SCATTERING OF HYDRIDE IONS
IN OXYGEN GAS
JOHN MURRAY McGUIRE
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
The author wishes to thank his research director, Dr. E. E.
Muschlitz, Jr., for his advice and assistance in carrying out the
research and writing of this dissertation. He also wishes to thank
Dr. J. H. Simons for his advice and encouragement and Dr. T. L.
Bailey for his help in collecting the data embodied herein. Finally, the
Physics Branch of the Office of Naval Research is thanked for the
Assistantship which the author held during the course of this research.
John Murray McGuire
Gaine ville, Florida
TABLE OF CONTENTS
LIST OF TABLES . . . . . .
LIST OF PLATES AND FIGURES .....
* . .
* . .
I NEGATIVE IONS . . . . . .
II SCATTERING . . . . .... .
III DESCRIPTION OF APPARATUS .......
IV EXPERIMENTAL PROCEDURE .......
V TREATMENT OF DATA . ........
VI CONCLUSIONS ....... .......
APPENDIX . .. . . . . . .
APPENDIX . . . . . .. .
DEFINITION OF SYMBOLS
BIBLIOGRAPHY ........ .. . . . .
VITA .. .... ...... .. .
LIST OF TABLES
I Fluxneter Data .. . .. .. . 21
1I Data for Cross-section Evaluation . . . . 29
LIST OF PLATES AND FIGURES
I. Photograph of Apparatus, Front View . 10
II. Photograph of Apparatus, Rear View . . 11
1. Schematic Diagram of Apparatus . . . 12
2. Electrolysis Cell .. . ...... . . 16
3. Fluxmeter Graph . . . . . ...... 22
4. Pressure Dependency Graph . . . . . 24
5. Lid Potential Graph ..... ........... 25
6. Cross-sections as Functions of Ion Energies . . 31
7. Potential Law Graph for Low Energies . . . .36
The existence of negative ions was postulated for solutions in
the early days of electrolytic theories of conductance. Consideration
of equilibrium phenomena such as precipitations in solutions has long
demonstrated, as has electrolysis of solutions, the existence of nega-
tively charged particles, other than electrons, in solution. The term
"particles" is used in preference to "ions" due to the presence, in gen-
eral, of solvated clusters in solution rather than simple ions.
Studies of the crystal diffraction of X-rays have also established
that many crystals are ionic in their structure and consist of free ions
held together by electro-static forces. It was later shown that the par-
ticles which carried negative electricity in gaseous discharges were not
restricted to electrons; thus negative ions were shown to exist in the
As might be expected, certain ions which have been postulated
as organic intermediates have been found by mass spectrometric means
while others have not. In cases where the postulated intermediate is
actually confirmed by mass spectrometric data, its existence may
generally be regarded as having been established as a free ion. How-
ever, this does not mean that it exists in any other form. A number of
ions, postulated in mechanisms, have been reported by this method;
e. g., O, NOZ, NO3Q(), OH', Li'(2), and Czr(3).
Many of the elements have been shown to have a tendency to
form negative ions, the tendency increasing with increasing electron
affinity. The rare gases are noteworthy in that there are apparently
no stable negative ions formed. This has been attributed to the necessity
for the added electron having a higher quantum number than the outer
atomic electrons, due to the Pauli exclusion principle.
The hydride ion has been the subject of a number of calcula-
tions due to its relative simplicity. It has been shown by Hylleraas (4)
that while the undisturbed state of the atom is not sufficient to produce
stability, the strong interaction in the Is state rearranges the charge
distribution enough to cause stability in the resultant ion. The presence
of this ion in the solar spectrum has now been definitely established (5).
In brief, negative ions are known to exist in all physical
states. They are stable in the solid state due to electro-static equilib-
rium (as in ionic crystals). In the liquid or gaseous state they will
migrate under the influence of applied potentials (as in electrolysis of
solutions). In the gaseous state, they may be produced by such diverse
means as natural processes in the upper atmosphere or, in the laboratory,
by thermal dissociation of ionic salts.
The general equations for negative ion production may be
(1. 1) A+ B" A + B
(1.2) AB + C" = AC + B"
(1.3) A + e + M A + M
(1.4) A2 + e = A" + A
(1.5) A + e = A" + hv
(1.6) A + B" A" + B
(1.7) A + B" AB"*
(1.8) A+ + S = A + S
(1.9) AB + e A" + B+ + e
Equation (1. 1) represents the dissociation of two ions held by electro-
static forces; this process occurs on solution of an ionic salt or upon
the very strong heating of one. Equation (1.2) may be illustrated by the
ion exchange process in solution. In equation (1. 3), M represents any
atom which serves to stabilize the negative ion by removing excess
energy. Dissociative attachment by electron collision is represented
by equation (1. 4). Equation (1. 5) is an example of radiative attachment
by an atom of an electron with simultaneous emission of a photon.
(1. 6) is an important equation demonstrating charge exchange between
a negative ion and a neutral atam. This process has been reported by
Dukelskii and Zandberg (6). In equation (1.7) the formation of an ex-
cited negative molecular ion from collision of an atom and a negative
ion is illustrated. Equation (1. 8) represents negative ion formation by
,reflection of a positive ion from a metal surface.
Negative ion destruction may proceed according to one of the
(1.10) A" + B = A + B + e
(1.11) A" + B = A + B"
(. 12) A" + B = AB + e
(1.13) A" + S = A + S"
(1.14) A" + B+ AB + h
(1.15) A + B = A* + B*
(1.16) A" ++ B+ M = AB + M
(1.17) A + e = A + 2e
(1.18) A' + hy = A + e
(1.19) A" + B* = A + B + e
The symbols, S and M, used in equations (1. 10-1. 19) have the same
meaning as in equations (1. 1-1. 9). In low velocity scattering, the two
most important destructive processes are illustrated by equations
(1. 10) and (1. 11); the first shows electron detachment from a negative
ion by collision with an atom; the second, shows charge exchange
between the ion and the atom.
Simple scattering theory is based on four assumptions: first,
the system is a conservative one; second, the scattering particle may
be considered initially at rest with respect to the scattered particle;
third, that the force field is a central one; and, a non-essential as-
sumption to simplify the treatment, that the scattering angle is small.
Since the analysis, based on this last assumption, gives results within
the experimental error of an exact treatment (7, 8), it appears to be
a justifiable assumption. The application of this theory to the problem
of this dissertation is given in Chapter V.
Scattering of beams is a powerful tool which has been used for
different purposes in a number of cases. In one of the most simple
experimental forms, it has been used to determine the equivalent wave
length of electrons (9). In recent years, it has been used for such
seemingly diverse purposes as determining the force law between rela-
tively high velocity atoms and molecules (10) and measuring electron
exchange cross-sections (11, 12).
Scattering is also applied in determination of nuclear structure,
and has been so used over a wide range of elements. The nucleus to be
studied is bombarded by a particular type of particle (e. g., alpha par-
tides or neutrons) and, through determination of the scattering behavior,
conclusions are drawn as to the structure of the nucleus itself.
The scattering of atoms by gases is similar in principle to the
present work; however, the great difficulties encountered in measuring
the intensities of neutral scattered particles have restricted the measure-
ments to considerably higher energies than those used in the determina-
tion of low velocity ion interaction with gases.
The investigation of elastic scattering of low velocity ions may
be said to originate with the work of Russell, Fontana, and Simons (13)
in 1941. This paper did not make provision for the quantitative separa-
tion of inelastic scattering from elastic scattering, and, consequently,
the results were less definite than later experimental work based on the
apparatus described by Simons, Francis, Fontana, and Jackson (14).
Results obtained in these experiments cover the scattering of H+, H2,*
and H+ ions in a variety of gases. The.potential laws operating over
a given energy range between the ion and the scattering particles have
been evaluated. In many cases the nature of the interaction may be de-
duced from the potential law. Naturally, this technique presents an un-
ambiguous method for determining proton affinities and has been used
for such a purpose (13).
The field of negative ion scattering is one in which little work
has been done (16). With the exception of the negative ion source, the
apparatus used in this dissertation was similar to that used by Simons
and co-workers (8, 12, 14, 15) in the study of positive ion scattering.
Since the low velocity scattering of positive ions in gases has
contributed useful information as to lon-molecule interactions, it was
felt that the same type of experiments with negative ions would also
rield useful results. This view was strengthened by the present lack of
definite knowledge in the field of negative ions. In the light of these
acts, the research for this dissertation was undertaken in order to
contribute to the understanding of the laws governing such behavior.
DESCRIPTION OF APPARATUS
The apparatus used for the scattering measurements was es-
sentially the same as that described by Muschlitz, Bailey and Simons
(16) in Technical report #2 to the Office of Naval Research under con-
tract Nonr 580(01) with several minor changes.
This apparatus utilizes electro-static focusing combined with
magnetic field selection of the desired negative ions. The ions, pro-
duced by collisions of gas molecules with an electron stream, are col-
limated into a well-defined pencil by empirical focusing of electro-
static elements concentric with the desired pencil. The beam then
passes through a magnetic chamber where a particular charge-to-mass
ratio may be selected by varying the magnetic field. The beam is bent
through a ninety degree angle and continues to the final set of electro-
static focusing elements. These elements are, in principle, the same
as the ones before the magnet and are used to refocus the beam which
is tending to diverge as it leaves the magnet chamber. A schematic
drawing of the apparatus is shown in Figure 1, while the exterior is
shown in Plates I and II.
Photograph of Apparatus, Rront View
Photograph of Apparatus, Rear View
FIG. I NEGATIVE ION APPARATUS
The beam leaving these final collimating elements is well-de-
fined, of constant charge-to-massratio, and approximately mono-ener-
getic. It is allowed to enter a scattering region containing the scattering
,gaa at some accurately measured low pressure and then passes into a
Faraday cage, C in Figure 1, which collects the transmitted current.
The scattered current is collected on the scattering elements, S and SL.
The scattering elements are a plate, SL, and a cylinder, S,
both of which are concentric with the beam. The plate may be either at
the same potential as the cylinder (and also the Faraday cage) or may
have a small potential applied so as to either remove slow particles
formed by inelastic collisions or to repel all negative particles.
The currents to the scattering cylinder, and collecting cage
are determined by the amplifiers described by Searcy (17). The maxi-
mum full-scale sensitivity is 5 x 10"14 ampere. These currents are
related to the effective cross-sections through the familiar Beer-
Lambert equation expressed in terms of cross-section, path length,
and pressure (see Appendix 1).
Several important modifications in the apparatus have been
made. The filament is now of iridium cataphoretically coated with
thoria (18). Each of three coatings was applied at 30 milliamperes for
15 seconds. Before coating, the ends of the filament were spot-welded.
to 5 mil nickel sheet for support in the filament clamps. The filament
emission to the anode is now regulated by a circuit adapted from the
thyratron control portion of a standard ion gauge circuit (19). Finally,
the repeller, R in Figure i, potential has been stabilized using two
voltage regulator tubes (type OA-3) in series.
The thoriated filament was used in preference to the tungsten
previously used in order to obtain a higher ion beam intensity. It has
been found possible to obtain more than ten times the beam intensity
from water vapor at 3 mm. backing pressure (at N in Figure 1) than
from hydrogen gas at 8 mxm. pressure. However, the life of the tungsten
filament in water vapor was reduced to less than the time for an average
run, whereas the thoriated iridium lasted about ten times this long.
The hydride ions produced from water vapor gave the same cross-
sections as those obtained using hydrogen in the ion source.
The water vapor was in dynamic equilibrium with carefully de-
gassed, distilled water maintained at approximately 0* C. by a Dewar
flask containing an ice-water mixture. It was not necessary to maintain
the temperature perfectly constant since small changes in the backing
pressure affect the total current rather than the fractions scattered or
One additional improvement was made. The oscilloscope used
in preceding work was replaced with a General Radio type 1231-3B null-
detector, for locating the null point in the determination of the magnetic
field used to bend the hydride beam. The field is determined by balancing
the potential developed by a constant-speed coil placed in the field
against a reference voltage.
Figure Z shows schematically the electrolytic cell, gas train,
and leak system used for producing and introducing the oxygen into the
scattering region. The electrolyte used in the cell was a 5% solution of
barium hydroxide (Baker and Adamson reagent grade). This was chosen
in preference to the potassium hydroxide used in previous work to pre-
vent possible formation of carbon dioxide with the oxygen. Other possi-
ble impurities were presumed to be hydrogen, water. ozone, and hydro-
gen peroxide. To remove these from the oxygen, the gas was passed
through a chain consisting of a platinum filament electrically heated to
redness (to catalyze water formation between the oxygen and any hydro-
gen present), a silver-foil packed tube (to catalyze peroxide and ozone
decomposition), and a phosphorous pentoxide tube (to remove water
The oxygen was then passed into an intermediate pressure region
constructed so that it could be pumped on through either or both of two
capillaries of different lengths. These were connected to a mercury
diffusion pump through a trap cooled by liquid air. From the interme-
diate pressure region, another capillary permitted the oxygen to enter
the scattering region. It was possible with the arrangement used to
Fig. 2-Electrolysis Cell
obtain equilibrium pressures in the range 0. 5 3. 0 x 10-3 mm of
mercury. This was adequate for the purpose.
In order to remove residual traces of gas from the apparatus
before scattering measurements were started, the entire apparatus was
evacuated by four high speed mercury diffusion pumps as described by
Muschlitz, Bailey and Simons (16). The pumping was continued until the
pressure was of the order of 2 or 3 x 10-5 mm. of mercury as measured
by the ionization gauge before the measurements began. The coolant
used on the pump traps was liquid air which expedited rapid removal of
condensible gases from the apparatus and prevented mercury vapor from
entering the system. When the pressure reached this value, the ion
source chamber was opened to the water vapor, and the filament and
anode voltages applied. The ion beam was focused to the Faraday cage
by varying the magnetic field and adjusting the focusing elements for
maximum ion intensity.
The method used in the past to determine ion velocity, i. e.
a retarding potential on the final elements (20), was found impractical
in the present apparatus. An energy correction, AV, applied to the
potential difference between magnet chamber and cathode was determined
in two ways. The first method utilized the mass spectrometer equation:
(4. 1) M HZrZ or, since r is an apparatus
constant, and m/e is constant for H",
(4. ) V = kH2. Since V = V + AV, it
follows from (4. Z) that
(4.3) VM kHZ + AV
Thus, plotting the square of the magnetic field needed to focus
the hydride beam for maximum intensity against the potential of the
magnet chamber, as measured in volts, gives a straight line with inter-
cept equal to the velocity correction in volts. The magnetic field is ex-
pressed in arbitrary units.in terms of the reading, D, of a Helipot which
is placed across a known reference potential, E^. D = Em which
is proportional to H. Therefore, the actual plot was based on the rela-
(4.4) V( = k'Em + AV = k'li D2 + AV
Equation (4.4) will give the same intercept as would equation
(4. 3). The different slope is immaterial to the determination of the
energy correction. Table I together with Figure 3, illustrate a plot
of the data as used in this research.
The second method of determining the energy correction was
to measure the total current at a number of voltages in the vicinity of
zero ion energy and extrapolate these voltages to zero total current.
Since the energy resolution at the magnet potential used was 3 e. v.,
the extrapolated voltage was increased to 3 ev. and this value was
averaged with that obtained by the first method to give a value of AV,
This process was then repeated at the end of each run to account for
any changes in conditions. The average of the two values obtained was
taken to correct all points during the run. These values were self-
consistent within 0. 5 e. v. over a twenty hour run.
As a compromise between maximum beam intensity (with an
energy resolution of 6 e. v. ) obtained with the magnet chamber main-
tained at 395 v. positive with respect to the cathode and smaller beam
intensity (with better resolution), the magnet chamber was maintained
at approximately 250 e.v. positive with respect to cathode during the
scattering measurements. This gave the 3 e.v. resolution mentioned
Source conditions were maintained essentially constant during
a given set of data in order to preserve the constancy of the ion energy
in so far as possible.
TA LE I
(e. ,v) (arbitrary units)
* Method of least squares.
Fig. 3--Fluxmeter Graph
To investigate the possibility of multiple scattering in the pres-
sure range covered,; data for RO/RS were plotted as a function of
10 1 s -3- !
This plot is given in Figure 4.. The linearity of the plot shows agS/a.
to be pressure independent in the pressure range 0 3 x 10'3 mm.,
One important change in technique used in this experiment was
the evaluation of the RT ratios at a negative lid potential. This potential
was -4 v. for higher energies than 50 v. and -2 v. for lower energies.
That this is justified may be seen by referring to Figure 5 in which
R/RS is plotted as a function of the lid potential. Below -1 v.: the
ratio is seen to become essentially constant. Since the negative potential
prevents slow negatively charged particles from reaching the lid, this
ratio is very nearly equal to R/RT.. The total ratios taken in this
fashion are probably more accurate than those obtained by the more labo-
rious method used in the past. This was to connect the lid, SL,. to the
scattering cylinder, and measure the current to the two while they were
maintained at amplifier ground potential. Under these conditions it is
possible that slow particles produced in the defining cylinder will be
measured on SL. The present method enables the elastic cross-section
to be obtained under the same experimental conditions, except for this
.o __ 0. 00
Fig. 4-- Pressure Dependency Graph
Fig. S-Idd Potential Graph
lid potential, as the total cross-section. Therefore, it offers largc
advantages over the prior nmettod, as re~ardE. accuracy and convenience
A firat-ordcr correction factor was applied to the R /R(
data. in view of the constancy of the slope at high positive potentials in
Figure 5. This was the extrapolation of the experimental ratios to zero
lid potential to account for loss in the beam due to the defocusing action
exerted by this potential on the elastically scattered ions. The action of
this correction was to increase the value of the elastic cross-section by
about 10 per cent. The extrapolation was carried out on the assumption
that the slope of the linear portion of the curve was independent of the
ion energy. This was found approximately true by experiment, The
same behavior of RO/Rg has been observed for the previous work on
the scattering of hydride ions in hydrogen (16). It is therefore reason-
ably certain that this is an apparatus effect rather than a physical phe-
The method followed in taking the current ratios was to admit
the oxygen to the scattering cylinder and to allow the pressure to equi-
librate before slowly filling the McLeod gauge. The pressure measure-
ments have been estimated to be accurate to t 1, 073 in the range
measured. After equilibrium was established, from three to eleven
ratios were taken at each energy. The average value of these reading
was taken as the "R" value for that energy. The oxygen was then shut
off from the scattering region and, after establishing equilibrium,
readings were taken for the "empty tube" or i"RO" values at the lowest
apparatus pressure attainable.
The pressure measurements were obtained in a carefully
calibrated McLeod gauge of 500 ml. capacity in the following way:
the average capillary correction for the pressure range employed was
determined as an average of from three to twelve readings; this value
was then subtracted from the scattering pressure reading which was
obtained in the same fashion. The capillary correction was taken at
the same pressure as the "empty tube" ratios. The scattering pres-
sure measurement was made each time gas was admitted to the scat-
tering tube. All measurements of cross-sections were made in the
vicinity of 1. 4 x 10-3 mm. of mercury. This pressure gave ratios
which were in the optimum range of 0. 6 0. 8, i. e. about 30 per cent
of the beam scattered. If more than 60 per cent of the beam is scat-
tered multiple collisions will seriously affect the results.
When shutting down, the system was pumped to a pressure of
the order of 5 x 106 mm. of mercury by the diffusion pumps. The
pumps were then shutdown and the system slowly filled with nitrogen
to one atmosphere pressure.
TREATMENT OF DATA
The results listed in Table 2 for the scattering investigated
in this research show the interaction cross-section of the hydride ion
and oxygen gas for three types of scattering. The data contained in this
table are plotted in Figure 6 over the energy region investigated. The
top curve shows the variation of the total cross-section with energy;
the center one, that of the elastic cross-section; and the lowest one,
the difference which is equated to the inelastic cross-section as dis-
cussed in Appendix I.
Analysis of Data for Elastic Scattering
This analysis is based on the four classical assumptions men-
tioned in Chapter II. The first of these, that interaction is attributable
to a central force field, is expressed as an attractive inverse n'th
power law of the particle distance.
(5.1) V = .K
DATA FOR CROSS-SECTION EVALUATION
0 0 Pc x 103 aT as
V(e.v.) AV(e.v.) RS RS RT RT T C VSLe. v.) (mm. Hg) (cm2/cm3) (m2 /cm3
S5 (mm. Hg) (cm /cm ) (cm /cm )
0 0 Pc xO 104 M as
V(e.v.) AV(e. v. RS R) RT R T* C VSL,(e.) (cm 1m3)
s s_(rnm. Hg) (cm /cra3 (cm /erm3)
The second assumption is that the oxygen molecule is initially at rest
with respect to the hydride ion. Since one has thermal energy and the
other has energy of at least several volts, this is certainly valid. The
third assumption is that the system is conservative. This assumption
is probably true at low energies, but becomes less safe at higher ener-
gies. As previously stated, the simplifying assumption is made that
the scattering angle is small. This assumption is certainly not true for
those ions which were scattered as they were about to leave the scatter-
ing cylinder; however, most of the scattering occurs in the upper half
of the scattering cylinder since the beam intensity drops in an exponen-
tial fashion. It has been shown by Kells ( 7) that a treatment of scattering
data based on the assumption that the cosecant of the relative scattering
angle equals the cotangent of that angle gives results which agree within
one per cent of an exact treatment described by Simons, Muschlitz,
and Unger (8 ).
For the system hydride ion oxygen molecule, the hydride ion
may be considered as having an initial velocity v0 and mass ml; while
the oxygen molecule will be treated as a single particle of mass m2,
initially at rest.
The distance between the particles is defined as r, with the
distance of closest approach defined as r0. The relative scattering
angle referred to the molecule is taken as .
Since the sys-icn io a conservative one, defining as the
relative radial angle,
(5.2) / rA = vob = a conasant, where / is the re-
duced mass and b is the impact parameter. The rclativc total
energy of the system is
(5.3) E M
Substitution of (5.1) givec
(5.4) Er A (i2 + r2Z*) & K I _v
a ran 0
Stice r = 0 when r = r equation (5.4) may be solved for
the irnpact parameter, b,
(. z 2 2.n
(5) b = r + r, and, averaging over the
scattering length, "I",
r. dx + 1 r dx
(5. 6) bas .o
= ~~" r" dx
K 2 Z-n
+ --,1 r dx .
Since a = NTba
(5.7) a N T + rN-+E / r2-n
"ro 0 E"o1" 0 d
Relating the minimum absolute scattering angle,
S= tan a where a is the radius of the scattering cylinder
hole, to the relative energy, E,
(5.8) tane =
KrZ r (n/2 + 1/z)
- -Tr-(8) and (7)
applying the assumption of small angle scattering to the equation
relating the relative and absolute scattering angles
(5.9) coto = cot
+ L cSc lWC cot 8
The measured ion energy is W = 1/2 mlv0
Equation (5. 7) now becomes
(5,. 10) a, S V = NC
Sn +2 aW
+ Nr a2n nm (:n1"C)(2/n) ( 12 /n
(5.11) a WZ/ = Q NnKC"" I a
S a ( 2 ""*
Upon taking logarithms of the first equation (5. 11),
(5. 12) log Q + 2/n log W = log Q follows.
This last equation shows that a plot of the logarithm of mS
against the logarithm of the ion velocity should be a straight line with
slope of *Z/n and intercept of the logarithm of Q. The intercept, log Q,
is used in evaluating the constant of the potential law from equation
(5. 11). It should be stressed that this law will hold only for the region
where all assumptions are valid. A plot of (5. 12) for hydride ions in
ogygen from 3420 volts is given in Figure 7. From this plot, the
slope was found to be -0.072 and the intercept, 1. 79. By utilizing the
preceding equations, the potential function for this range
(ba = 2.0 2.3 A ) has been evaluated as
(5.13) V - 10 2Z2 with r expressed in centi-
meters, and V in electron volts,
The previous work on the scattering of low velocity hydride
ions was in hydrogen; in which, there is no tendency for compound
formation. The H" -- H2 interaction appeared to follow an ion-dipole
() cm ocr 141 n
~ 0 0
Fig. 7--Potential Law Graph at Low Energies
(n = 4) type of attraction at low energies. The high value of the ex-
ponent in the potential law observed for oxygen at these low energies
is attributed to a short range attractive force arising from a valence
(or exchange) interaction between H"'and OZ. It would seem that the
data indicate IHOz to be a stable structure. This ion has been known to
exist in solution for some years as the anion of hydrogen peroxide (20).
The minimum in the elastic curve may be a typical property of
negative ion scattering since it has now appeared in both cases investi-
gated. For the hydride-hydrogen case, the minimum appeared at
about 55 volts and showed a gradual rise thereafter. In the present
case, as is shown in Figure 6, the minimum occurred at about 20
volts ion-energy and the cross-section then showed a gradual rise to
about 100 volts. Since this minimum was found at an ion energy of
twenty volts, it is quite possible that excitation of the oxygen molecule
takes place at higher energies. An excited electronic state, 'Ag,
having an energy of about 1 e. v. above the ground state is known for ox-
ygen (21). It would appear that this is a probable explanation of the
rise in the cross-section at higher energies. Such an excitation would
appear in the elastic measurement since the inelastically scattered ion
will still retain considerable kinetic energy.
The excitation postulated in the case of the hydride ion hydro-
gen molecule scattering (16) involves the excitation of the molecule to
the 1 u state. As this lies about 11 e.v. above the ground state of
hydrogen, the maximum probability for this excitation should occur at a
higher ion energy than should the excitation of oxygen postulated in the
present research. Comparison of the present data with those obtained
in the scattering in hydrogen bears out this theory. The elastic cross-
section for the hydride hydrogen system is still increasing at 400 e. v.
A search for positive ion formation was made at 400 e. v. ion
energy by measuring the current to the scattering lid when various neg-
ative potentials were applied to the lid. Since the ionization potential
of oxygen is 12. 5 e. v., no detectable ionization was anticipated at this
energy and no current was measured to the lid.
The inelastic cross-section curve qualitatively resembles that
obtained in the hydrogen scattering data above twenty volts. In the light
of the conclusions reached in that experiment, and also those obtained
by Hasted (11), it seems reasonable to conclude that the process taking
place above twenty volts is principally "ionization" of the hydride ion
according to the general type of equation (1. 10).
The inelastic cross-section for the system H -- H2 tended to
approach zero at lower energies. A similar behavior was noted by
Hasted (11) for the inelastic scattering of H" in the rare gases. In the
present research, the inelastic cross-section appears to rise below
twenty volts ion energy. Because of the rise in cross-section at lower
energies and because of the low absolute value of AE for the exchange
process, it seems reasonable to assume that the primary process which
is taking place in this neighborhood is the charge exchange phenomenon,
(6.1) H- f+ = H + 02 + aE.
AE is the energy difference between the electron-affinity of atomic
hydrogen and that of molecular oxygen. The probability of such a process
increases with a decrease in the absolute value of 4E. Values for the
electron affinity of molecular oxygen vary, but, assuming an approxi-
mate value of 0. 9 e. v. and using 0. 75 e. v. for the electron affinity of
the hydrogen atom there is obtained an absolute value of 0.15 e.v. for
the energy difference.
An estimate of the probable error in the determination of the
potential law, equation (5. 13), may be made assuming 0. 5% error in
RT/RT and RS/AS, 1% error in the pressure measurements, and
I 1 e. v. uncertainty in the ion energy. The probable errors in 0T,
aS and av are 1.5%, 4%, and 8% respectively. The resulting
probable error in the exponent, n, is L7. It is difficult to make an
estimate of the error due to the influence of inelastic scattering on
the elastic cross-section. This will be appreciable at ion energies
above 20 e.v. At very low ion energies the assumption that the
initial momentum of the oxygen molecule is negligible is questionable.
It is felt, however, that the overall uncertainty in the exponent is not
much more than 10.
Investigations have been made of the scattering cross-sections
of hydride ions in gaseous molecular oxygen in the energy range of three
volts to one-hundred-and-sixty volts incident ion energy. The cross-
sections for both the elastic and inelastic types of collision were investi-
gated and the results point to the following conclusions.
In light of the elastic scattering results, it seems quite reason-
able to assume that in the low energy region there is a tendency for
formation of the OaH ion, The potential expression which best fits
the data is
(7.1) V = 10
over the range of interaction 2. 0 2. 3 A. It is possible that this type
of behavior is a general one for negative ion collisions in which
valence or exchange forces are involved. Above twenty volts, the inter-
pretation is complicated due to a rise in cross-section which is attri-
buted to excitation of the molecular oxygen at the expense of kinetic
energy from the ionic beam.
The inelastic cross-section goes through a much sharper mini-
mum in the same region as the elastic scattering minimum. The lower
energy cross-sections are attributed to charge exchange between the ion
and the molecule while the higher energy cross-sections are primarily
the result of simple detachment of the electron from the ion to form a
(1) I Io exp (-* cr p) whore X is the current at dis-
tance x along the scattering path compared to Q, the current at the
start of the scattering paths crT ia the total cross-section for the
interaction (which may be attractive or repulsive); and p is the pres-
sure of scattering gas. Since all collisions are either elastic or inelas-
tic the total cross-section may be equated to the sum of the elastic and
inelastic cross-sections. Differentiation of (1) gives
(2) -dlx = dLT a Ix O pdx which, together with additivity
of the cross-sections, gives rise to
(3) -dk d( a td a gpdx and
(4) *dl1 a f(a Updx.
Division of (3) by (2) results in
d Ix( a) as
Integrating over the scattering length and indicating the total scattered
current by IT and the elastically scattered current by Ig, gives
(6) E = a Similarly, 1,= a where
T T T T
II is the inelastically scattered current.
Due to the design of the scattering region, all particles which
are not sufficiently deflected from the beam path will go to the Faraday
collecting cage and be counted as transmitted current. Essentially all
other ions will be collected on either the scattering cylinder or the lid
of the cylinder. The currents to the-collecting cage, the scattering
cylinder and the lid of the scattering cylinder may be represented by
IC, IS- ISL, respectively when the total ratios are being taken. When
the retarding potential is applied to the lid, IS and ISL become
Is' and LL'.
Since it is far simpler to work with fractions of the total cur-
rent than it is to calculate absolute currents as indicated by a potential
drop across a high resistance, it is found convenient to define the
(7) a) RT = =
IS + Ic + sL I C+ IT
b) Rg IC IC
S+ IC I + IE
c) RI = C 'C
ISL +C c C c +
Equations (7) may be: rearranged to give
S1 S L 1T
(8) a) ___ s
RT IC IC
I C IC
In light of equations (6),
(9) s -.---- aT and aI = .R. .
Rearrangement of (1) gives, with the length "1",
In L = pa "1"
(10) a In 0 1 In -L
P'"1" IC P11" RT
The actual equation used, however, was one based on scattering at two
different pressures; to account for small errors in alignment, beam
spread due to space charge effects. One of these pressures was the
lowest vacuum pressure obtainable and gave a scattering ratio, RO
very close to one. The corrected equatiir becarne'
(11) =_________. In
T (273. 2 K) "1" (P P) RT
= F log ; where
P P0 RT
S T(Z. 303)
273. 2*K "1
This same correction is applied to the RS values; i. e. R
ratios which were close to unity were used to obtain aS. Since
aI- a T ag it is not essential to measure I1L separately in
order to evaluate a
Definition of Symbols
C (Ion Source)
Collecting Cylinder or Faraday Cage
Electron Repelling Elements
Scattering Cylinder Lid or Plate
V (pp. 28-41)
V (p. 19)
Total Cross-Section in cm-1
Elastic Scattering Cross-Section in
Inelastic Scattering Cross-Section in
Relative Ion Energy in e.v.
Scattering Function (See p. 46)
Magnetic Field Strength
Length of Scattering Region
Scattering Pressure in mm of g
Radius of Curvature
Scattering Ion Current Ratios
(see pp. 44-46)
Total Ion Current Ratios (See pp. 44-46)
Distance of Closest Approach between
Ion and Molecule
Scattering Cylinder Lid Voltage
Ion-Molecule Interaction Potential
Ion Energy in Magnet Chamber in e.v.
Ion Energy in Scattering Region in e.v.
L. Ttxen, Z. Phys. 103, 463, (1936)
2. Sloan and Love, Nature, London, 159, 302, (1947)
3, Bailey, McGuire, and Muschlitz, in press, J. Chem. Phys,
4. Hylleraas, Z, Phys. 60, 624, (1930)
5. Massey, "Negative Ions, The Cambridge University Press,
Cambridge, 1950, pp. 122-128
6, Dukelskii and Zandberg, Doklady Akad. Nauk S. So S. R 82, 33,
7. Kells, J. Chem. Phys. 16, 1174, (1948)
8; Simons, Muschlitz, and Unger, J. Chem. Phys. 11, 322,
94 Davisson and Germer, Phys. Rev,, 30, 705, (1927)
104 Amdur and Pearlman, J. Chem. Phys. 8, 7, (1940)
11. Hasted, Proc. Roy. Soc., A205, 421, (1951); A222, 74, (1954)
12. Muschlitz and Simons, J. Phys. Chem, 56, 837, (1952)
13. Russell, Fontana, and Simons, J. Chem. Phys., 9, 381, (1941)
14. Simons, Francis, Fontana, and Jackson, Rev. Scl. Instruments
13, 419, (1942)
15. Simons and Cramer, J. Chem. Phys., 18, 473, (1950)
16. Muschlitz, Bailey, and Simons, "Technical Report Number Two
to the Office of Naval Research under contract 580(01)," 1955
17. Searcy, "A Study of Electronic Methods for the Measurement of
Small Direct Currents," M. S. E. Thesis, University of
18, Randolph, "A Study of Cataphoretically Coated Cathodes, "
M. S. E. Thesis, University of Florida, 1954
19. Nelson and Wing, Rev. Sci. Instruments, 20, 541 (1949)
20. Joyner, Z, anorg. Chem., 77, 103, (1912)
21. Herzberg, "Spectra of Diatomic Molecules, D. Van Nostrand
Co., Inc., New York, 1950- p. 446
22. Hasted, Proc. Roy. Soc., A212, 235, (1952)
23. Evans and Uri, Trans. Faraday Soc., 45, 217, (1949)
24. Kazarnovski, C. IR Acad. Sci. UR.. S., 59, 67, (1948)
25. Massey, op. cit., p. 28
John Murray McGuire was born in New Bedford, Massachu-
setts, on 15 May, 1929. He is the son of Mary Murray McGuire and the
late Thomas Christopher McGuire, Jr. He graduated from Belmont
Abbey Preparatory, Belmont, North Carolina, as Valedictorian in
1945. He then entered the University of Miami, Coral Gables, Florida,
from which he received the Bachelor of Sciencce degree cum laude in
1948 and the Master of Science in 1951. He then entered the Graduate
School of the University of Florida, Gainesville, Florida.
This dissertation was prepared under the direction of the
chairman of the candidate's supervisory committee and has been approved
by all members of the committee. It was submitted to the Dean of the
College of Arts and Sciences and to the Graduate Council and was ap-
proved as partial fulfilment of the requirements for the degree of
Doctor of Philosophy.
January 29, 1955
Dean, College 6f Arts and Sciences
Dean, Graduate School
A 1-44/ ?
? / x j6. I
/Y: c< ^,^*c <-
Internet Distribution Consent Agreement
In reference to the following dissertation:
AUTHOR: McGuire, John
TITLE: Scattering of hydride ions in oxygen gas. (record number: 559221)
PUBLICATION DATE: 1955
I, joA/6 Q /I (1 /re as copyright holder for the
aforementioned dissertation, hereby grant specific and limited archive and distribution rights to
the Board of Trustees of the University of Florida and its agents. I authorize the University of
Florida to digitize and distribute the dissertation described above for nonprofit, educational
purposes via the Internet or successive technologies.
This is a non-exclusive grant of permissions for specific off-line and on-line uses for an
indefinite term. Off-line uses shall be limited to those specifically allowed by "Fair Use" as
prescribed by the terms of United States copyright legislation (cf, Title 17, U.S. Code) as well as
to the maintenance and preservation of a digital archive copy. Digitization allows the University
of Florida or its scanning vendor to generate image- and text-based versions as appropriate and
to provide and enhance access using search software.
This/nt flpe missions pro ibits use of the digitized versions for commercial use or profit.
S atre oCopyright I-lder
Printed or Typed Name of Copyright Holder/Licensef
Personal information blurred
Dat of Sinature
Please print, sign and return to:
UF Dissertation Project
University of Florida Libraries
P.O. Box 117008
Gainesville, FL 32611-7008