KOTES
April, 1963 7ch
(see.)
=
2 X
exp(-27,700/RT)
(2)
It will be shown that it is easy to deduce the specific rate constant for the cleavage of the urethan linkages in sample 11. The following relation is clear3 j ( t )/ j ( O ) = lY(t)/N(O) (3: I n eq. 3, N ( 0 ) is the number of network chains per cc. originally present in the sample and N ( t ) is the iiumber of network chains per cc. which have not undergone any cleavage up to time t. Khether or not the irethan linkages re-form is immaterial in discussion of stress decay at constant extension. Each network chain of sample I1 contains two urethan linkages. The rate law for K(t) is dLIT - = 27c&(t) (4) dt uhcre 7 ~ 1is the specific rate constant for cleavage of the urethan linkage. (We assume all urethan linkages ill sample I1 to be equivalent.) Integrating eq. 3 one obtains
N(t)
=
N ( 0 ) exp(-2klt)
it?)
Inserting eq. 5 into eq. 3 there results
f(t)/.f(O) = exp(-2ht) (6) Comparing with eq. 1 and 2 , me obtain an expression for the specific rate constant for urethan cleavage k.1
= 1/27oh = 2.5
x
10’oexp[-27,700/RT]
It is to be noted from this expression that
(7)
is independent of the actual concentration of network chains and hence independent of the molecular weight of the original polyester. Of course, expression 7 is valid only for the urethan linkages of sample 11, i.e., those produced by treating aliphatic primary hydroxyl uith methyl triplienyl trisiocyanate. Also, eq. 7 is written as if the cleavage of the urethan linkage is unimolecular. It may be that the cleavage is actually a bimolecular reaction involving the urethan linkage and an unknown concentration of catalyst (e.g., amine). If this is so IC1 is to be regarded as a pseudo first order constant. Another urethan rubber (sample IT’) was discussed in ref. 1. This was obtained by treating the same ethylene-propylene polyester with excess 2,4-toluene diisocyanate to form isocyanate terminated “prepolymer.” This prepolyrner was cured to a three-dimeizsional network by reaction with 1,2,6-hexanetriol. I n this network (sample IT’) each network cha,im contains four urethan linkages. If the urethan linkages of sample IV mere identical with those of sample 11, it is clear from equations analogous to ( 5 ) , (B), and (7) that T , ~ ( I I ) / T ~ ~ (should IV) equal 2.3. I n actual fact, the ratios of 7ch(II)/T&(Iv) at 100, 120, and 140’ were 1.32, 1.44, and 1.44, respectively. We have no right to expect that the cleavage rates of the slightly different urethan linkages formed in sample I€and sample IV are exactly identical, but the fact that T c h ( I I ) / 7 c h ( I v ) is so close to the ideal value of 2.0 is very gratifying. (3) A. V. Tobolsky, ref. 2, pp. 208-209, 233.
~~k~
93 1
Acknowledgment.-We appreciate valuable discussions with Drs. R . Gobran and M.Berenbaum. The partial support of the Office of Kava1 Research is gratefully acknowledged. PHOTOLYSIS OF ALKYL NITRITES. THE PRIMARY PROCESS IS t-BUTYL XPTRITE AT 3660 8. BY G. R.M C ~ \ / ~ I L L A N ~ Celanese Chemzcal Co., a Dzuzszon of Celanese Corporation of Amerzca, Clarkzccod, Texas Recemed October I S , 1962
Recent interpretations of photolysis of alkyl nitrites include decomposition to alkoxy radicals and nitric oxide as an important primary proeess.2-6 I n the supposedlx continuous region of ahsorptioli (strong below 3000 A), the primary process in t-hiitpl nitrite is best written
(CHB)&O?;O
+ hv
----f
ao(CH3),CO*
+
[l - oiol(CH3)3CO
+ NO
with a probable primary quantum yield, p, of unity.G The excited t-butoxy radicals decompose rapidly to acetone and methyl. The primary quantum yield of alkyl nitrite photodecomposition following ab;orption of light in the banded region (3200-4000 A.) is less certain, but is known to be high.2nS Gray and Style2 estimated a primary quantum yield of 0.32 for methyl nitrite but it was not possible to take into account recombination of methoxy radicals and nitric oxide. Excited molecule mechanisms have not been considered, possibly because the bands were described by Thompson and Purkis’ as diffuse. However, the appearance of the spectrum of a polyatomic molecule is no sure guide to photochemical behavior, and it seemed possible to obtain a direct estimate of the primary yield by measuring the photochemical exchange between t-butyl nitrite and nitric oxide. Experimental &Butyl nitrite was prepared as before.0 Nitric oxide cbontaining 96%) NI60 was obtained from Isomet Corporation, Palisades Park, New Jersey. The volume of the quartz photolysis celi was 210 ml. The volume of cell plus connecting tubing was 252 ml. A 5-mm. thickness of Corning Glasq 5860 isolated the mercury lines near 3660 A. Light of 2537 A. was obtaiGed with the filter used before.6 The light intensity at 3660 A. was meas. ired with the potashiurn ferrioxalate actinometer,s taking = 1.21 The intensity absorbed by ths nitrite was calculated by the usual rr1ethods.Q The light beam was monitored with a Gl? 935 phototube and a pen recorder. Isotopic analysis was based on the 88/89 peak ratio in the mam spectrum of the nitrite. (1) Evans Chemical Laboratory, The Ohio State University, Columbus l o i Ohio. ( 2 ) J. A. Gray and D. W. G. Style. Trans. Faraday Soc., 46, 1137 (1952). (3) P. Tarte, Bull. 80c. roy. sci. Liege, 22, 226 (1953). (4) P. L. Hanst and J. G. Calvert, J . Phys. Chem., 63, 2071 (1959). ( 5 ) P. Kabasakalian a n d E. R. Townley, J . A m . Chem. Soc., 84, 2711 (1962). (6) G. R. AfchIillan, ibzd., 84, 4007 (1962). (7) H. W. Thompson and C. H. Purkis, Trans. Faraday SOC.,32, 1466 (1936). ( 8 ) C. G. Hatchard and C. A. Parker, Proc. Roy. SOC.(London), A235, 518 (1956). (9) A. Farkas and H. W. hfelville, “Experimental Methods in Gas Reactions,” Maomillan and Co., London, 1939, p. 247; R. E. H u n t and T . L. Hill, J . Chem. Phys., 16, 111 (1947).
NOTES
932
ilt low pressures of t-butyl nitrite and nitric oxide, no exchange of nit,rite was observed in blank experinients carried out in a darkened room. The nitric oxide did show dark exchange, yet the N16 did not appear in the nitrite. Evidently the nitrite contains an impurit,y (0.5 t,o 274) which undergoes fast dmk exchange with nit,ric oxide but not with t-butyl nitrite. The impurity could not be removed by dist2illation and was not, cJr1,erl~edhy gas chromatography. In the experiments :~1, 2,537 A . , a small correction pas necessary for exchange due to tho small amount, of 4047 A. light transmitted by the filter.
Results A fsw experiments were carried out at’ 25’ and 2537 A, using pressures of nitrite and nitric oxide of 14.5 mm. The quantum yield of combination of t-butoxy radicals and nitric oxide should be 0.16 under these conditions if the proposed mechanism6is correct. The observed quant’umyields of exchange vere 0.3, 0.7, and 0.7. The excess may be due t’o dark exchange beOween t-butyl nit,rite and some reaction product, nit’rosomethane, for example, derived from methyl radicals formed by decomposition of excit’ed t-butozy. The quant’um yield of acetone at 3660 A. (temperature, 25’; nitric oxide, 14.5 mm.; nitrite, 14.5 mm.) wa,s found to be 0.03, 0.04,0and 0.04. This may be compared with 0.84 at, 2537 A. under these conditions.6 Dark exchange with roeaction products may thus be a small effect a t 3660 A. The observed photocohemical exchange under different, conditions at 3660 A. is included in Table I. TABLE I PHOTOLYSIS OF (CH&COXO IN THE PRESESCE OF W50 A, 3660 T , 25’; incident light intensity, -5 X
w.;
quanta/sec. Psiitrite,
Time,
% Exch.
mm.
mm.
8ec.
found
% Exch. theoret./q
L*
14.5 14.5 14.5 14.5 14.5 40.5 44.0
14.5 15.5 17.5 32.5 33.5 14.5 16.5
540 600 900 1200 1200 540 900
1.68 1.65 2.44 2.07 2.14 1.65 2.58
1.72 1.85 2.44 2.15 2.09 1.66 2.61
0.98 .89 1.00 0.96 1.02 0.99 .99
Pxo,
Discussion Ignoring for the present the small amount of aceto2e formed, the main reactions in the system a t 3660 A. are
+ h~ +(CH3)3CO + 3 0 (CH3)sCO + NO + (CH,)&OXO
(CH,),COKO
(1)
(2)
Re-formation of nitrite accounts for the observation that prolonged illumination of t-butyl nitrite with 3660 8.radiation led to no net decomposition.1° Decomposition of primary and secondary nitrites did occur,1o presumably because disproportionation is possible between primary or secondary alkoxy radicals and nitric oxide. Based on reactions 1 aiid 2, the quotient % exchanged (theoretical)/p may be calculated aiid divided into the observed yo exchange to give the primary quantum yield. The mean value of 9 is 0.98 with an uncertainty not easily estimable. Possibly the acetone quantum yield, 0.04, should be added to the photo-exchange quantum yield. But if the view is correct that rapid (IO) H. I\-.Thompson and F.
(1937).
S Dainton, Trans. Faraday
Soc., 88, 1646
T‘ol. 67
dark exchange takes place between t-butyl nitrite and nitrosomethane (or derivatives thereof). the small portion of the primary process ultimately yielding acetone may already be counted in the photo-exchange. KO firm conclusions can be drawn about the source of the small amount of acetone, In consistency with the low wave length results,Gthe acetone may be ascribed to decomposition of excited t-butoxy radicals. This does not require that excited radicals be formed by absorption of light in the bands; the few excited radicals may be formed by absorption in the weak continuum underlying the bands. The results on t-butyl nitrite niight be summarized by noting that the primary process following light absorption in the bands or in the “continuum” is dissociation to a t-butoxy radical and nitric oxide, with a probable primary quantum yield of unity. The net photochemistry is different in the two regions because of the increased importaiice of decomposition of excited t-butoxy radicals at lower wave lengths.
DIEFUSIOS OF TRITIATED WATER ( ~ 3 I N AGAR GEL AKD WATER BY F. S. XAKAYAXA.4XD R.D. JACKSO&
~ 1 0 1 9
U . 5’. Water Conservation Laboratory, Tempe, Arizona Received October 86,1968
The “self-diffusion” coefficients of liquid water are of interest in that they may be used to ascertain the structural properties of liquid .ivater.l These coefficients are by necessity determined by isotopic tracer techniques. Wang, et a1.,2 by use of a diffusion capillary technique, determined the diffusion coefficient of tritiated water (H3H1016)in ordinary water at 25’. set.-' is Their value of 2.44 f 0.057 X lov5 widely accepted, although it is the only reported value known to the authors. We have determined the coefficient using a different technique and obtained a see.-’, which agrees value of 2.41 =I= 0,055 X with Wang’s value within the error of the experiment. Experimental The method consisted of determining the diffusion coefficient of tritiated water in low concentrations of gel and getting the coefficient in liquid water by extrapolating to infinite dilution of gel. Working u-ith gel material minimized error caused by non-diffusional movement of the liquid resulting- from mechaniral shock m d vibrations. Four low concentrations 10.3. 0.5. 0.75. and 1.0%- bv ” weight) of steam-sterilized agar-agar solutions in duplicate were allowed t o set in cylinders 1.9 cm. i.d. hy 1 2 cm. long. The cylinders were constructed so tshat.incremental 1-em. sections could be separated easily a t the end of a diffusion run. A 1.9-cni. filterpamperdid< was placed in contact with one end of the sample and treated with 0.01 ml. of a 100 pc./ml. t,ritiated water solutfion. The sample was settled and stored a t 25 f 1’ for 4s hr., and then sectioned. Water was ext,racted from the gel by vacuum dehydration, and the activity of the HZH1016 in the exbracted m-ater was determined by the liquid scintillatlon t e ~ h n i q u e . ~ The diffusion coefficient was calculated by obtaining the ratio of the activity (itx) in the column from z = 0 t o x = x to the tot,al activity (AT) in the column. Assuming a semi-infinite column, this ratio is equal to the integral of the instantaneous plane 8ource solution of the general diffusion equation. That) is
A,:AT
=
erf z(UIt)-’/*
J. H. Wang, J . A m . Chern. Soc., 73, 510 (1951). (2) J. H. Wang, C. V. Robinson, and I. A. Edelman, ibid.. 75, 466 (19.53). (3) F. E. Kinard, ‘‘A Liquid Scintillator for the Analysis of Tritium in Water,” Atomic Energy Comm. Rept. DP-190. 1956. (1)