Oct., 1961
MECHANISM OF
THERMAL ~ ~ i . : c O M P O S I T I OOF N
many colleagues in this Laboratory for discussions regarding this work and for providing samples. Dr. R. E. Merrifield was particularly helpful in discussing the origin of the dimerization of cyanocarbon anions. Dr. D. W. Wiley was especially helpful in synthesizing a number of the compounds studied, and this assistance is very gratefully
l'LUTONIUM
HEX.IFLUOKIDE
1843
acknowledged. The author is indebted to Mr. D. W. Marquardt and Miss A. Sheldon of the Engineering Research Laboratory, du Pont Experimental Station, who very skillfully carried out the computer calculations. He is also grateful to Professor Fuoss and Dr. Lind for making their manuscript available to us prior to publication.
THE KINE'TICS AND MECHANISM OF THE THERMAL DECOMPOSITION OF PLUTONIUM HEXAFLUORIDE' BY J. FISCHER, L. TREVORROW AND W. SHINN Chemical Engineering Dzvisim, Argonne National Laboratory, Argonne, Illinozs Receioed March 50. 1961
The kinetics and mechanism of the decomposition of plutonium hexafluoride have been investigated, and the rate of decomposition has been formulated as a concurrent first and zero-order reaction with respect to plutonium hexafluoride p r e s sure. It has been inferred that the decomposition, in the temperature range studied, proceeds by both a homogeneous and heterogeneous unimolecular decomposition. The heterogeneous decomposition occurs on the surface of plutonium tetrafluoride.
Introduction Plutonium hexafluoride undergoes thermal decomposition to plutonium tetrafluoride and fluorine. Both the stoichiometry and the equilibria involved in the decomposition of plutonium hexafluoride have been investigated previously. In the present work, the rate of thermal decomposition of plutonium hexafluoride vapor has been studied by a static method at initial pressures of 14 to 100 em. and at temperatures of 140, 161 and 173'. Experimental The physical and chemical properties of plutonium hexafluoride are such that special apparatus and techniques must be employed in experiment.ation with this compound. The triple point of plut,onium hexafluoride is 51.59 O , 533 .O mm ., and the boiling point is 62.16°.a Plutonium hexafluoride is a very strong fluorinating agent that reacts vigorously with many chemicals and must be handled in closed systems under anhydrous conditions. The radioactive decay of plutonium present,s some very serious hazards to the laborat,ory worker. Decay of plutonium-239, the most prevalent isotope, is accompanied by the emission of highly energetic a-particles. The 239 isotope has a high specific activity and a long biological half-life. It is extremely toxic. The maximum permissible body burden has been set a t lo-' g. There is an additional radiation hazard associated with the fluorides of plutonium-239. The a-particles emitted by thr plutonium isotope react with the fluorine nucleus to produce fast neutrons. This neutron hazard must be seriously considered by the experimenter when dealing with plutonium fluoride samples weighing more than 10 g. It is clear that special precautions in the way of equipment and procedure are necessary to protect the laboratory worker from ingestion of this isotope. The study of plutonium hexafluoride chemistry in our laboratories has been carried out in metal systems contained in glove boxes. Materials.-Commercial fluorine of high purity was used in preparing plutonium hexafluoride. The plutonium hexafluoride used in the decomposition rate studies was prepared by reacting fluorine, a t 500 to 550", with plutonium dioxide or plutonium tetrafluoride, obtained from ,4EC sources. Fluorine, a t a pressure of about one atmosphere, was circulated through the preparation equipment by means (1) This work was performed under the auspices of the U. S. Atomic Energy Commiseion. (2, L. E. Trevorrow, W. .4. Shinn and R. K. Steunenberg, J . Phgs. Chem.. 66, 398 (1961). (3) B. Weinstock, E. E. Weaver and J. G . Malm. J . Inore. d N u d e a r Chem., 11, 104 (1959).
of a magnetic piston pump. The fluorine flowed through a nickel preheater at 500'. Hot fluorine flowed from the preheater through a tubular nickel reaction furnace in which it passed across t.he surface of powdered plutonium tetrafluoride or plutonium dioxide contained in a nickel boat. This resulted in a mixture of plutonium hexafluoride, fluorine and oxygen, when the oxide was used, from which the plutonium hexafluoride was condensed in nickel traps cooled wit,h Dry Ice. Volatile impurities were removed from the plutonium hexafluoride by cyclic sublimation, condensation and vacuum distillation. Equipment.-All parts of the equipment, exposed to fluorine and plutonium compounds, were constructed with nickrl and Monel. Component parts of the apparatus were connect.ed by manifold systems to a high vacuum system consisting of mechanical and oil diffusion pumps, to a helium supply, to a fluorine supply, and to pressure measuring devices such as Bourdon gages and diaphragm pressure transmitters. Nickel valves, flare fittings, silver soldered and welded joints served to connect the various comp0nent.s of the apparatus. All of the equipment was pretreated with fluorine before use. Experimental Procedure.-There is no change in the number of gas molecules during the thermal decomposition of plutonium hexafluoride vapor according to equation 1
+
PuFdg) --+ fiF,(s) Fz(g) (1) Therefore, it is not possible t o follow the rate of decomposition of the vapor by pressure measurements in a static system. The deposition of plutonium tetrafluoride would interfere with measurement of the partial pressure of plntonium hexafluoride by many of the common physical methods. Thns a number of physical measurements which might be used to follow the reaction continuously are rendered inapplicable. Rates of decomposition were obtained from initial and final compositions after heating a sample of plutonium hexafluoride for a given period. Prior to each experiment t h r storage vessel containing the plutonium hexafluoride was evacuated to a pressure of about 2 X lo-* mm. at -196" to remove fluorine accumulated from radiation decomposition. The plutonium hexafluoride then was transferred by vacuum distillation to a supply vessel where i t was condensed, and any rcmaining fluorine was removed by evacuation at -196O. The 50-ml. supply vessel was ronnected by a manifold to a 50-ml. decomposition vessel which had been weighed previously. The supply vessel and associated lines were heated to about 70 to SO", at which temperat,ure the rate of decomposition was extremely low, in order to aid the transfer of adequate quantities of plutonium hexafluoride. The vessel was preheated to the experimental temperature in a thermostated aluminum block wound with Nichrome wire. The t,emperaturr, of the decompoPition
1844
J. FISCHER, L. TREVORROW AND W. SHINN
vessel was held constant to f 0.2". An experiment was initiated by opening the valve to the evacuated decomposition vessel, and allowing the warm plutonium hexafluoride to expand into it. The approximate amount of plutonium hexafluoride transferred to the reaction vessel waa controlled by P V T measurements. When the expansion waa complete, the valve on the reaction vessel was closed, and the decomposition of plutonium hexafluoride waa allowed to proceed, isothermally, for a measured time interval. A t the end of the experiment the furnace wae lowered, and the decomposition vessel was quenched in liquid nitrogen. The vessel wae cooled rapidly to approximately 100" in one minute and to room temperature in two minutes. Thus, the error introduced by decomposition during the quenching period was small since the quenching period was short. The rate of fluorination of plutonium t,etrafluoride has been shown to be rela.tively slow at temperatures below 170°.' Thus, during the quenching period the composition of the gas mixture was not sufficiently altered by changes in the rates of either the forward or the reverse reaction to cause a significant error in the results. After quenching, the decomposition vessel waa warmed to room temperature, and weighed to determine the starting amount of plutonium hexafluoride. The vessel then W M evacuated a t 25' to remove plutonium hexafluoride and the fluorine formed during the reaction. The vessel ww weighed again to find the weight of plutonium tetrafluoride formed. The weights were used to calculate the amount of plutonium hexafluoride whi.ch had decomposed. Partial pressures of plutonium hexafluoride were calculated from the weights using the ideal gas law. A single reaction vessel wm used for several consecutive experiments. Therefore, each experiment was carried out in the veRsel containing plutonium tetrafluoride which had accumulated from previous experiments. A small correction waa made of the volume of the vessel to ticcount for the accumulation of plutonium tetrafluoride.
Results and Discussion The values of the equilibrium constants for (:quation l are quite small.2 At 140" the equilibrium a t 170" the constant, (PuE'e)/(Fz), is 3.18 X equilibrium constant, is 5.26 X From consideration of the equilibrium constants and the extent of displac'ernent from equilihrium in these experiments, it it; concluded that the rate of fluorination is negligib1.ecompared to the rate of decomposition of plutonium hexafluoride. In the interpretation of the result,s on the rate of decomposition of plutonium hexafluoride, the effect of the reverse reaction was considered to be unimportant,. The r:it,e of decomposition of plutonium hexafluoride was studied by the st,at,ic met,hod in two types of vessels a t 161". The volumes of the vessels were almost equal, but the ratios of surface to volume were different. In one casc the vessel \vas packed wi1;h nickel wool t,o increase the specific area. For the non-packed vrsscl, the surface t,o volume ratio was 1.8 cm.-' and the volume was 5 2 ml. For -the packed vessel, the surface t,o volume ratio w a y 14 cm.-' and the vohime was 51 ml. Thr surfncr tCJ volume ratios are valiics bawd on geometric me:Lsuremeiits. The first few itxperinionts in 110th vessels yielded leery low dcc80niposit,ionrates. The rat,es iiicrcascd rapidly in thr first few expcriments for the non1. and t,hcn remained almostJ coust a n t i I I subsequcrit experinienls. In the packed vessel t t w rates inrrcascd to R value murh grcnter than t how in t,hc :.ion-pack(:d vwscl containing atmiit the sami. a,r::oiint of plutoiiium t,etrnfluoridr. T n ( 4 ) hl. J Stt,.,xlIcr. D. V. Sa.and Enu.. 6 933 I iUSS!.
Pteidl and R. li. Steunenberg, iVuclear
Vol. 65
the packed vessel the plutonium tetrafluoride probably was deposited over a greater area on the surface and in the interstices of the nickel wool. It was concluded that the rate of decomposition of plutonium hexafluoride is dependent upon the surface area of plutonium tetrafluoride. One of the reaction vessels was cut open and the plutonium tetrafluoride was found as a coating on the walls of the vessel and as pieces on the bottom of the vessel. The distribution of the plutonium tetrafluoride was not uniform. After the deposition of a small quantity of plutonium tetrafluoride, in the non-packed vessel, the rates appeared to be independent of the quantity of plutonium tetrafluoride in the vessel. This was true when the quantity of plutonium tetrafluoride in the vessel varied from about 1 to 3 grams. Because of the manner in which plutonium tetrafluoride accumulated in the vessel, the active surface area of plutonium tetrafluoride did not increase linearly with the weight of plutonium tetrafluoride. It is assumed that for this reason the active solid surface area did not increase significantly throughout the set of experiments in the non-packed vessel reported here. Thus it was possible to obtain a set of consistent results even though the weight of PuF4 in the reaction ~ e s ~ e l was different for each experiment in the set. The results of the rate experiments at 140, 161 and 173' in the non-packed vessel are showii in Table I. The rate was found to be dependent on the pressurc of plutonium hexafluoride in the non-packed vesqel. The experimental results can be expressed by a rate equation which is of concurrent first and zero orders with respect to plutonium hexafluoride pressure -dp/dt
=
ko
+ Xlp
(2)
The experimental procedure used in this work yields integrated rates of decomposition ; therefore, the integrated form of the rate equation was used to correlate the data obtained. The integrated form of equation 2 is
+
p = e - k l t po e - k l t (ko/k,) - k o / k l where po = initial pressure p = partial pressure after reaction time, t
13)
Each cxperlment yielded values of p and pc If the reaction times are held constant for a set of experiments, then equation 3 shows that p is a linear function of po. Therefore, most of the expcriments a t a given temperature were carried out for the sane reaction time but a t various initial pressures. The use of constant reaction times permitted the values of k" and k , to be calculated convrnimtly. Thc \-alucs of k , and k" for equation 3 uere crleulatcd ironi the data of Table I by using the method of least squares. The rste constants obtained in this inaniier are listed in Table 11. The Arrlicnius eqilatioll and the method of least squarcs were used to correlate the rate constauts with temperature. The equations obtained are 3 . m x 103 Ing ko -= "121 - - -o - ,, 7'1 It.
:Uld
4)
MECHANISM OF THERMAL DECOMPOSITION OF PLCTOS~UM HEXAFLUORIDE
Oct.. 1961
1845
The experimental activation energies are 15.9 i TABLE I T ~ A T E O F ' 3 E c o m c x m w s OF PLUTONIUM ~ I E X A F L U O R ~ D1.5 E : kcal./mole for the zero-order reaction, arid 19.6 f 0.7 kcal./mole for the first-order reaction. VAPOR A few experiments were carried out a t 150.3'. Static technique; initial volume of reaction vessel. 32.0 ml.: The rate constants a t
[email protected]', calculated from equanoii-packed reaction vessel. Reaction time (111 in
.
Initial
Obsd. final
pressurr:
press.
PllF6 (Clll.
101.8 99.5 116.7 76.3 70 . 0 64.0 56.4 48.5 41.4 I4 0
120 120 120 120 120 120 120 120 120 120
98.1 59.6 51.2 32.6 16.8
90
90 90 90 90
108.7 92.5 75.5 75.4 68.9 63.7 53.0 44.9 31.2 25.7 22.2
60 60 60 60 60 60 60 60 60 60 GO
102.5 49.8 14.3 14.0
120 I20 90 90
PUFB (cm.)
Ohsd. integral rate (cm./hr.)
Temp. 140.1' 86.8 i.5 85.6 7.0 73.6 6.6 62.2 7.1 50.1 7.0 52.2 5.9 45.7 5.4 37.4 5.6 30.0 5.1 7.1 3.4
Temp. 160.6" 70.4 18.5 37.9 14.5 31.5 13.1 10.5 10.7 5.9 7.3
Temp. 173.1' 74.2 34.5 59.9 32.9 43.8 31.7 46.3 29.1 39.5 29.1 35.4 28.3 27.0 26.0 20.0 24.9 13.1 18.1 8.3 17.4 6.0 16.2
Temp. 150.3' 77.9 12.3 31.9 8.9 6.0 5.5 6.1 5.3
Cnlcd .a find press.
P 11 F6
(cm.1
Dev. (cm.)
+o.
86.9
-
84.8
t,ions 4 and 5 , were used to calculate the final pressures a t 150.3' using the initial pressures. The observed and calculated final pressures were in agreement as shown in Table I. The rate constant for the zero-order reaction is dependent on the surface tto volume ratio. Since measurements of absolute surface areas were not obtained in the present work, the values for ko derived from the experimental results must be considered to hold true only for the particular surface to volume ratio of the reaction vessel used in these experiments. The results of experiments in the packed reaction vessel are listed in Table 111. The rates obtained in the packed vessel a t 161' increased with the weight of plutonium tetraflyoride in the vessel. The rate constant, kl for the homogeneous reaction, derived from experiments in the non-packed vessel a t l G l ' , was assumed to hold for the rat>esin both the packed and non-packed vessels. The value of k , then was used together with the results obtained in the packed vessel and equation 3 to calculate values of ko for the rates in the packed vessel. The values of Ic0 thus calculated were not constant, but increased in order of the average weight of plutonium tetrafluoride present during an experiment. This relationship can be seen from the data in the last two columns of Table 111.
1
.x
73.1 63.6 55.9 52.3 45.3 38.1 31.3 6.4
.5 +1.4 -0.2
Av. dev.
10.,1
+ .L - .'i + .7 +
.4
- .7 --
69.2 38.6 31.9 17.1 4.5
-1.2 $0.7 +0.4 +O.G -1.4
-
Av. dev.
*0.9
71.7 59.3 45.9 45.8 40.7 36.7 28.4 22.1 11.4 7.2 4.5
-2.5 -0.6 +2.1 -0.5 +0.9 +1.3 +1.4 +2.l -1.7 -1.1 -1.5
Av. dev.
f1.4
78.9 33.0 5.4 5.1
$1 .o +1.1 -0.6 -1.0
TABLE I11 RATEOF DECOMPOSITION OF PLUTONIUM HEXAFLUORIDE VAPOR
Static technique: initial volume of vessel, 51.1 ml.; vessel packed with nickel wool; temperature, 160.6'. Reaction time (min.)
___
--
10.9 Calculated from the equation p = poe-'If + e - k i t (kn/ki) - k o / k : iising: log ko = 7.124 - (3.469 X IO3)/ i " ( " k . ) l o g $1 = 7.260 - (4.292 X 103)/T("Iirnut,es3-j and two experimentally determined values. , i This papctr 'describes the determinat,ion of the heat evolved from t,he combustion of weighed samples of XC,, where 5 varied from 0.489 to 0.984. l'he m'pthod, using a bomb calorimeter a t a known initial pressure of oxygen, has been deThc energy equivalent. of the calorimetcr was 9988.0 f 3.3 j . / O as determined by the combustion of standard bcnzoic acid. Niobium Carbides.--In Table I are listed the aniLlyscs of t#he niobium carbides, which were prepared in this 1,nboratory. l'he method of (1) This work wit?. done linrler the auspices of the htoniic Energy Commission. ( 2 ) E. K. Storiiis and 5 . 11. Krikorian, J . I'hyls. Chem.. 64. 1 4 7 1 iiwn). (3) 0. €1. Krikorian, " l l i d i Temperatiire Sttidies.
P a r t 11. Therof the CarhirlPs," U n i r . of California Radiation CRL-2888, April, 19.55. ( 4 ) C . V. Saniiionov, .I. Phys. Chem. ( U . S . S . R . ) , 30, 2057 (1956). 65) B. F. Ormont, ibid., 33, 1 4 5 5 11959). (fi) .4. D. M a h and B. J. Royle. J . A n . Chem. SOC., 77, 6512 (1955). ( 7 ) I'. G. Kusenko and P. V. Gel'd, Izcsst. S i b i r . Oldel. A k a d . .Vauk S.S.S.R.. 2, 46 (1960). (81 E. .J. I l u h ~ r .Jr., , C. 0. Matthrws and C . E. Holley, J r . . J . A m . C h e m . S o c . , 77, 6193 (195,:).
preparation has been describedg and the niobiumniobium carbide system has been discussed.2 Yiobium metal also is included. TABLE I ANALYSESOF NIOBIUMCARBIDESA N D METAL,WEIQHTyo z in NbCz
c,
c,
N 0 H Fe 0,094 0.430 0,004 0.005 0.489 93.86 .. ,081 ,180 ,004 ,03 .500 93 84 .. ,100 .06 . . ... ,686 91.91 8.08 .. , . . ... . . . .01 ,699 91.74 8.25 .. ... ... ... .O1 ,786 90 68 9.20 .. ... ,022 .01 861 89.98 10.01 .. .005 ,002 .. ... ,889 89.64 10 24 .. ,021 ,024 ,004 . ,935 89 18 10 72 .. ,005 ,046 ,001 ,005 ,979 89.56 10.33 . ,021 ,024 ,047 ... ,984 87.05 12.92 1.87 ,024 ... ... . . Nb
Metal 99.45
total
0.016 5.81 6 00
free
..
,
,
The niobium metal powder used was -325 mesh which probably accounts for tmherelatively large oxygen impurity. Metal powder from the same bat'ch was used to make the carbides. The oxygen in each of the samples was determined by a diffusion-extraction method. lo The free carbon material was determined chemically in the NbCo,984 and also was seen in the X-ray pattern. T'anadium and zirconium were detected in two of the samples in very small amounts. At the temperature of preparation of the carbides, the iron impurity in the niobium should have been removed by volatilization. It is therefore assumed that the iron impurity in the carbides was introduced as metallic iron when the carbides were ground to a powder for the combustion experiments. Combustion of the Niobium Carbides.-Each sample, including the niobium metal, was burned in oxygen a t 25 atm. pressure on sintered discs of P-Nb205. Ignition was made by passing an electrical current through a 10 mil diameter niobium fuse wire. Although each material was in a finely divided form, no weight increase was found upon exposure t80 oxygen for one hour. The average (9) E. K . Storms and N. €1. Krikorian, J . Phva. Chem., 63, 1747 (1959). (10) W. R . Hansen and
(1957).
RI. W. Mallett, Anal. Chem.,
29, 1868