SOLIDS FORMED FROM C.4RBOP; M0P;OXIDE BY RADIATIOP;
391
THE SOLIDS COKDESSED FRO11 C.1RBOS MOSOXIDE BY ALPHA P-IRTICLES JOHX H. L. WATSON Edsel B . Ford Institute for Medical Research, Detroit 2 , Michigan
MARCEL T'AKPEE' Graduate School, l..niversity of Minnesota, .Uinneapolis, Minnesota AND
S . C . LIKD Carbide and Carbon Chemicals Corporation, Oak Ridge, Tennessee Received February 9 , 1949 ISTRODUCTIOS
When radon is mixed with carbon monoxide the action of the a-particles (3) produces one gaseous and tiyo solid substances: carbon dioxide, carbon, and a suboxide of carbon. The overall reaction is approximately given in equation 1.
6CO --", 2COz
+
C
+
c302
(1)
There is some evidence l o suggest that the reaction occurs in several successive steps. First, ionization of carbon monoxide causes the reaction in the gas phase represented by equation 2 . 3CO + CO?
+ C?O
(2)
A triple collision is not required if the COf ion can collect tlvo neutral carbon monoxide molecules before it is neutralized by a free electron. While the experimental data do not give -Mco/AVco higher than 2, there is evidence of a reverse reaction which may well lower the yield from the required value of 3. The hypothetical gaseous suboxide would then undergo the reaction in equation 3.
2czo + c + c302
(3)
It is thought that the carbon condenses and falls by gravity to the bottom of the vessel and that the suboxide (C30z)diffuses to the wall and polymerizes as a thin, evenly distributed film which can be removed intact (see photograph mentioned in reference 2 ) . With some variations this picture is supported by the present work. For some undetermined cause it has not been possible to reproduce the polymerized film of suboxide in the present experiments. Despite several attempts the carbon and the suboxide have fallen to the bottom together, with some attachment also to the neck and sides of the vessel, and no separation has been effected. However, the electron microscope has revealed two distinct solids in the mixture, and four lines of graphite in hexagonal structure have been identi1
gan.
Present address: The Edsel B. Ford Institute for Medical Research, Detroit 2 , Michi-
392
J O W H. L . WATSOK, MARCEL V A N P E E AND S. C. LIKD
fied by x-ray diffraction, along with lines of a new, unidentifiable solid which is presumed to be the suboxide. ELECTRON MICROSCOPY
A . S a m p l e preparatzon The sample is a dark brown powder adhering to the neck, sides, and base of a glass flask of 8 cm. diameter. The carbon monoxide gas, mixed directly with radon, was irradiated for more than a month with an initial activity of about 100 millicuries of radon. After the bombardment period the carbon dioxide and residual monoxide mere pumped off and the flask was immediately filled with nitrogen and sealed. For the purposes of electron microscopy the flask was broken and samples were prepared and examined in some cases immediately by rubbing the dry material upon prepared Formvar films. Other specimens were prepared in the same manner at later dates to discover what effects exposure to air might have upon the substance. One plate was taken a t a magnification of 13,880 X to show the fine structures, but the rest were taken at 4500 X, a magnification which showed up all pertinent detail and provided a t the same time a sufficient number of particles for statistical accuracy in the counting procedures. Shadow-cast specimens, shadowed with chromium a t an angle with tangent of 2, were also micrographed at the lower magnification.
B. Qualitative observations The material is particulate, and separate particles are recognized readily in the micrographs. The projected images have nearly the same dimensions in all directions in most cases, and shadow casting indicates that the particles are a t least as high as they are wide. The edges are not smooth and from considerations of image intensity appear to have small flakes projecting from them. From these observations it is concluded that the particles are agglomerations of solid flaky material and have grown about equally in all directions. In the electron microscope, upon the fluorescent viewing screen a considerable number of the particles are observed to be hexagonal in shape (see figures l a and l b a t the arrows and elsewhere). When the micrographs are enlarged and reproduced in prints, and when the morphology of the ragged edges becomes visible, this hexagonal appearance is not seen so strikingly, although the impression is still given. l’n shadow-cast specimens many of the shadows are straight-edged and show up as from hexagonally shaped particles (figures 2a and 2b). For these reasons, a differential count was made on the particles so that the hexagonal and the nonhexagonal ones could be studied separately as well as in totality. Examples of straight-edged shadows are noted at the arrows in figure 2. The flakes which project from the particle edges (see a r r o w in figure 3) are straight-edged and seem to be crystalline. In a dark field the samples exposed to very light electron bombardment in the microscope (figure 4a) show no crystal
396
J O H S H. L. TVITSOS. MARCEL V 4 S P E E A S D S. C. L I S D
Even under prolonged intense bombardment from a biased gun in a vacuum the material does not completely sublime or undergo changes any more startling than those already discussed. Electrons are less effectiyely scattered and absorbed by the body of the particles after bombardment (compare figures 4s and 4b), and this fact coupled with the noticeable shrinkage suggests that there are two components in the material, one of which is affected more radically than the other by the heating effect of the beam. Light-field microscopy suggests further that one of these components is a viscous liquid material which lies over the particles and probably penetrates them. I n figure 3 a neck, ivhich is more easily penetrated by electrons than the main body of the particle, is seen to be joining one particle to another. These necks can be stretched before breaking and resemble in this way the necks which join particle to particle in a-ray cuprene (6, 7 , 10) (alpenel). The difference is that in alprene the neck and the particle are one, but here, in the radiation product from carbon monoxide, the neck is one component and the r a i n body of the particle a second. TABLE 1 Data f r o m the total distribution of particles f o u n d in the radiation product f r o m carbon monoxide bombarded by a-particles TOTAL R A S G E
I
August August August August August
5 , 1948
10, 1948 21, 1948 21, 1948 21, 1948
-~
1
Base
1
TBOajle Side T~~
I'
' ~
i
388
iZ 144 206
li
P
LI
0.54 0.43 0 61 0.42 0 42
0.16
0.2-1 . 0 0.2-0.8 0.24-0.72 0.32-0,56 0.2-0.6
0.16 0.14 0.13
C . Quantitative observations The particles were counted and measured by methods already published ( 8 ) , and the data were used statistically to give the observations recorded in tables 1, 2, 3, and 4. Figure 5 shows typical distribution curves for the material plotted upon logarithmic probability paper. Several observations may be made from these data. I n table 1 it appears that. there has been a slight decrease in particle size during a 16-day exposure to air, from a mean diameter of 0.54 micron to one of 0.51 micron, or a difference of about 5 per cent. This is within the order of accuracy claimed for the experiments, and therefore is not conclusive. Hon-ever, the difference of over 16 per cent between the mean diameter of a sample taken from the base and that of one taken from the neck of the flask is significant and indicates that there is a spectrum of particle sizes over the surface of the flask with the larger and presumably heavier particles tending to deposit upon its base.
* Alprene is a polymer of acetylene formed by the bombardment of acetylene gas by a-particles from radon.
397
SOLIDS FORMED F R O M C.1RBOS Y O S O X I D E BY RADI.iTIOS
TABLE 2
Data front the distribution of the hezagonal particles alone zn the radiation product f r o m carbon monoxide bombarded by a-particles
'
DATE OF P R E P a R A T I O E
o
F L ~ TL*SK
-
August August .lugust August August
5 , 1948 10, 1948 21, 1948
TOP Base Side TOP
21, 1948 21, 1918
,
, ~ S ~U Y B~E R~OF ~
~
I ARITHMETIC E
PARTICLES
I
i4
56 25 15 38
1
IOTAL RAYGE
YEXS
____
'
Rase
~-
~
'
P
I
P
0 46 0 40 0 46 0 39 041
,
0 2-1 0 0 2-0 8 0 24-0 72 0 32-0 56
1
1 0 2 - 0 6
~
TABLE 3
Comparison of the m e a n diameters of the hexagonal and nonhezagonal particles i n the five preparations D ~ T COF P R E P m n I o s
August August August .August August
DIFFERESCE I \ DIU6ETERS
P E R C E S T A G E DIFIEREHtE
5 , 1948 10, 1948 21, 1918 21, 1948 21, 1948
,
A
$0 ccnf
620
11.5
340 280
8.0
5.5 7.0 4.0
290 270
TABLE 4
T h e percentage b y which the hexagonal particles make lip the whole distribution DATE O F P R E P A R A T I O S
August August August August August
,
S G U B E R OF HEXAG-
P E R C E E T A G E OF
o'-ILS
HEXIGONALS
5 , 1945 10, 1948
i4 56
388
21, 1948 21, 1948 21, 1948
25 15
248
38
19 17 10
326
111 206
'
10 14
In table 2 the hesagonal component follows the same trends as the total distribution does, but in tablc 3 there is an indication that the hexagonal particles tend to have a smaller mean size than those which appear nonhexagonal in the micrographs. Table 4 is included to show that the frequency of occurrence of hexagonal particles is appreciable and varies between 10 and 20 per cent of the count totality. hlmost certainly this percentage should be even higher, since from shadow-casting esperiments it is obvious that many particles which do not appear to be hexagonal in silhouette do throw shadows as from hexagonal particles.
398
JOHN H . L. WATSOK, MARCEL V A X P E E AKD 8. C. LIND X-RAY D I F F R b C T I O X
X-ray diffraction analysis was made upon the sample. The experiment was repeated and the same results were achieved each time. h 114.59 mm. Norelco TABLE 5 X - r a y diffraction data f o r the residue formed by w r a y bombardment of carbon monozide LIhT NO.
i '
BIASDAPD GXAPEITE d-VALUES'
M A S U X E D I-VALUE
Ill0
A.
4.. . . . . . . . . . . . . . 5.., ., , , . , , , ., .. 6.. . . . . . . . . . . . .
2.695 2.53 2.21
0.7 1.0 0.2 0.6 0.8 0.8 0.3 0.4 0.1 0.1
12. . . . . . . . . . . . . . . 13.. . . . . . . . . . . . . . 14.. . . . . . . . . . . . . .
15... . . . . . . . . . . . . 16. . . . . . . . . . . . . . .
1.52 1.33 1.26
3.37
1.0
2.11 2.03 1.81 1.690
0.30 0.60 0.01 0.02
1.560
0.02
0.6 0.1 0.1 1.227
,
0.35
* Values taken from Hull: Phys. Rev. 10, 661 (1917).
Fro. 5. Typical particle-size distribution curves for carbon monoxide residue powder camera was used with nickel-filtered copper radiation for a 2-hr. exposure at 35 kv. and 25 ma. The specimen was scraped from the reaction flask and broken with a knife on a glass plate to as fine a powder as possible. A 0.3 mm. glass fiber was dipped into collodion and rolled in the sample until it was evenly covered. During the expesure the sample was rotated continuously. Data for the lines recorded are given in table 5 along with the values for a
samemeans so far (6,7, 10). It islike alprene and the hydrogen cyanide polymer in that it has recognizable particles sometimes joined by necks, but where the particles in these two substances are completely smooth, those in the carbon monoxide residues are rough and composed of a flaky material. I t is similar in appearance to the cyanogen polymer made by a-ray bombardment, in that both materials hare rough particles which appear crystalline by bright-field microscopy, but necks are not demonstrable in the cyanogen polymer and the mean particle size of this polymer is about three to four times less. Comparisons of this type would be more significant, hon-ever, if they irere made between samples formed under more nearly identical and controlled conditions. There are several indications from this work that there is more than one phase present in the carbon monoxide residues. The difference in particle size between material taken from t,he base and the top and the double shape distribution are suggestive but not conclusive. l l o r e significant is the liquid component, rhich is distinguishable from the solid component that it encases. I t is recognized in bright-field microscopy where it joins particle to particle and in dark-field microscopy irhere it gives evidence of an area of high scattering pon-er around and between them. X-ray diffraction studies indicate that the solid component is graphitic. I t is unlikely that graphite \vi11 he affected by the beam, and the shrinkage of the material under intense electron bombardment can only be explained by assuming that a second substance, presumably the polymerized suboxide, is present and is evaporated someivhat by the heating effect of the elwtron beam in vacuum Dark-field microscopy indicates diffuse rather than crystalline scattering from the unbombarded material. However, since fairly strong. sharp lines are obtained in s-ray diffraction patterns, it may be that the discontinuities or "spikes" seen along the image edzes in the dark field are similar in nature to thojn described by Hall (1) for carbon black and that crystalline reflections do originate from the unbombarded residues. Since the substance is csmposed largely of graphite, it is reasonable to expect that such reflections do occur, and t o conclude that the spikes are crystal reflections. Dark-field microscopy also shows that the particles are very opaque to electrons through most of their volume, but that there is a region a t the edge and between particles which scatters ne11 and is quite thick electron optically, but which is thin enough so that wide-anzle, diffuse scattering is not predominant. After bombardment the opacity of thz whole particle is decreased considerably, and yery wide angle scattering no lonzer predominates over the field. The hexagonal nature of an apprxiable number of the particles is a reproducible observation from preparation to preparation, but it is difficult to account for it. Hexagonal forms are often observed in heat-treated (4) and untreated ( 5 , 9) carbon blacks, where it is xell known that the ultimate structure is an hexagonal nct. I t appears that particle groivth has occurred in the gas phase, since the particles, though rough, are developed to a similar extent in all directions. This is also indicated by the fact that when the particles grow large enough they fall
SOLIDS FORMED FROM CARBOK MOKOXIDE BY R A D I A T I O S
401
out under gravity so that the larger ones collect a t the bottom and the smaller ones near the top, there still being a probability that some of either extreme ivill be found at any location within the flask. It also appears that, unlike alprene, the ultimate building units are much larger so that, instead of being round and smooth, particles are rough. Measurements made upon the projecting flakes give a mean size of about 615 .i.. a value which is the order then to 73-hich some first-forming unit grows before the secondary growth commences, whereby these unite to form the much larger particles shown here. BUMMART A S D COXCLUSIOPI-S
1. Carbon monoxide, when polymerized by a-rays from radon, exhibits a particulate structure. The particles have recognizable shape and have nearly the same dimensions in all directions. 2 . There are two components in the residue. The first consists of solid particles which are rigid aggregates of graphitic flakes. The second is a viscous liquid material ivhich surrounds the particles, may exist ivithin them, and often joins them one to the other by short necks. 3. The mean diameter of the material taken from the base of the flask is about 0.5 micron and that of the material from the neck of the flask about 0.41 micron. 4. -1bout 15 per cent of the silhouette images of the particles appear hesagonal in shape. The percentage of particles having this shape is undoubtedly even higher than this, since nonhesagonal images often are accompanied by hesagonal shadows. 5. The approximate uniformity of particle dimensions in all directions and the increase in particle size toivard the base of the flask indicate that the particles are formed and grow in the gas phase, and that the larger ones fall out under gravity. Many, both largc and small, also diffuse to the walls,
Shadow casting was accomplished through the kind cooperation of the Physics Instrumentation Department of the Research Laboratories Division, General Motors Corporation. I n this regard the authors are indebted particularly to 111.. William L. Grube, of that company. Mr. *Jonathan Parsons, of the staff of the Edscl B. Ford Institute for Medical Research, took the x-ray diffraction patterns. REFERESCES (1) HALL,C. E . : J. Applied Phys. 19, 271-i (194S), (2) LISD,8 . C . : The Chenizcal Efects of A l p h a Particles and Electrons, 2nd edition, p. 153 (photograph). The Chemical Catalog Company, Inc., S e w 'I'ork (1928). (3) LISD, 8 . C., A N D BARDWI:LL, D . e.:J. rim. Chem. Soc. 47, 4682 (1925). (4) SCHOOS. TH.,ASD KOCH,H. W :Kautschuk 17, 1 (1941). ( 5 ) KATSOS, JOHN H. L . : J. Applied Phys. 17, 121 (1946). (6) WATSOS,JOHSH. L . : J. Phys. Br Colloid Chem. 61, 654 (1917). (i) WATSOS, JOHSH. L.: J. Phys. & Colloid Chem. 62, 470 (1948). (8) WATSOS,JOHHH. L.:h a 1 . Chem. 20, 5 i 6 (1948). (9) WATSON,JOHN H. L . : J. Applied Phys. 20, 7-17 (1949). (10) IVATSOS, JOHSH. L., A S D KAvF.\raNs, K . : J. Applied Phys. 17, 996 (1946).
