HAROLD S. JOHNSTON
4254
A N D JULIAN
HEICKLEN
Vol. 86
TABLE V POSSIBLE RADICAL-RADICAL RECOMBINATIOX REACTIONS R2---
7
R1
CH, CH,O CHaCO CH,COO HCO HCOO CH,COCH, CH,COCH,O HO
CHa
X
CHaO
?
0
CHCO
CHCOO
? ?
X
?
disproportionation reactions, two direct oxygen-radical reactions, and the heterogeneous decomposition of CHaOOH. Water could be formed by 20 different radical-radical reactions involving HO, plus the heterogeneous decomposition of CHaOOH, and possibly heterogeneous destruction of HO,. Methanol could be formed in 13 different radical-radical reactions. Methyl hydroperoxide is possible from 13 such reactions. There appears, however, to be only one source of CH300CH3, and its rate of formation is a qualitative measure of the CH30 radical concentration. Undoubtedly these 140 possibilities could be reduced by other arguments, based on energetics, the literature, and aesthetics. However, products were observed a t only 1%mass peaks; the initial rates plus curves of growth give a t most 24 data. It appears clear that in spite of the great detail in which the products are followed, this investigation is not sufficient to give the first approximation to a definitive mechanism. Furthermore, the analysis of all molecular products is insufficient to establish the mechanism. On the other hand, this investigation does give some mechanistic conclusions. The low yield of CH300CH3 and the high yield of C H 3 0 0 H indicates some very easily abstracted hydrogen and an unusually abundant source of free radicals formed either by (i) oxygen molecule attack on excited acetone molecules or (ii) oxygen attack on the acetyl radical. Subsequent Reactions.-As the reaction proceeds the products enter the reaction. For example, the
[CONTRIBUTION FROM
THE
? ? ?
X X X
X X
?
?
X X X ? X
X X X X X X
X X X X X X X X
HO
0 0 0 X
0 0 0 X X
half-life data clearly indicate that H,CO, CHsCOCHO, CH3COCH20H,and CHaOOH are removed. Since the first three products are stable in the dark, their removal must be due to free radical attack
+ radicals products + radicals --+ products CHzCOCHzOH + radicals -+- products H2CO
----f
CH3COCH0
(21) (22)
(23)
where the products probably include Hz0, CO, and (20,. The methyl hydroperoxide decays both in the dark and when illuminated, the decay when illuminated being somewhat greater than in the dark under the same conditions. Thus, the reactions indicated arelo wall
CHBOOH+HzCO CHPOOH
+ H20
+ radicals -+- products
(24)
(25)
Apparently, reaction 21 is more important than reaction 24, and the net effect is a diminution of the relative HzCO pressure with time. The products of reactions 21-23 presumably include H20, CO, and COZ, and thus these products should be examples of case 4 of article I. Analysis of the data does show that all of these products are greater a t the steady state than would be predicted solely from the initial reactions. Acknowledgment.-This work was supported by the Division of Air Pollution, Bureau of State Services, Public Health Service, Grant AP-104.
DEPARTMENT O F CHEMISTRY, UNIVERSITY
Photochemical Oxidations. BY HAROLD S. JOHNSTON
CHCOCH? CHCOCH20
HCOO
HCO
IV.
OF
CALIFORNIA, BERKELEY, CALIFORNIA]
Acetaldehyde
A N D JULIAN
HEICKLEN
RECEIVEDFEBRC'ARY 28, 1964 The room-temperature photooxidation 0: acetaldehyde (0.4 to 18 m m . ) in oxygen (1.0 t o 9.2 mm.) with continuous ultraviolet radiation above 2200 A. has been studied by t h e method outlined in part I of this series. Observations were made by leaking the reaction mixture directly into the electron beam of the mass spectrometer during photolysis. The principal products of reaction were CHaOH and presumably CO and CO,; other products were H 2 0 , CH.0, HCOOH, CHBOOH, CHSCOOH, CH300CHa, and probably CHIC(O)OOH. Because of the cracking pattern of the reactants, it was impossible t o establish the presence or absence of CHI, CH2C0, CO, and C o n . Oxidation of the primary radicals, CH: and HCO, leads t o CI1300, CHaO, HO. and H 0 2 , and probably HCO(O0) and HCOO. There are a t least six radicals in this system t h a t can disproportionate in 36 ways and undergo other reactions. Thuc this reaction is much too complicated for its mechanism to be revealed merely by analysis of all products.
Introduction The room-temperature photooxidation of acetaldehyde in the vapor phase, in liquid, and in solution was first studied by Bowen and Tietz.' They found ?he
product(s) of reaction to be a peroxide(s) formed in a chain length of about 1000. Presumably, the Peroxide(s) was peracetic acid and/or diacetyl peroxide. (1) E J Bowen and E. L T i e t z , J . Chem. S O L ,234 (1930).
Oct. 20, 1964
PHOTOCHEMICAL OXII).\TION
4255
O F ,ACETALDEHYDE
TABLE I IYITIAL RATESOF PRODLTT FORMATIO\I __-_---__ heiles--------1
mm. [CH3CHO], mm. [021,
A,,,
A.
Seutral density screens RO X 103/[CH3CHO],/L of prod./mm of CHZCHO-sec. H20 (obsd.) CHiO (obsd.)
1 5 045 2200 Sone
2
9 2 040 2200 Sone
3
4a
1 2 2200 Sone
1 0 1 5 2200 Sone
7
1 0 4 0 2200 Sone
6
8 9 7.3 2200 Sone
--_-_ 7
090 7 0 2200 Sone
8
0 9 7 0 2800 None
9
1 3 18 2200 None
10
1 3 18 2200 8 870
32 17 0
... 11 8 ... 5 9 3 8 , . . 17 2 (-13) 14 3 17 3 5 0 ,.. 172 (calcd.)" (-4) ... 85 97 24 ... 60 6.0 191 CHPOH (obsd.) (57) 60 103 27 5.6 63 226 (calcd.) (64) (5.9) .. 2 7 ... 3 5 0 7 1 4 ... 4 5 HCOOH (obsd.) .. 2 6 4 2 ( 0 . 3 2 ) (1 3) (>2 3) 3 5 (calcd. ) (31 ( 0 61) 12 5 ... , . . 19 8 3 3 114 18 2 18.5 CHaOOH (obsd.) .. 13 3 18 7 ( 1 3 . 9 ) ( 9 8) 18 1 3 3 110 19.5 (calcd.) 4 2 ... 2 2 3 5 , . . 4 8 1.0 20 4 CHaCOOH (obsd.) 4 9 (4.2) 2 4 4 0 (4 2) 5 4 1.03 21 0 (calcd.) (1.7) 0 27 ... 0 14 ... 0 30 0 082 ... CHaC(0)OOH (obsd.) 0 46 0 12 0 29 ... 0 099 0 34 (calcd. ) (0 34) ( 0 14) 0 29 ( 0 11) ... 1 46 ... 2 2 ... 1 5 0 41 7 3 (CH30)z (obsd.) 1 3 0 47 (0.07) 6 6 (calcd. ) (1 1) (1 2Id 1 5 (1.4) c Rx0(calcd) = 0 69 P X 8 * / 7 ~d Used Numbers in parentheses are uncertain a Series 4, 5 , 9, and 10 done with different gold foil T L of 200 sec
575 162
16.5 ...
58 19 6
Carruthers and Norrish2 photooxidized formaldehyde and acetaldehyde. Except for the length of the chain, their results with acetaldehyde were consistent with those of Bowen and Tietz. They also found the photooxidation of formaldehyde to proceed via a short chain a t 100°, but neither the performic acid nor diformyl peroxide was observed. The products were HCOOH and H2, COS,CO, and H20 which were believed to come from the forniic acid. In 1941, Mignolet3 photooxidized acetaldehyde and obtained results in agreement with those of ref. 1 and 2. More recently, McDowell and his c o - ~ o r k e r s ~have - ~ confirmed the presence of peracetic acid as a product in the photooxidation of acetaldehyde. Calvert and Hanst' studied the photooxidation a t lower pressures (42 mm. of CH3CHO) and found COZ, CHaOH, HCOOH, and CO as well as CH,COOH and CH3C(O)OOH. The surprising feature is that little or no CHzO was found. In these experiments, peracetic acid was still the major product of reaction. In keeping with the program of photochemical oxidations undertaken by US,^ the acetaldehydeoxygen system was studied a t total pressures of less than 20 mm. to suppress even further the formation of peracetic and acetic acids. In this way it was hoped to obtain information concerning the oxidation, of CH, and HCO radicals. Experimental T h e apparatus and experitnental procedures have been described fully in the preceding articles of this series.* (2) J . E . Carruthers a n d R . G. W. liorrish, J . C h m Sot.. 1036 (1936). (3) J. hlignolet, Bull. SOC. roy. sci. L i e p e . 343 (1941). (4) C . A. Mc1)owell and J. B. F a r m e r , " F i f t h (International) Symposium on Combustion," Reinhold Publishing Corp., New York, N. Y., 19.55, p. 453. ( 5 ) C . A. McUowell and L. K . Sharples, C a n . J . C h e m . , 36, 251 (1968). (6) C. A hIcDowell and S. Sifniades, i b i d . , 41, 300 (1963). ( 7 ) J. 0. Calvert and P. L. H a n s t , %bid., 37, 1671 (1959). (8) J. P Heicklen and H . S. Johnston. J . A m . Chem. SOL.,84, 4030. 4394 (1962).
,
.
...
... ...
30 14.3
Ten series of runs were made with variation of t h e oxygen and acetaldehyde pressures, the incident intensity, and the incident radiation. All series except series 8 were made with Corning 9-54 filter, which eliminates all radiation below 2200 A. In series 8, a Corning 0-53 filter was used t o eliminate radiation below 2800 A. Replications of series 1 through 3 and 6 through 8 were made; agreement was t o within 10 t o 15% except in a few scattered cases. In series 4, 5 , 9, and 10, initial rates were not measured, and only the steady-state concentrations were measured. Matheson tank oxygen was used, and irnpurities were 0 . 3 5 argon and O . i % nitrogen. Mallinckrodt acetaldehyde was used. A number of minor impurity peaks are present in its mass spectrum. Except in the case of diethyl ether, which is present a t an impurity level of about 0.17;, the impurity peaks are all less than 0 . 0 1 % .
a
Results During irradiation, product peaks were observed a t m/e = 17, 18, 30, 31, 46. 47, 48, 60, 61, and 62. The first eight peaks are readily identified with H2O (17 and 18), CHZO (30), CHaOH (31), HCOOH (46), CH3OOH (47 and 38, also part of 30 and 31), and CH3COOH (60) The 62 peak corresponds to CH1OOCH,. The small peak a t m / e = G1 is not so easy to identify but probably corresponds to a major cracking peak of CH,C(O)OOH, because i t is an odd number, it cannot be a parent peak The absolute pressures of the products were estimated irom calibrations, comparison with the literature, and interpolation among related compounds However, it is reemphasized that some of the calibration values may be in error by as much as 30 or 407, It was observed also that very minor product peaks occurred a t m/e = 73, 77, 87, and 92, whereas the impurity peak a t m/e = 72 diminished during irradiation Presumably, impurity is attacked giving rise to additional products. However, these are very minor, are not connected with the primary oxidation, and consequently will be ignored. Because of background cracking peaks, analyses could not be made for COn (44), CH,CO (42), CO
HAROLD S. JOHNSTON
4256
Vol. 86
HEICKLEN
A N D JULIAN
TABLEI1 STEADY-STATE PRESSURES OF PRODUCTS _-_-___-_________-_Series 1 2 3
[Ozl, mm [CH3CHO],mm Xm,m
A
h e u t r a i density screens P / [ C H , C H O ] , p of prod / m m of CHzCHO Hz0 CHzO CH30H HCOOH CHZOOH CH3COOH CHaCO( OOH) (CHaOh m/e = 72 m/e = 73 m / e = 77 m/e = 87 m/e = 92
4
1 5 045 2200 None
9 2 040 2200 None
1 0 1 5 2200 None
1 2 2200 None
131 41 59 1.2 13 5 5.0 0.10 1.92
14 6
20 3 5 2 23 0 36 3 5 1 67 0 044 0 47
21 8 (-1) 18 4 0 8 2 6 1 25 0 10 0 4 0 0 0 0
1 7 0 09
2 4 0 30
... ... ... ... ...
5
(-3 3) 16 6 >o 7 1 85 1 25 0 042 0 32 - 0 19 0 04 0 0 0
0
S
7
6
1 0 4U 2200 None
8 9 7 3 2200 Kone
090 7 0 2200 None
0 9 7 0 2800 None
10 2 3 5 16 0 75 3 4 1 36 0 069 0 35
11 8 4 0 29 0 66 2 6 1 85 0 089 0 39
3 1 6 0 0 0 0 0
9
10
1 3 18 2200 None
1 3 18 2200 8 8%
8 5 0 31 (-3 4 ) 4 22 18 0 7 1 27 77 1 25 86 030 0 13 115 0 34 -0 19 0 0 13 0 05 0 0%
-1
7
0 50 0 031 0 020
0 035 0 02
T A B L E 111 HALF-LIVES OF PRODUCTS
______ [04,mm. [CH3$HO], mm. A. X’eutral density screens TD, sec. Hz0 CHz0 CHiOH HCOOH CHaOOH CHaCOOH CH3C(0)OOH (CHa0)z r ~sec. , Hz0 CHz0 CH3OH HCOOH CHaOOH CHaCOOH CHaC( 0 ) O O H (CH30)z
Xmin,
1
2
3
1 5 0 45 2200 Sone
9 2 0 40 2200 None
10 1 5 2200 Xone
290 220 210 250 135 270 280 170
215 f 30
340 210 195 175 170 265 250 180
f 20 f 20 i 10 f 50 f 10 f 20
220 f 40 115 k 20 320 f 60
f 50 f 40
-
Series-
6
7
8
8 9 7 3 2200 Sone
0 90
0 9 7 0 2800 Sone
... . . ... ... .. .. .. . . ,
250 f 20 170 f 20 180 f 10 190 f 20 140 f 10 185 f 10 170 f 10 250 f 20
235 f 10 160 f 10 I75 f 5 180 i 10 160 3= 10 160 f 10 180 f 10 165 f 10
., ... .., ...
285 f 20 170 f 10 185 f 15 150 f 30 130 f 10 172 f 5 165 f 10 105 f 5
550 f 30 160 f 5 195 f 5 180 f 10 135 f 10 260 f 10 210 f 10 165 i 15
4
5
1 2 2200 Sone
1 0 4 0 2200 None
...
f 50 f 20 f 10 f 40
... ... ...
& 20
..
f 20 f 30 k 40
.. .. ...
390 f 40 210 f 10 250 f 25
.. ... ...
.. f 10 f 20 f 40
...
i 40
..
7 0 2200 Xone
Average 330 f 50 275 f 43 270 f 40 210 f 32 200 i. 16 . 200 f 26 150 f 10 145 f 15 210 f 10 235 f 50 205 f 20 215 i 38 170 f 20 190 zk 27 Average 210 31 400 i 20 500 f 240 180 =t30 180 rt 15 165 f 20 200 zk 23 ... 185 f 18 160 i 20 120 f 24 250 f 20 220 i 43 210 f 20 210 f 20 170 f 20 ___. 180 i 38 ,
+
195 f 20 1200 f 100 165 f 20 180 f 10 210 f 30 200 k 100 85 f 5 85 f 5 165 f 10 200 f 50 200 f 100 200 f 30
130 285 250 250
... ... ..,
, ,
,
.. ... ... ,
.4verage
(28), and CH4 (16). Carbon dioxide and carbon monoxide are surely present.’ The experimental results are listed in Tables I through 111. For all of the products, there are listed the initial rates, Ri ; the steady-state partial pressure, P s s ;the half-time build-up to the steady-state pressure, T [ < ; and the half-time of decay from the steady-state value after the light was turned off, T D . In Table I, the observed rates are those from the initial slopes of the curves of growth of the products. The calculated initial rates are found from the relationship
Xi (calcd.)
=
0.69Pss/T~~
(1)
as described in part I of this series. The steady state is reached when the rate for formation of the product equals its rate of removal, both by chemical processes and by effusion through the pinhole into the mass spectrometer.
224 f 52
If the only unobserved products are CO and COz, their sum can be estimated from the carbon-hydrogen mass balance Ri(C0)
+ Ri(C0z)
=
O.jRi(ZH) - Xi(ZC)
(2)
where Xi(ZH) is the sum of the initial rates of hydrogen atom production in the products and Ri(ZC) is the corresponding quantity for carbon atoms. These quantities are listed as entries 3 through 5 in Table IV for series 1 through 3 and 6 through 8. It is readily apparent that CO and COZ,along with CH,OH, are the major products of the reaction. The results can be summarized in phenomenological terms, from the classification of part I, with no reference to the mechanism. (1) The observed initial products are HzO, CH20, CH30H, HCOOH, C H 3 0 0 H , CH3COOH, CH300CH3, and probably CH3C(O)OOH. ( 2 ) Water, the major product a t low C H 3 C H 0pressures,
PHOTOCHEMICAL O X I D A T I O N O F
Oct. 20, 1964
ACETALDEHYDE
4257
TABLE IV DATA
Ah-ALYSIS O F
decreases in importance as the C H 3 C H 0 pressure is raised. (3) CHZO is always in smaller amounts than is C H 3 0 H . At the low pressure of CHSCHO (0.4 mm.), the rates of production of these two products a r e similar, but a t higher pressures the ratio Ri(CHz0)/1 Ri(CH30H) is about 0.2. (4) The rate of CHZOOH formation relative to C H 3 0 H is markedly enhanced as the oxygen pressure is raised. ( 5 ) HCOOH, CH3COOH, CH3C(0)OOH, and CH300CH3 are minor products. (6) The half-lives of products in Table I11 give diagnostic information about products in the system as discussed in terms of curves of growth in part I of this series. The half-life for emptying the reaction cell by diffusion through the pinhole is about 210 sec. In the dark CH300H disappears much faster and HzO disappears much slower. This behavior indicates heterogeneous destruction of CH300H to form HzO as a product. The other product is presumably C H 2 0 , b u t its half-life in the cell is normal. Thus formaldehyde from CHaOOH presumably remains as a polymer on the surface. In the illuminated cell the same pattern applies: water appears slowly (that is, it is formed by secondary processes as well as initial processes) and methyl hydroperoxide appears rapidly (it undergoes secondary destruction). Formaldehyde seems to have a short half-life in the light and also CH300CH3 and HCOOH, but these effects are not large or certain.
Discussion The first problem is that of the primary photochemical products. Apparently acetaldehyde decomposes both by a molecular and a free radical split.g CH3CHO
+ hv +CH, + CO CHI + H C O
+
0 2
--+ CH&* +H O
f CHzO
1 0 1.5 233 108 125 13
8
7
6
8 9 7.3 265 134 131 6 5(12)
0 90 7 0 276 135 141 12
0 9 7 0 74 39 35 12
sumably undergoes similar processes
(MI ---f
HCO(O0)
(8)
plus a direct bimolecular hydrogen atom abstraction HCO
+ On +HOs + CO
(9)
The concensus seems to be in favor of reaction 9 although there may be conditions favoring (7) or (8). As a part of the series of initial fast radical reactions, peroxy methyl radicals are believed to be converted to methoxy radicals ZCH300
+2CHsO
+
0 2
( 10)
Thus one expects important amounts of the following radicals promptly to follow the primary photochemical reaction: CH300, CH30, HO, HOZ, and perhaps HCO(O0) and HCOO. With six hydrogen-containing radicals one expects up to 36 different disproportionation reactions, but a t most 12 different products. Many of these mathematical possibilities can be eliminated, but the mechanism seems to be quite inaccessible to any method that follows only products. Even so, one or two quantitative statements can be made. The primary source of methyl peroxide is believed to be 2CH30
(4)
(5)
(M)
CHSOO
3
+CH300CHa
(11)
(3)
The quantum yield of (3) varies with wavg length: 0.013 a t 3130; 0.15 a t 2804; and 0.28 a t 2654 A. Thus reaction 3 presumably contributes heavily to the destruction of acetaldehyde, but it does not contribute to the oxidations here studied-except as a nonobserved parallel reaction. Oxidation of methyl radical is believed t o go by way of CH3
2
9 2 0 40 68 29 39
(6)
where HO is favored by high temperatures and low pressure and C H 3 0 0 is favored by high pressure and low temperature. These studies were made a t low temperature and low pressure and steps ( 5 ) and (6) are both important. The formyl radical HCO pre(9) F . E. Blacet and J, D. Heldman, J . A m . Chem. S o c . . 64, 889 ( 1 9 4 2 ) , F. E. Blacet and D . E. LoeWer, ibid., 64, 893 (19.12).
but these same reactants also give a disproportionation reaction 2CHaO
--+ CHzO
+ CHaOH
(12)
If reaction 11 is the primary source of peroxide and if (12) is the principal source of formaldehyde, the ratio of rate of formation of CHzO and CH300CH3 should be constant Ri(CHzO)/Xi(CH300CH3)
= kit/kii
(13)
This relation is tested in Table IV. The relation is not particularly constant, but the significant thing is that the values 6.5 to 22 are the same order of magnitude as t h a t found with methyl iodide (8.5 to 22). I n the methyl iodide system, reaction 12 was regarded as the principal source of methanol, and the magnitude of the value for eq. 13 indicates that this reaction is the principal source of formaldehyde in the photooxidation of acetaldehyde.
4258
D. J. C. TATES AND P. J. LUCCHESI
If CH300CH3 is produced only by reaction 11. then the relative yield of methyl peroxide in various systems is a qualitative measure of the CHjO radical abundance. Siniilarly CH300H is formed only by abstraction of a hydrogen atom from other radicals by CHdOO,and i t is thus a qualitative measure of CH300 steady-state concentration. The average relative quantum yields are
[CONTRIBUTION FROM T H E PROCESS
Vol. 86
Reactant
CH3I CHiCHO
Q(CH300CHd
O(CH3OOH)
0 037 0 011
0 12 0 21
Thus the steady-state concentration of CHnO is suppressed relative to CHjOO in this system. Acknowledgment.-This work was supported by the Division of -4ir Pollution, Bureau of State Services, Public Health Service, Grant AP-104.
RESEARCH DIVISIOY,ESSO RESEARCH AXD ESGIXEERISG COMPAXY, LISDEN,NEW-
JERSEY]
Effects of 7-Radiation on the Surface Properties of Silica as Studied by the Infrared Spectra of Adsorbed Molecules BY D . J. C. YATESAND P. J. LUCCHESI KECEIVEDMAY14. 1964
The effect of y-radiation on high area oxides has been mainly studied by indirect means. For example, the rate of H2-Dz exchange on silica has been shown to be markedly affected by radiation. T h e advent of infrared spectroscopy as a tool for the study of adsorbed molecules offers a more direct approach to studying the effect of radiation on surface properties. I t is shown that dissociative ethylene Chemisorption can be induced in silica by y-radiation. Careful control experiments showed t h a t no ethylene \?as adsorbed on the surface, without prior irradiation. The nature of the adsorbed species, the mechanism of growth of the adsorbed species, and the nature of the active sites induced by the radiation are discussed.
I. Introduction While a considerable literature exists on the effect of radiation on the surface properties of oxides, direct measurements of these effects have been few. For instance, with regard to catalytic properties, many experiments have been r e p o r t e d ' ~ on ~ the induced H2-Dz exchange activity due to y-ray bombardment on a series of oxides. However, in most cases the oxides used had some activity before irradiation. and it is possible t h a t the radiation had changed an already existing property of the oxides rather than having induced any new surface properties in the material. Another problem in the interpretation of H2-Dz exchange experiments is t h a t i t is unknown what particular surface property, or surface site, is responsible for the exchange. Hence, i t is difficult to interpret the effect of radiation on exchange activity in terms of changes in the surface properties of the solid. This point has been discussed by T a y l ~ rwho , ~ has done a great deal of pioneering work in this field, and he had concluded "there is a basic difficulty in the characterization of defects that influence catalysisnamely, physical methods normally see features in the interior of the solid rather than on the surface simply because of the ordinarily much greater number of the former." However, there are exceptions to this rule. For instance, infrared spectroscopy can show directly the existence of some defects (such as OH groups) on the surfaces of oxides. I t can also detect the presence of as little as I . 1"() of a monolayer of adsorbed hydrocarbons.fi The effect of irradiating silica, in naci/o. with Co6" y-rays has been studied by measuring the infrared '1) li H. T a y l o r and J . A . Wethington. J . A m . C h c m Soc., 76, 971 (1954' ( 2 ) H V;. Kohn and E . H. Taylor, J . P h y s Clrrm.. 63,300 (1959). (3) H . X.. Kohn and E . H . T a y l o r , zbi,i., 63, 966 (13591. ( 4 ) ?I. W . R o h n arid E. H . T a y l o r , T r a m . C t i o w d inluvn C o n g v . C i z t a l y s i s , a , 1461 1~1961). ) E H. Taylor, S u d e o ? i i c s , 20, X? (1062). 1 N . Sheppard and D. J. C . Yates, P ~ i i r R . O Y .S w I T ~ n d o n I .A238, R9 6)!
spectra of molecules subsequently interacting with the irradiated surface. T h e majority of work has been done with ethylene. After evacuation of unirradiated silica, ethylene can be left for very long times above the surface, a t room temperature, without adsorption. After irradiation, in marked contrast, adsorbed saturated species are very rapidly formed. This provides direct evidence t h a t centers, very active for the strong chemisorption of unsaturated hydrocarbons, have been induced in this oxide by y-radiation. Experiments to elucidate the nature of these sites, and their relation with the color centers5 induced in silica, have been performed and are discussed in detail.
11. Experimental ( a ) Materials.-Porous glass was obtained from the Corning Glass Works, by the courtesy of D r . M . E. Xordberg. Tubular material, of varying radii b u t wall thickness of about 1 t o 2 mm., was used, although some experiments were done with flat plates. I n an attempt to ensure the maximum degree of uniformity of the samples, they were c u t from 15-in. lengths of the tubing. In the preparation of porous glass, a n acid-soluble component is leached out, leaving a silica matrix.' The thoroughness of this leaching determines, to a considerable extent, the impurities left on the surface of this material. Although the normal S a content of this material is about 400 parts per million ( p . p . m . ) , we were able t o obtain material with 50 p . p , n i , of Xa. In a n effort t o reduce the latter still further, we leached this material in boiling 1 iV nitric acid for 1 5 h r . , then rinsed most of the acid out, and stored the glass in distilled water.8 Ah attempt was made t o reduce the boron content of the material by leaching, calcining in air a t a higher temperature than subsequently used, and then releaching. Ethylene of 99.0% purity was supplied by the Matheson Co., J . Specially purified ethylene was also used, East Rutherford, 9. supplied by the Phillips Petroleum Co., Rartlesville, Okla. This had a purity of 99.92yc, and gas chromatographic analyses showed the main impurity to be a small amount of ethane. Hydrogen was obtained from the Linde Co., of Linden, N . J . , and had a purity of 99.98T0. Deuterium was supplied by the General Dynamics Corp., with a purity of 99.5%. Oxygen, S O , and Ar were supplied by the Matheson Co., with respective purities of 99.6, 99.0, and 99.99'T). In the case of the noncondensable gases, they were dried by passage through traps a t 7 7 ° K . -.
( 7 ) M E . Nordherg, J . A m . Ceram. SOL..27, 299 (10-14). ; 8 ) Leaching procedure suggested by Dr. M. E. Nordberg:
