J O U R N A L OF THE AMERICAN CHEMICAL SOCIETY (Registered in I;. S . P a t e n t OfEce)
VOLUME
(‘a
Copyright, 1960, by t h e American Chemical Society)
JUNE 16, 1960
88
N U M B E11 R
PHYSICAL, A N D INORGANIC CHEMISTRY [COXTRIBalIoS S O .
15(%,STERLISG CHEMISTRY LABORATORY, \’ALE
c S I V E R S I T I ’ , NE\V
HAVES,C O X N E C T I C L T ]
The Effect of Moderators on the Reactions of Hot Hydrogen Atoms with Methane BY PEDER J. ESTRUP’ AND RICHARD WOLFG.WG RECEIVED JULY 23, 1959 The reaction of recoil tritium with methane has been examined in further detail. The previous hypothesis that 1 his system involves a hot displacement reaction of high kinetic energy hydrogen to give CH3T,CH?T.and HT is confirmed. The effect of moderator on this process is studied b y the addition of noble gases. -1s pxedicted these gases inhibit the hot reaction, their efficiency in this respect being H e > Ne > A > S e . The data are quantitatively in accord with a theor>-of hot atom kinetics. T h e mechanism of the hot displacement process is briefly discussed.
Introduction T h e reactions of gaseous methane with high kinetic energy hydrogen atoms-produced in the form of tritium (T) b y the nuclear process He3(n,p)Thas been the subject of several recent s t u d i e s - j This Laboratory has postulated t h a t the majority of the tritium atoms by virtue of their high energy undergo “hot” displacement reactions to form H T , C H B Tand CH2T.. T h e remainder, about 30%, of the tritium loses energy in collisions without combination and then undergoes relatively slow thermal reactions. These reactions are mainly abstraction of H from CHd to form H T , and combination with radiation produced species to form higher tritiated hydrocarbons. In support of this hypothesis the following evidence was cited : (1) The formation of hot products is unaffected by the addition of small concentrations of radical (and ion) scavengers such as I? while the thermal products are almost eliminated. Insensitivity of hot processes to minor concentrations of scavenger is expected since, if they occur a t all, they must do so in a relatively few collisions before losing their excess energy. ( 2 ) T h e hot products appeared to be insensitive (1) C E R N , Geneva, Switzerland. Work performed in partial iulfillment of t h e requirements for t h e P h . D . degree a t Yale University. (2) hl. Amr. El-Sayed, P. J. E s t r u p and R . Wolfgang, J . P h y s . Chem., 62, 1356 (1958). (3) M. Amr. El-Sayed a n d R . Wolfgang, THISJ O U R N A L , 79, 3280 (1957). (4) A. Gordus, RI. Sauer a n d J . Willard, i b i d . , 79, 3284 (19.77). ( 5 ) hl. Amr. El-Sayed, Ph.11. lliesis, Florida S t a t e University (1859).
to temFerature, in accord with the expectation that the required energy is supplied by the hot atoni. (8) In a preliminary finding it was reported t h a t the yields of hot products were sharply reduced by the presence of excess helium, a substance acting as a moderator in the sense of removing the excess kinetic energy of the hot atom. The alternative suggestion t h a t the reaction of recoil tritium actually involved T+ rather than atoms did not conform to the results and can also be rejected on the basis of independent evidence. T h e present paper reports on a more detailed study of the hot reactions occurring in this system; a study made in order to confirm and add detail to the reaction model postulated and to provide quantitative data t o test the theory of hot atomic kinetics set forth in the accompanying paper.6 The effect of the addition of various amounts of He, Ne, Ar and X e as progressively less efficient nioderators has been investigated. Further information has also been obtained on the effect of scavenger and temperature and on the fate of thermalized hot atoms. Experimental The experimental procedure has previously been dcscribed . 2 Samples containing He3, methane and various additives were irradiated in the Brookhaven reactor and the resulting products subsequently analyzed by gas chromatography. The irradiation time was short, usually a few minutes, with a neutron flux of 1012-1013 n./cm.* sec. A number of samples were cooled during irradiation. Here the quartz ampoules were placed in polystyrene containers, which were theu filled with an appropriate organic liquid. ( 6 ) P.
1. E s t r u p and R Wolfgang, THISJ O G R X A L , 82, 2iiti.i (1900).
The assembly was frozen mid injected into the pile when the compound began t o re-melt. T h e melting process kept the samples a t approximately constant temperature for the s h i r t irradiation time. In this manner t h e temperature of samples could be varied from - 100” t o room (pile) temperature. After irradiation the samples were, in general, stored in Dry Ice. Effect of Scavengers.-Iodine, bromine and nitric oxide were used as scavengers. Excess of the halogens were added to the samples (in form of a crystal or a droplet) and their \-itpor pressure determined by control of the temperature during irradiation. The results are shown in Table I , in which the yields of the \aricriis products are given as function of the pressure of scavenger P,. The yields are given by the ratio between the observed activity A i and t h e total available activitL-, . I s (see below).
Fnr samples containing large amounts < I f moderator (e.g., helium), it is necessary t o cm-rect also for those tritium atoms which become thermalized and diffuse t o the wall before reacting in the gas phase. T h e magnitude of this diffusion 10SS is estimated in L1ppendix I1 on the assumption t h a t the quartz walls of the ampoule a c t as a “scavenger” for diffusing (thermalized) tritium atoms. Table I11 shows a comparison between A o / A , and A,/A,, where A , is t h e calculated activity with corrections for both recoil and diffusion losses. -1s seen the variations in the observed yield with helium concentration, with pressure, temperature and dimensions of the ampoule can be explained qualitativell- by the corresponding changes in t h e diffusion loss. T h e quat)titative agreement is actually somewhat better than cxpected. ’LAIILE I11 CUMPARI>OX 01; OUFEKVED A N L I CALCULATED Aci IVITIES,
TABLE I
DIFFUSION Loss
REAC L‘IOSS O F RECOIL T R I T I U M WITH hIBTIfANE EFFECTO F SCAVEXCERS k d Y .
cnger
Temp
i’=, mni.
-~ .‘iHT -
+€lLT
A,
4.
Z~CHers
A.
A.
.. ?Room I1 0.535 0.325 0 0 .14 Ur2 --lOO 4.01 42 .32 0 0 1 8 .11 Hr2 GO 41 6 .3T ,345 ,042 ,015 11 >12oo111 -1 .35 .32 .(‘20 < , ( I 1 flr, 45 1-2 .:%? ,:I6 .(I27 < . 0 1 S O >Rooin - $ I .;i13 ,365 ... < .O1 Drn ,--I5 -2) . -, ‘)(I .33 .(I15 < . 0 1 Ijr2 >F.ooin Large .23 .30 ,042 < .01 A-
--
T h e large influence of vrry small aniounts o f scavenger is cvident from this table. Higher hydrocarbons (i.e., labeled ethane, propane, butane, etc.) essentially disappear a t pressures larger than 1 mm. and at this pressure t h e yield of labeled methyl halide levels off t o remain constant with increasing pressure of scavenger. T h e yield of labeled methane is seen to be unaffected by the scavenger within the experimental error. The yields of H T initially decreases sharply b u t for P , 5 1 m m . fall off very slowly. There are IIO indications t h a t the scavengers used act with different eficiency. “Saturation” then occurs quickly, namely, for P , 1 nim., which corresponds t o a mole fraction of scavenger of about 0.001. T h e use of I2 as scavenger with no cooling during irradiation then gives a satisfactory (Ps 1 m m . ) and convenient system. Total Activity in the System.-The total activity A produced in a sample during irradiation can be calculated from a kiiowledge of the He3 content, t h e irradiation time and the neutron flux. This activity is, however, invariably found t c i be larger than t h e experimeutally observed activity, Ao (in samples without scavenger). T h e discrepancy disappears when corrections are made for recoil and diffusion losses of tritium particles. Some tritium is lost from t h e system because i t is formed near the wall and recoils into i t instead of being stopped in thc gas phase. This recoil loss L depends on t h e recoil range of the tritons and of the dimensions of t h e ampoule. T h e calculation of L is discussed in appendix I . When correctioiis are made for such loss, we obtain A , , the total activity available for reaction. Table I1 shows a comparison of calculated and experimental values for t h e activity in samples containing methane with no additives. T h e activities are given in d.p.s. and the composition of tlic samples by the partial pressures of the compounds (in ciii.). R is the diameter of t h e ampoule and r t h e reco,il range of the tritons. T h e agreement between Ao a n d A , 1s good even for considerable variations of t h e dimensions of the ampoule arid t h e total pressure.
-
-
TABLE I1 COMPARISON OF EXPERIMENTAL AND CALCULATED ACTIVITIES ii,
L‘uniIw\ition
1 5 ~ ~ P3 C H ~ 2 9 2 8 .i2
110
’41 31,0(IO
(.),J.v) 2 0 1 0 0 1 0 (i ,-18,101)
.4 03.000 .3T,TAIO 10:3,00Il
A“
I,
L 0175
iiii~i~nrn.
3 2
160 I !IO
i
1
.
Rooin >Room >Room
2 Riioni
--45 --l(lll
>K(XJIII
5 5 5 5 3
lOO(j5 47 29 99 93
.J
73
2
100
087; 40 ‘3’’
I
I
113 91
-. la
95
I t has been mentioned2 previin!sIy that a teniperatare dcpendent re-entry of tritium froin t h e Tvalls can take place. This re-entry occurs not only during irradiation but continues until t h e sample is analyzed. This effect was eliminated by making t h e irradiation time short (1-2 min.) and by storing t h e samples in Dry Ice until they were analyzed. Effect of Temperature, Pressure and Surface.--?is seen above the changes in the total activity with variations o f the temperature, pressure and dimensions of reaction vessel can be accounted for. The corresponding dependence of the thermal production of H T can be seen from the same considerations. T h e yield of labeled methane, however, is unaffected by both temperature (see Table I ) and pressure.2 The production of labeled methyl radicals is observed only indirectly in the presence of I 2 or Brz, and changes in the temperature are accompanied by changes in the halogen concentration. Thc effect was seen in Table I. Changes in the dimensions of t h e ampoule gave the variations expected from calculations of recoil and diffusion losse5. S o effect was seen in the production of labeled methane or methyl halide. Moderated Systems.-Some results for systems contaiiiing a moderating, noble gas are shown in Table IV and \.. Helium was used most often b u t the effect of neon, argciii and xenon also was investigated. In Figs. 1, 2 and 3 the yields of the various products are plotted against the mole fraction of methane, x , in the sample. Figure 4 shows a plot of the total yields of all observed products, Aol.41, for samples with scavengcr. (The yield of labeled hydrogcn halide is not included.) T h e yields of H T are subject t o e r r w due t o a side reaction which may become relatively important in highly moderated scavenged systems. Radiation will produce a small amount of hydrogen halide. This HX will react efficiently with thermalized tritium atoms t o form H T . T h e amount of H X is sufficiently small so t h a t only a small proportion of the thermalized tritium will react in this manner instead o f n-ith X2. However, in highly moderated systems only 2 ( , , of the tritium may react hot t o form H T and the yield of H T thus becomes very sensitive t o this side reaction of the YSrO thermal tritium. The increasing ratio of HT/CHzT in scavenged and moderated systems probably is due t o this spurious effect.
Discussion T h e most striking feature of the formation of a product such as labelctl iiitlthatie is the indeiwitlence the yield shows, t o the presence of scavengers, to chariges in temperature :uid pressure and to irractiaticiii flus. A behavior oi t h k 1;iiitl can be uiitler-
TABLE IV
pri.3
PCHP
IRRAnIArION O F h l E T H A N E IVITH TRITILX. hIODERATOR PRESENT l:lux, Irrad hlodera- n./cm.: sec. time, ---Yields in r$ of total activity 0 b s d . F CaHs High. C H4 min. Hz X 10-12 PMod tor
2.6 2.7 3.7 2 3 2 2 2 7
62 13 3.8 2 2
96 112 94 105
1 1
107
----Cornposition--
107
12 18
2 6
12 12 8 12 12 12 in
He He He He He A Se
9.5
74.1 65.0 89.9 83.9 90.5 51.8 66.8
2 2 2.5 2 2
-1
1.5
24.7 21.0 7.1 " ' 1 . 1
'7.8 42.0 32.0
TABLE l' IRRADIATION OF METHASEWITH TRITITM. ?rfODERrlTOR
-
----Compo?iiimPH.3 Pr1rr
3.3 3.4 3.3 3.3 3.1 2.3 3.4 2.8
79 60
31 18 5 0 2.3 13 8 8
.
-
lloiiera-
P M ~ t~o r
18 54 65 94 105 105 113
IIe He He He He
81
Se
He Se
Flux, n./ctn.*, sec X 10-12
Irrad. tiine, niin.
10 10 10 10 10 12 12 10
15 1.5 1.5 1.5 1.5 2 2 1.5
7--
H?
15.3 46.1 48.4 59 5 65.0 71.8 47.0 45.2
stood if it were formed by a hot reaction of the tritium atom. By contrast t h e higher hydrocarbons which are postulated to be formed by thermal reaction involving radiation produced species exhibit a complex dependence on these factors.
- Yield.
c HI 48 5
46 6 44 6 33 0 26 8 24 2 44 3 45 9
I
iii
0.9 12 2 2 2 6.5 2 1 7, N
0.6
0.3 1.8 0.8 1.9 0 2.6 0 6
AQ/A.
0.91 .3i .47 .2?
.20
. 39 1.n(i
AXD 1 2 PRESENT
7 of
total activity obsd.-C2Ha Higher CHsS
.1
0.1
5.8 6 4 6.5 6.85 5.25 4.0 8.7
.2
0.8
7.9
0 2
0 2
.5
.4 .3 .3 1.2
.2 .a
.9
..
..
EtS , .
..
.. .. 0.; .. ..
..
An/ 1. 0 88 ,636
481 .315 .117 .0 i 3 32 46
tion of ions is smaller than the methane concentration by a factor of los. Furthermore, yields from ion-molecule reactions should be sensitive to the presence of a substance such as Is, with a comparatively low ionization potential. However, as seen the yield of methane is not affected by IS. Similarly, the yield in moderated systems is found to decrease in the series, Xe, A, Ne, He. The opposite behavior would be expected for ionic reactions. Thus, in studies on radiation induced exchange of T? and CHP, X e has a quenching effect attributable to its low ionization potential which permits i t to discharge the reactive ions.g
Fig. l,-Plot of the relative yield of labeled methane as function of the mole fraction of methane; moderator\: helium 0, neon 0, a r g o n v , xenon A.
We may briefly examine some other mechanisms which might be advanced : X It has been suggested4,' that the reaction prodI . 1 - _ 1 > 0 5 I O ucts were direct consequences of ion-molecule reactions. However, as mentioned, T+ cannot be conFig. 2.-Plot of the relative yield of labeled methyl halide sidered a probable reagent.2 It is also doubtful as function of the mole fraction of methane; scavenger whether Towill react with ions such as CHI+, CH5+ present; moderators: heliuni 0, neon [?. and CH3+. These species are most certainly formed during irradiation, b u t their density will be (8) T. H. Pratt, Yale University, Ph.D. Thesis, 1960. (9) In the present system it is interesting to note that the higher very small. It can be estimated t h a t under the exhydrocarbons, which are formed by thermal reactions with radiperimental conditions the steady-state concentra- labeled ation produced species, are sharply reduced by the X e (see Table V). (7) J . B . Evans, J. E. Quinlan, M. C. Sauer and J. E. Willard, J. Phrs. Chem., 62, 1351 (1958).
This is in accord with the explanation that the radiation produced species are themselves formed by processes involving ions.
. . AHT
i?,
0 41-
0 3
i
T
T
1
1
1
0 5
X
1
I
I
I
1
I
I
1
X I
I
I
1
I
l
l
I,
0.5 1.0 Fig 3 --Pl!it i i f the relative yield o f labeled hydrugen as fuiictiim i i f the ~ n i ~fraction lc of meth:rne: scavenger present ; mot1er:itrrs: helium 0, neon %,
Formation of hot products a t the surface of the reaction vessel cannot be considered important, since changes in t h e surface to volume ratio did not affect the yields of methane or methyl halide. Xnother argument is the calculation made on the diffusing, thermalized tritium atoms (Table I11 and appendix 11). In pure methane only about 10‘; of these atoms would reach the walls, a fraction not nearly large enough to account for the yields observed. Furthermore, a mechanism involving diffusion would make the yields temperature and pressure dependent, in contrast to the observations. Sonie radiation induced exchanges leading to production of labeled compounds probably will occur in the system. However, with data from investigations of the self-induced tritium labeling of metliane,8 i t can be estimated t h a t under the experimental conditions less than 104 of the actual amounts found can be formed in t.his manner. This is well within the experimental error. The postulate of high-yield hot reaction between methane and tritium atoms is, however, in accord with all the data presented here. In particular, the noble gases do not act merely b y diluting the methane (Fig. 1, 2 , 31. Samples with the same mole fraction of one of these gases gave different yields, the yields decreasing in the order Xe, Ar, S e , He. This is the expected behavior if the idea of hot reactions is adopted, since the efficiency of these gases in moderating the recoiling tritium atoms in creases through this series. Furthermore, as is shown in the accompanying paper6 a kinetic reaction model applicable only to hot reactions provides a satisfactory quantitative description of this system. The niechanisiri of this hot hydrogen atom reaction appears to involve a fast displacement process.’J3 An intermediate metastable complex of the type CH,T in which all the hydrogens are equivalent appears most unlikely. By purely statistical consideration such a complex should give a yield ratio of H T to CHnT.of 2 3 . Even making allowance for a possible isotope effect this is completely incompatible with the observed ratio of 9 1 . The
1
1
0 5
1
,
[
I
>
I0
Fig. -1.---Pliit i l f the total relative yield of laheled products as func?icin o f the moie fraction of nietliarie: vxvenger prcwnt; riiodcrators: helium 0, neon 0.
detailed mechanism of the fast displacement is being investigated in other systems. Appendix I Recoil Loss.-Immediately after formation a frrrction o f the tritons will recriil into t h e walls with no piissibility f o i reacting with niolccules in the gas. This recoil loss c:tn l)? calculated from purely geometrical considerations. -In ampoule has the shape of a cylinder, with radius K . If a triton is formed a t a point 0 and if its recoil range is Y, it may a t t h e end of its track be a t any point on the surface of sphere with center a t 0 and radius T . T h e probability t h a t the triton will hit the cylinder wall is then equal t o t h a t fraction of t h e spherical surface which is outside the cylinder. This fraction must then be averaged over all positions of 0 inside t h e cylindcr. For a cylinder of infinite length the recoil l i m , L,is found to he
> 812: L
+
(2)
(;)I
[(CZ 1) E - (C’ - 1 ) K 37r where C = r!%R, anti k‘ arid E the complete elliptic integrals of the first and second kind respectively. These expressions need rmly small corrections t o apply to 10, h being the length o f cylinders of finite length if hli. the cylinder. In calculation of the recoil range, T , the following values cmployed for 0.192 niev. tritons (in gases a t i 6 cm., O0)’”~” were CH?, fl.24; He, 1 .45; > \ r ,0 . 2 i t j : Nc, 11.82, anti St., 11.17 c m . Y
=
>
Appendix I1 Diffusion Loss.--The formulas given in appendix I can be used t o find the diffusion loss, when an estimate has beeti made of the “diffusion range.” -4 treatment of random diffusion combined with simple gas-kinetic theory, leads t o t h e expressiunls: 9 2 ’-. vht where A? is the mean square displacement of a particle from the origin, v the average \velocity, h the mean free path and t the time taken for the displacement. (Considering the uncertainty in these paramcters, there is little point in using formulas more refined than the above.) Neglecting other thermal reactions, t will be determined by the rate i ) f the reaction T CHI +HT CH313j’4
+
(10) C. (1953)
+
J. Cook, E. Jones and T. J@rgensen, Phys. Rev., 91, 1417
(11) u’. WhalinR, “Handbuch d e r Physik,” Vol. XXXIV, JR58, p. 208. (12) E. A. hIoelwyn-Huqhes, “Physical Chemistry,” 2 n d E d . . London, 1957, p. 67. (13) hl. R . Betlie a n d D . I. I x R o y , C a n . J . Chem., 32, ti50 (1954). (14) R. Klein, J. R. LIcSesby, h l . D. Scheer and I.. J , Schoen, J ( h c n z Phys., 30, ,58(19.59).
June 5 , 1900
REACTION
OF
TRITIUM WITH
v and A were calculated from t h e standard gas-kinetic expressions. T h e loss found in this way must be multiplied by the probability of a tritium atom reaching thermal energies. This probability is approximately equal t o 1 - Ao/A,, values of which can be found from Fig. 4.
Acknowledgments.-The authors wish t o thank Professor L. Onsager for stimulating discussions of the mechanisms of hot reaction and for help with
[CONTRIBUTIOS
h-0 15F;i,
STERLING CHEMISTRY
26F3
?\.rETHANE
the calculations of the recoil loss. Suggestions of Professor F. S . Rowland regarding recoil ranges are Fatefully Comparisons with the work of T. Pratt on radiation induced exchange reactions was very useful. Mr. P r a t t also provided much help on problenis of technique. This work was performed under the auspices of the United States Atomic Energy Commission.
LABORATORY, YALE
GSIVERSITY, S E W
HAVEN, CONNECTICITT 1
Kinetic Theory of Hot Atom Reactions : Application to the System H
+ CH,
BY PEDER J. ESTRUP’ AND RICHARD WOLFGANG RECEIVED JULY 23, 1989 -4 model has been developed for t h e kinetics of hot atom reactions in the gas phase. .4n expression is derived giving the total probability t h a t a hot atom will react before losing its excess energy, in terms of the average collisional energy loss and the efficiency of reaction upon collision. The model predicts the relative effect of inert moderating compounds and provides a measure of the relative energy at which various products are formed in reactions with a given substance. T h e theory is tested using experimental d a t a on t h e action of moderators on the reactions of hot hydrogen with methane and is found t o provide a good representation of this system.
Introduction Atoms having much higher kinetic energy than provided b y the thermal motion of their surroundings are usually termed “hot” atoms. If such atoms take part in reactions before they have lost their excess energy, the processes are called “hot” reactions. In recent years several systems have been found in which hot atoms, produced b y photochemical or nuclear processes, could be unambiguously shown to undergo hot reaction^."^ These hot reactions were identified by one or more of several characteristic features which serve to distinguish them from thermal reactions : (1) Since the energy required for the reaction is almost exclusively supplied b y the hot atom rather than b y the environment, the hot reactions are to a first approximation temperature independent. ( 2 ) The hot species rapidly loses its excess energy to the surroundings. Thus, any reaction must take place in one of the relatively few collisions that it makes before i t is thermalized. Hot reaction vields are therefore independent of minor constitLents, in particular scavengers (such as In) which may be added to remove reactive thermal species. (3) H o t reactions are sensitive to moderators (e.g.! helium) which, though chemically inert, will efficientlyremove energy from the hot species. The detailed experimental investigation of the effect of moderators on the hot reactions of hydrogen atoms with methane5 has provided a stimulus for attempting to develop a quantitative model of hot atom kinetics. Evidently models involving an equilibrium distfbution of thermal energies, such as that leading to the -4rrhenius equation k = (1) C E R N , Geneva, Switzerland. Work performed in partial fulfilment of the requirements for the P h . D . degree a t Yale University (2) H. A. Schwarz, I E > E1 and zero elsewhere. ip,(E) corresponds to the collision efficiency pe--aE’RT in normal kinetic terminology. However, since for hot atoms E > A E and e--AE/RT 1, p,(E) beconies a high energy equivalent to the “steric factor,” p.) Then if .lis is the total number of hot atoms available for reaction and if ,f, is the relative probability of collision with compound j , the number of hot combination products is
-
where n ( E ) d E is the number of collisions suffered between E d E and E.
+
(G) A. F. Trotman-Dickenson, “Gas Kinetics,” Academic Press, Inc., X e w York, iY.Y,, 1955.