3874
JUI
H. WANGA N D EVERLY R.FLEISCHER
cal.) and to 298.13’R. ( - 10 cal.) leads to AI120815 = -394.1 f 0.8 kcal./iiiole for the standard heat of combustion. The estimated uncertainty takes into account the impurities in the sample as well as the uncertainties involved in the combustions and calibration. Again, combining with the heat of formation of niobium pentoxide gives AH29815 = - G 1 . 1 =t 1.0 kcal./mole as the standard heat oi formation of niobium subnitride froni the elements. No previous heat of formation value of niobium subnitride exists. The present result is 4.3 kcal / mole more negative than the heat of formation of NbN (-56.8 =t 0.4 kcal./niole).d Because of uncertainty regarding the coinposition of the sample, only a tcntative value is offered for tantalum subnitride. A tentative value
[CONTRIBUTION NO. 1455 F R O M THE
1701. SO
appears justified :IS no previous value exists. Four combustions gave a mean of 1123.5 i 2.0 cal./g., after correction for inconipletion of combustion (average, 36.8 cal.) and for impurities (39.5 cal.). This corresponds to = -422 7 kcal. ’mole for combustion under bomb conditions. CorrecJ ~ ~ ~ tions to standard conditions gives l I l ~ = -424.1 i. 3.9 kcal./mole. Combining with Ilul11phrey’s? heat of formation of tantalum pentoxide (-4S8.S i. 0.5 kcal./mole) gives AH2Jd 15 = - 04.7 =t 3.0 ltcal./mole as the tentative heat of formation from the elements. This value appears reasonable in magnitude, being 2.7 kcal. mo!e more iietative than the heat oi forinition of T‘iS (-60.0 i: 0.6 kcal./mole).3 I~LRKEI.CY4, CALIFORNIA
DEPARTMEST O F
CIIEMISTRY, YALE USIVEKSI’TY]
Tracer Studies on the Mechanism of Combustion of Carbon, Sulfur and Mercuric Sulfide Bs JUI 13. WANGAX‘D EVERLY U.FLEISCHER RECEIVED FEBRUARY 21, 1958
+
+
The mechanisms of the reactions C Op C 0 2 and S O1 -+ S3., respectively, were studied with 018018 as tracer. It was concluded that the two oxygen atonis i n each COI or SO, i n ~ l ~ c ~produced ilr came from differelit 0,muleculei. --f
The elucidation of the mechanism of combustion of solid fuels through kinetic studies is often hampered by the difficulty of reproducing the surface conditions. However, valuable information on the mechanism of combustion can sometimes be deduced from the result of tracer studies. For example, when solid carbon, sulfur and mercuric sulfide, respectively, are burned in excess of I)iire oxygen, the main over-all reactions are
c + 0 2 +co* s + +so,
HgS
+
0 2
0 3
+H g
+ SO,
One may ask, “Are the two oxygen atoms in each of the COZ or ,302molecules produced in the above reactions from the same oxygen molecule or from two different oxygen molecules?” The answer was deduced from the result of the present study with 018018 as tracer. The principle of our method, which has already been described,2 is very simple. Suppose oxygen gas is prepared by the thermal decomposition of 018-enriched silver oxide. If the isotopic atomfraction of 0l8in this oxygen gas is X , the molefraction of 01*01* must be approximately equal to X 2 . (The small isotope effect due to the difference in zero point energies of light and heavy 0-atoms in O2 may be neglected for the present purpose.) When each mole of this 018-enriched oxygen gas is blended with q moles of ordinary oxygen gas, the isotopic atom fraction of Oi8and the in the mixed 0 2 become mole fraction of 018018 [X g(0.0020)]/(1 q) and [ X 2 q(0.0020)*]/
+
+
+
(1) This work was supported in part by a research grant (USPHSRG-4483) from the Division of Research Grants, Public Health Service. ( 2 ) K . C laruayin and J H \Vdng, THISJ O U R N A L , 80, 786
(1Q.58)
(! + y ) , respectively, where 0.0020 is the aton1 fraction of Old in ordinary oxygen gas. Tlie isotopic distribution in this blended oxygen gas is unnatural, because according to natural probabilities q(O.O020)]’ if the atom fraction of OlS is [ X (1 q ) , the mole fraction of Oi80i5 should be q(O.O02G)]’, (1 q)’. Xow approxiniately [X if this 0’8-enriched oxygen gas with unnatural isotopic distribution is usecl to burn solid carbon, sulfur and mercuric sulfide, respectively, the determination of O’*-distribution in the combustion product (CO, or SO,) could lead t o the answer of the above question. Thus if both 0-atonis in each COn (or SOy)molecule produced are from the same O? molecule, the isotopic mole fraction of Cc>lyO” (or SO1sO”) sliould sl.i!l be approximately equal to [ X 2 4(0.0020)’],/(1 (1). But if the two 0 atoms in each CO,(or SOn) molecule arc froni different 0. molecules, the reaction would involve a reshuffling of 0-atoms and yield an isotopic mole fraction of C015018(or S01801s)approxiIllately equal to [ X -/- q(0.0020)]2,’(1 q)’ in accordance with natural probabilities. Experimental
+
+
+
+
+
+
+
Preparation of the Labeled Oxygen Gas.-Labeled oxygen gas was prepared by therinal decomposition of 0’8-labeled silver oxide. The labeled silver oxide was precipitated froni its saturated solution in 0’8-enriched water (10.7 atom % in 0 1 8 ) with a concentrated solution of KOH in 018-enriched water. The precipitate, about 1.4 g., was washed with 20 ml. of ordinary water ten times, vacuunl dried and :hen lcft in a drying oven a t 105” for three days. It was stored in a desiccator before use. The oxygen gas so prepared contained 5.8 atom 70 0 ’ 8 and 0.336 mole yo of O‘80’8. Combustion Experiments.-The combustion was carried out in a n enclosed Pyrex tube which was approximately 25 cm. long and had a total inside volume of about 12.5 ml. I n each experiment about 120 to 140 mg. of labeled silver uxidc prepared ahovi. I V : I ~wi:ighed out and placed xt one
Aug. 5 , 1958
hfECI-IANISM OF COMBUSTION 0 1 ' CARBON, SULFUR AND MERCURIC SULFIDE
3875
There was no experimental proof for the absence of 0-atom end of the combustion tube. .4t the other end of the tube exchange between the pairs SOrSOz or COz-CO, under the was placed a weighed amount of ordinary silver oxide which present experimental conditions. However, b y diluting is equal t o 1.05 times the weight of the labeled silver oxide. The fuel, 12 to 15 mg. of carbon, sulfur or mercuric sulfide, the doubly 018-labeled SO2 and CO2 samples with ordinary was placed near the center of the tube. The combustion SO2 and COn gas, respectively, it was shown t h a t there was no 0-atom exchange between the above said pairs in 4 hours tube then was evacuated slowly. The silver oxides were then decomposed in turn by heating with a micro Bunsen at room temperature. It seems unlikely that such 0-atom exchange had taken place in the combustion experiments burner. The blended oxygen gas so produced had an unsince only a small spot of the combustion tube was heated natural 0 ' 8 distribution as explained above. The combustion was started by gently heating the center portion of the t o a higher temperature for a few seconds in each experiment. tube, where the fuel was placed, with the micro burner. Brandner and Urey3 found no detectable 0-atom exchange This caused the fuel to spark, and the combustion was between CO and COS below 900'. The rate of 0-atom usually finished in a few seconds. The fuels used were: char- exchange between two Cot molecules in the absence of water should be even slower from structural considerations. coal, Fisher, activated, pulverized; graphite, Fisher, Acheson S o . 38, pulverized; mercuric sulfide, Fisher M-195, powder. I n the burning of powdered charcoal or graphite, considResults and Discussion erable amount of solid carbon was left unoxidized after each The results of mass spectrometric analysis of the combustion was finished. The gas mixture left after combustion was found by mass spectrometric analysis t o con- combustion products are listed in Tables I and 11. tain 1 part of CO, 1 to 8 parts of 0 2 and 2 to 16 parts of CO:. Even in some blank experiments in which powdered charcoal TABLE I was burnt in 100 to 1000 times excess of ordinary oxygen, 1 COMBUSTION OF SOLID CARBON IN OXYGEN GAS mole of CO was produced for every 2 or 3 mole of CO, formed. Presumably because of the elongated shape of the Atom % of Ol8 in the oxygen gas used = 2.93%; mole % of combustion tube, the CO molecules formed could rapidly 018018 in the oxygen gas used = 0.164% diffuse to colder parts of the tube and thus escape further Isotopic mole % of C0180'8 produced ovidation by 02. However, this finding itself cannot prove 7 Calcd. 0-atoms in 0-atoms in that the Con molecules were formed through the stepwise each COz each COz oaidation of carbon by molecular oxygen. molecule molecule Found by Mass-spectrometric Analysis .-The gaseous mixture left from mass-specfrom trometric the same 01 different Oz after each combustion was analyzed by means of a Consolimolecules analysis Sample molecule dated-401 mass-spectrometer for all the isotopic species of 0 2 , CO and COn (or SOn). In some experiments the COZwas Powdered charcoal 0.164 0.086 0.080 frozen out of the gaseous mixture by means of liquid nitrogen Powdered charcoal .164 .OS6 .075 and analyzed separately. The results are consistent with Powdered charcoal .164 ,086 ,081 those analyses without this separation step. The calcula.164 .086 .OS4 tion of the isotopic mole fractions in the case of 0 2 and COn Powdered graphite from the mass-spectra was simple because 0 1 8 0 1 8 and Powdered graphite .164 .086 ,087 C12018018 were the only molecular species in the mixture Powdered graphite .164 .OS6 ,081 which contributed t o the m/e = 36 and 48 peaks, respectively. (The contribution of C1301i018 t o the m/c = 48 TABLE I1 peak was negligible as compared t o that of C12018018.) The evaluation of the isotopic mole fraction of S01801s COMBUSTION OF SULFUR AND MERCURIC SULFIDE I N O X Y G E N was more complex because of the presence of 4.270 of V4. GAS Thus the m / e = 68 peak was_contr&utedby both S3201n01s Atom % of 0 ' 8 in the oxygen gas used = 2.93%; mole % of and S3401601s. If we use (64), (66), (68) t o denote the 01s018 in the oxygen gas used = O.164YO heights of the peaks for m/e = 64, 66, 68, respectively, we Isotopic mole % of SOW11 produced have 0-atoms i F 0 . a t o m s in = (S32018016) (S3401601G) each SO2 each SOz molecule molecule = (S32018018) (S3401*016)
(a)
(m)
+ +
Sample
from t h e same 0 2 molecule
from Found by massdifferent 01 spectrometric molecules analysis
Powdered sulfur 0.164 0.086 0.086 Powdered sulfur ,164 ,086 .089 Powdered sulfur .164 . 086 .OB where 95.1 is the yo abundance of S32in nature. Solving Powdered sulfur ,164 ,086 .081 these equations simultaneously and assuming that (S33017018) Powdered HgS .164 ,086 ,105 is negligible, we get Powdered HgS .I64 ,086 .lo2 Powdered HgS ,164 ,086 ,097 Powdered HgS .22" .I5 .I8 Thus the mole fraction of S018018,which includes both Powdered HgS .22O .15 .14 S32018018 and Sa4018018,can be computed readily by the T h e oxygen gas used t o burn this sample contained usual procedure. 3.86 atom 'j?~ of 0l8and 0.22 mole 70of 0 ' 8 0 1 8 . Evidences for the Absence of 0's-Exchange between the Gaseous Species.-Since the present method of tracing reaction mechanisms is based on the assumption of the abThe data in Tables I and I1 show t h a t when sence of 0-atom exchange, it is imperative t o verify this powdered charcoal, graphite and sulfur, respecassumption experimentally before definite conclusions can tively, are burnt in an atmosphere of pure oxygen, be drawn. This mas done by analyzing the unreacted oxygen gas left after each combustion experiment for the rela- the two 0-atoms in each COS or SO2 molecule are from different 0 2 molecules. I n the case of the tive amounts of O15010, 0 1 6 0 ' 8 and 0 ' 8 0 ' 8 , respectively. I n every case it was found that the mole fraction of 0 1 8 0 1 * combustion of mercuric sulfide, the present results in the unreacted oxygen gas was, within experimental error, are inconclusive because of the larger experimental equal t o that of the original blended oxygen gas of unnatural isotopic distribution. This shows that there was uncertainties. no 0-atom exchange between the pairs 0 2 - 0 2 , OZ-COZ, NEW HAVEN,CONNECTICUT 02-CO and Oa-SO?, respectively, under the present experimental conditions. (3) J. D. Drandner and Ii. C . Urey, J. Chem. Phys., 13,361 (19.13).
