depending on the sign of
2
k,[B]k2[D] - (2irfn) , the phase shift of the fundamental of [Y] may lie
anywhere between 0° and -180°. The dependence of the concentration
modulation of Y on flashing frequency is determined by k,[B], k_[D], and
f . A convenient way of looking at the modulation of Y is to plot the
amplitude and phase shift of the fundamental as a function of T/T-^ for
several values of T2/T,. This is done in Figure 7 and 8. When T2/t,
is large, at frequencies where the primary radical X has a phase shift
close to 0°f the secondary radical Y behaves like a primary radical.
Under such conditions the life-time of Y can be easily obtained. When
Tp/T, is small, however, determination of the life-time will be
difficult because the phase shift of the secondary radical Y is deter-
mined, for the most part, by the phase shift of the preceding radical.
At flashing frequencies high enough to impart a substantial phase shift
to the secondary radical due to its own inherent life-time, the
modulation amplitude of the preceding radical is very low. As a result,
the modulation amplitude of the secondary radical is also very small
making detection difficult.
RADICAL WHICH DECAYS BY A PROCESS SECOND-ORDER IN RADICAL
CONCENTRATION
The differential equation for the second=order radical Z is .
4£L = k^[D] [Y] - 2k,[Z]2. (14)
v* t_ -3
24
-------
o
Q_
10
T/T,
100
1000
XBL 6910-5905
Figure 7 -
The dependence of the modulation of a secondary radical on
the ratio of its life-time to the life-time of the
preceding radical. a) ^2/i:l = °-01' b) T2/T1 = °-10''
c) TA = 1.00; d) T/T = 10; e) ^/^ = 100.
25
-------
1.0
0.9
0.8
-g 0.7
5.0.6
o>
0.5
-2 0.4
a*
or
0.3
0.2
O.I
O.I
10
T/T,
100
1000
XBL 6910-5907
Figure 8 - Modulation amplitude relative to the limiting amplitude as
T/T, approaches infinity. a) T/T, = 0.01; b) T/T, = 0.10;
2,
2,
=1 nn-
---
= in
= inn
26
-------
This equation is intractable because the equation is non-linear, and
[Y] and [Z] are both functions of t. A special case of (14) is of
considerable interest
o
(15)
where P and Q are constant. This case can arise from (14) if the
radical Y is very fast and is in phase with the exciting light. Another
case occurs when the radical Z is formed directly from the primary
photolysis of the reactant. A numerical solution to (14) can be readily
obtained if the equation is written in two parts, one corresponding to
the lamp on and one to the lamp off.
^df-=2^f < p - ^ZJ2 > -* < e < o (16)
^£r- = -o-r* Q[Z]2 0 < 9 < TT (17)
Each of these equations can be integrated; the solution with the lamp
on is
[Z], - ( I )* tanh 0+*) o/2
27
-------
the coefficients a and b can be found by the numerical integrations
n n
1 m
a = J [Z] sin(n0.)A6
Ti ** Q 1
n i=l ei x
= i ?
n ^1=1 i
where m is the number of increments of 9. Such an analysis has been
done over a wide range of flashing frequencies. The amplitude and phase
shift behavior of the fundamental as a function of the ratio of flash-
ing period to radical life-time is shown in Figure 9. Because the life-
time of a second-order species depends on concentration, we define "the
radical life-time" to be the half-life of the radical from its steady-
state concentration. The steady-state concentration of the radical Z
is
/20I [A] V"
= (P/Q)T (18)
and the life- time becomes
*
Comparison of the life-time formulae for the first- and second-order
radicals indicated how these kinetically different radicals can be
distinguished experimentally. The life-time of a first-order radical,
given by
T = 1/k.jjB], (20)
is independent cf radical formation rate and inversely proportional
to the concentration of a reactant involved in the radical's decay.
The life-time of a second-order radical given by (19) is inversely
proportional to the square-root of the radical formation rate. Changes
in reactant concentration affect the life- time only be affecting the
radical formation rate. Thus varying reactant concentrations and the
28
-------
T/T = T(PQ)'
1000
XBL 691O-59O9
Figure 9 - Concentration modulation of a second-order radical as a
function of the ratio of flashing period to radical life-
time. The amplitude is relative to the limiting
amplitude as T/T approaches infinity.
29
-------
photolytic light intensity can provide the information necessary to
determine whether the radical decay reaction is first or second order in
radical concentration.
PRODUCT OF RADICAL REACTIONS
The differential equations describing the behavior of the products, C,
E, and F all have the same general form:
'-af = li? (G + H I an sin Sin(n9)
n \
n2] J.
(Bn + «) cos(ne)]/[f'/2irfV)2 + n'] >. (22)
From this equation it is apparent that if f /2irfV « 1, the product
fundamental lags the radical fundamental by 90°. Note also the inverse
dependence of the product modulation amplitude on flashing frequency.
This means that a product amplitude falls off faster with increases in
flashing frequency than does that for a radical. The amplitude and
phase shift behavior of the product of a second-order radical is
presented in Figure 10.
REACT ANT ATTACKED BY A RADICAL
The differential equation for a reactant B which is attacked by a radical
is of the same form as the equation in Section 5 with two minor
30
-------
48
44
40
36
^28
o
h_
524
<
-^20
o>
-o
16
8
4
0
O.I
I 10 100
T/r = T(PQ)l/2
- 90
-100
-110
-120
-130
-140
-150
-160
-170
-180
-o
o
0>
1000
XBL691O-5906
Figure 10 - Concentration modulation of the product of a second-
order radical as a function of the ratio of the flashing
period to life-time of the radical forming the product.
Note the strong dependence of amplitude on T/T.
31
-------
differences: there is a flow-in term (f'/V) [B] ) and a change in sign
because molecules are being lost through reaction. The solution shows
that the reactant modulation leads the radical by 90°. Also the reactant
amplitude has the same frequency dependence as a product, i.e., it falls
off faster than a radical with increases in flashing frequency.
32
-------
SECTION VI
CHLORINE OXIDE RADICALS
A. INTRODUCTION
The C1OO radical was first proposed by Porter and Wright as an
intermediate in the flash photolysis of chlorine-oxygen mixtures.
C12 + hv' -* Cl + Cl
Cl + O2 + M » C10O + M
Cl + C10O > CIO + CIO
Cl + C1OO - * C12 + O2
CIO + CIO » C12 + O2
From absorption spectroscopy in the ultraviolet (2600-3000 A) they
observed the intermediate diatomic radical CIO. Although not observed,
the peroxy radical was postulated to be a short lived precursor to CIO
in their system. Since then, the ultraviolet absorption spectrum of CIO
25
1424
has been observed repeatedly, however, until recently no spectro-
graphic evidence existed for the presence of C10O. Arkell and Schwager"
published an infrared spectrum attributed to matrix-isolated C10O. This
system was chosen originally as a test case for the molecular-modulation
method; but new, interesting results were obtained and the spectra of
ClOO and CIO was explored by infrared, ultraviolet, and mass spectrometry
B. INFRARED SPECTROSCOPIC MEASUREMENTS
Exploratory scans were carried out over a fairly wide spectral region.
The region from 800 to 200 cm" was scanned using a 100-sec time constant.
A reaction mixture of 4 Torr of C12 and 756 Torr of O2 flashed at 1 Hz.
From these preliminary measurements, a single chemically modulated
absorption band was detected at 1410-1490 cm" . This absorption ex-
hibited the phase shift expected for a primary intermediate species, and
05
on this basis, as well as its position, the band was tentatively
attributed to the ClOO radical.
33
-------
The region of 1400-1520 cm" was subjected to a more intensive
investigation by scanning repeatedly at a speed of 8 on" /min. A
spectrometer slit width of 6 mm was used which corresponds to a. spectral
width of 13 cm . Since a time constant of 10 sec was used, a spectral
slit width was covered in about ten time constants. Voltage measure-
ments were made at 1.1-cm intervals. These studies were made with
4 Torr of C12, 189 Torr of O,, and 567 Torr of He. These reactants were
photolyzed with square-wave radiation at 2 Hz. The cumulative results
of 40 such scans are shown in Figure lla as a plot of amplitude and phase
of the modulated signal vs. infrared frequency. To test reproducibility,
another set of 40 scans of the 1400-1520 cm" region was recorded. The
cumulative results of the second set are given in Figure lib. The two
sets of scans were then combined to obtain an improvement in signal-to-
noise ratio. The combined results of 80 scans were curve-smoothed and
corrected for background to obtain true absorbance. The resulting
spectrum is shown in Figure lie. Curve smoothing consisted of applying
a simple three-point moving average to the raw data. Since a spectral
slit width is covered by 12 data points, a three point average will not
introduce distortion.
Although a chemically modulated absorption spectrum that could be
attributed to the CIO radical was not detected in the exploratory spectro-
scopic measurements, attempts were made to observe its spectrum by means
of multiple scanning over narrower spectral regions for longer periods
26
of time. On the basis of Porter's prediction of the CIO ground-state
vibrational frequency (868 cm ), the observed frequencies (936, 945
cm ) of a matrix-isolated species tentatively identified as (CIO)_ by
93 ^
Rochkind and Pimentel, and the strong peak we previously observed
_ i
between 930 and 980 crn *", it can be expected that the absorption band of
CIO most probably lies in the region of 850-980 cm . Therefore, this
region was scanned extensively, over prolonged periods of time and under
a variety of experimental conditions, in an effort to observe a modulated
infrared absorption spectrum of CIO. . However, these attempts proved
fruitless, even under seemingly the most favorable conditions.
34
-------
1430 1450 1470 14901430 1450 1470
IR Frequency (cm'1)
1430 1450 1470 1490
Figure 11 - C1OO infrared molecular modulation spectrum observed with
4 torr C12, 189 torr O_, 567 torr He, 2 hertz square-wave
ultraviolet photolysis. a. 40 scans b. Duplicate, another
set of 40 scans, c. Smoothed cumulative measurements, 80
scans (a and b). | 1 indicates spectral slit width.
35
-------
C. ULTRAVIOLET SPECTROSCOPIC MEASUREMENTS
Chlorine at 1.8 Torr in the presence of 1 atm of oxygen was photolyzed
with square-wave excited lamps at 1 Hz. The rate of light absorption
14 3
was estimated to be 2.4 x 10 photons/(cm sec). A modulation was
o
observed at wavelengths between 2250 and 2900 A, Figure 12. The measured
phase angles lie in the quadrant expected for intermediates; however,
the change in phase observed with wavelength indicates the presence of
two or more species with overlapping absorption spectra. The small
o
phase angle at 2300 A represents a "fast" intermediate, that is, one
whose lifetime is short compared to the flashing frequency of 1 Hz. The
o
larger phase angle at 2800 A indicates a much slower intermediate. The
observed signal is a vector sum of the various species present.
14-24
Using literature values for the rate of recombination of CIO and
21
for its cross section, we estimate that this species should be seen in
the modulation experiment as a slow intermediate. From flash photolysis
o
studies, CIO is known to have a banded structure from 3000 A to the
° 26
dissociation limit at 2630 A. Porter has published the locations of
the vibrational band heads in this region. Experimentally it was found
that reducing the total pressure to 50 Torr and flashing at 0.25 Hz in-
creased the modulation amplitude. This allowed the spectral slit width
o
to be reduced from 13 to 2 A. Figure 13 shows the experimentally ob-
served modulation amplitude together with the positions of the CIO band
heads as determined by Porter. The two are in excellent agreement.
Thus one of the species being observed is definitely identified as the
CIO radical.
It was desired to obtain a spectrum of the fast intermediate with a
minimum of interference from CIO. Since CIO is a slow intermediate, its
modulation amplitude can be suppressed by increasing the flashing fre-
quency of the photolyzing radiation. As can be seen from Figure 14, the
amplitude of CIO modulation begins to fall off at a lower frequency
than does the unknown, because of the difference in 1 i-Fetl??, If the
flashing frequency is chosen approximately equal to the lifetime of the
36
-------
d>
o
CL
-20°
-40°
-60°
3x|0
-4
OJ
T3
13
<
c
o
2x10
-4
-g IxICT4
I
o
o
ro
OJ
O
O
«d-
CJ
O
O
ir>
OJ
o
o
CO
OJ
o
o
r<-
OJ
o
o
oo
OJ
o
o
en
OJ
Wavelength
Figure 12 - CIO modulation spectrum in ultraviolet showing presence of
j£
two or more intermediates by virtue of change of phase
shift with wavelength.
square wave photolysis,
1.8 torr Cl,
1 atm 02, 1 hertz
37
-------
2796.0 -
2771.6 -
2749.5 -
o
-------
O.lh
O.I 0.25 |
40 100
F sec"
Figure 14 - Calculated relative modulation amplitude for the two CIO
j£
intermediates observed in this system (based on rate
constants deduced later in this article, presented here to
demonstrate how unwanted overlap of CIO can be reduced to
zero at high frequencies and how phase of fast radical can
be reduced to known zero value at low photolysis
frequencies).
39
-------
fast intermediate, this is much faster than the CIO lifetime. A
spectrum taken at 32 Hz is given in Figure 15. The modulation spectrum
o o
extends from 2250 to 2700 A with a maximum about 2475 A. The phase
angle is almost constant (-60 to -65°) over this range, indicating CIO
modulation was effectively suppressed.
At lower frequencies the modulation is no longer simple but a vector sum
of several amplitudes and phases. If this complex spectrum consists of
only two species and if the phase of each is known, then the complex
modulation spectrum can be factored into two components. An experiment
was carried out at 0.25 Hz, Figure 16. The observed modulation shows a
o o
maximum at 2600 A and the phase angle varies from -12° at 2300 A to -34°
o
at 2800 A. From Figure 14 it can be seen that if an intermediate has a
phase of -60° at 32 Hz, it would have essentially zero phase, and the
o
plateau of -34° at 2800 A is taken as the phase of CIO, then the
observed amplitude can be decomposed into two separate spectra, one for
CIO and one for the faster species, Figure 16. The spectrum of the
fast radical is shown in Figure 17.
A detailed comparison of the kinetics of the fast radicals with that
observed for C10O in the infrared system identified the fast radical as
C1OO. The cross sections for absorption of ultraviolet radiation by
CIO and C1OO are given in Table 1.
D. MOLECULAR MODULATION MASS SPECTROMETRY
Matheson research grade Cl- and 02 and high purity Ar were used directly
without further purification. Gas mixtures of appropriate proportions
were stored in a 5 liter glass bulb, painted black to prevent photo-
reaction frcir. room lights. Grease-free stop cocks were used throughout
the system. Except for stainless steel in the flow-control valves and
certain fittings, all surfaces were glass and Teflon from storage bulb
to reaction cell. The pressure of the gas reaction mixture in the
reaction cell was controlled by a Granville-Phillips series 9100
variable leak valve and was monitored by a Pace model CD-25 pressure
40
-------
o>
co
0
_c
Q.
-50°
-70°
-------
Wavelength
Figure 16 - Resolution of complex spectrum into two components of
constant phase at low (1/4 hertz) .photolysis frequency.
(Compare Figure 14). The 34° component has been identified
as CIO (compare Figure 13).
42
-------
to
"c
o>
'o
o>
o
o
o
to
_o
0>
en
o 1/4 cps
32 cps
I
I
o
o
OJ
OJ
o
o
ro
OJ
O
O
^r
OJ
o
o
m
OJ
o
o
CD
OJ
o
o
r-
OJ
o
o
CD
OJ
Wavelength
Figure 17 - Ultraviolet spectrum of C1OO radical as obtained by two
different procedures.
43
-------
Table 1. Absolute cross section for absorption of
ultraviolet radiation by the free radical C10O and CIO in
18 2
units of 10 cm .
o
A, A
2250
2300
2350
2400
2450
2500
2550
2600
2650
2700
2750
2800
C100, a2(A)
2.60
4.91
7.80
10.5
12.7
13. 3b
12.4
10.0
7.28
5.11
3.40
2.28
CIO, ^(X)0
0.641
0.846
1.33
1.91
2.67
3.56
4.45
5.27
5.68
5.60
4.85
4.71
a-d In I = a[X]L, where concentrations are in particle/cc,
L is path length in cm, and In is logarithm to base e.
64.
21
bThis work, eq 64. cBased on 4.83xlO~18 cm2 at 2577 A by
Clyne and Coxon
44
-------
transducer. At a cell pressure of 10 torr the typical gas flow rate
was 0.008 standard cc/sec or 2x10 molecules/sec, and the leak rate
from the pin hole was 1.5x10 molecules/sec, of which about one per
cent passed through the skimmer for analysis. Five chopping speeds for
the photolytic light were used: 16, 8, 4, 2 and 1 sec/cycle. For each
chopping speed, from 3 to 30 measurements were taken, and each measure-
ment involved 160 to 1600 seconds of integration time. Data were taken
for both ascending and descending chopping speeds, and no significant
difference was found. The photolyzing light flux was determined by NO_
photolysis, and by a detailed comparison of absorption cross section be-
tween NO- and C12 over the range of photolyzing light. The effective
cross sections for light absorption in this range were found to be
4.65xlO~19 cm2 for N00 and 1.44xlO~19 cm2 for Cl_. The photolyzing
17 2
light flux was 1.35x10 photons/cm sec.
A preliminary experiment was done with a reaction mixture of C12/02 =
1/19 at 20 torr flashed at a speed of 16 sec/cycle. Results are given
in Table 2. Using the ion-counting method, we observed strong modula-
tion signals at mass-to-charge raties (m/e) of 51 and 53. The phase
shifts were identical for both signals, and the amplitude at m/e = 51
was three times that at m/e = 53, very strongly indicating it to be CIO
37
and CIO. The phase of the signal at m/e = 51 did not change when the
ionization energy was reduced from 40 to 20 volts, indicating that the
signal was not a super-position of one mass number from two different
molecules. To check against the possibility of thermal-acoustical modu-
lations at all mass numbers, we added one-tenth per cent of Ar and
searched for a. modulation signal at m/e = 40. As can be seen in Table 1,
the amplitudes were very small and the phases were random. A weak
modulation signal was found at m/e = 67, Table 3. The small phase
shift, -6.7 degrees, is kinetically correct for the C1OO radical.
35 35
Attempts were made to detect C1202 (m/e = 102) and C12° (m/e = 86>'
but no modulation signals were obtained.
Two series of systematic experiments were performed to measure the
pressure dependence and inert gas effect on the CIO disappearance rate.
45
-------
Table 2. Comparison of modulated signal of
CIO and unmodulated signal of Ar.
m/e Electron
Energy
volts
51 40
53 40
51 20
40 40
(Ar) No. In-Phase
Cycle Counter
torr eye count
0 10 31490
31099
33639
29815
0 10 11083
9506
10696
11885
0 20 23213
25668
23775
24010
0.02 10 -272
-163
+ 75
+ 336
-139
Out-Phase
Counter
count
9961
8068
9463
8212
2958
3231
2571
3325
5953
7748
7038
8575
-301
421
+ 198
-219
+284
Ampli-
tude
count/sec
206
201
218
194
Ave 205
73
64
69
77
Ave 71
77
91
79
84
Ave 83
1.8
2.8
1.2
1.3
2.0
Ave 1 . 8
Phase
degree
-17.5
-14.5
-15.7
-15.4
-15.8
-15.2
-17.9
-13.3
-16.2
-15.7
-14.3
-16.8
-15.4
-17.9
-16.1
+131.7
-111.2
-69.5
+ 33.0
-116.1
-26.5
[C12] - 1 torr, [Q2] = 19 torr, chopping period = 16 Hz,
46
-------
Table 3. Comparison of modulated signal of C100 and CIO.
m/e Electron
Energy
volts
67 40
51 40
[C12] = 2 torr,
(Ar) No. In-Phase
Cycle Signal
torr eye count
0.04 800 14206
16983
24733
11825
0.04 200 37295
36142
36998
35872
[O2J =38 torr, chopping
Out-Phase
Signal
count
1569
2442
2195
1496
48872
49102
47767
49413
period = 1 Hz.
Ampli-
tude
count/sec
18
21
24
15
Ave 19
307
304
295
306
Ave 303
Phase
degree
-6.3
-8.2
-5.1
-7.2
-6.7
-52.7
-53.5
-52.2
-54.3
-53.2
47
-------
In one series the ratio C12/02 was maintained at 1/19, the total
pressure was varied from 40, 20 10 to 5 torr, and the chopping speeds
were 16, 8, 4, 2 and 1 sec/cycle. In the second series the reactants
were held constant (C12 = 0.75 torr, O~ = 6.75 torr), and Ar was varied
as 0, 5, 17.5, and 42.5 torr to give total pressure of 7.5, 12.5, 25,
and 50 torr. The phases for both of these series are given in Table 4
and some of the data are plotted in Figures 18 and 19.
48
-------
Table 4. Observed phases as a function of
chopping frequency and concentrations.
Pressure, torr
Total
40
20
10
5
7.5
12.5
25.0
50.0
'C12
2
1
0.5
0.25
0.75
0.75
0.75
0.75
°2
38
19
9.5
4.75
6.75
6.75
6.75
6.75
Ar
0
0
0
0
0
5
17.5
42.5
Flashing period, sec/cycle
16
-10.8
-15.9
-21.6
-40.7
-20.5
-17.0
-13.7
- 6.4
8
-15.5
-24.3
-34.5
-63.1
-30.1
-23.5
-17.8
-11.7
4
-23.4
-36.3
-53.8
-74.5
-45.5
-39.8
-33.1
-26.5
2
-36.7
-60.9
-77.1
-91.3
-66.0
-61.3
-54.8
-46.1
1
-53.1
-76.3
-96.4
-
-80.5
-78.8
-75.4
-67.0
49
-------
-100
-90
-80
-70
r -eo
O
o -50
*o
o>
| -40
Q.
-30
-20
- 10
O
I
I
I
IS 8 42 i
Flashing Period sec/cycle
0.5
Figure 18 - Observed points and calculated curves for the phase shift
of CIO in a mixture of 5% C12 and 95% O*. Total pressure
in torr: ^, 40; 4> , 20; +, 10; A, 5.
50
-------
-100
-90
-80
-70
r -so
§-50
O)
8 -40
-30
-20
-10
0
I I I I
I I
I
I
I
16 8 4 2 I
Flashing Period sec/cycle
0.5
Figure 19 - Observed points and calculated curves for the phase shift
of CIO in a mixture of C12 =0.75 torr, 02 = 6.75 torr, and
Ar in torr: 0, 17.5; A, zero.
51
-------
SECTION VII
SPECTRA AND KINETICS OF THE HYDROPEROXYL FREE RADICAL
A. INTRODUCTION
The occurence of HOO as an intermediate has been postulated in kinetic
studies of the hydrogen-oxygen reactions, ~ in photochemical studies
2437
of ozone in the presence of water or in the presence of hydrogen
3 8 39
peroxide, and in the photolysis of hydrogen peroxide. In a study
of the flash photolysis of hydrogen peroxide with direct observation of
the hydroxyl ]
by photolysis
40
the hydroxyl radical, Greiner showed that hydroxyl is formed directly
H2°2 + hv * 2H°
and that it decays by a process first order in HO and first order in
H2°2' Presumably
HO + H202 -» H2O + HOO
o
Greiner searched for an absorption by HOO in the region 2500 - 10,000 A
but found nothing. The overall kinetics of the photolysis of hydrogen
peroxide indicates the following reaction as the radical terminating
step
HOO + HOO -» H2O2 + 02
41
Direct observation of HO2 was first accomplished by Forier and Hudson
who succeeded in producing hydrcperoxyl radicals by the reaction
H + O2 + M -» H02 + M
and detecting them by mass spectrometry. This radical has also been
42
observed mass spectrometrically by Robertson who added O0 to a stream
of H atoms, Ingold and Bryce who reacted 02 with H atoms and with
52
-------
44
methyl radicals, and Fabian and Bryce who studied the reaction of
41
methane with oxygen molecules. Foner a-nd Hudson have reported observ-
ing the mass spectrum of H02 formed in six different ways: the reactions
of H atoms with O2 and H2O2, of 0 atoms with H2O2, of OH radicals with
H2°2' tlie Pnot°lysis °f H2^2' an(^ a l°w~P°wer electrical discharge of
H202.
45
Spectroscopic detection of HO2 has been achieved by Milligan and Jacox
using the matrix isolation technique. They photolyzed an HI-O2 mixture
in an argon matrix at 4°K and obtained infrared absorption peaks in the
regions 1040 - 1101 cm"1, 1380 - 1390 cm"1, and at 3402 and 3414 cm"1.
These absorptions were attributed to the O-O stretching, HOO bending,
and H-O stretching vibrations, respectively. The spectrum has been con-
firmed by Ogilvie in an argon-:
quency assignments are reversed.
firmed by Ogilvie in an argon-neon matrix at 4°K, but his low fre-
The transient ultraviolet absorption spectrum of H02 has been observed
following the pulsed electron irradiation of oxygenated aqueous solutions
47
by Czapski and Dorfman. The spectrum of the radical in solution begins
o o
at approximately 3000 A and has a maximum at 2300 A with a molar ex-
tinction coefficient a = 1150 M~ cm Subsequent experiments in pulse
radiolysis and flash photolysis of aqueous solutions have confirmed the
47
u.v. spectra.
48
Troe observed the ultraviolet spectrum of gaseous HOO between 2100 and
o
2800A at 1100°K in the thermal decomposition of hydrogen peroxide in a
49
shock tube. Hochanadel and coworkers have observed the ultraviolet
spectrum of HOO and the rate constant for the reaction
HOO + HOO -» H2O2 -I- 02
Also, Foner and Hudson have obtained a value for this rate constant.
B. THE INFRARED SPECTRUM AND KINETIC RESULTS
Molecular modulation absorption spectra were recorded while an ozone-
hydrogen peroxide mixture in helium was being illuminated by the
53
-------
photolytic lamp flashing at 1.4 Hz. The ozone concentration was about
14 15
5 x 10 molecules/cc; the peroxide concentration was about 5 x 10
molecules/cc. The two reactants were diluted by helium at atmospheric
pressure. The spectrometer slits set to 6 mm correspond to an average
resolution of 12 cm" . Figure 20 shows the modulated absorption of
ozone from 1050 to 1075 cm ; the phase shift of the ozone modulation
is +85°, which is in the quadrant proper to a reactant destroyed by both
the photolytic light and radical attack. Absorption by a second species
displays a maximum at 1127 cm . The phase shift of the second species
is -35° which is proper for a radical intermediate. The breadth of the
region with radical phase shift (1080 to 1140 cm~ ) indicates that the
absorption by the second species is quite broad and may well extend be-
low 1080 cm , but the ozone absorption is so intense that it obscures
all else.
The photolysis of ozone and hydrogen peroxide at 1.4 Hz under the same
chemical conditions as above was studied in the infrared region 1340 to
1500 cm (Figure 21) . The resolution in this region with the slits
opened to six mm is 12 cm . Another spectrum of this region, obtained
under a resolution of 8 cm , is shown in Figure 22. Both spectra show
a strong absorption with a radical phase shift between 1350 and 1440
cm" . An absorbance minimum occurs at 1390 cm . The sharp peaks at
1408 and 1425 cm"1 in Figure 21 and at 1368, 1378, and 1425 cm in
Figure 22 are narrower than the resolution of the spectrometer and can-
not be regarded as spectral features. The spectrum in this region then
consists of a pair of peaks centered at 1390 cm and separated by about
42 cm"1.
The last region in the infrared which shows modulated absorption is
between 3300 and 3605 cm" . This region was studied under the same
chemical conditions as the previous two at a flashing frequency of 1.4
Hz. The resolution in this region varied from 18 to 25 cm , at 3300
cm" to at Jbuo cm ~. Tne modulated absorption spectrum shown in
Figure 23 is the result of sixteen scans and a three point curve smooth.
54
-------
80
60
40
20
0
-20
-40
-60
-80
Phase Shift vs. Infrared
Frequency
Amplitude vs. Infrared
Frequency
A,
_L
i
1050
1075
1100 1125
1150
Figure 20 - The modulated infrared absorption spectrum (1050 to 1150
""" ) obtained during the photolysis of ozone in the
cm
presence of hydrogen peroxide.
55
-------
IT
CO
CO
a:
in
Q_
CJ
CE
CD
OC
O
cn
CD
a:
CE
_
cc
-180
1380 1400 1420
WRVE NUMBER
1-00
.20
0.00
1500
1340 1360 1380 1400 1420
13140 136° WRV£ NUMBER
1440
1460
1480
1500
XBL 6910-5762
Figure 21 -
The modulated absorption spectrum obtained during the
photolysis of ozone in the presence of hydrogen peroxide
_ 1 A
at 1.4 Hz. Six repetitive scans at 12 cm
and a three-point curve smooth.
resolution
56
-------
180
CO
cr
X
0
z
cr
m
cc.
o
en
CD
CE
-IBO
-so
1.00
.80
.60
1340 1360 1380 1400 1420 1440 1460
WRVE NUMBER ECM**-!!!
uj .go
CC .20
UJ
cc
o.oo
1480
1500
1340 1350 1380 1400 1420
WRVE NUMBER
1440
1460
1480
1500
XBL 6910-5801
Figure 22 - The modulated absorption spectrum obtained during the
photolysis of ozone in the presence of hydrogen peroxide
at 1.4 Hz. Eleven repetitive scans at 8 cm"1 resolution
and a three-point curve smooth.
57
-------
CO
UJ
CO
UJ
cc
CD
cc
D
CD
CD
CE
CE
_J
UJ
CC
-90
-180
1-00
3300 3325 3350 3376 3401 3427 3152 3177 3503 3528 3554 3579 3605
WflVE NUMBER
0.00
.20
3300 3325 3350 3376 3401 3427 3152 3177 3503 3528 3554 3579 3605
WRVE NUMBER ECMxx-13
XBL 6910-5761
Figure 23 -
The modulated absorption spectrum obtained during the
photolysis of ozone in the presence of hydrogen peroxide
at 1.4 Hz. The spectrum is tne average or sixteen scans
and a three-point curve smooth.
58
-------
The absorption maximum at 3600 cm" is very similar to the hydrogen
peroxide peak and has a +90° phase shift due primarily to hydrogen
peroxide. At lower infrared frequencies new absorption peaks are evi-
dent which cannot be assigned to either hydrogen peroxide or water
because the new absorption is strongest where both water and peroxide
absorption is weak, i.e., between 3350 and 3450 cm"1. Furthermore, the
phase shift of the new absorption is displaced toward the radical
quadrant, but the phase shift never reaches a constant value typical of
single species absorption so there must still be some peroxide component
in the signal. In addition to the main peak which extends from 3380 to
3440 cm" , there are several smaller peaks at 3550, 3505, 3470, 3345,
and 3312 cm . These peaks are not present in the spectrum of water or
hydrogen peroxide. The pattern of peaks does not fit the position of
45
Q and P branch lines of the OH radical which is known to be present
in this chemical system. Later it will be shown that these peaks are
consistent with the best current estimates of the structure of HOO.
During the photolysis of hydrogen peroxide the dependence of the
modulated absorption peaks at 1075, 1120, 1373, and 1420 cm on flash-
ing frequency was studied. In these experiments the infrared spectro-
meter was set at a fixed I.R. frequency and the modulation signal was
recorded by the computer every 10 sec for a period of ten to thirty min.
The hydrogen peroxide concentration was 5 x 10 molecules/cm . The
helium carrier gas at a pressure of one atmosphere flowed through the
cell at 8.4 1/min. Room temperature during these runs was 22°C. Flash-
ing frequency was varied from 1/4 Hz to 4 Hz.
The observed phase shifts, each the result of from ten to thirty min.
of averaging, are plotted in Figure 24. The average phase shift at each
flashing frequency is indicated in Figure 24 by an arrow. The solid
curve in the figure is the calculated curve for a second-order radical.
The position of the calculated curve on the time axis gives the quantity
(PQ)1/2; it is 5.55 sec"1.
59
-------
10
0
-10
£ -20
8 "3°
£ -40
| -50
| -60
-70
-80
-90
J I I
O.I 0.2 0.5 1.0
T (sec)
10
XBL 691 0-5899
Fiaure>
of flashing period.-
= 1373 cm"1; A =
peas as a. uncton
'1 cm"1;
x = 1075 cm"1; O= 1120
-1 *»» ^"^
1420 cm ;O = Average of all points.
-60-
-------
C. THE ULTRAVIOLET SPECTRUM
-ak. ».
Two series of experiments were carried out in obtaining the ultraviolet
o
spectrum of HOO between 1950 and 2500 A. In one series H0O0 in one
o 22
atmosphere of He was photolyzed at 2537 A. The mechanism is presumably
reactions (1), (2), and (3). In the other system, chlorine was photo-
o
lyzed between 3200 and 3800 A in the presence of H,O9. The expected
& £t
elementary reactions are
C12 + hv -» 2C1
Cl + H_O -»- HC1 + HOO
HOO + HOO -> H2O2 + O2
In a preliminary experiment, a modulation absorption spectrum was
O l /-
obtained from 2450 to 2000 A during the photolysis of H202 at 1 x 10
molecules/cc, at 1 Hz photolysis frequency, in an atmosphere of helium
o
flowing at 8.4 1/min. The modulation absorption was measured at 50 A
o
intervals with a spectral resolution of 13.3 A. The spectrum showed an
o
absorption band with a peak at 2100 A, with phase shift between -15.8°
and -24.0° (which is in the radical quadrant), and a variation of phase
shift with wave length that suggests the presence of two modulated
species. The same spectral region was examined at 1/4 Hz, and the re-
sults are shown in Figure 25. At this frequency the phase shift of the
modulation displays a strong dependence on wave length, clearly indi-
cating the presence of two species one of which is relatively more
important at 1/4 Hz than at 1 Hz.
Since the phase shift at 1/4 Hz is in a reactant quadrant and since
reactant amplitudes show a stronger dependence on flashing frequency
than do radical amplitudes, the species gaining importance at 1/4 Hz
must be the reactant hydrogen peroxide. The shape of the hydrogen
peroxide absorption spectrum is well-known ' and agrees closely with
the spectrum obtained in our laboratory. It is evident that hydrogen
peroxide could be contributing to the modulated spectrum.
61
-------
40
30
20
10
o
£
10 12
b
CD
10
=5. 8
E
o
"o
2000 2200 2400
Wavelength (A)
XBL 691 0-5897
Figure 25 -
The modulated ultraviolet absorption spectrum obtained
during the photolysis of hydrogen peroxide by one lamp at
1/4 Hz.
62
-------
o
The response of the modulated absorption at 2200 A to variations in
flashing frequency was studied from iy4 to 32 Hz. The conditions of the
experiment were set as closely as possible to the conditions under which
the spectrum (Figure 25) was obtained; the hydrogen peroxide concentra-
tion was 1.13 x 10 molecules/cm . The phase shift is plotted against
log T where T is flashing period in Figure 26; the theoretical second-
order radical curve is included in the figure for comparison with the
observed results. At low frequencies the observed phase shift is too
positive for a pure radical signal. Apparently a reactant is becoming
progressively more important as the flash period increases, agreeing
with the previous results.
Two modulation spectra that overlap in wave length can be resolved if
the kinetic behavior (phase shift) of the two species is substantially
different. The mechanism is rewritten in the following way
H2°2 + hv "*" 2HO
2HO + 2H202 -> 2H2O + 2HOO (fast)
2HOO + H2O2 + O2
Three molecules of H2O2 disappear with absorption of one photon, and
two molecules of HOO are formed. Two molecules of HOO disappear accord-
ing to the third reaction, and one molecule of H202 is reformed. From
this mechanism the differential equations are
d[H70 ]
£* £ A r v* s~\ 1 T i i_
= 2a2[H202] I - 2k3[HOO]2
The steady-state relations are
[H00]
gs
Tss =
63
-------
The values of light intensity I, average peroxide concentration, and
peroxide absorption cross section a. are known. For square wave photo-
lysis of period T, the concentration profiles of both H,0_ and HOO can
A 4*
be calculated uniquely for a given value of the ratio of period T to
steady-state half-life T
ss
p = T/TSS
These concentration profiles were computed for a wide range of values
of p, 0.05 to 333; and fourier analyses of the concentration profiles
gave the fundamental amplitude and phase shift. At flashing frequencies
of 8 Hz or greater, the reactant amplitude is so small that it may be
neglected entirely; and the resulting plot of phase shift against flash-
ing frequency (Figure 26) gives the value of p. With our tables of
phase and amplitude as a function of p, the observed signal can be de-
composed into radical amplitude and reactant amplitude from the observed
net phase shift. This procedure is tested for the 1/4 Hz data of
Figure 25, and the results are given in Figure 27. The observed resolu-
tion of amplitude from the experimental data agrees very well with the
theoretical curve, which justifies the application of this treatment of
the data. With the availability of this method of resolving the over-
lapping spectra, an extensive set of data were taken over a wide range
of wavelength and photolysis periods.
In another set of experiments, modulation spectra of the entire 2450 to
o
2000 A region were obtained at flashing frequencies from 1/4 to 16 Hz.
o o
The spectra were acquired at 50 A intervals with a resolution of 13.3 A.
The hydrogen peroxide was again carried into the reaction cell by a
stream of helium flowing at 8,4 1/min. Two photolysis lamps were used.
Because of the greater light intensity, the hydrogen peroxide concen-
tration was lower than in the previous experiment. The concentration
of peroxide in the two-lamp experiment was 8.5 x 10 molecules/cm .
The observed spectra acquired at the various flashing frequencies in
this experiment are shown in Figures 28 and 29. Each spectrum haa iLs
o o
most negative phase shift between 2150 and 2250 A usually at 2200 A.
o
The phase shift at 2200 A is plotted against log T and the model curve
is fitted to the high frequency data, i.e., 2, 4, and 8 Hz. The spectrum
at 16 Hz is not used here because it is quite noisy. The 1/4 Hz spectrum
64
-------
20
10
0
-10
-20
£-30
^-40
1-50
-60
-70
-80
-90
1/16 1/4 1.0 4.0
Flashing Period T(sec)
XBL 6910-590O
Figure 26 - The dependence of the phase shift of the modulation at
2200 A on flashing period. The curve is the calculated
curve for a primary second-order radical. Only one
photolysis lamp was used.
65
-------
1/16 J/8 1/4 1/21 24
Nominal Flashing Period T(sec)
XBL 691 O - 5904
.figure 2v - Comparison ot radical and reactant amplitudes extracted
from the data with the amplitudes predicted by the mechanism.
O = Radical amplitude; CU = Reactant amplitude. Solid
curves are the amplitudes predicted by the mechanism.
66
-------
50
40
30
20
10
0
-10
I -20
CO
1-30
f -40
-50
-60
-70
-80
-90
2000 2200 2300
Wavelength (A)
XBL 691 O- 5910
Figure 28 - The modulated absorption phase shifts from 2450 to 2000 A
obtained during the photolysis of hydrogen peroxide by two
lamps. O= V4 Hz; A = 1/2 Hz; H = 1 Hz; A = 2 Hz;
0=4 Hz; f = 8 Hz; O = 16 Hz.
67
-------
15
14
13
12
X
I 9
i. e
I 7
o
2000 2200 2400
Wavelength (A)
w
Figure 29 - The modulated absorption amplitudes from 2450 to 2000 A
obtained during the photolysis of hydrogen peroxide by
two lamps. O= V4 Hz; A = 1/2 Hz; = 1 Hz; A = 2 Hz;
= 4 Hz; = 8 Hz; CD = 16 Hz.
68
-------
is decomposed into -8.2° and 93.7° components. The 93.7° component is
fit by the method of least squares to the known hydrogen peroxide
spectrum to obtain the amplitude of the peroxide modulation at 1/4 Hz.
The phase shift and amplitude behavior of the reactant as a function of
flashing frequency is known from the model, so the reactant contribution
at each flashing frequency can be calculated and subtracted from the
experimental data to get the modulation signal of the radical at each
flashing frequency and measured wave length.
o
The average phase shift of the radical over the 2000-2500A region is
plotted against log T (Figure 30).
A curve results that fits the second-order radical function within the
1/2
experimental error. The position of the curve gives a value of (PQ) '
of 14.5 sec
D. THE DISPROPORTIONATION RATE CONSTANT
The differential equation for HOO radical formation and second-order
decay may be written (compare Eq. 16)
§t - p - a*2
The correlation of phase shift with photolysis period (Figures 24, 26,
30) gives the product PQ. The difference in H_02 decomposition in the
light and in the dark gives P. The rate constant k., is thus given by
t _ Q _ (PQ)
K3 2 ~ 2P
With respect to this rate constant, the results from the infrared
(Figure 24) and ultraviolet (Figures 26 and 30) studies are combined.
The value of the rate constant is
-12
k = (3.6±0.5) x 10 cc/particle-sec.
where k_ is defined by
69
-------
CO
o>
to
o
0
-10
-20
-30
-40
-50
-60
-70
-80
-90
1/16
1/4
I
Flashing Period T (sec)
XBL 6910-5902
- The phase shift of the radical VG the flashing period.
The curve is the calculated curve for a primary second-
order radical.
70
-------
E. ABSORPTION COEFFICIENTS OF THE HYDROPEROXYL RADICAL
The absorption coefficient a is defined by the equation -AI/I = aAnl
for small changes in the transmitted light intensity through an absorb-
ing medium of length 1 caused by small changes in the concentration of
the absorbers An. In our spectroscopic systems, the optical path
length is known and the modulated absorbance AI/I is measured. In the
limit of very long photolysis periods T the radical amplitude is given
by
An = [H001 = (P/Q)1/2
o o
The observed quantities are P and PQ. Thus
An = P/(PQ)1/2
The absorbance AI/I is measured at finite flashing frequencies and must
be corrected to the low frequency limit. The observed phase shift and
the model second-order radical curves specify the ratio of measured
absorbance at each flashing frequency to the absorbance at the low fre-
quency limit. The experiments in the ultraviolet cell provide a
considerable body of data from which to calculate the cross-section for
absorption. From the two-lamp experiment the spectra obtained at 1/4,
1, 2, and 4 Hz were used to determine the absorption coefficient. (The
amplitude of the spectrum at 1/2 Hz is not used because it is abnormally
small). Both the spectra at 1/4 and 1 Hz obtained from one-lamp experi-
ments were used for another determination of the absorption coefficient.
At the absorption peak of the hydroperoxyl radical its absorption
182
coefficient is to the base e: 4.5 x 10 cm /molecule; to the base
18 2
ten: 2.0 x 10 cm /molecule. The absorption coefficient of this
4 0
radical at its peak has been found in aqueous solution to be (base 10):
18 2
1.9 x 10 cm /molecule. The absorption cross section as a function
o
of wave length between 1950 and 2450 A is given by Figure 31.
The infrared absorption of HO_ is strongest at 1420 cm where AI/I is
A
observed to be 1.56 x 10 under conditions which give the radical a
71
-------
X
OJ
o
o
CO
to
2 2
o c
c
o
"QL
o
I I
b"
0
2000 2200 2400
Wavelength (&)
XBL69IO-5896
Figure 31 -
The ultraviolet spectrum of the hydroperoxyl radical,
-dl = alLdN. $= Average of all the ultraviolet spectra
obtained fro, the photolysis of hydroqen peroxide at
2537 A; A = The spectrum from the photolysis of chlorine
9
in the presence of hydrogen peroxide at 3600 A.
72
-------
phase shift of approximately -40°. In the low frequency limit AI/I =
-4
2 x 10
limit is
-4
2 x 10 . The concentration modulation of HO in the low frequency
A[H02] = (P/Q)1/2
12 3 \ 1/2
4.5 x 10 molecules/cm ,sec\ '
' -12 3 ' /
x 10 cm /molecules. sec/
= 8 x 10 molecules/cm
The optical path is 48 meters; the absorption coefficient to the base
-20 2
e is 5 x 10 cm /molecule. Because of uncertainties in the hydrogen
peroxide concentration when the spectra was obtained, this figure may
be in error by as much as a factor of two.
73
-------
SECTION VIII
PHOTOLYSIS OF NITROGEN DIOXIDE
A. INTRODUCTION
The photolysis of nitrogen dioxide is a key reaction in photochemical
52 53
air pollution , is important in stratospheric photochemistry, and is
often used in the laboratory to measure light intensity between 300 and
400 nm. The reaction is complex, and there is disagreement in the
literature as to the value of certain rate constants and rate-constant
ratios. There is agreement in the literature "~ that the following
mechanism describes the overall kinetics of the photolysis of small
amounts of NO_ in an inert carrier gas.
Reaction Rate
aI[N02]
a.
1.
2.
3.
4.
5.
6.
7.
8.
3
NO2 + hv -» NO + O( P)
NO + O + M + NO2 + M
NO0 + O -» NO + O0
2 2
NO2 + O + M -> NO3 + M
NO3 + NO * 2NO2
N03 + N02 + M -> N205 +
N.0O- + M -» NO« + NO- +
^ O ^ j
02 + O + M^-O3 + M
NO + 03 * N02 + O2
M
M
[NO] [O]
k2[N02] [O]
k3[M] [N02] [O]
k4[N03] [NO]
k?[M] [02] [O]
kg [NO] [03]
The purpose of this study was to examine this system with the molecular
technique, and in particular to focus on the rate constants for
reactions 2, 3, and 4 at room temperature. ~
Quantities measured in this study were the decay of N02, the build-up
and decay of N2O5/ and the molecular modulation of NO2 during intermittant
illumination of a steady flow system. In terms of the mechanism these
74
-------
measurements give the following information: At low pressures (below
50 torr) the M-gas dependent reactions (1 and 3) become negligible com-
pared to reaction 2, and the decay of nitrogen dioxide can be described
by
/d £n[N0]
(23>
At one atmosphere total pressure and with small initial NO concentration,
the disappearance of NO is given as a good approximation by
[N021 _ _2alt
- "^ (24)
where k is the empirical, first-order, decay constant. If the initial
a
concentration of nitric oxide is close to zero, the initial rate of
formation of dinitrogen pentoxide is
al[NO_]
(25)
dt / n (1 + k,A,[M])
t. \J £ «J
At an early stage of the reaction (t=t*) dinitrogen pentoxide goes
through a maximum, at which point
aKK/kJ [NO,] *
2 (26)
[MIT TNO]
where K = k5/k,. The modulation amplitude of nitrogen dioxide in a
steady flow system illuminated with square-wave photolyzing light of
frequency f is
[NO,]/ k, [NO]
1 + r-
(27)
It can be seen that Eq. 23 gives al; with I so determined, Eqs. 24
and 25 give two independent measures of the ratio k_/k ; the combina
tion of Eqs. 23, 24 and 27 gives k,/k2.
75
-------
Thus by solving the five experimental relationships simultaneously, it
is possible to obtain values for al, k./K, k,/k2, and k^/k2, with one
equation redundant to use as a check. By taking the literature values
for k, and for the equilibrium constant K, it is possible to obtain
the elementary rate constants k-, k~, k..
B. RESULTS
A series of experiments was carried out in the closed cell with low
total pressure, 19 to 56 torr; the decay of NO2 was first-order through-
out an experiment. When NO2 is photolyzed in a closed cell at one
atmosphere, the initial decay is first-order, but deviations from first-
order kinetics appear at later times, as illustrated by Figure 32 (The
concentration of NO2 was followed by infrared absorption at 1600 cm" ) .
The build up and decay of N20_ is illustrated by Figure 33 for condi-
tions at one atmosphere total pressure (N9) and 24°C. The concentra-
-1
tion of N,O_ was followed by infrared absorption at 1238 cm . A plot
^ J j. j.
* *
of ( [N2O5]/[NO_]) against ([NO2]/[NO]) is given as Figure 34, compare
Eq. 26.
Along with these conventional measurements, a molecular-modulation study
of NO2 photolysis was carried out. The molecular-modulation method
consists of monitoring the periodic concentration fluctuations induced
in the reacting species by flashing the photolysis lamps on and off.
In this work the NO2 modulation was studied in a steady flow system
with the photolysis lamps flashing at 1 cps. Under these conditions
the NO- radicals and oxygen atoms will be at their steady state concen-
trations and N2O_ will be equilibrium.
The behavior of the concentration modulation of N02 can be described by
a differential equation with the flashing lamps represented by the
Fourier series for a square wave.
1C.
-------
5.0
O Experimental points
Calculated decay curve
40 50
Time (seconds)
70
SO
XBL728-6843
Figure 32 - NQ~ decay; comparison of experimental points and calcu-
lated curve based on rate constants in Table (5).
Initial concentrations:
14
2.05x10 .
15
[N02] = 4.2x10 , [NO] =
77
-------
2.0
in
3
O
I
i
O
in
-iiiiiiiir
O Experimental points
Calculated decay curve
I
I
10 20 30 40 50
Time (seconds)
60 70
80
XBL728-6845
Figure 33 -
N2Og decay as observed by infrared absorption at 1237 cm ;
comparison of experimental points and calculated curve
based on rate constants in Table (5). Initial concentrations:
NO_ = 2.8x10 molecules cm" , [NO] = 0.
78
-------
CM
O
o"
1.0
0.8
Q6
0.4
m
O
CM
*L 0.2
Slope = 1.08±.O2 KID'3
8
The * indicates the concentration at the time of the N205
concentration maximum, t*
XBL728-6842
Figure 34 -
Concentration ratios where nitrogen pentoxide is a
maximum, Equation 4.
79
-------
2] F f k [NO ]2
* F rr_ , _ [NQ , , _ | 22
IkijjNO] [M] + (k2+k3TM]) [N02]
dt V Ll 2Jin l" 2Jout
2 Z^i :r sin (nut) + oD) (28)
11 odd n
With: F = flow rate of chemicals
V = volume of the reaction cell
[NO_]. = the NO_ concentration flowing into the cell
[NO2] . = the NO_ concentration flowing out of the cell
a) = 2TT/T, where T is the period of the square wave
This equation neglects the ozone reactions, since the system contains
no initial oxygen and the 0_ buildup will not be significant. Under
these conditions the flow system has reached a steady state with respect
to reactants and products, so the unmodulated or DC terms in Eq. 28
cancel leaving only the modulation or AC terms.
~Gt I £» ** VJUUAlll /OQ\
d6 2, k.. [NO] [M] + (k0+k,[M] [NOT]
IT f 1 2. 3 2
where d6 = 2irfdt; 9 = wt; and f = flashing frequency.
Selecting the conditions so that the modulation will be less than one
part per thousand leaves the reactant concentrations essentially
constant, allowing Eq. 29 to be integrated in closed form to give:
ctHNCU2 2k^ 4ri .. n"2 cos'(ne)
i T n
^
2AC ~
f k, [NO] [M] + (k-+k_[M] [NO,]
TT £ 1 £ 3 *
80
-------
By electronically filtering out all but the first Fourier component
of the modulation signal, and taking the peak to peak amplitude gives
the experimental relationship:
4al k [NO,]2
[NO,] , = z - - (31)
^ moa TT^f(k [NO] [M]-f-k +k [M]) [NO ])
Comparison of Eqs. 30 and 28 shows the NO2 modulation waveform to be
triangular, phase shifted 90° from the exciting light. The amplitude
of this waveform and its phase shift from the exciting light can be
determined experimentally and related to Eq. 31.
A summary of the steady state concentrations with the phase shift and
amplitude measurements are recorded in Table 5. The photolytic light
intensity for the modulation experiments was found to be 11.2% greater
than in the initial slope experiments, when measured by the photo-
transistor mounted in the cell, and the values- of I have been normalized
by this factor.
The observed rate constant ratios are assembled in Table 6. The two
independent values of k_/k_ agree within 10 per cent; the value based
on Eq. 24 was more precise and is preferred. This study gives only
rate ratios, k,/k2, k-./k2, and k4/K. To evaluate absolute values of
rate constants, either k, , k_, or k- must be taken from the literature.
Of the three, the most extensive and most precise data are those for
k,. From the literature, the value 6.9 x 10 cm molecule sec was
taken for N2 as foreign gas at room temperature. The observed quantity
k^/K is reduced to k. by the literature value of K. With these values
of K and k,, the elementary rate constants k,, k.,, and k. are evaluated
and listed in Table 7 with the other 8 elementary rate constants in this
system.
With these 8 elementary reactions, a computer program was written that
integrated the simultaneous rate equations to give NO, and N,05 as a
81
-------
Table 5. Data for NO2 modulation at one atmosphere total
pressure, 24°C.
Carrier
gas
N2
N2
N2
N2
N2
N2 .
Ar
Ar*
Ar
Steady state
concentration
[N02] [NO]
-14 -15
xlO ** xlO *-°
1.02
1.19
0.68
1.13
1.87
1.68
1.55
2.12
1.34
0.74
1.22
1.49
2.04
2.12
1.60
1.38
1.30
1.34
ka
sec
xlO3
8.3
8.3
8.3
8.3
8.3
7.2
8.3
4.6
8.3
phase
shift
degrees
93.6
89
70
82
80
94
88.2
87.8
91.8
A[N02]
mod.
1.43
1.31
1.40
0.83
1.62
1.68
1.78
1.65
1.41
VM]
k2
0.163
0.179
0.189
0.185
0.190
0.178
0.190
0.206
0.198
82
-------
Table 6. Observed rate constant and rate constant ratios
Rate constant
function
al
k
a
k3[M]/k2
Vk2
k3[M]/k2
k3/k2
VK
k1[M]/k2
Vk2
k1[M]/k2
kl/k2
Observed Based on
value equation
7.45xlO"3 sec"1
1.22xlO~2 sec"1
0.22
91
8.9x10
0.20
91
8.1x10
0.71 sec"1
0.18
7.3xlO~21
0.20
8.1xlO~21
(1)
(2)
(2)
(2)
(3)
(3)
(4)
(5)
(5)
(5)
(5)
Condition
1 atm N2
1 atm N2
N2 = M
1 atm N2
N2 = M
1 atm N2
N2 = M
1 atm Ar
Ar = M
83
-------
Table 7. Elementary rate constants for nitrogen dioxide
photolysis, 297°K, 1 atm N0
Constant
kl
k2
k3
K
k4
k6M
k5M
k7
k8
Value
6.9xlO~32 cm6 sec"1
9.2xlO"12 cm3 sec"1
8.2xlO~32 cm6 sec"1
1.24X10"11 cm"3
8.7xlO~12 cm3 sec"1
0.104 sec"1
1.29xlO~12 cm3 sec"1
6.24xlO~34 cm"6 sec"1
2.1xlO"14 cm3 sec"1
Source
Literature
1' 1 2
k2' k3//k2
Literature
k, k4/K
Literature
K/ k6M
Literature
Literature
Reference
57-63
Table 5
Table 5
[64]
Table 5
[65]
[66]
[19]
84
-------
function of time. In Figures 32 and 33 the circles represent observa-
tions and the smooth curves are those.calculated from the elementary rate
constants in Table 7. There is good agreement between the calculated
and observed curves over the full course of observations.
85
-------
SECTION IX
DISCUSSION
Solar radiation in the visible and near ultraviolet spectrum is "weak"
relative to the radiation intensity from flash photolysis. The method
of flash photolysis has been highly successful in obtaining the spectra
and kinetics of free-radical intermediates in high intensity photo-
chemical experiments. In a typical, complex photochemical reaction,
there is competition between radical-molecule reactions (chain propaga-
tion) and radical-radical reactions (chain terminating, usually). The
.balance between these two competing processes is quite different under
conditions of strong and weak light intensity. There is a distinct
possibility that a given photochemical reaction will follow a different
mechanism in the real sunlit world and in the laboratory using flash
photolysis.
The molecular modulation.method was designed to detect and follow free
radicals in photochemical reactions at light intensities comparable to
sunlight in intensity. VJith the first test of this method new radicals
were seen in the HO system (later shown to be the elusive HOO radical)
J±
and in the CIO system (still not positively identified but probably a
X
complex of Cl, O_, and CO). These radicals were detected at concentra-
11 -3
tions of about 10 molecules cm or less than one part per hundred
million in the atmosphere. Design calculations based on light sources,
light through-put in optical instruments, radical light-absorption cross
sections, and detector sensitivity promised substantial improvements
over this level of detection and routine analysis of reaction mechanisms
at this level of radical concentration. Especially during the last phase
of this grant, a new model of the molecular modulation apparatus was
developed that (on paper) should have had greatly improved sensitivity,
convenience, and flexibility. The light path was greatly extended, the
monitoring infrared source was stabiliw*; the infrared detector was
cold-shielded, and the electronics was set up in a digital mode, such
that results could be recorded on paper tape or sent directly to a small
computer.
86
-------
The testing of this new improved system revealed a new, limiting source
of error that overrides all of the improvements that were designed and
made. The goal was to detect changes in the monitoring light beam at
fractions of 10 to 10 . The infrared measuring beam has a long,
folded, optical path of 100 meters or more. As the photolyzing lights
are turned on and off, the light is absorbed by the primary reactant and
heat is released to the carrier gas as free radical recombinations and
reactions occur. These thermal effects, introducing time-varying and
spatially non-uniform density gradients, appear to introduce noise to the
long-path infrared beam. This noise restricts the usefulness of the
molecular modulation method.
In the presence of this "accoustical noise", it is very time consuming
and difficult to obtain molecular modulation data on free radical reactions
at low intensities of photolyzing light. Even so, this report gives
successful use of this method for the radicals HOO, C1OO, and CIO. The
results for HOO have been confirmed by other studies48'49, which used
different experimental methods. The results for CIO and ClOO are quite
different from those obtained by other methods (flash photolysis). There
remains an unresolved conflict between these two methods. It may be that
mechanism is quite different at low and at high light intensities, and
there could be one or more unidentified elementary reaction responsible
for the conflict. In view of the expected and observed difference in
mechanisms at low and at high light intensities, it may be important to
pursue the study.of radical intermediates by the molecular-modulation
method. However, it is so expensive in time and effort to make
molecular-modulation studies that this method should be pursued only
where there is a specific need for examing low-intensity photolysis.
According to verbal reports, other groups (North American Research Center
and Shell Research Laboratory of England) have adopted the molecular
modulation method and used it with better success then that obtained here.
In each case, they shortened the optical path to a meter or so and used
photolysis lights 10 or 100 times more intense than sunlight. Such light
intensities are still very small compared to flash photolysis. With a
short optical path there is much less accoustical noise and there can be
greater throughput of infrared radiation. With a chamber of small cross
87
-------
section and small volume, there can be a much higher light intensity
for photolysis.
In this laboratory, further work is being done on this method, but it
is in the direction of shorter optical path and higher light intensity.
The large apparatus built under this grant has been disassembled and
rebuilt with the incorporation of these changes.
The goal of this project was to detect the CH300 and CH3(CO)OO radicals
and to study their kinetics. This goal was not attained. This project
did characterize spectra and kinetics of the HOO radical, and these data
are useful for model calculations in the stratosphere and in the tropo-
sphere. This project characterized the C100 and CIO radicals. This
study was started simply as a calibration of the new method. Recently
there has been a surprising development that makes CIO chemistry very
J^
much a matter of atmospheric chemistry. As is now well established,
the oxides of nitrogen catalyze the destruction of stratospheric ozone
by catalytic cycles including
NO + O3 + NO2 02
0- + hv -> O2 + 0
NO2 + O -» NO +
net:
In this cycle at typical stratospheric temperatures, NO_ is 10,000
times as effective in destroying ozone as ozone itself. Recently it
has been found that Cl and CIO undergo a similar cycle
Cl + O3 + CIO +
O3 + hv -» O2 + O
CIO + O -> Cl + O
2 G_ * 3 0~
88
-------
The rate-determining step in this cycle has a rate constant 5 or 6 times
larger than the rate determining step in NO_ cycle. Thus one part of
CIO is about 50,000 times as effective in destroying ozone as ozone
fi Q
itself. Recently it has been proposed that freons and other organic °
chlorides released into the troposphere will eventually (decades) perco-
late into the upper stratosphere where they will be photolyzed to
produce chlorine atoms. Thus the CIO chemistry studied here may have
J^
applications to atmospheric problems after all.
89
-------
SECTION X
REFERENCES'"
Johnston, H. S., McGraw, G. E., Paukert, T. T. , Richards, L. W.
and Van den Bogaerde, J., Proc. Natl. Acad. Sci. U.S., 57 (1967)
1146.
2. Van den Bogaerde, J., Ph.D. Thesis, 1968, Univ. of Calif. Berkeley.
3. McGraw, G. E. and Johnston, H. S., J. Chem. Kinetics, 1 (1969) 89.
*
4. Paukert, T. T., Ph.D. Thesis, 1969, Univ. of Calif. Berkeley.
5. Morris, E. D., Jr., Ph.D. Thesis, 1969, Univ. of Calif. Berkeley.
6. Johnston, H. S., Morris, E. D., Jr. and Van den Bogaerde, J.,
J. Am. Chem. Soc., 91 (1969) 7712.
7. Morris, E. D., Jr. and Johnston, H. S., J. Am. Chem. Soc., 90
(1968) 1918.
8. Morris, E. D., Jr. and Johnston, H. S., Rev. Sci. Instr., 39 (1968)
620.
9. Wu, C. H., Ph.D. Thesis, 1969, Univ. of Calif., Berkeley.
10. Wu, C. H. and Johnston, H. S., Bull. Soc. Chem. Beiges. 81 (1972),
135.
11. Bier, K. and Hagena, O., "In Rarefied Gas Dynamics," ed. J. A.
Laurmann (1963), Academic Press, New York.
12. Margeneau, H., and Murphy, G. S., "The Mathematics of Physics and
Chemistry", D. Van Nostrand Co., Inc., Princeton, N, J, (1956) p. 41.
13. Porter, G. and Wright, F. J., Disc. Faraday Soc., 14 (1953) 23.
14. Lipscomb, F. J., Norrish, R.G.W. and Thrush, B. A., Proc. Roy. Soc.
A233, 455 (1956).
15. Edgecombe, F.H.C., Norrish, R.G.W.. and Thrush, 5. A., Proc. Roy.
Sou. A243, 24 (1957).
16. McGrath, W. D., and Norrish, R.G.W., Proc. Roy. Soc. A254, 317
(1960).
90
-------
17. Clyne, M.A.A., and Coxon, J. A., Trans. Faraday Soc. 62, 1175
(1966).
18. Durie, R. A. and Ramsay, D. .A., Can. J. Phys. 36, 35 (1958).
19. Burns, G., and Norrish, R.G.W., Proc. Roy. Soc. A271, 289 (1963).
20. Benson, S. W., and Buss, J. H., J. Chem. Phys. 27, 1382 (1957).
21. Clyne, M.A.A., and Coxon, J. A., Proc. Roy. Soc. A303, 207 (1968).
22. Benson, S. W., Anderson, K. H., J. Chem. Phys. 31, 1082 (1959).
23. Rochkind, M. M., and Pimentel, G. C., J. Chem. Phys. 46, 4481
(1967).
24. Eachus, R. S., Edwards, P. R., Subramanian, S. and Symons, M.C.R.,
Chem. Communications 1967, 1036; J. Chem. Soc. (A) 1968, 1704.
/
25. Arkell, A., and Schwager, I., J. Am. Chem. Soc. 89, 5999 (1967).
26. Porter, G., Discussions Faraday Soc. 9, 60 (1950).
27. Marshall, A. L., J. Phys. Chem. 30, 34, 1078 (1926).
28. Taylor, H. S., Trans. Faraday Soc. 21, 560 (1926).
29. Lewis, B., and von Elbe, G., Third Symposium on Combustion and
Flames and Explosion Phenomena, Williams and Wilkins Co.,
Baltimore, Md., (1949) p. 484.
30. Lewis, B., and von Elbe, G., "Combustion, Flames, and Explosions
of Gases, Academic Press, New York (1961).
31. Baldwin, R. R., Mayor, L. and Doran, P., Trans. Faraday Soc. 56,
93 (1960).
32. Baldwin, R. R., and Mayor, L., Trans. Faraday Soc. 56, 80, 103
(1960).
33. Burgess, R. H., and Robb, J. C., Chem. Soc. Spec. Pub. No. 9, 167
(1957).
34. Heidt, L. J., and Forbes, G. S., J. Am. Soc. 56, 1671, 2365
(1934).
35. Volman, D. H. , J. Am. Chem. Soc. 73, 10.18 (1951).
91
-------
36. Norrish, R.G.W., and Wayne, R. P., Proc. Roy. Soc. A288, 200
(1965).
... ^.
37. McGrath, W. D., and Norrish, R.G.W., Proc. Roy. Soc. A254, 317
(1960).
38. Volman, D. H., Advances in Photochemistry 1, 43, Interscience
Publ., New York (1963).
39. Volman, D. H., J. Chem. Phys. 17, 947 (1949).
40. Greiner, N. R., J. Chem. Phys. 45, 99 (1966); J. Phys. Chem. 72,
406 (1968).
41. Foner, S. N. and Hudson, R. L., J. Chem. Phys. 36, 2681 (1962);
Foner, S. N., and Hudson, R. L., J. Chem. Phys. 21, 1608 (1953).
42. Robertson, A.J.B., Applied Mass Spectrometry, Institute of
Petroleum, London (1954).
43. Ingold, K. U., and Bryce, W. A., J. Chem. Phys. 24, 360 (1956).
44. Fabian, D. J., and Bryce, W. A., Seventh International Symposium
on Combustion, Butterworths Scientific Publications, London (1959)
p. 150.
45. Milligan, D. E., and Jacox, M. E., J. Chem. Phys. 38, 2627 (1963)
40, 605 (1964).
46. Ogilvie, J. F., Spectrochimica Acta 23Af 737 (1967).
47. (a) Czapski, G. and Dorfman, L. M., J. Phys. Chem. 68, 1169 (1964);
(b) Bielski, H. J. and Schwarz, H. A., J. Phys. Chem. 72, 3836
(1968); (c) Behar, D. and Czapski, G., Israel J. Chem. S, 699
(1970).
48. Troe, J., Ber. Bunsengesellschaft Phys. Chem. 73, 946 (1969); also
recent private communications.
49. Hochanadel, C. J., Ghormley, J. A., and Ogren. P = J., J. Chem.
Phys. 56. 4426 (1S72).
50. Holt, R. B., McLane, C. K., and Oldenberg, 0. J. Chem. Phys. 16,
225, 638 (1948).
92
-------
51. Urey, H. C., Dawsey, L. H., and Rice, F. O., J. Am. Chem. Soc.
_51, 1371 (1929).
52. Hagen-Smit, A. J., Ind. Eng. Chem. 44, 1342 (1952).
53. (a) Nicolet, M., J. Geophys. Res. 70, 679 (1965). (b) Crutzen,
P. J., Quart. J. Roy. Met. Soc. 96, 320 (1970). (c) Johnston,
H. S., Science 173, 517 (1971).
54. Hall, T. C., Jr. "Photochemical Studies of Nitrogen Dioxide and
Sulfur Dioxide", Doctoral Thesis, University of California,
Los Angeles.
55. Ford, H. W. and Endow, N., J. Chem. Phys. 27, 1156, 1277 (1957).
56. Troe, J., Ber. Bunsenges. Physik Chem. 73, 906 (1969).
57. Kaufman, F., Proc. Roy. Soc. A247, 123 (1958).
58. Hartek, P. Reeves, R. R., and Mannella, G. G., Air Force Cambridge
Research Center, Technical Report, AFCRC-TR60-264 (1960).
59. Clyne, M.A.A., and Thrush, B. A., Proc. Roy. Soc. A269, 404
(1962).
60. Ogryzlo, E. A., and Schiff, H. I., Can. J. Chem. 37, 1690 (1959).
61. Takahashi, S., and Miyazaki, S., Mem. Def. Acad., Jap. 8, 611
(1968).
62. Takahashi, S., Mem. Def. Acad., Jap. 8, 777 (1968).
63. Klein, F. S., and Herron, J. T., J. Chem. Phys. 41, 1285 (1964).
64. Schott, G., and Davidson, N. , J. Am. Chem. Soc. 80, 1841 (1958).
s
65. Mills, R. L. and Johnston, H. S., J. Am. Chem. Soc. 73, 938 (1951)
66. Johnston, H. S., "Gas Phase Reaction Kinetics of Neutral Oxygen
Species", NSRDS-National Bureau of Standards 20.
67. Johnston, H. S., and Crosby, H. J,/t J. Chem. Phys. 22, 689 (1954).
68. Molina, M. J., and Rowland, F. S., "Stratospheric Sink for Halo-
genated Hydrocarbons (Freons) - Chlorine Atom Catalyzed Destruc-
tion of Ozone", submitted to Nature, January 1974.
93
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-650/4-74-004
2.
3. RECIPIENT'S ACCESSION1 NO.
4. TiTLE ANDSUSTITLE
Molecular Modulation Spectrometry for Observation
of Free Radicals
5. REPORT DATE
March 1974
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Harold S. Johnston
8. PERFORMING ORGANIZATION! REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Chemistry
University of California
Berkeley, California 94720
10. PROGRAM ELEMENT NO.
A11Q08
11. CONTRACT/GRANT NO.
801120
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency
National Environmental Research Center
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report contains a description of the methodology that has been developed for
measuring free radicals intermediates. The molecular modulation method developed
was designed to detect and follow free radicals in photochemical reactions that
are known to occur at light intensities equal to that of sunlight. The goal of
the project was to detect CH^OO and CH3(CO)00 radicals and study their kinetics.
This goal was not attained. However, reaction intermediates that were studied
were C100, CIO, HOO, and ^Oc. The method developed can measure radicals at
concentrations as low as lO^i particles/cc.
7.
KCV VvuHDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Free Radicals
Photochemistry
Atmospheric Chemistry
Spectrometry
3. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
aggi f i prl
21. NO. OF PAGES
93
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
94
------- |