Citation
Radiation chemistry of systems containing phosphine

Material Information

Title:
Radiation chemistry of systems containing phosphine
Creator:
Buchanan, James Wesley, 1937-
Publication Date:
Language:
English
Physical Description:
vi, 80 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Radiochemistry ( lcsh )
Phosphine ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida, 1968.
Bibliography:
Bibliography: leaves 77-79.
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright James Wesley Buchanan. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
030419815 ( ALEPH )
17009595 ( OCLC )

Downloads

This item has the following downloads:


Full Text
RADIATION CHEMISTRY OF SYSTEMS
CONTAINING PHOSPHINE
By
JAMES WESLEY BUCHANAN
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FORTH
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA 1968




ACKNOWLEDGEMENTS
Deep appreciation is expressed to Dr. R. J. Hanrahan, whose guidance was essential in the prosecution of this research. He has been a good friend, as well as an excellent research director, for many years, during which time the author took an M.S. degree and taught for several years before returning to work toward the Ph.D. in 1966.
Acknowledgement is also made to Gwen, who has been a good wife for many years, and who heroically typed throughout many nights to complete this manuscript.
ii




TABLE OF CONTENTS
Page
ACKNOWLEDGL4ENTS ...................................................... ii
LIST OF TABLES .................................................... iv,
LIST OF FIGURES ................................................... v
Section
I. INTRODUCTION ............................................. 1
Review of Earlier Work ........................... ........ 1
Status of Present Work ..................................... 8
II. EXPERIMENTAL METHODS ......................................... 9
Apparatus .......................... ..................... 9
Reagents .................................................20
Preparation of Samples .................................. 21
Product Analysis ........................................ 23
III. EXPERIMENTAL RESULTS ..................................... 27
Dosimetry ..................................... 27
Pure Phosphine............................................. 27
Pure Ammonia ..................................... 32
Phosphine-Amnmonia Mixtures ...................... 35
Pure Methyl Iodide Vapor ........... ...... 43
Phosphine-Methyl Iodide Mixtures........................ 46
IV. DISCUSSION AND INTERPRETATION ............................ 53
Pure Phosphine and Pure Ammonia .......................... 53
Phosphine-Ammonia Gas Mixtures ............................. 58
Methyl Iodide Vapor and Phosphine-Methyl Iodide Gas
Mixtures .................................................. 64
V. SUMMARY AND CONCLUSIONS .................................. 69
APPENDIX .......................................................... 71
BIBLIOGRAPHY ...................................................... 77
BIOGRAPHICAL SKETCH ............................................... 80
iii




LIST OF TABLES
Table Page
1. Mass Balance Results....................................... 33
2. Phosphine-Ammonia Mixtures................................. 41
3. Phosphine-Methyl Iodide Mixtures.......................... 52
iv




LIST OF FIGURES
Figure Page
1. High vacuum manifold ....................................... 10
2. Toepler pump-McLeod gauge combination ...................... 11
3. Sample analysis system ..................................... 12
4. Radiolysis vessel .......................................... 14
5. Radiolysis vessel .......................................... 15
6. Cross section of cobalt-60 gamma ray source ................ 17
7. Gas chromatograph sampling module .......................... 18
8. Gas chromatographic sampling loop ......................... 19
9. hydrogen yields from ethylene at 60 cm pressure, as a
function of irradiation time ............................ 28
10. Hydrogen yields from phosphine at 55 cm pressure, as a
function of radiation dose .............................. 30
11. Hydrogen yields as a function of phosphine pressure,
irradiated to a total dose of 4.4 x lO18 ev per cm PH3.. 31
12. Hydrogen yields from ammonia at 55 cm pressure, as a
function of radiation dose .............................. 34
13. Nitrogen yields from ammonia at 55 cm pressure, as a
function of radiation dose ............................... 36
14. Nitrogen yields as a function of ammonia pressure,
irradiated to a total dose of 4.4 x 1018 ev per cm NH3.. 37
15. G-values of hydrogen from wmmonia-phosphine mixtures as a
function of the fraction of energy absorbed by
phosphine ............................................... 39
16. G-values of nitrogen from ammonia-phosphine mixtures as a
function of the fraction of energy absorbed by
phosphine ................................................ 40
V




17. Methane yields from methyl iodide at 26.8 cm pressure, asa function of radiation dose ...........................44
18. Hydrogen yields from methyl iodide at 26.8 cm pressure, as afunction of radiation dose ...........................45
19. Methane yields as a function of methyl iodide pressure, irradiated to a total dose of 1.1 x 10ol9 ev per cm CH 3 47 20. Hydrogen yields as a function of methyl iodide pressure, irradiated to a total dose of 1.1 x 10 19 ev per cm CH 3I 48 21a. Yields of hydrogen and methane as a function of the
fraction of energy absorbed by phosphine ..................49
21b. Yields of hydrogen and methane as a function of the
fraction of energy absorbed by phosphine (expanded
abscissa scale)......................................... ....... 51
vi




I. INTRODUCTION
The object of this study was to carry out an investigation
of the gamma radiolysis of pure phosphine and to compare its decomposition with that of its nitrogen analogue, ammonia. The irradiation of gas mixtures of ammonia and phosphine was also done. Results obtained from radiolysis of these mixtures indicated that phosphine was acting as an efficient radical scavenger. Confirmation of this was sought from radiolysis of mixtures of methyl iodide vapor and phosphine.
Review of Earlier Work
Radiolysis and Photolysis of Phosphine
The only report of a study involving the irradiation of
1
phosphine is that of Sellers, Sato and Strain, who irradiated a large number of phosphorus-containing compounds with neutrons plus gamma rays as well as with intense gamma radiation. Their work with phosphine was not quantitative, but they did observe that with both the mixed radiation and pure gamma rays the only detectable products were hydrogen and an orange-red deposit on the walls of the quartz tube in which the irradiations were carried out. Also, the irradiation of yellow phosphorus resulted in conversion to the red form, which was stable to neutrons and gamma rays even in the presence of hydrogen gas. This work was a very minor part of their study, which concentrated on the
1




2
radiolysis of crystalline salts of phosphoric, phosphorus and hypophosphorous acids.
A mass spectrometric study of phosphine and diphosphine by Wada and Kiser2 indicates that bombardment of phosphine with 70 ev electrons probably results in the following two major processes:
PH3 4 PH+ + H2 + e- ()
PH PH 3+ + e (2)
33
To a lesser extent, the reactions
PH 3 P+ + H. + H2 + e- (3)
PH 3 PH 2+ + H* + e (4)
also may occur. Two similar studies, by Neuert and Clasen3 and
4
Saalfeld and Svec, agree reasonably well with respect to the relative abundances of ions in the mass spectrum of phosphine.
Halmann and Platzner5 have demonstrated that an important reaction occuring when phosphine at increased pressures is bombarded with electrons in a mass spectrometer is
PH 3+ + PH3 PH4+ + PH2" (5)
The pressure dependences of the intensity of the primary ion PH 3+ and
+
the secondary ion PH4 were measured, as well as the appearance potentials of the ions PH4+, PH3+, PH2+ and PH+.
Several studies of the photochemical decomposition of
phosphine have been made. The classical work of Melville6'7'8'9 provides a study of the overall reaction as well as an investigation of




possible mechanisms. lie found that the decomposition followed the reaction
PH3 + hv Pred + 3/2 H2 (6)
and concluded (from pressure versus quantum yield studies) that wall reactions were probably important secondary steps in the photolysis. His mechanism involved a primary dissociation
PH3 + hv PH2. + H. (7)
and (after diffusion to the walls) the secondary reactions:
PH2. + PH 2 P 2(red) + 2112 (8)
H.+H. H2 (9)
PH2+H. PH3 (10)
Temperatures up to 3000C had little effect on the yields, implying that the heats of activation of the surface reactions were small.
A study of the flash photolysis of phosphine by Norrish and
OldershawI0 has shown that when high radical concentrations are created the secondary reactions are homogeneous. The overall reaction again results in red phosphorus and hydrogen, with the phosphorus being produced in the form of small particles which remain suspended in the gas phase for several minutes. The spectra observed in this work were those of PH2., PH, P2 and a continuous absorption which was taken to be the suspended solid phosphorus. The major secondary reaction of PH2" was assumed to be
PH2. + PH 2 PH + PH (11)




4
rather than reaction (8) as in Melville's mechanism. Both this study and a similar flash photolytic study by Ramsay 11show that the major primary process in phosphine photolysis is reaction (7). Their work at lower wavelengths (below 2000 A) indicates that some primary dissociation to PH also occurs.
Work on Irradiated Ammonia
A large number of studies exist which deal with both the photochemistry and radiation chemistry of gaseous ammonia. What is included here is a summary of some of the more recent and definitive investigations.
The primary steps in the photolysis of ammonia are the following:
NH 3+ hv -~NH 2* + H. (12)
NH 3+ hv -~NH +2HorH 2(3
According to McNesby, Tanaka and Okabe, 12the major process is reaction
(12), but reaction (13) increases in importance at lower wavelengths.
An interesting mass spectroscopic study by Melton 13shows that when ammonia at one torr is bombarded with 100 ev electrons the predominant positive ion, is NH 4 + formed most probably in the ionmolecule reaction
NH 3+ + NH 3 NH 4+ + NH (14)
Measurements at 2 x 10-7 torr gave no measurable m/e =18, but NH 3+ accounted for over one-half the total positive ions formed. At intermediate pressures the percent abundance of NH 4 +increased from zero




5
to 83, and the parent ion abundance fell from 60% to 12%. At pressures approaching atmospheric virtually all the positive ions must be ammonium and the neutralization of NH4+ must constitute a very important secondary step in the radiolysis of ammonia.
14
Toi, Peterson and Burton at Notre Dame have studied the
effect of gas density on the radiolysis of ammonia in a stainless-steel autoclave at slightly above the critical temperature of 132 0C. They found sharp decreases in G(H 2) and G(N 2) in the density region 0.05 to 0.15 grams per cc. In radiation chemistry, the G-value is defined as the yield from a radiolysis, in molecules changed (formed or destroyed) per 100 electron volts of absorbed energy. At densities less than 0.05 grams per cc, G(H 2) = 6.2 and G(N ) = 2.0 and at densities exceeding 0.15 grams per cc, G(H 2) and G(N 2) leveled off at about 1.5 and 0.4, respectively. The sharp drop in yields occurs well below the critical density of 0.235 grams per cc. The results are interpreted in terms of the formation of ion "clusters," which reduce the probability of dissociative neutralization by permitting the energy of neutralization to be spread over the molecules of the cluster. They consider that other effects, such as deactivation of excited species, may also be involved.
Meaburn and Gordon15 studied the rate of disappearance of NH radicals in irradiated ammonia, using the technique of pulse radiolysis. They found that the radical forms with a G-value of 0.4 and decays by a second-order process. Also, the addition of oxygen, ethylene and propylene decreased sharply the half-life of NH, but did not reduce it to zero. The explanation given for the persistence of some NH radicals even at high scavenger concentrations was that the radicals were




6
regenerated by further reactions of some of the products formed in the scavenging steps.
Two very recent studies of the electron radiolysis of gaseous ammonia and ammonia mixtures by Jones and-Sworski16 and Jones, Sworski and Williams17 have thrown considerable light on the nature of some of the secondary processes involved and particularly on the "scavengeable" and "nonscavengeable" yields of hydrogen. The G-values of H2 and N2 at 230C and 400 torr were found to be 4.5 and 1.5, respectively. Both yields increased with an increase in temperature to limiting values of G(H 2) = 15 and G(N 2) = 5. Very low concentrations of ethylene drastically reduced the hydrogen yield to a minimum of 0.8 at [C 2H 4] NH3] =
0.02. The yield increased linearly with further increase in [C 2H 4/[ NH3]. Extrapolation of the subsequent increase back to [ C2H] ] NH3] = 0 gave 0.75 0.05 molecules per 100 ev as the minimum "molecular" H2 yield in ammonia. The slope of this straight line gave G(H 2) = 1.5 0.5, which is the hydrogen yield in ethylene, within experimental error. These workers also found a pressure effect between 100 and 400 torr, with the yields of both hydrogen and nitrogen maximized at just greater than 100 torr. They found no evidence of appreciable hydrazine formation in irradiation of ammonia in a static system.
The effect of hydrogen atom and electron scavengers on the gas phase gamma radiolysis of ammonia was investigated by Nishikawa, Shinohara and Matsuura.18 Carbon tetrachloride vapor depressed G(H2 from 4.3 (in pure ammonia at 50 cm pressure) to 0.64 at 2 mole percent CC14. The authors suggest that carbon tetrachloride may be acting both as an H atom and electron scavenger, by the reactions




7
CC14 + e- CCl 3 + Clf (15)
CC14 + H. Ccl 3 + HC (16)
Both of these reactions are energetically favored. Other additives such as ethylene and propylene were also effective in reducing G(H 2) to less than one, with the exception of nitrous oxide. Minimum G(H 2) with added N20 was 2.5. This depression of only 1.8 G-units with added N20 is attributed to the change in the neutralization process of ammonium, from
NH + + e H- and/or H2 (17)
to a reaction such as
NH 4 + 0- or 02- no H* or H2 (18)
Nitrous oxide does not scavenge hydrogen atoms. The ratio of the decrease in G(H 2) on addition of N 20 to G(H 2) in pure ammonia, 0.42, is considered to be representative of the contribution to the total hydrogen yield from processes involving ions. Radiolysis and Photolysis of Methyl Iodide
The reactions leading to methane formation in irradiated
liquid methyl iodide have been studied by Gillis, Williams and Hamill19
20
and by Petry and Schuler. There are probably two reaction routes, one involving a spur reaction between methyl radicals and hydrogen iodide
CH + HI CH4 + I. (19)




and another in which "hot" radicals attack the substrate
CH + CH3I 3 CH 4 + CH 2I. (20)
Most of the radicals back-react with iodine
CH + 1 CH 3I + I- (21)
so that the competition for methyl radicals is represented by reactions (19), (20) and (21). No information on the vapor-phase radiolysis of methyl iodide appears to be available, but a photolytic study by Souffie, Williams and Hamill21 shows that, in the vapor, methane is probably formed exclusively via the "hot" radical step. The use of hydrogen iodide as a scavenger in liquid CH 3I results in a 139
large increase in methane production.19 The radiolysis or photolysis of methyl iodide vapor with added HI does not appear to have been reported.
Status of Present Work
A quantitative study of the decomposition by gamma radiation of both pure phosphine and pure ammonia has been carried out. Also, the irradiation of mixtures of ammonia and phosphine and of methyl iodide and phosphine has shown that phosphine is an efficient radical scavenger, through the donation of a hydrogen atom. Further studies of the radiolysis of systems containing phosphine are indicated.




II. EXPERIMENTAL METHODS
Apparatus
Vacuum System
All samples were prepared on the vacuum line shown in Figure 1. The design of the line was standard, consisting of a Welch Duo-Seal forepump, two liquid nitrogen cold traps and a two-stage mercury diffusion pump. To the main manifold were attached the following: a mercury manometer (M), inlets for phosphine or ethylene and ammonia (both fitted with barium oxide drying tubes) (I), gas storage bulbs (each with a cold finger trap for degassing) (B), a fitting with stopcock for a vacuum thermocouple gauge (G), a detachable trap with o-ring fittings (DT), a submanifold (SM) and a Toepler pump-McLeod Gauge combination (Figure 2). The submanifold was fitted with a small calibrated volume (V) and a tubulation to which the irradiation cell (C) was attached for loading. A stopcock isolated the Toepler pump from the main manifold. On the back side of the main manifold were three detachable U-tube cold traps arranged in series (UT), each fitted with two Teflon stopcocks and o-ring joints, and another cold trap (TT) which was situated between the Toepler pump and the tubulation for the attachment of the irradiated cell (C) via a glass breakseal (Figure 3). All greased stopcocks and joints were greased with Apiezon N and L vacuum greases.
The entire vacuum system was constructed inside a stainless9




GK
SM B B
IT\V
M DT
Figure 1. High vacuum manifold




jI
Sample
loop (see I
Figure 8)
II
L ---------- M
Figure 2. Toepler pump-McLeod gauge combination




TT
UT
C
....Fu..r 3 Sa ls.......is.
Figure 3. Sample analysis system




13
steel "California" hood with sliding glass doors, so that the system could be effectively isolated from the remainder of the laboratory. Irradiation Cells
Two types of irradiation cells were used, as shown in
Figures h and 5. The one-liter spherical cells which were used for most of the irradiations were fitted with a cold finger trap and breakseal. The small annular cells were of approximately 10 ml capacity.
Cobalt-60 Gamma Ray Source
All irradiations were carried out in a "Wisconsin" type cobalt-60 gamma irradiator described previously.22 Figure 6 is a cross-sectional view of the irradiator. The absorbed dose rate was measured relative to G(H 2) of 1.2 for ethylene, as will be described subsequently. Irradiation cell geometry was such that the maximum gamma ray intensity per unit volume was obtained. All irradiations were done at room temperature.
Microtek Gas Chromatograph
A model GC-2000-R Microtek Research Gas Chromatograph was used, with the sampling module modified as shown in Figure 7. This modification was done by Mr. R. E. Marcotte of this laboratory. The sampling loop for the gas samples is shown in Figure 8. Bendix Time-of-Flight Mass Spectrometer
Product identifications were made with a Bendix model 14-107 Time-of-Flight Mass Spectrometer. Samples were introduced by attaching U-tube traps with an o-ring joint to a semi-direct inlet. The inlet




7
Figure 4. Radiolysis vessel




15
Figure 5. Radiolysis vessel




Figure 6. Cross section of~ cobalt-60 gamma ray source
Legend = (A) counterweight; (B) upper support;
(C) control rod handle; (D) extra top shielding;
(E) storage turret; (F) 400 curie C6osource;
(G) shutter shown open; (H) rear wall; (I) door;
(J) downward shielding; (K) door carriage;
(L) door crank; (M) door frame




17
A
C
D E L
F
G ,x
%kORKING SPACE' Al




18
GAS C1-11-?OX,-7'A- TOG17,4PI-1 SA"PLIAIG All-ODUIZIIVlL-C7IOIV POR7- C', ax S
fj t
70
///8hV S7 441VZZ--,5S R07A RY VA r:-L TCIDIIVG
., i OIV7- A451- DC)X-;
%TZ D141 0 A Z- VIIC-.* /IV
0 tI r
Figure 7. Gas chromatograph sampling module




19
Figure Gas chromatographic sampling loop




20
was connected to the ion source through a metering valve. The large ballast volume of the inlet system was not used.
Reagents
Ammonia
Matheson Company anhydrous ammonia (99.99% minimum purity) was used as obtained from the manufacturer, for all experiments. Phosphine
Matheson Company phosphine (99.5% minimum purity) was used without further purification.
Ethylene
Matheson Company C.P. grade ethylene (99.0% minimum purity) was used as obtained.
Methyl Iodide
Mallinckrodt Analytical Reagent grade methyl iodide was
passed through an 11-inch column of Alcoa activated alumina, grade F-20, and its purity ascertained via gas-liquid chromatography. Only one lower boiling impurity was found, whose concentration was significantly reduced by passage-through alumina. Although it was not identified, it was estimated that its concentration could not be greater than one or two parts per million. Two higher boiling impurities, also of low concentration, were present. The material was further purified by distillation (see section on preparation of samples).




21
Preparation of Samples
Dosimetry
After evacuation of the main manifold, mercury manometer, ethylene inlet and two calibrated storage bulbs to a pressure of about 0.1 micron, the first bulb was closed off and the main manifold was isolated from the pumping station. Ethylene was admitted slowly through the barium oxide drying tube until the desired pressure was obtained. The ethylene to be used in filling the cell was closed off in the second storage bulb, whose volume was calibrated with distilled water before attaching to the line. Excess ethylene was then frozen down with liquid nitrogen in the first bulb, to be used in a later run.
After three cycles of freeze-evacuate-thaw degassing, the
ethylene was allowed to come to room temperature in the bulb while the submanifold and irradiation vessel were pumped down for a minimum of 20 minutes. The main and submanifolds were then isolated from the pumping station and the ethylene was transferred to the liquid nitrogen cold trap in the irradiation cell. Fifteen minutes was allowed for complete condensation, after which the cell was sealed off (while open to the pump) by collapsing the constricted part of the neck with a gasoxygen torch. The neck was allowed to cool to room temperature before removing the liquid nitrogen from the cold finger.
After each irradiation and analysis the cell was evacuated, removed from the line and annealed in an oven at 5750C. After attachment to the submanifold for another run, it was thoroughly flamed out with a bush gas-oxygen flame while open to the pumping station.




22
Pure Phosphine
Sample preparation was similar to that described above for ethylene except that the cell was washed out with concentrated nitric acid and then distilled water before annealing, to remove the elemental phosphorus left on the glass surface by the previous experiment. Pure Ammonia
Sample preparation was very similar to that for ethylene
except that only one storage bulb was used. The volume of the main manifold, mercury manometer, ammonia inlet and storage bulb was calibrated with a known amount of gas and, by a suitable proportion, the main manifold pressure which would correspond to a given cell pressure could be determined. Thus in these runs all the gas taken from the cylinder was used in filling the cell.
Pure Methyl Iodide Va por
A sample of 1 or 2 mls of methyl iodide was attached to the main manifold via a detachable trap with 0-ring joint, and the liquid was degassed in the usual way. It was then expanded into a calibrated one-liter capacity gas storage bulb until the desired pressure was obtained, when the stopcock was closed and the excess methyl iodide was frozen down in the trap and discarded. In a typical run only about one-half of the methyl iodide was needed to fill the bulb, so that the purity of the material (with respect to higher boiling substances) was further enhanced by this distillation process.
After thorough evacuation of the manifold and cell the methyl iodide was transferred to the cold finger trap on the cell and the cell was sealed off under vacuum.




23
Phosphine-Ammonia Mixtures
Phosphine and ammonia were, in turn, admitted to the manifold to the desired pressure and then frozen down and degassed in the appropriate storage bulbs. The gases were then transferred to the irradiation cell and sealed off in the usual manner.
Phosphine-Methyl Iodide Gas Mixtures
Loading of these cells was carried out in the same manner as with the phosphine-ammonia mixtures.
Product Analysis
Dosimetry Products
Irradiated ethylene samples were attached by means of a breakseal fitting to the vacuum line, as shown in Figure 3. All materials noncondensible at -196C were collected via a Toepler pump and the gas pressure measured on a McLeod gauge. Each sample was degassed through approximately 18 cycles or until pressure increments were 0.1 mm or less, as read on a meter stick attached to the McLeod gauge. Intermittant thawing of the condensible materials was carried out, to remove traces of noncondensibles trapped in the frozen material. After collection, the gas samples were transferred via the Toepler pump to the gas chromatographic sampling loop (see Figure 2).
Analysis for methane was made with the Microtek Research Gas
Chromatograph, using a 0.75 meter 100/110 mesh silica gel column at 400C and a hydrogen flame ionization detector. The carrier gas was dry nitrogen at a flow rate of approximately 80 cc per minute. Methane standards were injected with a Hamilton 500 Ul gas-tight syringe.




Hydrogen analysis was by difference, the only materials
present in the Toeplered gas mixture being hydrogen and methane. Estimated error in hydrogen analysis was 1%. Phosphine Irradiation Products
Hydrogen analysis was via Toepler pump and McLeod gauge, as previously described. A mass spectrometric analysis of the material collected and measured in this way showed it to be only hydrogen.
Two U-tube cold traps were used to fractionate the condensible materials. These traps were maintained at -1270C and -1960C. After fractionation, the traps were removed and their contents analyzed mass spectrometrically. No quantitative analysis for elemental phosphorus was made.
Phosphine Mass Balance
Analysis for hydrogen was as described earlier. Phosphine
pressure before irradiation was determined directly from the main manifold mercury manometer. The phosphine was then sealed off via a Teflon stopcock on a calibrated submanifold volume. Phosphine pressure after irradiation was determined with the McLeod gauge, using a large calibrated volume attached above the stopcock via a standard taper joint. A blank run was made to verify the consistency of the volume calibrations. Estimated error was 1 1%.
Ammonia Irradiation Products
After measuring with the McLeod gauge the total gas collected via the Toepler pump, the gas was transferred to the gas chromatographic sampling loop and attached to the sampling module of the Microtek gas




25
chromatograph, as shown in Figure 7. A mechanical fore pump was used to evacuate the volume between the loop stopcocks and the sliding valve. The valve was then pushed in to direct the carrier gas toward the sampling loop and through the loop bypass (refer to Figure 8). After the resulting pressure surge had subsided, as evidenced by the recorder pen returning to its normal baseline, the sample loop stopcocks were opened and the bypass stopcock closed. The sample passed through a 0.75 meter 100/110 mesh silica gel column at room temperature, and a thermal conductivity detector which was calibrated for nitrogen analysis with standard injections from a 500 V1l gas-tight Hamilton syringe. The carrier gas was helium at a flow rate of about 25 cc per minute. Estimated error was 2%.
Hydrogen analysis was by difference, since the products noncondensible, at -1960C could only be hydrogen and nitrogen.
No attempt was made to analyze for hydrazine, it being assumed that no measurable amount was present in the irradiated samples. 13,14,17
Methyl Iodide Vapor Irradiation Products
Products noncondensible at -1960C were collected with the
Toepler pump and measured with the McLeod gauge. The gas was then pumped into the gas chromatographic sampling loop and attached to the Microtek sampling module. It was passed through a 0.75 meter 100/110 mesh silica gel column at room temperature, and a hydrogen flame ionization detector calibrated for methane analysis in the usual way. The carrier gas was helium at a flow rate of about 65 cc per minute.
Hydrogen analysis was by difference, since in this system




26
the only products noncondensible at liquid nitrogen temperature are methane and hydrogen. This was verified by a mass spectrometric analysis, which showed an intense peak at mass 2, in addition to the spectrum of methane. No attempt was made to analyze for products other than methane and hydrogen.
Phosphine-Ammonia Irradiation Products
Analysis of the materials which were noncondensible at -1960C was identical to that for the ammonia irradiation products.
In a search for other products, a series of three U-tube cold traps was used (sae Figure 3). After pumping off the permanent gases, the remaining materials were fractionated among the traps, which were immersed in baths at 00C, -780C and -196C. The traps were then removed from the vacuum line and their contents analyzed mass spectrometrically.
Phosphine-Methyl Iodide Irradiation Products
Quantitative analyses for methane and hydrogen were as described in the preceding section.




III. EXPERIMENTAL RESULTS
Dosimetry
The results of the dosimetry with ethylene are summarized in Figure 9. In this and all subsequent graphs, the units for the ordinate and abscissa are considered to include any multiplicative factor, such as 10 -6. From this graph it can be seen that the hydrogen yield is directly proportional to irradiation time, from three hours to 24 hours, at 60 cm ethylene pressure and 23 20C. The longer time corresponds to l20
a total absorbed dose of 5.4 x 10 ev per gram. The calculation and its theoretical justification are found in the Appendix. The absorbed dose rate was found to be 5.68 x lo14 ev per mm of ethylene pressure per liter of cell volume per minute of irradiation time. This rate was corrected at two-week intervals to account for the decay of the cobalt-60.
Pure Phosphine
Although from previous photochemical investigations6,7,10 it was suspected that the only products of the gaLmma ray-induced decomposition of pure phosphine would be hydrogen and red phosphorus, an attempt was made to find other products. Mass spectra of the contents of the cold traps used to fractionate the condensible materials after radiolysis (see page.24) showed only trace amounts of masses 66 and 62 27




28
9
0 1
-o 8
E
._ 7
( 6 -j 5
Li!
4
z
(.5 o 0
C' 2 2: 1
6 12 18 24
IRRADIATION TIME (in hours )
Figure 9. Hydrogen yields from ethylene at 60 cm pressure, as a function of irradiation time




29
in the -1270C trap, in addition to the spectrum of phosphine. Virtually all the phosphine was collected in the -1960C trap.
The irradiation cell contained a fine deposit of solid
material which would not distill into the traps, even on warming with a hot air gun. Although its coloration was not precisely determined because of the brownish coloration of the irradiated glass of the cell, it appeared to be brownish-red or brownish-orange. One striking feature was the spacial distribution of the solid on the bottom of the cell. Emanating from the center (directly under the test tube well into which the cobalt-60 source fits), the solid showed a star-shaped or "exploding" pattern, extending several inches out from the center.
The hydrogen yields from pure phosphine at 55 cm pressure
and room temperature, irradiated at a dose rate of 3.25 x 1017 ev per liter per minute, are plotted as a function of dose out to approximately 1021 ev, in Figure 10. Beyond 5 x 10 ev there appears to be a slight drop-off in hydrogen production. Figure 11 shows that the hydrogen yields are directly proportional to phosphine pressure, from below 4 cm up to 76 cm. The absorbed dose was 4.4 x 1O18 ev per cm PH3 or 2.4 x 120 ev per gram. The irradiation time was held constant.
To illustrate the calculation of a G-value when stopping
powers are used, the determination of G i(H 2), the initial (or zero dose) hydrogen yield per 100 ev, from phosphine is shown here in some detail. The rate of hydrogen production as a function of irradiation time (corresponding to Figure 10, except that the abscissa in the figure has been converted from hours of irradiation time to absorbed dose) was
3.7 x 10i6 molecules per minute. Since the phosphine pressure was 55 cm the rate of hydrogen production was 6.7 x 1013 molecules per minute




0
E 0
.C
c 8 ..J
LU
-6
L 4 (9
0 fr2
I I I J 1 I I I
2 3 4 5 6 7 8 9
RADIATION DOSE( in ev x 102)
Figure 10. Hydrogen yields from phosphine at 55 cm pressure, as a function of radiation
dose




50
o
0
x 40
E
30
W 20
z
LUI
0. 10
I I I f I I I
10 20 50 40 50 60 70
PRESSURE OF PHOSPHINE (in cm)
Figure 11. Hydrogen yields as a function of phosphine pressure, irradiated
to a total dose of 4.4 x 1018 ev per cm PH3
3..




32
per mm of phosphine. The calculation of Gi(H 2) was as follows: Gi(H2) molecules H2 produced/100 ev absorbed dose (6.74 x 1013 molecules/min-mm PH3) x i00 (5.83 x 1014 ev/min-mm C2H4)(SpH 3/S c2H4)(Co decay constant)
6.74 x l013 x 100
114
5.83 x 10 (l.044)(0.9838)
= 11.3
In the above equation, S PH3/Sc2H4 is the ratio of electron stopping powers of phosphine and ethylene.
Results of the mass balance experiments are summarized in Table 1. The decrease in phosphine pressure during irradiation was found to correspond quite closely to two-thirds of the hydrogen pressure, after making temperature corrections. Although no analysis was made for phosphorus, these experiments clearly showed that no products other than hydrogen and phosphorus were formed, and that the decomposition closely followed the stoichiometry:
PH3 + Pred + 3/2 H2 (22)
The trace amounts of masses 62 and 66 in the -127 C trap probably indicate P2 and PH 2PH2, respectively.
Pure Ammonia
Hydrogen and nitrogen yields from irradiation of pure ammonia at 55 cm pressure and room temperature, at a dose rate of 2.06 x 1017 ev per liter per minute, are plotted as a function of dose in Figures 12




TABLE 1
Mass Balance Results
Run Irradiation Pi (PH3)a Pf(PH3)b AP(PH3) P(H2) -3/2 AP(PH3) % Decomposition Time (hours) (cm) (cm) (cm) (cm) (cm)
blank 0 20.60 20.61 ----- ---- ---- --1 143 19.01 17.48 -1.53 2.30 2.30 8.1
2 108 15.32 14.40 -0.92 1.39 1.38 6.0
pressure calculated for McLeod gauge, from initial phosphine pressure in calibrated
volume and the ratio of McLeod gauge and submanifold volumes.
pressure of phosphine in McLeod gauge after irradiation.




LO
0 0
0
E
Cr)
-J
w
uJI II
I2 3 4
RADIATION DOSE (in ev x 10 20)
Figure 12. Hydrogen yields from ammonia at 55 cm pressure, as a function of radiation dose




35
20 20
and 13. Both are linear from 1 x 10 ev out to 5 x 10 ev, and the line extrapolates back to zero yield at zero dose. Figure 14 shows that nitrogen yields increase linearly with increasing ammonia pressure, from 10 cm up to one atmosphere, the line extrapolating back to zero yield at zero pressure. Hydrogen yields also were independent of pressure and were always three times the nitrogen yield. For the study of pressure dependence, the absorbed dose was 4.4 x 1018 ev per cm NH3, and the irradiation time was held constant.
The G-values for nitrogen and hydrogen were found to be 1.5 and 4.5, respectively. A number of other workers13,14,17 have failed
to find significant amounts of hydrazine formation from the gamma radiolysis of ammonia in a static system. No attempt to find hydrazine was made in this work. However, the yield of nitrogen as found by gasliquid chromatography was checked against one-fourth of the total gas measured in the McLeod gauge, and the results agreed within 2% or less in each case.
Phosphine-Ammonia Mixtures
Mass spectrometric analysis of the materials noncondensible at -1960C showed peaks at masses 1 and 2 and enhancement of the background mass of 28. The 0C and -780C traps showed the presence of small quantities of masses 31 and 62, with the -78C trap having a trace of masses 79 and 91. The -1960C trap gave intense spectra of phosphine and ammonia.
The solid residue remaining in the irradiated cell was of
exactly the same appearance, including the unique spacial distribution,




X 10
(n 9
0
---7
C
'67
0
_o 6
- 5
z 4
x
(.9
0 3 0:
F
I
! 2 3 4 5
RADIATION DOSE (in ev x 10 20)
Figure 13. Nitrogen yields from ammonia at 55 cm pressure, as a function of radiation dose




O 9
8
o$ 7
C()5 06
0
u4
z
2
2
0
10 20 30 40 50 60 70
PRESSURE OF AMMONIA ( in cm )
Figure 14. Nitrogen yields as a function of ammonia pressure, irradiated to a total dose of
4.4 x lO18 ev per cm NH3




38
as that from the radiolysis of pure phosphine. This residue was still visible, although diminished in quantity, after radiolysis of samples containing as little as one mole percent phosphine.
The mixtures were made up to a total pressure of 55 cm at 230C 120 120
and irradiated to doses of either 5 x 10 ev or 1 x 10 ev. The lower dose was chosen for low phosphine mixtures in order to avoid using up all the phosphine during radiolysis.
G-values for hydrogen and nitrogen from these mixtures were plotted as a function of the fraction of the dose which was absorbed by the phosphine. These graphs are Figures 15 and 16. It should be noted that while the hydrogen yields are larger than predicted from the assumption of an "ideal" mixture in which each compound decomposes proportionately to the dose it absorbs (see dashed lines on figures), the nitrogen yields show a striking depression. This occurs even in the mixtures with very low mole percentages of phosphine. It is noteworthy that a "residual" yield of nitrogen persists even in mixtures containing relatively high phosphine concentrations. In mixtures 8 and
9 (see Table 2), where the mole percentages of phosphine were 0.2 and 0.02, respectively, a reduced dose was used. These runs, in addition to one of the runs with pure ammonia, are indicated by circles on the figures.
The composition of each mixture and the irradiation times are given in Table 2. The fraction of the dose which was absorbed by phosphine, [S PH3/(SNH3 + S PH3)], was found by first computing the ratio S NH3/SPH3 and then forming the quotient [1/(l + S NH3/S PH3)]. The required irradiation time was calculated from the dose which each component absorbed. The time for each mixture was checked by making




"0 !0 .' s,, ,, ." '
O,,
02
7 ..--- 5 x 102 ev
OO20 E -0 1 x 1o ev
c- 5 -.
04
fraction of energy absorbed by phosphine
U,
Z)
_I3
0.0c20 03504 .0 06 07 .0 09
FRCIO FENRYABOBD YPOSHN
Fiue1.Gvle fhdoe rmamni-hshn itrsa ucino h frcto 3feeg bobdb hshn




40
Absorbed Dose 14 0 -i x 1020 ev
\020
\- 5x 10 ev
3 1.0
E \
CD
.8\
wI .6I UJ
o
:
> 4
0\
0.20 0.40 0.60 0.80
FRACTION OF ENERGY ABSORBED BY PH3 Figure 16. G-values of nitrogen from ammonia-phosphine
mixtures as a function of the fraction of energy absorbed
by phosphine




TABLE 2
Phosphine-Ammonia Mixtures
Mixture P(PH3 )a P(NH )a Mole Fraction SPH Timeb
Number (cm) (cm) PH 3 (hours)
3
SNH3 + SH
NH PH
33
1 55 0 1 1 25.17
2 50 5 0.909 0.940 26.97
3 30 25 0.545 0.654 31.33
4 15 40 0.273 0.371 35.48
5 5 50 0.091 0.138 39.02
6 2 53 0.036 0.057 40.17
7 0.5 54.5 0.009 0.01o4 40.75
8 0.1 54.9 0.002 0.003 8.32
9 0.011 55 0.0002 0.0003 8.32




TABLE 2 (continued)
Mixture P(PH3 )a P(NH )a Mole Fraction SPH Timeb
Number (cm) (cm) PH3 (hours)
SNH3 + SPH3
10 0 55 0 0 ho.83
11 0 55 0 0 8.32
aAll pressures are corrected to 230C.
bTimes greater than 25 hours correspond to a dose of 5 x 1020 ev for that mixture;
8.32 hours corresponds to a dose of 1 x 10 ev.




43
the calculation twice, once for each component, and verifying that these times were the same.
The percent decomposition of phosphine with mixtures 6, 7, 8 and 9 was 3, 11, 18 and 90, respectively. These were calculated from the observed yields of hydrogen and nitrogen, assuming that no products were formed in significant amounts other than these and elemental phosphorus. The mass spectral analyses of the 00C and -78C traps would appear to bear out this assumption.
The extent to which the "ideal mixture" yields do not obtain is best indicated by Figure 16. The extremely sharp depression of the nitrogen yield corresponds to the sudden increase in hydrogen yield (see Figure 15), with a mixture in which (initially) less than 0.1% of the absorbed dose was absorbed by phosphine, and in which 90% of the phosphine was decomposed during radiolysis. The residual nitrogen yield appears to be decreasing proportionately to the energy fraction absorbed by phosphine. By extrapolating this curve back to the ordinate, a G-value for "molecular" nitrogen of 0.05 is obtained.
Pure Methyl Iodide Vapor
Methane and hydrogen yields from methyl iodide vapor, irradiated at a dose rate of 4.64 x 1017 ev per liter per minute, are plotted as a function of absorbed dose in Figures 17 and 18, respectively. Methyl iodide pressure was 26.8 0.1 cm at 23C. Beyond 1 x lu ev both products show rates of production proportional to dose. Hydrogen shows an initial rate of production greater than that observed after
lO20
1 x 10 ev, whereas methane production is depressed by a similar




14
0
x12
Co
010
C
-J
116
[ 4
2
I
2
I2 3 4
RADIATION DOSE (in ev x 1020)
Figure 17. Methane yields from methyl iodide at 26.8 cm pressure, as a function of
radiation dose




0
- 5
X
U)
0
E4
O0
c 3
Of
_Il
20
z
0 a: CI
II I I
I 2 3 4
RADIATION DOSE (in ev x 1020)
Figure 18. Hydrogen yields from methyl iodide at 26.8 cm pressure, as a function of
radiation dose
V1




46
amount over the same dose region (0 to 1 x 1020 ev). The initial G-value for methane production is 1.7. In the linear portion beyond
1 x 10 ev the 100 ev yield has increased to 2.3. These calculations are based on an electron stopping power ratio, SCH3 1 /S 2H4, of 3.185, as calculated by the Bethe equation. Hydrogen G-values were 1.h and
0.6 in the initial and linear portions of the curve.
Both product yields show a dependence on methyl iodide
pressure. Irradiations were carried out such that the absorbed dose was 1.1 x 1019 ev per cm CH 3I. The irradiation time was constant. The total yields are roughly proportional to pressure out to about 20 cm, beyond which the G-value for methane production shows a decided increase and that for hydrogen production falls off (see Figures 19 and 20). Because of the scatter in the hydrogen yields, it is difficult to determine the shape of the curve beyond 30 cm pressure. Measurements were taken out to the equilibrium vapor pressure of methyl iodide (38 cm) at room temperature.
Phosphine-Methyl Iodide Mixtures
These mixtures were made up to a total pressure of 27-30 cm at 23 C. Irradiation times were such that each mixture received a dose of 1 x 10 ev.
Methane and hydrogen yields from the radiolysis of these
mixtures are plotted as a function of the fraction of energy absorbed by phosphine (Figure 21a). The methane yield shows a sharp maximum close to 0.01, after which it falls off rapidly. The dotted line shows the way the yield should decrease if it was a function only of energy




-13 -o
2 16 14
0
E 2
10
J 8 >"6 2 4 12
,, I i l I .. i I. I I
4 8 12 16 20 24 28 32 36
PRESSURE OF METHYL IODIDE (in cm)
Figure 19. Methane yields as a function of methyl iodide pressure, irradiated to a total dose
of x l101 ev per cm CH3I
3--4




5
0
a)4 0
00
E
C
w
z
w
0
00
4 8 12 16 20 24 28 32 36
PRESSURE OF METHYL IODIDE (In cm)
Figure 20. Hydrogen yields as a function of methyl iodide pressure, irradiated to a total
dose of 1.1 x l019 ev per cm CH3I




(D
0
x 22
It)
76 20 --. E )
~H a 16
14 Z 12
w
0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.30 0.90 1.00
FRACTION OF ENERGY ABSORBED BY PHOSPHINE
Figure 21a. Yields of hydrogen and methane as a function of the fraction of energy absorbed
by phosphine
U0




50
absorbed by methyl iodide. Hydrogen yields increase linearly with the energy absorbed by phosphine, beyond a minimum phosphine content of about 30 mole percent. The hydrogen yield remains constant and the same as in pure methyl iodide vapor until just beyond the maximum methane yield. This is shown more clearly in Figure 21b, where the abscissa in this region has been greatly expanded.
Table 3 lists the composition and irradiation time for each mixture. Calculation of the maximum rate of methane production from Figure 21a was done by taking the yield of methane at the apparent maximum of the curve. G(CH4) at this point was 12.6. The mixture corresponding to this maximum was number three, containing 3.5 mole percent phosphine and in which 99% of the total radiation dose was absorbed by methyl iodide. The ratio between this maximum G and the G for methane in the pure vapor, at the same total dose, is 6.3. The minimum and maximum values for hydrogen G-values in the proportional region were 3.5 and 11.3. Hydrogen production in those mixtures which contained only a few mole percent phosphine was the same as in pure methyl iodide vapor, with a G-value of 1.0.




(1,
x 22
0
20
E
.E J8
1J
LCH
z 12 H4
LLJ
0
O 10
o~H2 06
~. 6
Lw 4
z
S2 A
Lw __ __ __ I
0.01 0.02 0.03 0.04
FRACTION OF ENERGY ABSORBED BY PHOSPHINE
Figure 2lb. Yields of hydrogen and methane as a function of the fraction of energy absorbed
by phosphine (expanded abscissa scale)




TABLE 3
Phosphine-Methyl Iodide Mixtures
Mixture P(PH3)a P(CH3 I)a Mole Fraction S Time
Number 33PH PHTm
Number (cm) (cm) PH3 S3 (hours)
SCHI31 + SPH3
0 0 26.7 0 0 3.73
1 0.19 26.7 0.007 0.002 3.73
2 0.50 26.9 0.018 0.006 3.57
3 0.97 26.9 0.035 0.012 3.53
4 2.10 26.9 0.072 0.023 3.50
5 2.66 26.7 0.091 0.032 3.50
6 10.16 19.6 0.341 0.145 4.22
7 19.60 9.6 0.671 0.399 6.02
8 24.06 5.1 0.844 0.605 7.45
9 27.56 1.9 0.936 0.826 8.88
*
10 55.00 0 1 1 4.85
aAll pressures are corrected to 23oC.
e
Taken from Figure 10.
N)T




IV. DISCUSSION AND INTERPRETATION Pure Phosphine and Pure Ammonia
From a comparison of the initial G-values of hydrogen which were obtained from the radiolysis of ammonia and phosphine, it is apparent that phosphine is considerably more susceptible to radiolytic breakdown than its nitrogen analogue. The most obvious explanation for this lies in a comparison of the available data on bond strengths in the two compounds. Cottrell23 lists the average bond energy terms, E, for ammonia and phosphine as 93.4 and about 77 kcal, respectively. The bond dissociation energy, NH2-H, is given as 102 kcal but PH 2-H is apparently not known. At any rate, it is safe to assume that dissociation processes such as
PH 3 PH + H" (23)
PH3 PH + 2H" (24)
are less endothermic than the corresponding steps with ammonia. Charged species such as PH4+, PH 3+, PH 2+ and PH+ are also probably less stable with respect to dissociation than the nitrogen species, and on neutralization should show a more pronounced tendency toward breakdown into (finally) hydrogen and phosphorus. The lack of dependence of the product yields on phosphine pressure indicates a probable lack of surface reactions or steps requiring collisional deactivation of an intermediate under the experimental conditions of this study.
53




54
If ion-molecule reactions figure prominantly in the radiolysis of phosphine, as they appear to in ammonia, it is reasonable to consider these processes as being analogous to those postulated for ammonia. Melton's work13 contains a detailed examination of these reactions in irradiated ammonia. Also, the electron impact study of phosphine by Halmann and Platzner5 has shown that the formation of phosphonium ion in phosphine occurs via an ion-molecule reaction. This is analogous to the formation of ammonium ion when ammonia is bombarded with electrons.
Reaction of hydrogen atoms must be predominantly by the abstraction reaction
H- + PH 3 PH2- + H2 (25)
rather than by recombination of atoms, by a simple kinetic argument. The rate of recombination of H atoms is given by
Rr = k H'][PH3]
and the rate of abstraction by reaction (25) is
Ra = [H.]1[PH3]
A value for kr of 10 liters2 per mole2 per sec (with NH3 as the third body) has been reported, but k does not appear to be known. It is
a
probably comparable to the rate constant for the reaction
H. + HBr H2 + Br- (26)
which is about 109 liters per mole per sec.25 The steady-state concentration of hydrogen atoms during radiolysis can be estimated as follows:




55
Rate H* generation = Rate H* removal G(-PH )
S_ x dose rate
100
(7.5 x lo-2 atoms/ev)(5 x 1015 ev/liter-sec)
4 x o101 atoms/liter-sec 6 x l-10 moles/liter-sec If reaction (25) is the step by which most H atoms react, then
Rate H- removal = k[ PH][ H-]
= 109 liters/mole-sec [10-2][ H] = 107 [H*]/sec Solving for the concentration of hydrogen atoms,
[H] = 10-16 moles/liter
Finally,
R k
a a
R k H.]
= 1015
It follows that the recombination of hydrogen atoms is insignificant in this system.
The phosphino radical may abstract hydrogen from phosphine, which is not a chemically detectable reaction. Recombination of the radicals
PH2. + PH* PH + PH (11)




56
is probably exothermic, since the PH2-H bond strength is probably 23
greater than that of PH-H., by analogy with ammonia.
Elemental phosphorus formation from recombination of PH 10
species is supported by the data from the flash photolysis, which showed that the formation of P2 was too rapid (15 Psec or less) to be accounted for by a three-body collision. The P2 must combine to form P 4.
The G(-PH3) of 7.5 suggests that a mechanism involving no chemical amplification of the primary ionization and excitation processes may be appropriate. For most gases the ion-pair yield, W, is approximately 30 ev, so that G for positive ions can be taken as about 3. Assuming that one to three excited molecules are also formed for each ion, the primary yield (ions and excited molecules) should be of the order of 6-12. The following reaction steps summarize a possible mechanism which takes account of these considerations:
PH 3 PH2. + H. (27)
PH3 PH + H orr 211" (28)
H- + PH 3 H2 + PH2" (25)
P12' + PH" PH3 + PH (11)
PH + PH P2 + H2 (29)
P2 + P2 P4 (30)
The mechanism suggested by Norrish and OldershawI0 for the decomposition of phosphine by flash photolysis is very similar to the one given here, although it was necessary for them to propose two




57
reaction schemes. One scheme was used to explain the results at room temperature, the other at higher temperatures. The formation of hydrogen was considered to occur by recombination of atoms at room temperature, with hydrogen abstraction from phosphine by hydrogen atoms occuring at higher temperatures. The necessity for the two schemes arose from the fact that the quantum yield was one-half at room temperature, but increased to one at high temperatures. In view of the previous calculation of relative rates for the recombination and abstraction reactions in this system, it seems improbable that recombination of hydrogen atoms is important, even during flash photolysis, where the concentration of radicals would be much larger than during gamma radiolysis.
Some diphosphine may be formed by the reaction
PH2* + PH 2 PH2PH2 (31)
Diphosphine, like hydrazine, would be susceptible to attack by various radical species, notably hydrogen atoms. If significant amounts of diphosphine form during the radiolysis, an experiment such as reported by Jones, Sworski and Williams17 would be informative. They found that the recovery of hydrazine from radiolysis of ammonia in a flow system depended directly on the flow rate of ammonia. At high flow rates, most hydrogen atoms are scavenged by ammonia and most amine radicals combine to form hydrazine. The present study cannot discount the production of diphosphine and its subsequent removal by radical attack in a static system. It is notable, however, that whereas hydrazine thermally decomposes to nitrogen and hydrogen, diphosphine breaks down at room temperature according to the following equation:




58
(3x y)P2H4 (4x 2y)PH3 + 2PX H (32)
26
where x > y. The mass balance results of this study have shown that no appreciable amounts of lower phosphorus hydrides are formed during or after the radiolysis. If the decomposition of diphosphine by radical attack yields the same products as the thermal reaction, and if significant quantities are formed during the radiolysis, the mass balance would show a hydrogen deficiency due to the presence of quantities of hydrides in the solid residue. Within the experimental error of the mass balance experiment, which was less than 1%, no such deficiency was observed; i.e., hydrogen produced was equal to three-halves phosphine decomposed.
Although this study and two others reported very recently 16,18 show quite close agreement for the G-value of hydrogen from the radiolysis of ammonia, earlier investigations have shown wide variation in the hydrogen yield. An examination of the experimental conditions in each study shows that the discrepancies are probably explainable in terms of 27
these conditions. In one instance, an "uncorrected" G-value has been given which was calculated using the Fricke liquid dosimeter. Recent workers, using gas phase dosimeters (and electron stopping powers rather than electron densities for dose rate calculations), are in good agreement.
Phosphine-Amonia Gas Mixtures
The most recent work done on radiolysis of ammonia in the gas
Melton (See Table 3, Reference 13) provides a summary of these G-values as obtained by various workers.




59
phase has been that of Jones and Sworski, who obtained a G(H 2) of 4.5 at room temperature, but observed a dependence on ammonia pressure below 40 cm in a continuation of their work.17 Nishikawa, Shinohara and
18
Matsuura, irradiating ammonia at 50 cm pressure, found the hydrogen yield to be 4.3. Finally, this study at 55 cm gives G(H 2) = 4.5, with no dependence on ammonia pressure from 100 torr up to one atmosphere, with the curve extrapolating to zero yield at zero pressure. The consistency of these results with pure ammonia appears to make a further comparison valid; namely, the results obtained by these workers with various radical scavengers, and the present study using phosphine.
The earlier workers used ethylene as a scavenger and obtained minimum hydrogen yields of 0.7-0.8, with ethylene concentrations of about 2 mole percent. Carbon tetrachloride decreased the yield slightly, to 0.64. These results indicate a yield of "molecular" hydrogen of about
0.75. In the present study the "molecular" nitrogen yield was measured directly and found to be 0.05. It is evident that phosphine is a very efficient scavenger of amino radicals, by the reaction
NH 2 + PH3 NH3 + PH 2 (33)
which is exothermic of NH 2-H > PH 2-H. It is probable that this reaction has a AH 2 -15 kcal. By reaction (33) and
H. + PH3 H2 + PH 2 (25)
which is probably exothermic by around 20 kcal, both hydrogen atoms and amino radicals could be efficiently scavenged. The fact that phosphine may react readily with both amino and nitrene (NH) radicals formed in




6o
the primary dissociations may account for its effectiveness in depressing the nitrogen yield.
Despite extensive investigations of both the photochemistry and radiation chemistry of ammonia, the mechanism for formation of nitrogen is not well established. The following reactions are often given as a possible route to N2:
NH2. + NH2. N2H4 (34)
N2H 4 + H- H2 + N2H3- (35)
N2H3' + N22H3' N2H4 + N2H2 (36)
N2H2 N2 + H2 (37)
It has also been suggested13 that the diimine may be formed by the elimination of hydrogen from excited hydrazine, as below:
N2H 4 N2H2 + H2 (38)
followed by the unimolecular decomposition of N2H2, as in reaction (37) above.
The formation of nitrogen by reaction steps not involving scavengeable species; i.e., which result in production of "molecular" nitrogen, can be rationalized by postulating the existence of a nitrene, NH, which undergoes an insertion into ammonia, as in the following reaction:
NH + NH N2H (39)
3 2 4
The excited hydrazine would break down to nitrogen and hydrogen via reactions (38) and (37). The formation of a significant amount of NH




161
in irradiated ammonia has been demonstrated by Meaburn and Gordon.15 They measured the absorption spectrum of the radical due to the A311 X3E transition at 3360A, after pulsing with 250 kev electrons. Because of the experimental technique used, no information on NH in the singlet state was obtained. Provided that some small fraction of the nitrenes could exist in a singlet state, such that one of the lower energy orbitals is unoccupied, it is possible to propose a transition
3
state in which bonding occurs through this orbital and the filled sp of ammonia. The only rearrangement then required is a protonic shift. The analogy with carbene insertion reactions on methane makes reaction (39) worthy of speculation as a route to formation of "nonscavengeable" nitrogen, since the singlet nitrene would not be reactive to phosphine except through a Lewis acid-base reaction. The much greater Lewis basisity of ammonia would not allow phosphine to compete with ammonia for the nitrene except in those mixtures containing relatively large concentrations of phosphine. In these mixtures, the production of nitrogen falls linearly toward zero, in pure phosphine.
The charge transfer reaction
PH+ NH+ PH 3+ + NH3 (ho)
may well occur in this system under certain conditions, since the ionization potentials of phosphine and ammonia are virtually identical
2
(10.2 ev). But unless the concentration of phosphine becomes comparable to ammonia this cannot be a significant reaction, in view of the very high efficiency of the reaction
NH 3+ + NH3 NH 4+ + NH2- (lh)




62
for which the rate constant has been estimated to be 3 x 10I moles per
28
liter per sec. It could be that, in mixtures containing fairly large concentrations of phosphine, reaction (40), or one such as
Nil 3+ + PH3 NH4+ +PH2. (41)
which may be as efficient as reaction (14), becomes important.
It is interesting to compare the product yields in this
system, at a concentration of PH3 just sufficient to depress the nitrogen yield to its limiting low value, with the yields in pure ammonia. Mixture number 6 (see Table 2) provides data in this region. It is possible to use these data and those from the irradiation of pure ammonia as the basis for a calculation of phosphine decomposed relative to ammonia protected. The data follow:
G(N 2) (Mixture) = 0.06 G(H 2) (Mixture) = 6.4
G(N2) (Pure NH 3 1.52 G(H 2) (Pure NH3) = 4.5
Since in the mixture G(H2) (from NH3 decomposition) =
(55/53) 3 [G(N2)(Mixture)] = 0.2, then G(H2) (from PH3 decomposition) 6.2. The factor (55/53) takes into account the difference in ammonia pressure in the two systems. It follows that G(-PH3), which is 2/3 of [G(H2)(from PH3)], is 4.1. The G(-PH3) must be corrected for direct decomposition by radiation, since in this mixture approximately 6% of the dose is absorbed by phosphine. This correction is made by assuming that the rate of direct gamma ray decomposition of PH3 in the mixture is the same as in irradiation of the pure substance: i.e., G(-PH3) =
7.5. This amounts to 11% of the total PH3 decomposed, and gives a G(-PH 3) from reaction with ammonia moieties of 3.7. What is desired




63
is a comparison of this yield with the yield of protected ammonia, which is the ammonia not decomposing in the mixture, due to the presence of phosphine.
Defining AG(N 2) as the difference in nitrogen yield in pure ammonia and the mixture, AG(N2) = 1.52 (55/53)(0.06) = 1.46. This represents the nitrogen which was not formed, due to the presence of PH3. The yield of protected ammonia is related by a factor of two to AG(N2); i.e., G(NH3) (protected) = 2 [AG(N2)] = 2.9.
From these results it is clear that, in a mixture where the
phosphine concentration is just large enough to be efficiently scavenging radical fragments from ammonia, less than four PH3 molecules are decomposed in protecting three ammonia molecules. This nearly one-to-one stoichiometry can be explained by examining the initial stages of ammonia decomposition without and with added phosphine. In pure irradiated ammonia, the following reactions occur:
NH3 -#+ NH2" + H" (42)
H- + NH3 H2 + NH2' (43)
H- + + 3 4 H2 + NH3 (44)
The rate constants for reactions (43) and (44) are known to be about
5 liters per mole per sec29 and about 8 x 109 liters2 per mole2 per sec.e2 In pure irradiated ammonia, however, reaction (43) should predominate over reaction (44), since the hydrogen atom concentration is so low (about 10-10 moles per liter) during radiolysis.
With added phosphine, reaction (43) would be unable to compete with the reaction




64
H" + PH H2 + PH2* (25)
for which the rate constant has not been reported. Because of the much weaker PH2-H bond, however, it can be estimated that the rate constant of this reaction is about 108 greater than that of reaction (43). Thus, via reactions (25) and
NH2" + PH3 NH3 + PH2" (33)
two ammonias are protected, one directly and one by re-formation, and two phosphines are destroyed. The excess of phosphine (3.8 PH3: 3.0 NH 3) may be an indication of the importance of reaction (44) in pure ammonia, since reaction (44) does not involve an ammonia molecule except as a third body, and suppression of this reaction by reaction (25) would not protect NH This argument is based on the assumption that phosphine is not reacting significantly with ionic species.
Methyl Iodide Vapor and Phosphine-Methyl Iodide Gas Mixtures
This part of the present study had as its major purpose the
verification of the radical scavenging ability of phosphine, rather than a detailed study of the radiolytic decomposition of CH 3I. For this reason, analyses of products were confined to those noncondensible at -196C, namely, methane and hydrogen. The yields of these products were determined in the pure vapor and in mixtures containing phosphine. Because of the limited scope of this investigation of the pure vapor radiolysis, it is not possible to formulate a decomposition mechanism. In addition, the radiolysis of methyl iodide in the vapor phase does not appear to have been reported, although photolytic studies have




65
been made by a number of workers. 21303 It has been shown 21that the
gas phase photolysis produces methane according to reaction (20) (page 8) involving a hot methyl radical, since the yield is unaffected by the presence of efficient thermal methyl radical scavengers and is also temperature independent.
It is likely that hot methyls are a major source of methane in the gamma radiolysis of methyl iodide, although it may be that OH 4 is also formed by other routes, such as via an ion-molecule reaction. The dependence of the yield on methyl iodide pressure (see Figure 19) is unlikely to be due to a wall effect. Approximate calculations of the number of collisions which a hot methyl undergoes with the substrate OH I and with the wall indicate that over the pressure range involved there is not a significant change in the number of wall collisions relative to substrate collisions,, especially in view of the high efficiency of reaction (20). If an ion-molecule reaction is a second source of CH 41 the pressure dependence of this reaction might explain the observed increase in G(CH 4) as the CH3I1 pressure is increased. Some of these reactions have been reported by Pottie, Barker and Hamnill. 3
A significant amount of hydrogen is formed. It is unlikely that reactions of thermal radicals lead to H production in this system, since-radical attack would occur more readily at the C-I bond. The present data is insufficient to suggest a mechanism, although the pressure dependence (see Figure 20) indicates that one possibility is molecular elimination from an excited molecule. Increasing pressure would cause more rapid deactivation and a reduced yield, as observed. The reaction of hot hydrogen atoms with methyl iodide could also lead




66
to H Finally, the work of Pottie, Barker and Hamill32 has shown that in the mass spectrometer some H2 may be formed by the reaction
CH 31 + e- Cl+ + H2 + H. + 2e- (45)
Since it is probable that the major primary processes in the radiolysis of methyl iodide vapor result in rupture of the C-I bond and formation of methyl radicals, the addition of even low concentrations of phosphine should result in an increased yield of methane, by the reaction
CH3* + PH3 4 CH4 + FH2' (46)
which is probably about 20 kcal exothermic. Phosphine would not be a good scavenger of iodine atoms, because the H-I bond is somewhat weaker than the PH 2-H bond.
The very sharp rise to a maximum yield of CH4 indicates
the efficiency with which phosphine is reacting with thermal methyls. Since virtually all (99%) of the energy is being absorbed by methyl iodide at the maximum of the curve (see Figures 21a and 21b), and the concentration of phosphine is only 3.4 mole percent, it is reasonable to suggest that the G for methane at this maximum is approximately the G for methyl radical production in the pure irradiated vapor. On this basis, a comparison of methane yields in the pure vapor and in the phosphine-methyl iodide system shows that about 85% of methyl radicals produced in the pure vapor during radiolysis do not react to produce methane.
If hydrogen is not formed by a process involving thermal radicals, the yield should not be affected by low concentrations of




phosphine. Examination of Figure 21b shows that, out to 7 or 8 mole percent phosphine, hydrogen production is constant and equal to that in the pure vapor. At higher phosphine concentrations the hydrogen yield is due largely to the decomposition of phosphine by direct radiation effect, as indicated by the linearity of the yield with respect to energy absorption by phosphine. The same linearity is evident in the ammonia-phosphine system when hydrogen yields are plotted as a function of the energy fraction absorbed by phosphine (see Figure 15).
The rapid drop-off in methane production immediately beyond the maximum coincides with a large increase in hydrogen yield. This may be due to an energy transfer process from methyl iodide to phosphine. It is probably not due to charge transfer, since the reaction
CH3I+ + PH3 CH3I + PH3+ (47)
is endothermic by 10-15 kcal, based on the ionization energies of
33
methyl iodide and phosphine. From 30 mole percent phosphine out to pure phosphine, the methane yield decreases approximately in proportion to the energy fraction absorbed by phosphine. Inspection of Figure 21a shows that this is the same region in which the hydrogen yields show a proportionate increase with energy absorbed by phosphine.
The data of Figure 21a can be thought of as resulting from the interaction of at least three processes: (1) efficient scavenging of methyl radicals by phosphine; (2) an energy transfer process from methyl iodide to phosphine; and (3) direct radiation effects on both phosphine and methyl iodide. The scavenging process produces a large increase in methane yield at very low phosphine concentration. At slightly larger concentrations of PH3, the energy transfer process




68
causes the CH4 yield to drop rapidly and the H2 yield to show a sharp increase. As the phosphine concentration becomes relatively large, both yields become functions only of the energy absorbed by phosphine. The energy transfer process apparently becomes complete at fairly low phosphine concentration, so that it exerts approximately a constant effect at higher concentrations; i.e., the system is saturated with respect to energy transfer at approximately 30 mole percent phosphine.




V. SUMMARY AND CONCLUSIONS
The investigations reported here may be summarized, along with the major conclusions drawn from them, as follows:
1. As expected, the decomposition of phosphine is more extensive than that of ammonia, when the two compounds are irradiated under
comparable conditions. Some similarities which are notable are
that both compounds break down to their elements, and neither
forms a dimeric X2H4 compound (where X is nitrogen or phosphorus)
to a significant extent in a static system. Under the experimental
conditions of this study, neither appeaitto have rate-determining third-body reaction steps. Although other workers have shown that
in a flow system ammonia radiolysis produces considerable quantities
of hydrazine, it is less likely that diphosphine would be formed
under comparable conditions in phosphine radiolysis. A logical
extension of the work reported here would include the flow photolysis or flow radiolysis of phosphine.
2. The irradiation of phosphine-ammonia mixtures at room temperature does not produce significant amounts of any nitrogen-phosphorus
compounds. The only products which are observed are hydrogen,
phosphorus and nitrogen. Drastic depression of the nitrogen yield,
even in mixtures containing very low concentrations of phosphine, leads to the conclusion that phosphine is protecting ammonia from
radiolytic decomposition by scavenging free radicals, through
69




70
donation of a hydrogen atom. In a mixture containing less than
4 mole percent phosphine, the decomposition of less than four molecules of phosphine resulted in the protection of three molecules of ammonia (per 100 ev of absorbed energy). This stoichiometry is explained with a simple reaction scheme involving the scavenging of both hydrogen atoms and amino radicals. It is reasonable to suggest that phosphine can act as a good free radical scavenger via reactions involving donation of a hydrogen atom. Although the PH 2-H dissociation energy is not known, it is probably about 20 kcal less than the CH 3-H or NH2-H bond energies. In general, phosphine should be reactive to many carbon and nitrogen-containing radicals generated in radiolytic systems, as well as to a number of atomic radicals such as H. and Cl. The only requirement would seem to be that the scavenging reaction should be exothermic.
3. The radiolysis of methyl iodide vapor results in the formation of methane and hydrogen, with G-values of 2.3 and 0.6, respectively. It is probable that methane formation in the irradiated vapor parallels that in the photolyzed material; i.e., it proceeds via the reaction of hot methyl radicals with the substrate. It may be that methane is also formed by other routes, including ion-molecule reactions. The mode of hydrogen formation is speculative but may involve a molecular elimination or hot atom process. In the presence of phosphine, the methane yield is increased by a factor of six, due to scavenging of methyl radicals. This indicates that the primary methyl yield in irradiated methyl iodide is probably 12-13 (per 100 ev absorbed energy), since it is probable that most methane is formed by processes involving methyl radicals.




APPENDIX
Calculation of the Dosimetry
In gas phase systems radiation dosimetry is more complex
than in condensed systems. Absorption by matter of gamma radiation in the mev range occurs principally by three processes: Compton scattering, photoelectron ejection and pair production. With systems containing atoms of moderately low atomic weight the most important process is Compton scattering, which involves the interaction of the gamma photon with an orbital electron. A lower energy photon and an ejected electron result. Since the electron density of condensed phases is usually comparable to that of the irradiation vessel, with most radiation geometries the bulk of the energy absorbed is via Compton scattering in the sample. It follows then that the absorbed dose as determined by a liquid dosimeter solution is related to the absorbed dose in the sample by the ratio of electron densities in the two systems.
In most gaseous systems the absorption of energy from gamma rays by the medium is insignificant relative to that absorbed from electrons ejected from the cell walls and traversing the cell. For these electrons the energy absorption by the gas is proportional to the electron stopping power of the gas, rather than its electron density, and the absorbed dose in the sample and dosimeter are related by the ratio of their stopping powers. The stopping power of
71




72
a medium is defined as the energy loss per unit path length of a particle passing through the medium.
The use of stopping powers for the calculation of dose rates is justified if the irradiation cell and its contents constitute a Bragg-Gray cavity. 34,35,36 Spiers 37has outlined the basic conditions required, as follows:
1. The cavity must have dimensions such that only a very small portion of the electrons ejected from the walls have ranges less
than the distance they must traverse through the cell medium.
2. A negligible amount of energy absorption must occur from the
radiation interacting directly with the gas.
3. The cavity must be surrounded by a sufficient thickness of
solid material so that all of the electrons traversing the cavity
originate in the solid material. Although theoretically this means that the thickness must be equal to or greater than the
range of the electrons in the solid, in practice a considerably
thinner wall is sufficient'. due to scattering in the solid.
4. The energy dissipation in the solid medium must be reasonably uniform all around the cavity. Ordinarily, this implies that the cavity should be placed far enough from the radiation source that
the divergence of the field is slight.
The cell geometry used in most of the irradiations (see
Figure 14) can be shown to satisfy the first three requirements. The average energy of Compton electrons produced by cobalt-60 gamma rays has been shown to be 0.587 mev. 38 According to Nelms, 39 the range of 0.587 mev electrons in ethylene is 0.20'7 grams per cm 2(by an interpolation of the values given for 0.55 and 0.60 mev electrons). At the




73
ethylene pressure used in the dosimetry (60 cm), the range was then about 2.3 meters. The cell diameter was 15 cm.
The mass densities of glass and ethylene at 60 cm pressure are about 2.2 x 103 milligrams per cc and 0.9 milligrams per cc, respectively. The efficiency with which these two materials attenuate a beam of gamma radiation can be approximately compared by taking the product of their mass densities and the path length of the radiation, since the molecular electron densities of glass and ethylene are roughly comparable. The path length in the glass is 4 mm and in the ethylene about 15 cm maximum. Thus the relative absorption efficiencies are at least 75 and one in this system, satisfying the second requirement. A glass wall thickness of 1 mm or more ensures electronic equilibrium since the range of the Compton electrons generated in glass by cobalt-60 gammas is roughly 1 mm. The walls of the cells used in this work were 2 mm thick.
The fourth requirement is a good deal more difficult to
analyze with the cell geometry used. For the determination of comparative rather than absolute dose rates, however, this should not be important.
Bethe 4o.l derived an expression for the calculation of stopping power, shown in the following equation:




74
dE 27Ne4Z mu 2 E
- = 2 n (2 2 1 + a2 in 2
dx m 212 0+ (1-)2 + 1/8 ergs per cm ()
where N is the number of molecules per cubic centimeter, e is the charge on an electron in electrostatic units, m is the rest mass of
0
an electron in grams, u is the velocity of the electron in cm per second, Z is the number of electrons per molecule, 8 is U/c (where c is the velocity of light), I is the mean excitation potential for the molecules of the stopping material in ergs and E is the kinetic energy of the electron in ergs. Insofar as dosimetry calculations are concerned the stopping power ratio of the sample gas and the dosimeter gas, rather than the absolute values of the stopping powers themselves, are needed. This facilitates the calculations, as shown below:
dE
dx sample N Z k-ln I 2
___s s s(2
dE NdZd k-ln Id22
dx dosimeter
mu E2 2
where k = in 0 (2 1 + ) in 2 + 1 -a
2(1-2
+ 1/8 (1 a2)
A value for 0 of 0.8765 for a 0.587 mev electron was found by interpolation of values provided by Johns and Laughlin. The mean excitation
potentials for the atoms C and H were taken directly from Spiers 37




75
and a value for P was found by a linear interpolation of the values for Al and Cl. An average excitation potential for compounds can be calculated, assuming simple additivity, according to the relation: ZN.Z. in I.
in I = (3)
IN.Z.
i11
where I is the potential for the compound and the i subscripts indi42
cate values for the atoms. Calculated in this way, I values for ethylene and phosphine were found to be 51 and 112 ev, respectively. Substitution of these values (in the proper units) into equation (2) and solving leads to the relation:
SPH3 PPH3
SC2H4 PC2H4
where pressures have been substituted for number densities. Similar calculations lead to the following analogous equations:
NH3 = 0.661 __NH_3(5) SC2H PC211
SC2 H4 PC2 H4
SCH3I PCH31
= 3.185 (6)
SC2H4 PC2H4
Equations (4), (5) and (6) were used in the calculation of dose rates.
The use of ethylene as a dosimeter gas has been recommended




76
43
by Yang and Gant, who showed that the hydrogen yield from radiolysis was highly reproducible without extensive purification of the ethylene. Figure 9 (page 28) shows that the rate of hydrogen production is constant with radiolysis time. The slope of the graph was 0.419 Pmoles per hour. Assuming that the G-value of hydrogen is 1.2 molecules per 100 ev, which appears to be the best value,44'5 the absorbed dose rate was 3.50 x 1017 ev per minute in the cell. The volume of the cell was calibrated with distilled water as 1,027 5 ml and the ethylene pressure was 60 cm. The dose rate was thus found to be 5.68 x l014 ev per mm of ethylene pressure per liter of cell volume per minute. Although several cells were used in subsequent work, their geometries and volumes were very similar. For example, the three cells used for the ammonia-phosphine mixtures had volumes of 1,035, 1,027 and 1,020 ml. The error involved in using these different cells, insofar as dose rate differences were concerned, was probably negligible.




BIBLIOGRAPHY
1. P. A. Sellers, T. R. Sato and H. H. Strain, J. Inorg. and Nuc.
Chem 5, 31 (1957).
2. Y. Wada and R. W. Kiser, Inorg. Chem. 3, 174 (1964).
3. H. Neuert and H. Clasen, Z. Naturforsch. 7a, 410 (1952). 4. P. E. Saalfeld and H. J. Svec, Inorg. Chem. 2, 50 (1963).
5. M. Halmann and I. Platzner, J. Phys. Chem. 71, 4522 (1967).
6. H. W. Melville, Proc. Roy. Soc. A138, 374 (1932). 7. H. W. Melville, Proc. Roy. Soc. A139, 541 (1933).
8. H. W. Melville and J. L. Bolland, Proc. Roy. Soc. A160, 384 (1936).
9. H. W. Melville, J. L. Bolland and H. L. Roxburgh, Proc. Roy. Soc.
Al60, 406 (1936).
10. R. G. W. Norrish and G. A. Oldershaw, Proc. Roy. Soc. A262,
1 (1961).
11. D. A. Ramsay, Nature 178, 374 (1956). 12. J. R. McNesby, I. Tanaka and H. Okabe, J. Chem. Phys. 36, 605
(1962).
13. C. E. Melton, J. Chem. Phys. 45, 4414 (1966). 14. Y. Toi, D. B. Peterson and M. Burton, Rad. Res. 17, 399 (1962). 15. G. M. Meaburn and S. Gordon, J. Phys. Chem. 72, 1592 (1968). 16. F. T. Jones and T. J. Sworski, Trans. Farad. Soc. 63, 2411 (1967). 17. F. T. Jones, T. J. Sworski and J. M. Williams, Trans. Farad. Soc.
63, 2426 (1967).
18. M. Nishikawa, N. Shinohara and N. Matsuura, Bull. Chem. Soc. Jap.
40, 1993 (1967).
19. H. A. Gillis, R. R. Williams and W. H. Hamill, J. Am. Chem. Soc.
83, 17 (1961).
77




78
20. R. C. Petry and R. H. Schuler, J. Am. Chem. Soc. 75, 3796 (1953). 21. R.C. Souffie, R. R. Williams and W. H. Haill, J. Am. Chem. Soc.
78, 917 (1956).
22. R. J. Hanrahan, Intern. J. Appl. Radiation Isotopes 13, 234 (1962). 23. T. L. Cottrell, The Strengths of Chemical Bonds, 2nd ed., Academic
Press, Inc., New York (1958).
24. C. Kretschmer and H. L. Peterson, J. Chem. Phys. 39, 1772 (1963). 25. W. Jost, Z. Phys. Chem. B3, 95 (1929). 26. J. R. Van Wazer, Phosphorus and Its Comnounds, Vol. 1, Interscience Publishers, Inc., New York (1958).
27. R. C. Schiek, Ph.D. Dissertation, Univ. of New Hampshire (1963. 28. L. M. Dorfman and P. C. Noble, J. Phys. Chem. 63, 980 (1959). 29. M. Schiavello and G. Volpi, J. Chem. Phys. 37, 1510 (1962). 30. R. D. Doepker and P. Ausloos, J. Chem. Phys. 41, 1865 (1964). 31. R. D. Schultz and H. A. Taylor, J. Chem. Phys. 18, 194 (1950). 32. R. Pottie, R. Barker and W. H. Hamill, Rad. Res. 10, 644 (1959). 33. R. W. Kiser, Introduction of Mass Spectrometry, Prentice-Hall,
Inc., New Jersey (1965).
34. W. H. Bragg, Studies in Radioactivity, Macmillan, New York (1912). 35. L. H. Gray, Proc. Roy. Soc. A156, 578 (1936). 36. L. H. Gray, Brit. J. Radiol. 10, 600, 721 (1937). 37. F. W. Spiers in Radiation Dosimetry, eds. G. J. Hine and
G. L. Brownell, Academic Press, Inc., New York (1956),
Chapter 1.
38. R. J. Hanrahan, Ph.D. Dissertation, Univ. of Wisconsin (1957). 39. A. T. Nelms, Nat. Bur. St. (U.S.) Circ. 577 (1956). 40. H. A. Bethe, Handbuch der Physik 24, 273 (1933). 41. H. A. Bethe and J. Ashkin in Experimental Nuclear Physics, ed.
E. Segre; Vol. 1, John Wiley and Sons, New York (1953).




79
42. H. E. Johns and J. S. Laughlin in Radiation Dosimetry, eds.
G. J. Hine and G. L. Brownell, Academic Press, Inc., New
York (1956), Chapter 1.
43. K. Yang and P. L. Gant, J. Phys. Chem. 65, 1861 (1961). 44. M. C. Sauer, Jr., and L. M. Dorfman, Abstracts, 137th Meeting
Am. Chem. Soc., April 1, 1960.
45. F. W. Lampe, Rad. Res. 10, 691 (1959).




BIOGRAPHICAL SKETCH
James Wesley Buchanan was born in Rutherfordton County,
North Carolina, on May 5, 1937, and received his early education in several western North Carolina towns. He graduated from Claremont Central High School of Hickory, North Carolina, in 1955.
He studied at the University of North Carolina in Chapel Hill, where he was a Morehead Scholar, and received the Bachelor of Arts degree in chemistry in 1959.
His first graduate degree was awarded in 1962, when he received the Master of Science in chemistry. After several years of teaching chemistry at the undergraduate level, he returned to the University of Florida to complete his Ph.D. He has held graduate teaching and research assistantships, and during the last two years of study was an NDEA Title IV Fellow.
He is a member of the American Chemical Society and Alpha Chi Sigma chemical fraternity.
8o




This dissertation was prepared Under the direction of the chairman of the candidate's supervisory committee and has been approved by all members of that committee. It was submitted to the Dean of the College of Arts and Sciences and to the Graduate Council, and was approved as partial fulfillment of the requirements for the degree of Doctor of Philosophy.
August, 1968
Dean, College of Arts and Sciences
Dean, Graduate School Supervisory Committee:
Chafrm )
J<




Full Text

PAGE 1

RADIATION CHEMISTRY OF SYSTEMS CONTAINING PHOSPHINE By JAMES WESLEY BUCHANAN A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1968

PAGE 2

ACKNOWLEDGEM E NTS Deep appreciation is expressed to Dr. R. J. Hanrahans whose guidance was essential in the prosecution of this research. He has bee n a good friend, as well as an excellent research director for many years, during which time the author took an M .S. degree and taught for several years before returning to work toward the Ph.D. in 1966. Acknowledgem e nt is also made to Gwen, who has been a good wife for many years, and who heroically typed throughout many nights to complete this manu sc ript ii

PAGE 3

TABLE OF CONTE N TS Page ACKNO~lLF~DGD4ENTS. . . . ii LIST OF TABLES. . . . . . . . . . . i V LIST OF FIGURES. . . . . . . .. . . . . V Section I. II. III. IV. v. INTRODUCTION ............................................ Review of Earlier Work Status of Present Work EXPERir.IBNTAL METHODS Apparatus ............ Reagents ............. Preparation of Samples. Product Analysis EXPERir1ENTAL RESULTS Dosimetry Pure Phosphine Pure Ammonia Phosphine-Ammonia Mixtures. Pure Methyl Iodide Vapor Phosphine-Methyl Iodide Mixtures DISCUSSION A.ND INTERPRETATION Pure Phosphine and Pure Ammonia Phosphine-Amr:ionia Gas Mixtures Methyl Iodide Vapor and Phosphine-Methyl Iodide Gas Mixtures ... .................. ................... SUMMARY .AND CONCLUSIONS l 1 8 9 9 20 21 23 27 27 27 32 35 43 46 53 53 58 64 69 APPE1ilDIX 71 BIBLIOGRAPHY. . . . . . 77 BIOGRAPHICAL SKETCH. . . . . . 80 iii

PAGE 4

Table 1. 2. 3. LIST OF TABLES Mass Balance Results ...................................... Phosphine-Ammonia Mixtures Phosphine-Methyl Iodide Mixtures iv Page 33 41 52

PAGE 5

LIST OF FIGUR ES Figure Page 1. High vacull.In n 1 anifold .. ., . . . . . . . . 10 2. Toepler pump-McLe od gauge co mb in at io n 11 3. SaJD.ple analys is system..................................... 12 4. Radiolysis vessel. c; ........................................ 14 5. Radiolysis vessel. . . . . . . . . . . 15 6. Cross section of cob alt-6 0 g a mma ray source................ 17 7. Ga s chro mat o graph sampling module.......................... 18 8. G a s chro matograph ic sampling loop.......................... 19 9. Hydrogen yields from ethylene at 60 c m pressure, as a function of irradiation time. 28 10. Hydrogen yields from phosphine at 55 cm pressure, as a functio n of rad ia t i on dose.............................. 30 11. Hydrogen yields as a function of phosphine pressure, irradiated to a total dose of 4.4 x 101 8 ev per cm PH 3 31 12. Hydrogen yields fro m ammonia at 55 cm pressure, as a function of radiation dose.. . 3l i 13. Ni tro gen yields fr om ammonia at 55 c m pressure, as a function of radiation dose.............................. 36 14. Nitrogen yields as a function of ammonia pressure, irradiated to a total dose of 4.4 x 10 18 ev per cm NH 3 37 15. G-valu es of hydrogen from a'ilmonia-phosphine mixtures as a function of the fr ac tion of energy absorbe d by phosphine. . . . . . . . . . . . 39 16. G-values of nitrog e n from ammonia-phosphine mixtures as a function of t he fraction o f energy absorbed by phos phine. . . . . . . . . . . . 40 V

PAGE 6

17. Methane yields fro m methyl iodide at 26.8 c m pressure as a function of radiation dose.......... 44 18. Hydrogen yiel ds from meth yl iodide at 26 .8 c m pr essure as a function of r adiati on dos e ......................... 45 19. Methane yiel ds as a fu n c t io n of methyl io dide pressure, irradiate d to a total dose of 1.1 x 10 19 ev per c m CH 3 I 47 20. Hydro gen yields as a function of methyl iodide pressure, irradiated to a total do se of 1.1 x 10 19 ev per cm CH 3 r 48 21a. Yields of hydro g en and methane as a fu n ction of the fraction of energy absorbed by phosphine................ 49 21b, Yields of hydrogen and methane as a function of the fraction of energy absorbed by phosphine (ex panded abscissa scale)................................. ... ....... 51 vi

PAGE 7

I. INTRODUCTION The object of this study was to carry out an investigation of the gamma radiolysis of pure phosphine and to compare its decomposi tion with that of its nitro gen analogue, amm onia. The irradiation of gas mixtures of ammonia and phosphine was also done. Results obtained from radiolysis of these mixtures indicated that phosphine was acting as an efficient radical scavenger. Confirmation of this was sou g ht from radiolysis of mixtures of methyl iodide vapor and phosphine. Review of Earlier Work Radiolysis and Photolysis of Phosphine The only report of a study involving the irradiation of phosphine is that of Sellers, Sato and Strain, 1 who irradiated a large number of phosphorus-containing compounds with neutrons plus gamma rays as Yell as with intense gamma radiation. Their work with phosphine was not quantitative, but they did observe that with both the mixed radia tion and pure gamma rays the only detectable products were hydro gen and an orange-red deposit on the walls of the quartz t~be in which the irradiations Yere carried out. Also, the irradiation of yellow phosphorus resulted in conversion to the red form, which was stable to neutrons and gamma rays even in the presence of hydrogen gas. This work was a very minor part of their study, which concentrat e d on the 1

PAGE 8

radiolysis of crystalline salts of phosphoric, phosphorus and hypo phosphorous acids. A mass spectro metri c study of phosphine and diphosphine by Wada and Kiser 2 indicates that bombardment of phosphine with 70 ev electrons probably results in the followin g two major processes: PH + + e 3 To a lesser extent, the reactions + PH++ H + e 2 also may occur. 3 1'wo similar studies, by Neuert and Clasen and (l) (2) (3) (4) 4 Saalfeld and Svec, a g ree reasonably well with respect to the relative abundances of ions in the mass spectrum of phosphine. Halmann and Platzner 5 have demonstrated that an important reaction occuring when phosphine at increased pressures is bombarded with electrons in a mass spectrometer is + PH++ PH 4 2 (5) + The pressure dependences of the intensity of the primary ion PH 3 and the secondary ion tials of the ions + PH 4 were measured, as well as the appearance poten+ + + + PH 4 PH 3 PH 2 and PH. Several studies of the photochemical decomposition of phosphine have been made. The classical work of Melville 6 7 ,B, 9 pro vides a study of the overall reaction as well as an investigation of 2

PAGE 9

possible mechanisms. He found that the decomposition followed the reaction -+ Pred + 312 H2 (6) and concluded (from pressure versus qu a ntum yield studies) that wall reactions were probably important secondary steps in the photolysis. His mechanism involved a primary dissociation -+ PH + H 2 and (after diffusion to the walls) the secondary reactions: -+ H 0 + H -+ -+ (7) (8) (9) (10) Temperatures up to 300c had little effect on the yields, implying that the heats of activation of the surface reactions were small. A study of the flash photolysis of phosphine by Norrish and 10 Oldershaw has shown that when high radical concentrations are created the secondary reactions are homogeneous. The overall reaction again results in red phosphorus and hydrogen, with the phosphorus being pro duced in the form of small particles which remain suspended in the gas phase for several minutes. The spectra observed in this work were those of PH 2 PH, P 2 and a continuous absorption which was taken to be the suspended solid phosphorus. The major secondary reaction of PH 2 was assumed to be PH+ PH 3 (11) 3

PAGE 10

rather than reaction (8) as in Melville's mechanism. Both this study 11 and a similar flash photolytic study by R amsa y show that the major primary process in pho s phi ne photolysis is reaction (7). Their work at lower wavelengths (below 2000 A) indicates that some pri ma ry dissociation to PH also occurs. Wor k on Irradiated Ammonia A lar ge nu mb er of studi e s exist which deal with both the photochemistry and radiation chemi s try of gaseous ammonia. What is included here is a summary of some of the more recent and definitive investigations. The primary st eps in the photolysis of a.'llillonia are the following: (12) (13) 12 According to Mc.Nesby, Tanaka and 0kabe, the major process is reaction (12), but reaction (13) increases in importance at lower wavelengths. An interesting mass spectroscopic study by Melton 13 shows that when ammonia at one torr is bo m barded with 100 ev electrons the + predominant positive ion is NH 4 formed most probably in the ionmolecule reaction (14) Measurem en ts at 2 x 107 torr gave no measurable m/e = 18, but NH 3 + accounted for over one-half the total positive ions formed. At inter + mediate pressures the percent abund a nce of NH 4 incre a sed from zero 4

PAGE 11

to 83, and the parent ion abundance fell from 60 % to 12 % At pressures approaching atmospheric virtu a lly all the positive ions must be ammonium and the neutralization of NH 4 + must constitute a very important secondary step in the radiolysis of ammonia. 14 Toi, Peterson and Burton at Notre Dame have studied the effect of gas density on the radiolysis of a m monia in a stainless-steel autoclave at slightly above the critical temperature of 132c. They found sharp decreases in G(H 2 ) and G(N 2 ) in the density region 0,05 to 0.15 grams per cc. In radiation chemistry, the G-value is defined as the yield from a radiolysis, in molecules changed (formed or destroyed) per 100 electron volts of absorbed energy. At densities less than 0.05 grams per cc, G(H 2 ) = 6.2 and G(N 2 ) = 2.0 and at densities exceeding 0.15 grams per cc, G(H 2 ) and G(N 2 ) leveled off at about 1.5 and 0.4, respectively, The sharp drop in yields occurs well below the critical density of 0,235 grams per cc. The results are interpreted in terms of the formation of ion "clusters," which reduce the probability of dissociative neutralization by permitting the energy of neutralization to be spread over the molecules of the cluster. They consider that other effects, such as deactivation of excited species, may also be involved. Meaburn and Gordon 15 studied the rate of disappearance of NH radicals in irradiated ammonia, using the technique of pulse radiolysis, They found that the radical forms with a G-value of 0.4 and decays by a second-order process. Also, the addition of oxygen, ethylene and propylene decreased sharply the half-life of NH, but did not reduce it to zero. The explanation given for the persistence of some NH radicals even at high scavenger concentrations was that the radicals were 5

PAGE 12

regenerated by further re ac tions of some of the products form ed in the scaven g ing steps. Two very rec en t studies of the electron radiolysis of gaseous ammonia and ammonia mixtures by Jones and Sworski 16 and Jones Sworski and Williams 17 have thrown considerable li gh t on the nature of some of the second ary processes involved and particularly on th e "sc ave ngeable" and "nonscaven geable yields of hydrogen. 'l 'h e G-valu e s of n 2 and N 2 at 23c and 400 torr were found to be 4.5 and 1.5, respectively. Both yields increased with an increase in temperature to limiting values of G(H 2 ) = 15 and G(N 2 ) = 5, Very low concentrations of ethylene drasti cally reduced the hydrogen yield to a minim1.un of O .8 at [ c 2 H 4 ] /[ NE 3 ] = 0.02. The yield increased linearly with further increas e in [ c 2 H 4 J /[ NH 3 ]. Extrapolation of the subsequent increase back to [ c 2 H 4 ]/[NH 3 ] = 0 gave 0.75 0.05 molecules per 100 ev as the minimmn "molecular" H 2 yield in ammonia. The slope of this straight line gave G(H 2 ) = 1.5 0.5, which is the hydrogen yield in ethylene, within experimental error. These workers also found a pressure effect between 100 and 400 torr, with the yields of both hyurogen and nitrogen maximized at just greater than 100 torr. They found no evidence of appreciable hydrazine formation in irradiation of a.rn,~onia in a static system. The effect cf hydrogen atom and electron scavengers on the gas phase gamma. radiolysis of ammonia was investigated by Nishikawa, Shinohara and Matsuura. 18 Carbon tetrachloride vapor depressed G(H 2 ) from 4.3 (in pure ammonia. at 50 cm pressure) to 0.64 at 2 mole percent cc1 4 The authors suggest that carbon tetrachloride may be acting both as an H atom and electron scavenger, by the reactions 6

PAGE 13

cc1 4 + e CCl + Cl--:3 CC1 3 + HCl (15) (16) Both of these reactions are energetically favored. Other additives such as ethylene and propylene were also effective in reducing G(H 2 ) to less than one, with the exception of nitrous oxid e Minimum G(H 2 ) with added N 2 o w as 2. 5. This depression of only 1. 8 G-units with added N 2 o is attributed to the chan ge in the neutralization process of ammonium, from H and/or H 2 (17) to a reaction such as (18) Nitrous oxide does not scavenge hydrogen atoms. The ratio of the decrease in G(H 2 ) on addition of N 2 0 to G(H 2 ) in pure ammonia., 0.42, is considered to be representative of the contribution to the total hydro gen yield from processes involving ions. Radiolysis and Photolysis of Met h y l Iodide The reactions leading to m e thane formation in irradiated liquid methyl iodide have been studied by Gillis, Williams and Hamill 19 20 and by Petry and Schuler. There are probably two reaction routes, one involving a spur reaction between methyl radicals and hydrogen iodide (19) 7

PAGE 14

and another in which "hot" radicals attack the substrate CH 3 -+ CH 4 +CHI 2 Most of the radicals back-react with iodine -+ (20) (21) so that the competition for methyl radicals is represented by reactions (19), (20) and (21). No information on the vapor-phase radiolysis of methyl iodide appears to be available, but a photolytic study by Souffie, Williams and Hamill 21 shows that, in the vapor, methane is probably formed exclusively via the "hot" radical step. The use of hydrogen iodide as a scavenger in liquid CH 3 I results in a large inc;ease in methane production. 19 The radiolysis or photolysis of methyl iodide vapor with added HI does not appear to have been reported. Status of Present Work A quantitative study of the decomposition by gamma radiation of both pure phosphine and pure ammonia has been carried out. Also, the irradiation of mixtures of ammonia and phosphine and of methyl iodide and phosphine has shown that phosphine is an efficient radical scavenger, through the donation of a hydrogen atom. Further studies of the rauiolysis of systems containing phosphine are indicated. 8

PAGE 15

II. EXPERIMENTAL METHODS Apparatus Vacuum System All samples w e re prepared on the vacuum line shown in Figure 1. The design of the line was standard, consisting of a Welch Duo-Seal forepump, two liquid nitro g en cold traps and a two-stage mercury dif fusion pump. To the main manifold were attached the following: a mercury manometer (M), inlets for phosphine or ethylene and ammonia (both fitted with bariu m oxide drying tubes) (I), gas storage bulbs (each with a cold finger trap for degassing) (B), a fitting with stop cock for a vacuum thermocouple gauge (G), a detachable trap with o-ring fittings (DT), a submanifold (SM) and a Toepler pump-McLeod Gauge combi nation (Figure 2). The submanifold was fitted with a small calibrated volume (V) and a tubulation to which the irradiation cell (C) was attached for loading. A stopcock isolated the Toepler pump from the main manifold. On the back side of the main manifold were three detach able U-tube cold traps arranged in series (UT), each fitted with two Teflon stopcocks and o-ring joints, and another cold trap (TT) which was situated between the Toepler pump and the t ubulation for the attachment of the irradiated cell (C) via a glass breakseal (Figure 3). All greased stopcocks and joints vere greased vith Apiezon N and L vacuum greases. The entire vacuum system was constructed inside a stainless9

PAGE 16

G SM V DT C Figure 1. High vacuum manifold

PAGE 17

r----. ....... --, Sample loop (see Figure 8) Figure 2. 'l oepler pump-McLeod gau g e combination 11

PAGE 18

___ _. TT u U'l' Figure 3. Sample analysis system

PAGE 19

steel "California" hood with sliding gla ss doors, so that the system could be effectively iso late d from the re1 :1a inder of the laboratory. Irradiation Cells Two types of irradiation cells were used, as shown in Figures 4 and 5, The one-liter spherical cells which were used for most of the irradiations were fitted with a cold finger trap and break seal. The small annular cells were of approximately 10 ml capacity. Cobalt-60 G amm a Ray Source All irradiations were carried out in a 11 Wisconsin" type cobalt-60 gamma irradiator described previously. 22 Figure 6 is a cross-sectional view of the irradiator. The absorbed dose rate was measured relative to G(H 2 ) of 1.2 for ethylene, as will be described subsequently. Irradiation cell geometry was such that the maximum gamma ray intensity per unit volume was obtained. All irradiations were done at room temperature. Microtek Gas Chro m ato g raph A model GC-2000-R Microtek Research Gas Chromatograph was used, with the sampling module modified as shown in Figure 7, This modification was done by Mr. R. E. Marcotte of this laboratory. The sampling loop for the gas samples is shown in Figure 8. Bendix Time-of-Flight Mass Spectrometer Product identifications were made with ~ a Bendix model 14-107 Time-of-Flight Mass Spectrometer. Samples were introduced by attaching U-tube traps vith an o-ring joint to a semi-direct inlet. The inlet 13

PAGE 20

d\. i I I I I ;, ;j I //<('._~\,~ /2 \ ~ I / ,) \ ;i \ /,r ( ;~ ". y } J ) / .,, i ( ----_,,,/_ / 4 / / ~ : r: < :./ /. / -----..... : .::. ~"'};,',., ; ---.. .. -~ Figure !1. ~-;.: ... --i i ., Radiolysis vessel

PAGE 21

15 \ '( ,! \-----...... '1\ \ \ },y ~";;.\ \' I 1 1 -. --. I i \ \:. il ... l 1 1 I ., ( d 1: .; I i I, ': : ,' j i \ f ?) 'j t f ; \ \ \ l I \ f .f,1 Figure 5. Radiolysis vessel

PAGE 22

Figure 6. Cross section of cobalt-60 gamma ray source Legencl = (A.) counterweight; ( B ) upper support; (C) control rod handle; (D) extra top shielding; (E) storage turret; (F) 400 curie n 60 Go source; (G) shutter shown open; ( H) rear wall; (I) door; (,T) downward shielding; (K) door carri a ge; (L) door crank; (M) door frame

PAGE 23

17

PAGE 25

19 < ----c~ ->_---~-L r Figure 8. Gas chromatographic sampling loop

PAGE 26

was connected to the i on source through a metering valve. The large ballast volume of the inlet system was not u sed Re ag ents Ammonia Matheson Company anhydrous ammonia (99.99 % minimum purity) was used as obtained fro m the manufacturer, for all experiments. Matheson Company phosphine (99.5% minimum purity) was used without further purification. Ethylene Matheson Company C.P. grade ethylene (99,0 % minimum purity) was used as obtain ed Methyl Iodide Mallinckrodt Analytic al Reagent grade methyl iodide was passed through an 11-inch column of Alcoa activated alumina, grade F-20, and its purity ascert a ined via gas-liquid chromatoeraphy. Only one lower boiling impurity was found, whose concentration was significantly reduced by passage through alumina. Although it was not identified, it was estimated that its concentration could not be greater than one or two parts per million. Two higher boiling im p urities, also of low concentration, were present. The material was further purified by distillation (see section on preparation of samples). 20

PAGE 27

Preparation of Samples Dosimetry After evacuation of the main manifold, mercury manometer, ethylene inlet and two calibrated storage bulbs to a pressure of about 0.1 micron, the first bulb was closed off and the main manifold was isolated from the pumping station. Ethylene was admitted slowly through the barium oxide drying tube until the desired pressure was obtained. The ethylene to be used in filling the cell was closed off in the second storage bulb, whose volume was calibrated with distilled water before attaching to the line. Excess ethylene was then frozen down with liquid nitrogen in the first bulb, to be used in a later run. After three cycles of freeze evacuate-thaw degassin g the ethylene was allowed to come to room temperature in the bulb while the submanifold and irradiation vessel were pumped down for a minimum of 20 minutes. The main and submanifolds were then isolated from the pumping station and the ethylene was transferred to the liquid nitrogen cold trap in the irradiation cell. Fifteen minutes was allowed for complete condensation, after which the cell 'W'as sealed off (while open to the pump) by collapsing the constricted part of the neck with a gas oxygen torch. The neck was allowed to cool to room temperature before removing the liquid nitrogen from the cold finger. After each irradiation and analysis the cell was evacuated, 0 removed from the line and annealed in an oven at 575 C. After attachment to the submanifold for another run, it was thoroughly flamed out with a bush ga s -oxygen flame while open to the pumping station. 21

PAGE 28

Pure Phos p hine Sample preparation was similar to that described above for ethylene except that th e cell was washed out with concentrated nitric acid and then distilled water before annealing, to re move the elem enta l phosphorus left on the glass surface by the previous experiment. Pure Ammonia Sample preparation was very similar to that for ethylene except that only one storage bulb was used. The volume of the ma in man ifold, mercury manometer, ammonia inlet and stora g e bulb was calibrated with a known amount of gas and, by a suitable proportion, the main manifold pressure which would correspond to a given cell pressure could be determined. Thus in th es e runs all the gas taken from the cylinder was used in filling the cell. Pure Methyl Iodide Vapor A sample of l or 2 mls of methyl iodide was attach ed. to the main manifold via a detach ab le trap with o-ring joint, and the liquid was degassed in the usual way. It was then expanded into a calibrated one-liter capacity gas storage bulb until the desired pressure was ob tained, when the stopcock was closed and the excess methyl iodide was frozen down in the trap and discarded. In a typical run only about one-half of the methyl iodide was needed to fill the bulb, so that the purity of the material (with respect to higher boiling substanc e s) was further enhanced by this distillation process. After thorough evacuation of the manifold and cell the methyl iodide was transferred to the cold finger trap on the cell and the cell was sealed off under vacuum. 22

PAGE 29

Phosphine-A m monia Mixtures Phosphine and ammonia were, in turn, admitted to the manifold to the desired pressure and then frozen down and d e gassed in the appro priate storage bulbs, The gases were thep. transferred to the irradiation cell and sealed off in the usual manner. Phosphine-Methyl Iodide G a s Mixtures Loading of these cells was carried out in the same manner as with the phosphine-ammonia mixtures. Product Analysis Dosimetry Products Irradiated ethylene samples were attached by means of a brea.1<.seal fitting to the vacuum line, as shown in Figure 3. All materials noncondensible at -196c were collected via a 'l'oepler pump and the gas pressure measured on a McLeod gauge. Each sample was de gassed through approximately 18 cycles or until pressure increments were 0,1 mm or less, as read on a meter stick attached to the McLeod gauge. Intermittant thawing of the condensible materials was carried out, to remove traces of noncondensibles trapped in the frozen material. After collection, the gas samples were transferred via the Toepler pump to the gas chromatographic sampling loop (see Figure 2). Analysis for methane was made with the Microtek Research Gas Chromatograph, using a 0,75 meter 100/110 mesh silica gel column at 4o 0 c and a hydrogen flame ionization detector. The carrier gas was dry nitrogen at a flow rate of approximately 80 cc per minute. Methane standards were injected vith a Hamilton 500 l gas-tight syringe, 23

PAGE 30

Hydrogen analysis was by difference, the only materials present in the Toepl ered gas mixtu re bein g hydrogen and methane. Esti mated error in hydrog en analysis was 1 % Phos ph ine Irradiation Produc ts Hydrogen analysis was via Toepler pump an d McLeod gauge, as previously described. A mass spectrometric analysis of the material collected and measured in this way showed it to be only hydrogen. Two U-tube cold traps were used to fractionate the condensible materials. These traps were maintained at -127c and -196c. After fractionation, the traps were removed and their contents analyzed mass spectrometrically. No quantitative analysis for elemental phosphorus was made. Phosphin e Mas s Balance Analysis for hydrogen was as described earlier. Phosphine pressure before irradiation was determined directly from the main mani fold mercury manometer. The phosphine was then sealed off via a Teflon stopcock on a calibrated submanifold volume. Phosphine pressure after irradiation was determined with the McLeod gauge, using a large cali brated volume attached above the stopcock via a standard taper joint. A blank run was made to verify the consistency of the volume calibra tions. Estimated error was 1%. Ammonia Irradiation Products After measuring with the McLeod gauge the total gas collected via the Toepler pump, the gas was transferred to the gas chromatographic sampling loop and attached to the sampling module of the Microtek gas 24

PAGE 31

chromato g raph, as shown in Figure 7A mechanical fore pump was used to evacuate the volume between the loop stopcocks and the sliding valve. The valve was then pushed in to direct th e carrier gas toward the sampling loop and through the loop bypass (r e fer to Figu r e 8). After the resulting pressure surge had subsided, as evidenced by the recorder pen returning to its normal baseline, the s& ~ ple loop stopcocks were opened and the bypass sto p cock closed. The sample passed throu g h a 0.75 meter 100/110 mesh silica gel column at room temperature, and a then n al conductivity detector which was c a librated for nitrogen analysis with standard injections from a 500 l gas-tight Hamilton syringe. The carrier gas was helium at a flow rate of about 25 cc per minute. Esti mated error was 2%. Hydrogen analysis was by difference, since the products non condensible at -196c could only be hydrogen and nitrogen. No attempt was made to analyze for hydrazine, it being assumed that no measurable amount was present in the irradiated 1 13,14,17 samp es. Methyl Iodide Vapor Irradiation Products Products noncondensible at -196c were collected with the Toepler pump and measured with the McLeod gauge. The gas was then pUinped into the gas chromatographic sampling loop and attached to the Microtek sampling module. It was passed through a 0.75 meter 100/110 mesh silica gel column at room temperature, and a hydrogen flame ionization detector calibrated for methane analysis in the usual way. The carrier gas was helium at a flow rate of about 65 cc per minute. Hydrogen analysis was by difference, since in this system 25

PAGE 32

the only products noncondensible at liquid nitror,en temperature are methane and hydrogen. This was verified by a mass spectrometric analysis, which showed an intense peak at mass 2, in addition to the spectrum of methane. No attempt was made to analyze for products other than methane and hydrogen. Phosphine-A mm onia Irradi a tion Products Analysis of the materials which were noncondensible at -196c was identical to that for the ammonia irradiation products. In a searc:h for other products, a series of three U-tube cold traps was used ( s2e Figure 3). After pumping off the permanent gases, the remaining materials were fractionated among the traps, which were immersed in baths at o 0 c, -78c and -196c. The traps were then removed from the vacuum line and their contents analyzed mass spectro metrically. Phosphine-Methyl Iodide Irradiation Products Quantitative analyses for methane and hydrogen were as describ ed in the preceding section. 26

PAGE 33

III. EXPERIMENTAL RESULTS Dosimetry The results of the dosimetry with ethylene are summarized in Figure 9, In this and all subsequent graphs, the units for the ordinate and abscissa are considered to include any multiplicative factor, such -6 as 10 From this graph it can be seen that the hydrogen yield is directly proportional to irradiation time, from three hours to 24 hours, at 60 cm ethylene pressure and 23 2c. The longer time corresponds to a total absorbed dose of 5,4 x 10 20 ev per gram. The calculation and its theoretical justification are found in the Appendix. The absorbed 14 dose rate was found to be 5.68 x 10 ev per mm of ethylene pressure per liter of cell volume per minute of irradiation time. This rate was corrected at two-week intervals to account for the decay of the cobalt-60. Pure Phosphine Alth hf ht h 1. t t 6 7 ,lO it oug rom previous po oc emica inves iga ions was suspected that the only products of the gamma ray-induced decompo sition of pure phosphine would be hydrogen and red phosphorus, an attempt was made to find other p1 oducts. Mass spectra of the contents of the cold traps used to fractionate the condensible materials after radiolysis (see pa g e.24) showed only trace amounts of masses 66 and 62 27

PAGE 34

11 I.O I 10 0 X 9 CJ) C) 0 E 8 C: 7 ....... (J) 6 0 _J 5 w >4 z w 3 <..!) 0 Ct: 0 2 >:r: 6 12 18 24 IRRADIATION TIME ( in hours ) Figure 9. Hydrogen yields from ethylene at 60 cm pressure, as a function of irradiation time 28

PAGE 35

in the -127c trap, in addition to the spectrum of phosphine. Virtually all the phosphine was collected in the -196c trap. The irradiation cell contained a fine deposit of solid material which would not distill into the traps, even on warming with a hot air gun. Although its coloration was not precisely determined because of the brownish coloration of the irradiated glass of the cell, it appeared to be brownish-red or brownish-orange. One striking feature was the spacial distribution of the solid on the bottom of the cell. Emanating from the center (directly under the test tube well into which the cobalt-60 source fits), the solid showed a star-shaped or "exploding" pattern, extending several inches out from the center. The hydrogen yields from pure phosphine at 55 cm pressure and room temperature, irradiated at a dose rate of 3.25 x 10 17 ev per liter per minute, are plotted as a function of dose out to approximately 21 20 10 ev, in Figure 10. Beyond 5 x 10 ev there appears to be a slight drop-off in hydrogen production. Figure 11 shows that the hydrogen yields are directly proportional to phosphine pressure, from below 4 cm up to 76 cm. 18 The absorbed dose was 4.4 x 10 ev per cm PH 3 or 2.4 x 10 20 ev per gram. The irradiation time was held constant. To illustrate the calculation of a G-value when stopping powers are used, the determination of Gi(H 2 ), the initial (or zero dose) hydrogen yield per 100 ev, from phosphine is shown here in some detail. The rate of hydrogen production as a function of irradiation time (corresponding to Figure 10, except that the abscissa in the figure has been converted from hours of irradiation time to absorbed dose) was 16 3.7 x 10 molecules per minute. Since the phosphine pressure was 55 cm the rate of hydrogen production was 6.7 x 10 13 molecules per minute 29

PAGE 36

r---. L() 0 X (/') 0) 0 E C: .._... (/) 0 _J w >z w (.9 0 [( 0 >I 14 12 10 8 6 4 2 l 2 3 4 6 7 8 9 RADIATION DOSE ( in ev x 10 20 ) Figure 10. Hydrogen yields from phosphine at 55 cm pressure, as a function of radiation dose w 0

PAGE 37

(!) I 0 X ti", (D 0 E C ....... CJ) a _J w >z w (9 0 0:: 0 >:c 50 40 30 20 10 IO 20 30 40 50 60 70 P~ESSURE OF PHOSPHINE (in cm) Figure 11. Hydro g en yields as a function of phosphine pressure, irradiated 18 to a total dose of 4.4 x 10 ev per cm PH 3 w I-'

PAGE 38

per mm of phosphine. The calculation of Gi(H 2 ) was as follows: Gi(H 2 ) = molecules H 2 produced/100 ev absorbed dose (6.74 x 10 13 molecules/min-mm PH 3 ) x 100 = 6.74 X 10 13 X 100 = 5.83 X 10 14 (1.041)(0.983 8 ) = 11.3 In the above equation, SPH /SCH is the ratio of electron stopping 3 2 4 powers of phosphine and ethylene. Results of the mass balance experiments are summarized in Table 1. The decrease in phosphine pressure du.ring irradiation was found to correspond quite closeJ.y to two-thirds of the hydro ge n pressure, after making temperature corrections. Although no analysis was made for phosphorus, these experiments clearly showed that no products other than hydro g en and phosphorus were formed, and that the decomposition closely followed the stoichiometry: pred + 312 H2 (22) The trace 8lllounts of masses 62 and 66 in the -127C trap probably indi cate P 2 and PH 2 PH 2 respectively. Pure Ammonia Hydrogen and nitro ge n yields from irradiation of pure ammonia at 55 cm pressure and room temperature, at a dose rate of 2.06 x 10 17 ev per liter per minute, are plotted as a function of dose in Figures 12 32

PAGE 39

Run Irradiation blank 1 2 Time (hours) 0 143 108 20.60 19.01 15.32 TABLE 1 Mass Balance Results b Pf(PH 3 ) 6P(PH 3 ) ( cm) ( cm) 20.61 17.48 14.40 -1.53 -0.92 2.30 1.39 2.30 1.38 % Decomposition 8.1 6.o ~ressure calculated for McLeod gauge, from initi ~l p hosphine pressure in calibrated volume and the ratio of McLeod gauge and submanifold volumes. b Pressure of phosphine in McLeod gauge after irradiation. w w

PAGE 40

,.,...... L{) I 0 X 3 Cf> (1 ) 0 E C .._,, 2 (f) 0 _J w >z l.JJ (:) 0 e r:: 0 >:c 2 3 4 RADIATION DOSE ( in ev x 10 20 } Figure 12. Hydrogen yields from ammonia at 55 cm pressure, as a function of radiation dose

PAGE 41

and 13. 20 20 Both are linear from 1 x 10 ev out. to 5 x 10 ev, and the line extrapolates back to zero yield at zero dose. Figure 111 shows that nitrogen yields incre ase linearly with increasing ammonia pressure, from 10 cm up to one atmosphere, the line extrapolatin g back to zero yield at zero pressure. Hydrogen yields also w e re independent of pressure and were always three times the nitrog e n yield. For the study of pressure dependence, the absorbed dose was 4.4 x 10 18 ev per cm NH 3 and the irradiation time was held constant. The G-values for nitrogen and hydro g en were found to be 1.5 and 4.5, respectively. 13 14 17 A number of other workers have failed to find significant amounts of hydrazine formation from the gamma radiolysis of ammonia in a static system. No attempt to find hydrazine was made in this work. However, the yield of nitrogen as found by gas liquid chromatography was checked against one-fourth of the total gas measured in the McLeod gauge, and the results agreed within 2 % or less in each case. Phosphine-Ammonia Mixtures Mass spectrometric analysis of the materials noncondensible at -196c showed peaks at masses 1 and 2 and enhancement of the back8 0 80 ground mass of 2 The O C and -7 C traps showed the presence of small quantities of masses 31 and 62, with the -78c trap having a trace of masses 79 and 91. The -196c trap gave intense spectra of phosphine and ammonia. The solid residue remaining in the irradiated cell was of exactly the same appearance, includin g the unique spacial distribution, 35

PAGE 42

(D I 0 X (/) (!) 0 E C ......... (J) 0 _J w >z ld (!) 0 cc }z 12 11 10 9 8 7 6 5 4 3 2 2 3 4 RADL/\ TION DOSE ( in ov x 10 20 ) Figure 13. Nitrogen yields from ammonia at 55 cm pressure, as a function of radiation dose 5 w 0\

PAGE 43

(!) I 0 X (/) (l) 0 E c.: '-' (f) 0 _J w >z w (.9 0 0:: rz 9 8 7 6 5 4 3 2 10 20 PRESSURE Figure 14. Ni trogen yields 4 4 18 x 10 ev per cm NH 3 30 or40 50 AM M ON I A ( in 60 cm ) 70 as a function of ammonia pressure, irradiated to a total dose of

PAGE 44

as that fro m the radiolys is of pure phosphine. This residue was still visible, al tho ugh diminish e d in qu ant ity, after radiolysis of samples containi ng as little as o ne mole percent phosphine. The mixtures were ma de up to a t otal pr essure of 55 cm at 23C and irradiate d to dos es of either 5 x 10 20 ev or 1 x 10 20 ev. The lower dose was c h os en fo r lo w phosphine mixtures in ord e r to avoid usin g up all the phosphine during radiol ys is. G-values for hydrogen and nitrogen from these m ixtures were plotted as a function of the fraction of the dos e which was absorbed by the phosphine. These graphs are Figures 15 and 16. It should be note d that while the hydro gen yields are larger than predicted fro m the assumption of an "ideal" mixture in which each co mp ound decom p oses proportionately to the dose it absorbs ( see dashed lines on fi g ures), the nitrogen yields show a stri k ing depres s ion. This occurs even in the mixtures with very low mole percentages of phosphine. It is note worthy that a nresidual" yield of nitro ge n p e rsists even in mixtures containing relatively high phosphine concentrations. In mixtures 8 and 9 ( se e Table 2), where the mole percenta ges of phosphine were 0.2 and 0.02, respectively, a reduced dose was used. These runs, in addition to one of the runs with pure ammonia, are indicated by circles on the figures. The composition of each mixture and the irradiation times are given in Table 2. The fraction of the dose which was absorbed by phosphine, [SPH /(SNH + SPH )], was found by first computing the 3 3 3 ratio SNH /SPH and then forming the quotient ( 1/ ( 1 + SNH /SPH ) ] The 3 3 3 3 required irradiation time was calculated from the dose which each component absorbed. The time for each mixture was checked by making 38

PAGE 45

11 > !O 0 0 9 I... (\) 0.. 8 (/) <1) 7 :::, (.)
PAGE 46

> (l) 0 0 C C/) w ::) _J I 0 z w (..9 0 0:: 1z 14 .8 .6 :4 .2 Absorb ed Dose \ \, 0 1 x 10 20 ev \ \ 6 5 x 10 20 ev \ \ \ \ \ \ \ \\ \ \ \ \ ', ', \ \ ', \ \ \ \ \ \ \ \ \ '\ 0.20 0.40 0.SO 0.80 \ \ \ \ FRACTION OF ENERGY ABSORBED BY \ Pl-I t I 13 Figure 16. G-values of nitroeen from ammonia-phosphine mixtures as a funct ion of the fraction of en e r g y absorbed by phosphine 40

PAGE 47

Mixture N umber 1 2 3 4 5 6 7 8 9 55 50 30 15 5 2 0.5 0.1 0.011 TABLE 2 Phosphine-Ammonia Mixtures Mole Fraction 0 1 5 0.909 25 0.545 40 0.273 50 0.091 53 0.036 54.5 0.009 54.9 0.002 55 0.0002 1 0.940 o.654 0.371 0.138 0.057 0.014 0.003 0.0003 T b ime (hours) 25.17 26.97 31.33 35.48 39.02 40.17 40.75 8.32 8.32

PAGE 48

TABLE 2 (continued) Mixture Number 10 11 0 0 55 55 a o All pressures are corrected to 23 C. Mole Fraction 0 0 0 0 T b 1.me (hours) 40.83 8.32 bTimes greater than 25 hours correspond to a dose of 5 x 10 20 ev for t h at mixture; 20 8.32 hours corresponds to a dose of 1 x 10 ev.

PAGE 49

the calculation twice, once for each co mpo nent, and verifying that these ti mes were the sru ne The percent deco m position of phosphine with mixtures 6, 7, 8 and 9 was 3, 11, 18 and 90, re spe ctiv e ly. These were calculated from the obs e rved yiel ds of hydrogen and nitro ge n, assuming th a t no products were for med in significa n t am ounts other than thes e and el eme ntal 0 8 0 phosphorus. The mass spectr a l analyses of the O C and -7 C traps would appear to bear out this as sump tion The extent to which th e "ide al mixture" yields do not obta.in is best indicated by Figure 16. The extremely sharp depression of the nitrogen yield corresponds to the sudden increase in hydrogen yield (see Figure 15), with a mixture in which (initially) less than 0.1 % of the absorbed dose was ab s orbed by phosphine, and in which 90 % of the phosphine was decomposed during radiolysis. The residua l nitro ge n yield appears to be decreasing proportionately to the energy fraction absorbed by phosphine. By extrapolating this curve back to the ordi nate, a G-value for "molecular" nitrogen of 0.05 is obtained. Pure Methyl Iodide Vapor Methane and hydrogen yields from methyl iodide vapor, irradi ated at a dose rate of 4.64 x 10 17 ev per liter per minute, are plotted as a function of absorbed dose in Figures 17 and 18, respectively. Methyl iodide pressure was 26.8 0.1 cm at 23c. Beyond 1 x 10 20 ev both products show rates of production proportional to dose. Hydrogen shows an initial rate of production greater than that observed after 1 x 10 20 ev, whereas methane production is depressed by a similar 43

PAGE 50

< Cl) Ci) 0 E C ......... (J) 0 _j w >w z
PAGE 51

(.!) I 0 >< (/) Q ) 0 E C ........... if) 0 _J I tJ ->z w (.9 0 er: 0 >:c 5 4 3 2 2 3 4 RADIATION DOSE ( in ev x 10 20 ) Figure 18. H ydrogen yields from methyl iodide at 26.8 cm pressure, as a function of radiation dose

PAGE 52

amount over the same dose re g ion (0 to 1 x 10 20 ev). The initial G-value for methane production is 1.7. In the linear portion beyond 1 x 10 20 ev the 100 ev yield has increased to 2,3. These calculations are based on an electron sto p ping power ratio, SCH I/SCH of 3.185, 3 2 4 as calculated by the Bethe equation. Hydrogen G-values were 1.4 and 0.6 in the initial and linear portions of the curve. Both product yields show a dependence on methyl iodide pressure. Irradiations were carried out such that the absorbed dose 19 was 1.1 x 10 ev per cm CH 3 I. The irradiation time was constant. The total yields are roughly proportional to pressure out to about 20 cm, beyond which the G-value for methane production shows a decided in crease and that for hydrogen production falls off (see Figures 19 and 20). Because of the scatter in the hydrogen yields, it is difficult to determine the shape of the curve beyond 30 cm pressure. Measure ments were taken out to the equilibrium vapor pressure of methyl iodide (38 cm) at room temperature. Phosphine-Methyl Iodide Mixtures These mixtures were made up to a total pressure of 27-30 cm at 23c. Irradiation times were such that each mixture received a 20 dose of 1 x 10 ev. Methane and hydrogen yields from the radiolysis of these mixtures are plotted as a function of the fraction of energy absorbed by phosphine (Figure 21a). The methane yield shows a sharp maximum close to 0.01, after which it falls off rapidly. The dotted line shows the way the yield should decrease if it was a function only of energy 46

PAGE 53

.......... 13 (!) 'o 16 >< (/) 14 0 0 E '? ,_ C '-..J' 10 (J) 0 _J 8 w >6 w z 4
PAGE 54

(l) 'o 5 X en 4 Q.) 0 E C (/) 3 0 _j w >2 z w (!) 0 0 >I 0 -----------0 0 4 8 12 16 20 24 28 32 36 PRESSURE OF METHYL 100\DE ( In cm) Figure 20. Hydrogen yields as a function of methyl iodide pressure, irradiated to a total 19 dose of 1.1 x 10 ev per cm CH 3 I

PAGE 55

'----'----- '-----'-----L-. -,----'--0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.30 0.90 FH/\CTION OF ENERGY ABSORBED DY PHOSPHINE Figure 21a. Yields of hydrogen and methane as a function of the fraction of energy absorbed by phosphine 1.00

PAGE 56

absorbed by methyl iodide. Hydrogen yields increase linearly with the energy absorbed by phosphine, beyond a minimum phosphine content of about 30 mole percent. The hydrogen yield remains constant and the same as in pure methyl iodide vapor until just beyond the maximum methane yield, This is shmm more clearly in Figure 21b, where the abscissa in this region has been greatly expanded Table 3 lists the composition and irradiation time for each mixture. Calculatio n of the maximum rate of methane production from Figure 21a was done by taking the yield of methane at the apparent maximum of the curve. G(CH 4 ) at this point was 12.6. The mixture corr esponding to this maximum was number three, containing 3,5 mole percent phosphine and in which 99% of the total radiation dose was absorbed by methyl iodide. The ratio between this maximum G and the G for methane in the pure vapor, at the same total dose, is 6.3. The minimum and maximum values for hydrogen G-values in the proportional re gion were 3.5 and 11.3. Hydrogen production in those mixtures which contained only a few mole percent phosphine was the same as in pure methyl iodid e vapor, with a G-value of 1.0. 50

PAGE 57

'o X c,, 0 E C (/) 0 _, w >z w t!) 0 cc: 0 >I a 7.
PAGE 58

TABLE 3 Phosphine-Methyl Iodide Mixtures Mixture P(PH 3 )a P(CHi)a Mole Fraction 8 PH Time Number (cm) (cm) PH 3 3 (hours) 8 cH I+ 8 PH 3 3 0 0 26.7 0 0 3.73 1 0.19 26.7 0. OO'T 0.002 3.73 2 0.50 26.9 0.018 0.006 3.57 3 0.97 26.9 0.035 0.012 3.53 4 2.10 26.9 0.072 0.023 3.50 5 2.66 26.7 0.091 0.032 3.50 6 10.16 19.6 0.341 0.145 4.22 7 19.60 9.6 0.671 0.399 6.02 8 24.06 5.1 o.844 0.605 7.45 9 27.56 1.9 0.936 0.826 8.88 10 55,00 0 1 1 4.85 a All pressures are 0 corrected to 23 C. Taken from Figure 10. \.n f\)

PAGE 59

IV. DISCUSSiO N AND INTERPRETATION Pure Phosphine and Pure A mm onia From a comparison of the initial G-values of hydrogen which vere obtained from the radiolysis of ammonia and phosphine, it is apparent that phosphine is considerably more susceptible to radiolytic breakdown than its nitrogen analogue. The most obvious explanation for this lies in a comparison of the available data on bond strengths in the two compounds. Cottre11 23 lists the average bond energy terms, E, for ammonia and phosphine as 93.4 and about 77 kcal, respectively. The bond dissociation energy, NH 2 --H, is given as 102 kcal but PH 2 -H is apparently not known. At any rate, it is safe to assume that dissocia tion processes such as -+ PH + H 2 -+ PH+ 2H (23) (24) are less endothermic than the corresponding steps with ammonia. Charged + + + + species such as PH 4 PH 3 PH 2 and PH are also probably less stable vith respect to dissociation than the nitrogen species, and on neutrali zation should show a more pronounced tendency toward breakdown into (finally) hydrogen and phosphorus. The lack of dependence of the product yields on phosphine pressure indicates a probable lack of surface reactions or steps requiring collisional deactivation of an intermediate under the experimental conditions of this study. 53

PAGE 60

If ion-molecule re ac tions figure prominently in the radiolysis of phosphine, as th ey appear to in ammonia, it is reasonable to consider these processes as bein g analogous to those postulated for amm onia. Melton's work 13 contains a detailed exa m ination of these reactions in irradiated ammonia. Also, the electron impact study of phosphine by Halmann and Platzner 5 has sho wn that the formation of phosphonium ion in phosphine occurs via an ion-molecule reaction. This is analogous to the formation of ammonium ion when ammonia is bo mbar ded with electrons. Reaction of hydrogen atoms must be predominantly by the abstraction reaction (25) rather than by recombination of atoms) by a simple kinetic argument. The rate of recombination of H atoms is given by and the rate of abstraction by reaction (25) is A value for kr of 10 10 liters 2 per mole 2 per sec (with NH 3 as the third 24 body) has been reported, but k does not appear to be known. It is a probably comparable to the rate constant for the reaction H + HBr -+ H + Br 2 (26) 9 25 which is about 10 liters per mole per sec. The steady-state concentration of hy dr oge n atoms during radiolysi.s can be estimated as follows: 54

PAGE 61

Rate H generation= Hate H removal x dose r a te 100 = (7,5 x 102 atoms/ev)(5 x 10 15 ev/liter-sec) 4 x 10 14 atoms/liter-sec 6 -10 = x 10 moles/liter-sec If reaction (25) is the step by which most H atoms react, then Rate H removal = kJ PH 3 ][ H ] = 10 9 liters/mole-sec [ 102 ][ H) = 10 7 [ H ]/sec Solving for the concentration of hydrogen atoms, ( H] = 1016 moles/liter Finally, R k a a --R kj H) r = 1015 It follows that the recombination of hydrogen atoms is insignificant in this system, The phosphino radical may abstract hydrogen from phosphine, which is not a chemically detectable reaction, Recombination of the radicals PH+ PH 2 2 -+ (11) 55

PAGE 62

is probably exothermic, since the PH 2 H bond strength is probably greater than that of PH-H, by analogy with ammonia 23 Elemental phosphorus formation from recombination of PH sp e cies is supported by the data from the flash photolysis-, lO which sho we d t hat t he formation of P 2 was too rapid (15 11sec or less ) to b e accounted for by a three-body collision. The P 2 must combine to for m The G(-P H 3 ) of 7.5 su ggests th at a mechanism i nvolving no che m ical ampl ificatio n of the primary ionization and excitation pro cesses may be appropriate For most gases the ion-pair yield, W, is approxi mately 30 ev, so that G for positive ions can be taken as about 3. Assuming that on e to three ex c i ted molecules are also formed for each ion, th e primary yield (ions and excited mole cules) should be of the order of 6-12. The following r ea ction steps summarize a possible mechanism whi ch takes account of these c onsiderations : PH 3 PH + H 2 (27) P H 3 PH+ H2 or 2H (28) H + PH 3 -+ H2 +PH 2 (25) PH +PH -+ P H 3 + PH 2 2 (11) PH + P H -+ p2 + H2 (29) p2 + p2 -+ P4 (30) The mechanism suggested by Nor rish and 0ldershaw 10 for the deco mposition of phosphine by f lash photolysis is veri-J similar to the one given here, although it was necessary for them to propose two 56

PAGE 63

reaction schemes. One scheme wa.s used to explain the results at room temperature, the other at hi Gh er tempera t ures. The formation of hydrogen was considered t o occur by reco m bination of atoms at room temperature, with hydro ge n abstraction from phosphine by hydro ge n atoms occuring at higher temperatures. '!'he necessity for the two schemes arose from the fact that the quantum yield was one-half at room tempera ture, but increased to one at high temperatures. In view of the previous calculation of relative rates for the recombination and abstraction reactions in this system, it seems improbable that reco m bination of hy drogen atoms is important, even durin g flash photolysis, where the con centration of radicals would be much larger than during gamma radiolysis. Some diphosphine may be formed by the reaction PH+ PH 2 2 (31) Diphosphine, like hydrazine, would be susceptible to attack by various radical species, notably hydrogen atoms. If significant amounts of diphosphine form during the radiolysis, an experiment such as reported by Jones, Sworski and Williams 17 would be informative. They found that the recovery of hydrazine from radiolysis of ammonia in a flow system depended directly on the flow rate of ammonia. At high flow rates, most hydrogen atoms are scaven g ed by ammonia and most amine radicals combine to form hydrazine. The present study cannot discount the pro duction of diphosphine and its subsequent removal by radical attack in a static system. It is notable, however, that whereas hydrazine thermal ly decomposes to nitrogen and hydrogen, diphosphine breaks down at room temperature according to the following equation: 57

PAGE 64

-+ (4x 2y)PH 3 + 2P H X y (32) 26 where x > y. The mass balance results of this study have shown that no appreciable amounts of lower phosphorus hy drides are for med during or after the radiolysis. If the decomposition of diphosphine by radic a l attack yields the same products as th e ther mal reaction, and if signifi cant quantities are for med during the rad iolys is, the mass balance would show a hydrogen deficiency due to the presence of quantities of hydrides in th e solid r esidue Within the experimental error of the mass balance experiment, which was less than 1%, no such deficiency was observed; i.e., hydro gen produced was equal to three-halves phosphine decomposed. 16,18 Although this study and two others reported very recently show quite close agreement for the G-value of hydro gen fro m the rad iolysis of ammonia, earlier investi gat ions have shown wide variation in the hydrog e n yield. An examination of the experimental conditions in each study shows that the discrepancies are probably explainable in terms of these conditions. In one instance, 27 an "uncorrected" G-value has been given which was calculated using the Fricke liquid dosimeter. Recent workers, using gas phase dosimeters (and electron stopping powers rather than electron densities for dose rate calculations), are in good agreement. PhosphineAmmon ia Gas Mixtures The most recent work done on radiolysis of ammonia in the gas Melton (See Table 3, Reference 13) provides a summary of these G-values as obtained b y various workers. 58

PAGE 65

phase has been that of Jones and Sworski, 16 who obtained a G(H 2 ) of 4.5 at room temperature, but observed a dependence on ammonia pressure below 40 cm in a continu ation of their work. 17 Nishikawa, Shinohara and 18 Matsuura, irradiating ammonia at 50 cm p_ressure, found the hydrogen yield to be 4.3. Finally, this study at 55 c m give s G(H 2 ) = 4.5, with no dependence on ammonia pressure fro m 100 torr up to one atmospherei with the curve extrapolatin g to zero yield at zero pressure. The consistency of these r e sults with pure ammonia appears to make a further comparison valid; namely, the r es ult s obtained by these workers with various radical scavenge rs and the present study usin g phosphine. The earlier workers used ethylene as a scavenger and obtained minimum hydrogen yields of 0.1-0.8, with ethylene concentrations of about 2 mole percent. Carbon tetrachloride decreased the yield slightly, to 0.64. These results indicate a yield of "molecul a r" hydrogen of about 0.75. In the pres en t study the "molecular" nitro ge n yield was mea sured directly and found to be 0.05. It is evident that phosphine is a very efficient scavenger of amino radicals, by the reaction (33) which is exothermic of NH 2 -H > PH 2 -H. It is probable that this reaction has a MI -15 kcal. By reaction ( 3 3 ) and -+ H + PH 2 2 (25) which is probably exothermic by around 20 kcal, both hydrogen atoms and amino radicals could be efficiently scavenged. The fact that phosphine may react readily -with both amino and nitrene (NH) radicals formed in 59

PAGE 66

the primary dissociations may account for its effectiveness in depressing the nitrogen yield. Despite extensive investigations of both the photochemistry and radiation chemistry of ammonia, the mechanism for formation of nitrogen is not well established. The following reactions are often given as a possible route to N 2 : (34) + (35) (36) -+ (37) 13 It has also been suggested that the diimine may be formed by the elimination of hydrogen from excited hydrazine, as below: -+ (38) followed by the unimolecular decomposition of N 2 H 2 as in reaction (37) above. The formation of nitrogen by reaction steps not involving scavengeable species; i.e., which result in production of "molecular" nitrogen, can be rationalized by postulating the existence of a nitrene, NH, which undergoes an insertion into ammonia, as in the following reaction: -+ (39) The excited hydrazine would break down to nitrogen and hydrogen via reactions (38) and (37). The formation of a significant amount of NH 60

PAGE 67

15 in irradiated ammonia has been demonstrated by Meaburn and Gordon. They measured the absorption spectrum of the radic a l due to the A 3 rr + x 3 L t;ansition at 33 6 0A, after pulsing with 250 kev electrons. Because of the experimental technique used_, no information on NH in the singlet state was obtained. Provided that so m e small fraction of the nitrenes could exist in a singlet state, such that one of the lower energy orbitals is unoccupied, it is possible to propose a transition state in which bonding occurs through this orbital and the filled sp 3 of ammonia. The only rearrangement then required is a protonic shift. The analogy with carbene insertion reactions on methane makes reaction (39) worthy of speculation as a route to formation of "nonscaven g eable" nitrogen, since the singlet nitrene would not be reactive to phosphine except through a Lewis acid-base reaction. The much greater Lewis basisity of ammonia would not allow phosphine to compete with ammonia for the nitrene except in those mixtures containing relatively large concentrations of phosphine. In these mixtures, the production of nitrogen falls linearly toward zero, in pure phosphine. The charge transfer reaction (40) m.ay well occur in this system under certain conditions, since the ionization potentials of phosphine and ammonia are virtually identical 2 (10.2 ev). But unless the concentration of phosphine becomes comparable to ammonia this cannot be a significant reaction, in view of the very high efficiency of the reaction + (14) 61

PAGE 68

11 for which the rate constant has been estimated to be 3 x 10 moles per 28 liter per sec. It could. be that, in mixtures containing fairly large concentrations of phosphine, reaction (40), or one such as -+ + NH + PH 4 2 (41) which may be as efficient as reaction (14), becomes important. It is interesting to compare the product yields in this system, at a concentration of PH~ just sufficient to depress the nitrogen .:) yield to its limiting low value, with the yields in pure ammonia. Mixture number 6 ( see 'I'able 2) provides data in this region. It is pos sible to use these data and those from the irradiation of pure arr.monia as the basis for a calculation of phosphine decomposed relative to ammonia protected. 'rhe data follow: G(N 2 ) (Mixture) = 0.06 G(H 2 ) (Mixture)= 6.4 Since in the mixture G(H 2 ) (from NH 3 decomposition)= (55/53) 3 [ G(N 2 )(Mixture)] = 0.2, then G(H 2 ) (from PH 3 decomposition) = 6.2. The factor (55/53) takes into account the difference in ammonia pressure in the two systems. It follows that G(-PH 3 ), which is 2/3 of [G (H 2 )(from PH 3 )], is 4.1. The G(-PH 3 ) must be corrected for direct decomposition by radiation, since in this mixture approximately 6% of the dose is absorbed by phosphine. This correction is made by assuming that the rate of direct gamma ray decomposition of PH 3 in the mixture is the same as in irradiation of the pure substance: i.e., G(-PH 3 ) = 7.5. This amounts to 11% of the total PH 3 decor.1posed, and gives a G(-PH 3 ) from reaction with ammonia moieties of 3.7. What is desired 62

PAGE 69

is a co m p a rison of this yield with the yi e l d of protected ammonia, which is the ammonia not d e co mp o s in g i n the m i xture, due to the pres e nce of phosphine. Defining 6G(N 2 ) as the difference in nitrogen yield in pure ammonia and the mixture, 6G( N 2 ) = 1.52 (55/53)(0.0 6 ) = 1.4 6 This represents the nitro g en wh i ch was not for me d, du e to the presence of PH 3 The yield of protected a..'ll!llonia is related by a factor of two to L\G(N 2 ); i.e., G(NH 3 ) (protected) = 2 [6G (N 2 )] = 2.9. From these results it is clear that, in a mixture where the pho s phine concentration is just lar g e enough to be efficiently scavenging radical fra gm ents from a mm onia, less than four PH 3 molecul e s are decom posed in protectin g three ammonia molecules. This nearly one-to-one stoichio m etry can be explained by examining the initial sta g es of ammonia decomposition without and with added phosphine. In pure irradiated ammonia, the following reactions occur: (42) (J.3) (44) The rate constants for reactions (43) and (44) are know~ to be about 29 S 109 2 2 24 5 liters per mole per sec and about x liters per mole per sec. In pure irradiated ammonia, however, reaction (43) should predominate over reaction (44), since the hydrogen atom concentration is so low -10 (about 10 moles per liter) during radiolysis. With added phosphine, reaction (43) would be unable to compete with the reaction 63

PAGE 70

H + P H 3 + (25) for which the rate constant has not be en r eporte d. Because of the much weaker PH 2 -H bond, however, it can be estimated that the rate constant of this reaction is about 10 8 greater than that of reaction (43). Thus, via reactions (25) and + (33) two ammonias ar e protected, one directly and one by re-formation, and two phosphines are destroyed. The excess of phosphine (3.8 PH 3 : 3.0 NH 3 ) may be an indication of' the importance of reaction (44) in pure ammonia, since reaction ( 41,) does not involve an ammonia molecule except as a third body, and suppression of this reaction by reaction (25) would not protect NH 3 'l'his argument is based on the assumption that phosphine is not reacting significantly with ionic species, Methyl Iodide Vapor and PhosphineMe th y l Iodi d e Gas Mixtures This part of the present study had as its major purpose the verification of the radical scavenging ability of phosphine, rather than a detailed study of the radiolytic decomposition of CHi For this reason, analyses of products were confined to those noncondensible at -196c, namely, methane and hydrogen. The yields of these products were determined in the pure vapor and in mixtures containing phosphine. Because of the limited scope of this investigation of the pure vapor radiolysis, it is not possible to formulate a decomposition mechanism. In addition, the radiolysis of methyl iodide in the vapor phase does not appe a r to hav e been reported, althou g h photolytic studies h a ve 64

PAGE 71

21 30 31 been made by a number of workers. It has been shown 21 that the gas phase photolysis prod uces methane according to reaction (20) (page 8) involving a hot methyl radical, sinc e the yield is unaffected by the presence of efficient thermal methyl radical scavengers and is also temperature inde pendent It is likely that hot methyls are a major sourc e of methane in the gamma radiolysis of methyl iodide, although it may be that CH 4 is also formed by other routes, such as via an ion-molecule reaction. The dependence of the yield on methyl iodide pressure (see Figure 19) is unlikely to be due to a wall effect. Approximate calcul a tio n s of the number of collisions which a hot methyl undergoes with the substrate CH 3 I and with the wall indicate that over the pressure range involved there is not a significa.-rit change in the number of wall collisions relative to substrate collisions, especially in view of the high efficiency of reaction (20), If an ion-molecule reaction is a second source of CH 4 the pressure dependence of this reaction mi gh t explain the observed increase in G(CH 4 ) as the CH 3 I pressure is increased. Some of these reactions have been reported by Pot tie, Barker a.-rid Ha.mill. 32 A significant amount of hydrogen is formed. It is unlikely that reactions of thermal radicals lead to H 2 production in this system, since radical attack would occur more readily at the C-I bond. The present data is insufficient to suggest a mechanism, although the pressure dependence (see Figure 20) indicates that one possibility is molecular elimination from an excited molecule. Increasing pressure would cause more rapid deactivation and a reduced yield, as observed, The reaction of hot hydrogen atoms with methyl iodide could also lead

PAGE 72

to H 2 Finally, the work of Pottie, Barker and Hamil1 32 has shown that in the mass spectro me ter some H 2 may be formed by the reaction CH I+ e 3 (45) Since it is proba ble t hat th e major prinary process es in the radiolysis of methyl iodide vapor result in ru pture of the C-I bond and formation of methyl radicals, the addition of even low concentra tions of phosphine should result in an increased yield of methane, by the reaction (46) which is probably about 20 kcal exothermic. Phosphine wouJd not be a good scavenger of iodin e atoms, because the H-I bond is somewhat weaker than the PH 2 -H bond. The very sharp rise to a maximum yield of CH 4 indicates the efficiency with which phosphine is reacting with thermal methyls. Since virtually all (99%) of the energy is being absorbed by methyl iodide at the maximum of the curve ( sr::e Figures 21a and 21b) and the concentration of phosphine is only 3,4 mole percent, it is reasonable to suggest that the G for methane at this maximum is approximately the G for methyl radical production in the pure irradiated vapor. On this basis, a comparison of methane yields in the pure vapor and in the phosphine-methyl iodide system shows that about 85% of methyl radicals produced in the pure vapor during radiolysis do not react to produce methane. If hydrogen is not fonned by a process involving thermal radicals, the yield should not be affected by low concentrations of 66

PAGE 73

phosphine. Examination of Figure 21b shows that, out to 7 or 8 mole percent phosphine, hydrogen production is cons tant and equal to that in the pure vapor. At high e rphosph in e co ncentrati ons the hydrogen yield is due lar gely to the decomposition of phosphine by direct radi atio n effect, as in di c ated by the lin e arity of the yield with respect to energy absorption by phos p hine The s am e linearity is evident in the ammonia-phosphine system when hydrogen yields are plotted as a function of th e energy fraction absorbed by phosphine (see Figure 15). The rapid drop-off in methane production i mmediately beyond the maximum coincides with a large increase in hydrogen yield. This may be due to an energy transfer process from methyl iodide to phosphine It is probably not due to charge transfer, sir..ce the reaction CH I++ PH 3 3 -+ (47) is endother mic by 10-15 kcal, based on the ionization energies of methyl iodide and phosphine. 33 From 30 mole percent phosphine out to pure phosphine, the methane yield decreases approximately in propor tion to the energy fraction absorbed by phosphine. Inspection of Figure 21a shows that this is the same region in which the hydrogen yields sho w a proportionate incr ease with energy absorbed by phosphine. The data of Figure 21a can be thought of as resultin g from the interaction of at least three processes: (1) efficient scaven ging of methyl radicals by phosphine; (2) an energy transfer process from methyl iodide to pho~phine; and (3) direct radiation effects on both phosphine and methyl iodide. The scavenging process produces a large increase in methane yield at very low phosphine concentration. At slightly larger conc entrat ions of PH 3 the energy transfer process 67

PAGE 74

causes the cH 4 yield to drop rapidly and the H 2 yield to show a sharp increa se As the phosphine concentration becomes relatively large, both yields become functions only of the energy absorbed by phosphine. The energy transfer process apparently be co mes complete at f airly low phosphine concentration, so that it exerts approximately a constant effect at higher concentrations; i.e., the system is saturated with respect to energy transfer at approximately 30 mole percent phosphine, 68

PAGE 75

V. SUM M ARY AND CONCLUSIONS The investigations reported here may be summarized, alon g with the major conclusions drawn from them, as follows: 1. As expected, the deco m position of phosphine is more extensive than that of ammonia, when the two compounds are irradiated under comparable conditions. Some similarities which are notable are that both compounds break down to their elements, and neither forms a dimeric X 2 H 4 compound (where Xis nitrogen or phosphorus) to a significant extent in a static system. Under the experimental conditions of this study, neither appeais to have rate-determining third-body reaction steps. Although other workers ha v e shown that in a flow system ammonia radiolysis produces considerable quantities of hydrazine, it is less likely that diphosphine would be form e d under comparable conditions in phosphine radiolysis. A logical extension of the work reported here would include the flow photoly sis or flow radiolysis of phosphine. 2. The irradiation of phosphine-ammonia mixtures at room tempera ture does not produce significant amounts of any nitrogen-phosphorus compound~. The only products which are observed are hydrogen, phosphorus and nitrogen. Drastic depression of the nitrogen yield, even in mixtures containing very low concentrations of phosphine, leads to the conclusion that phosphine is protecting ammonia from radiolytic decomposition by scavenging free radicals, through 69

PAGE 76

donation of a hydro gen atom. In a mixture containin g less than 4 mole percent phos phine the decomposition of l ess than four molecules of phosph i ne res ulted in the prot ectio n of thre e molecules of ammonia (p e r 100 ev of ab s orbed energy). This stoichio me try is explained with a si mp le re a ction sche me invo lvin g the scav e n g ing of both hy dr o gen a to m s and a m in o r adical s. It is re asonabl e to suggest that phosphine can act as a good free radical sc aveng er via reactions involving donation of a hydro gen atom. Although the PH 2 -H dissociation energy is not known, it is probably about 20 kcal less than the CH 3 -H or NH 2 -H bond energies. In gen e ral, phosphine should be reactive to many carbon and nitro g en-containing radicals generated in radiolytic systems, as well as to a number of atomic radicals such as H and Cl. The only requir emen t would s e em to be that the scaven gi n g reaction should be exothermic. 3. The radiolysis of methyl iodide vapor r es ults in the formation of methane and hy dro g en, with G-valu e s of 2.3 and 0. 6 res pe ctively. It is probable that methane formation in the irradiated va p or parallels that in the photolyzed material; i.e., it proce e ds via the reaction of hot m ethyl radicals with t h e substrate. It may be that methane is also formed by other routes, includin g ion-molecule reactions. The mod e of hydrogen formation is speculativ e but may involve a molecular elimination or hot atom process. In the presence of phosphine, the m ethane yield is increased by a factor of six, due to scavenging of methyl radicals. This indicates that the primary methyl yield in irradiated methyl iodide is probably 12-13 (per 100 ev ab s orbed energy), since it is probable that most methane is formed by processes involving methyl radicals. 70

PAGE 77

APPENDIX Calculation of the Dosimetry In gas phase systems radiation dosimetry is more complex than in condensed systems. Absorption by matter of gamma radiation in the mev range occurs principally by three processes: Compton scatter ing, photoelectron ejection and pair production. With systems con taining atoms of moderately low atomic weight the most important process is Compton scattering, which involves the interaction of the gamma pho,ton with an orbital electron. A lower energy photon and an ejected electron result. Since the electron density of condensed phases is usually comparable to that of the irradiation vessel, with most radi ation geometries the bulk of the energy absorbed is via Compton scattering in the sample. It follows then that the absorbed dose as determined by a liquid dosimeter solution is related to the absorbed dose in the sample by the ratio of electron densities in the two systems. In most gaseous systems the absorption of energy from gamma rays by the medium is insignificant relative to that absorbed from electrons ejected from the cell walls and traversing the cell. For these electrons the energy absorption by the gas is proportional to the electron stopping power of the gas, rather than its electron density, and the absorbed dose in the sample and dosimeter are related by the ratio of their stopping powers. 'l'he stopping po;;er of 71

PAGE 78

a medium is defined as the ene r gy loss per unit path length of a particle passin g throu g h th e medium. The use of sto p pin g pow e rs fo r the calculation of dose rates is justified if the irradiation cell and its contents constitute a Bragg-Gray cavity, 34 35 36 Spiers 37 has outlined the basic conditions required, as follows: 1. The cavity must have dimensions such that only a very small portion of the electrons ejected from the walls have ranges less than the distance they must traverse through the cell medium. 2. A negligible amount of energy absorption must occur from the radiation interacting directly with the gas. 3. The cavity must be surrounded by a sufficient thickness of solid material so that all of the electrons traversing the cavity originate in the solid material. Although theoretically this means that the thickness must be equal to or greater than the range of the electrons in the solid, in practice a considerably thinner wall is sufficient, due to scattering in the solid. 4. The energy dissipation in the solid medium must be reasonably uniform all around the cavity. Ordinarily, this impli ~ s that the cavity should be placed far enough from the radiation source that the divergence of the field is slight. The cell geom e try used in most of the irradiations (see Figure li) can be shown to satisfy the first three requirements. The average energy of Compton electro n s produced by cobalt-60 gamma rays has been shown to be 0.587 mev. 38 According to Nelms, 39 the range of 0. 587 mev electrons in ethylene is O. 2o r grams per cm 2 (by an interpolation of the values given for 0.55 and 0.60 mev electrons). At th e 72

PAGE 79

ethylene pressure used in the dosimetry (60 cm), the range was then about 2.3 meters. The cell diam e ter was 15 cm. The mass densities of glass and ethylene at 60 cm pressure are about 2.2 x 10 3 milligrams per cc and 0.9 milligrams per cc, respectively. The efficiency with which these two materials attenuate a beam of gamma radiation can be approximately compared by taking the product of their mass densities and the path length of the radiation, since the molecular electron densities of glass and ethyle n e are roughly comparable. The path length in the glass is 4 mm and in the ethylene about 15 cm maximum. Thus the relative absorption efficiencies are at least 75 and one in this system, satisfying the second require ment. A glass wall thickness of 1 mm or more ensures electronic equilibrium since the range of the Compton electrons generated in glass by cobalt-60 gammas is roughly 1 nun. The walls of the cells used in this work were 2 mm thick. The fourth requirement is a good deal more difficult to analyze with the cell geometry used. For the determination of compara tive rather than absolute dose rates, however, this should not be important. Bethe 40 41 derived an expression for the calculation of stopping power, shown in the following equation: 73

PAGE 80

dE dx = 4 2nNe Z m J?. 0 l 2 m u E l n -21-~(-1 -82 ) 2 (211 -8 ~ 1 + 8 ) ln 2 ergs per cm (1) where N is the number of mol e cules per cubic c e nti me ter, e is th e charge on an electron in electr o static units, m is the re s t m ass of 0 an elect r on in gram s u is the velocity of the e l e ctron in c m per seco n d, Z is the nu mbe r of el e ctrons per mol e cule, 8 is u/c (where c is th e velocity of light), I is the mean excitati o n potential for the molecules of the stopping material in ergs and Eis t he kinetic energy of the electron in er g s. Insofar as dosimetry calculations are concerned t h e stopping power ratio of the sample gas an d t h e dosim e ter gas, rather than the absolute values of the stop pi n g powers themselves, are nee d ed. This facilitates the calculations, as sho w n below: where k dE dx sample dE dx dosimeter m u 2 E = ln -0 -2(1-82) + 1/8 (1 N Z k-ln I 2 s s s = NdZd k-ln I 2 d (2) A value for 8 of 0.8765 for a 0.587 mev electron was found by interpola tion of values provided by John s and Lau g hlin. 42 The mean excitation potentials for the atoms C and H were taken directly from Spiers 37 74

PAGE 81

and a value for P was found by a li near interpolation of the values for Al and Cl. An average exci t a tion pot e n tial for co mpounds ca n be calcul ated, assuming simple additivity according t o the relation: ln I= rn.z ln r. i 1 1 1 rn.z i 1 1 (3) where I is the potential for the compound and the i subscripts indi42 cate values for the atoms Calculated in this way I values for ethylene and phosphine were found to be 51 and 112 ev, respectively. Sub sti tution o f these v alues (i n the pro pe r units) into equation (2) and solv ing l ead s to t he r elation : = 1.041. (4) where pressures have been substituted for number densities. Similar calc ulati o ns lead to the follo wi n g analogous equations: SNH p 3 0.661 NH 3 = ( 5) s C2H4 p C2H4 SC H I p 3 CH 3 I = 3,185 (6) s C2H4 p C2H4 Equations (4), (5) and (6) were used in the calculation of d os e rates. The use of ethylene as a dosimeter gas ha s been reco m mended 75

PAGE 82

43 by Yang and Gant, who showed that the hydrogen yield from radiolysis was highly r e producible without extensive purification of the ethylene. Figure 9 ( page 28} shows t hat t he rate of h yd rogen produ.ct ion is constant with radiolysis time. The slope of the graph was O .1119 moles per hour. Assuming that the G-value of hydro ge n is 1.2 molecules 44 45 per 100 ev, which appears to be the best value, the absorbed dose rate was 3.50 x 10 17 ev per minute in the cell. The volum e of the cell was calibrated with distilled water as 1,027 5 ml and the ethylene pressure was 60 cm. The dose rate was thus found to be 5,68 x 10 14 ev per mm of ethylene pressure per liter of cell volume per minute. Although several cells were used in subsequent work, their geometries and volumes were very similar. For example, the three cells used for the ammonia-phosphine mixtures had volumes of 1,035, 1,027 and 1,020 ml. The error involved in using these different cells, insofar as dose rate differences were concerned, was probably negligible. 76

PAGE 83

BIBLIOGRAPHY 1. P.A. Sellers, T. R. Sato and H. H. Strain, J. Inorg. an d Nuc. Chem i, 31 (1957). 2. Y. Wada and R. W. Kiser, Inor g Chem .1, 174 (1 96 4). 3. 4. 5, 6. 7. H. P. M. H. H. Neuert and H. Clasen, Z. Naturf orsch. 7a, 410 (1952). E. Saalfeld and H.J. Svec, Inor g Chem g_, 50 (1 96 3). Halinann and I. Platzner, J. Phys. Chem. 71, 4522 (1967). w. Melville, Proc. Roy. Soc. Al38, 374 (1 932). w. Melville, Proc. Roy. Soc. Al39, 541 (1933), 8. H. W. Melville and J. L. Bolland, Proc. Roy. Soc. ~160, 384 (1936). 9, H. W. Melville, J, L. Bolland and H. L. Roxburgh, Proc. Roy. Soc. Al60, 406 (193 6 ). 10. R. G. W. Norrish and G. A. Oldershaw, Proc. Roy. Soc. A262, 1 (1961). 11. D. A. Ramsay, N atu r e 178, 374 (1956). 12. J. R. McNesby, I. Tanaka and H. Okabe, J. Chem. Phys. 36, 605 (1962). 13. C. E. Melton, J. Chem. Phys.~, 4414 (1966). 14. Y. Toi, D. B. Peterson and M. Burton, Rad. Res. 17, 399 (1962). 15, G. M. Meaburn and S. Gordon, J. Phys. Chem. 72, 1592 (1968). 16. F. T. Jones and T. J. Sworski, Trans. Farad. Soc. 63, 2411 (1967). 17. F. T. Jones, T. J. Sworski and J.M. Willia.ms, Trans. Farad. Soc. 63, 2426 (1967). 18. M. Nishikawa, N Shinohara and N. Matsuura, Bull. Ch e m. Soc. Jap. 40, 1993 (1967). 19. H. A. Gillis, R.R. Williams and W. H. Hamill, J. Am. Chem. Soc. 83, 17 (1961). 77

PAGE 84

20. R. C. Petry and R.H. Schuler, J. Am. Che m Soc. 12.., 3796 (1953). 21. R. C. Souffie, R. R. Williams and W. H. H amill, J. Am. Chem. Soc. 78, 917 (1956). 22. R. J. Hanrahan, Intern. J. Appl. R a diation Isotopes 13, 234 (1962). 23. T. L. Cot t rell, The Str e n g ths of Che m ic a l BC?_nds, 2nd ed., Academic Press, Inc., New York (1958). 24. C. Kretsc hm er and H. L. Peterson, J. Che m Phys. 39, 1772 (1963). 25. W. Jost, Z. Phys Che m B3, 95 (1929). 26. J. R. Van Wazer, Phospho r us and Its Co mn oun d s, Vol. 1, Interscience Publis her s, Inc., N ew York (195 8 ). 27. R. C. Schiek, Ph.D. Dissertation, Univ. of New Hampshire (1963; 28. L. M. Dorfman and P. C. Noble, J. Phys. Chem. 63, 980 (1959). 29. M. Schiavello and G. Volpi, J. Chem. Phys. 37, 1510 (1962). 30. R. D. Doepker and P. Ausloos, J. Chem. Phys.~' 1865 (1964). 31. R. D. Schultz and H. A. Taylor, J. Chem. Phys. 18, 194 (1950). 32. R. Pottie, R. Barker and W. H. Hamill, Rad. Res. 10, 644 (1959). 33. R, W. Kiser, Introductio n of M ass Spectro m etry, Prentice-Hall, Inc., New Jersey (1965). 34. W. H. Bragg, Studies in Radioactiv:L.!Y-, Macmillan, New York (1912). 35. L. H. Gray, Proc. Roy. Soc. Al56, 578 (1936). 36. L. H. Gray, Brit. J. Radiol. 10, 600, 721 (1937). 37, F. W. Spiers in R a diation Dosi m etry, eds. G. J. Hine and G. L. Brownell, Aca d emic Press, Inc., N ew York (1956), Chapter 1. 38. R. J. Hanrahan, Ph.D. Dissertation, Univ. of Wisconsin (1957). 39. A. T. Nelms, Nat. Bur. St. (U.S.) Circ. 577 (1956). 40. H. A. Bethe, Handbuch der Physik 24, 273 (1933). 41. H. A. Bethe and J. Ashkin in Exoerimental Nuclear Physics, ed. E. Segre; Vol. 1, John Hiley and Sons, New York (1953). 78

PAGE 85

42. H. E. Johns and J. S. Laughlin in Radiation Dosimetrv eds. G. J. Hine and G. L. Brownell, Acade m ic Press, Inc., New York (1956), Chapter 1. 43. K. Yang and P. L. Gant, J. Phys. Chem .L, 1861 (19 61 ). 44. M. C. Sauer, Jr. and L. 1L Dorf man Abstracts 2 137th Meeting Am. Chem. Soc April 1, 19 60 45. F. W. Lampe, Rad. Res. 10, 691 (1 959 ). 79

PAGE 86

BIOG R AP H ICAL S KET C H James Wesley Buchanan was born in Rutherfordton County, North Carolina, on May 5, 1937, and received hi s early ed u cation in several wester n N or t h Carolina towns. He gr a du a ted fr om Claremont Central High School of Hickory, North Carolina, in 1955, He studied at the University of North Carolina in Chapel Hill, where he -was a Moreh e ad Scholar, and received the Bachelo r of Arts degree in chemistry in 1959, His first gradu a te degree was awarded in 1962, when he re ceived the Master of Science in chemistry. After several years of teachi ng che m istry at the undergraduate level, he returned to the University of Florida to complete his Ph.D. He has held graduate teaching and research assistantships, and during the last two years of study was an NDEA Title IV Fellow. He is a member of the American Chemical Society and Alpha Chi Sigma chemical fraternity. 80

PAGE 87

This diss ertation was prepared und er the direction of the chairman of the candidate's supervisory co mmi tte e and has b ee n approv e d by all m em ber s of that co mm itt ee It was submitted to the Dean of the College of Arts and Scie n ces and to the Gr ad uate Council, and was approved as partial fulfillment of the require m ents for the degree of Doctor of Philosophy. A ug ust, 196 8 Dean, College of Arts and Sciences Dean, Graduate School