JOURNAL OF RESEARCH of the National Bureau of Standards - A. Physic s and Chemistry Vol. 74A, No.6, November- December 1970 Fluorine Flame Calorimetry III. The Heat of Formation of Chlorine Trifluoride at 298.15 K* R. C. King** and G. T. Armstrong Institute for Materials Research, National Bureau of Standards, Washington, D.C. 20234 (July 6, 1970) The standard heat 0(" formati on of chlorine triA uoride (gas) at 298.15 K has been determined to be - 164.65 kJ mol- I (- 39.35 kcal mol- I) with a n ove rall experimenta l uncertainty of 5. 14 kJ mol - I (1.23 kcal mol- I). This valu e is de rived from the enthalpies of the foll owing reactions whic h were measured directly in a flam e calorimeter operated at 1 atm pressure a nd 303.5 K, together with data from previous investigations. C!Fig) + 2H,(g) + 100H, O(l) -> [H CI·3HF·lOOH, OI(l) (1/2)C I2(g) + 1/2H,(g) + [3HF·100H, OJ(l ) -> [HC I·3 HF·100H,OI(l ) The e nthalp y of formation of [HCI· lOOH ,O](I) was a lso measured. The average CI-F bond e ne rgy in c hl orine triAuoride is calc ulated to be 160.1 kJ mol- I (38.26 kcal mol- I). Key wo rds: Bond e nergy (CI-F); chlorine reaction with hydrogen ; chl ori ne trifluoride, heat of forma· tion; fl ame calorim etry; fl ow calorime try; flu orine compounds; heat of formati on; heat of reacti on ; hydrogen chl oride (aqueous), he at of form ation; mixed acids: (HCl + 3HFL", heat of formation; reaction c alorimetry; reacti on with h ydrogen. 1. Introduction and - 38.869 ± 1.0 [6] kcal mol- I. It is difficult to analyze the reasons for the relatively small un cer­ Chlorine trifl uoride is a vigorous fluorinating agent tanties in these data and no further attempt will be whi ch combines spontaneously with many other made at this time, The principal reason for unde r­ compounds and elements. This fluorine-containing taking this work was that evidence from so me test oxidizer is easily liquefied, and for this reason, has rocket propulsion studies, and some unpublis hed some appli cations different from those of other fluorine th ermoche mi cal information s uggested that the compounds. However, as with the other fluorin e­ enthalpy of formation mi ght be in th e range of - 26 containing oxidizers, the special applications of kcal mo] - I, a difference of over 10 kcal mol- I from the chlorine trifluoride, such as its use in rocket propulsion, published values. require accurate thermochemical data. In the present study, the heat of formation of In the past, there have been few original studies on chlorine trifluoride is determined by fluorine flame the heat of formation of chlorine trifluoride. Possibly calorimetry, a technique which is very different from ). this is caused by its extre me reactivity, which is the techniques used in earlier studies on this substance. greatest obstacle to a definitive study of its heat of The advantages of this technique are discussed in formation. Schmitz and Schumacher [111 and Schafer detail in recent reports on the heat of formation of other and Wicke [2] determined equilibrium constants at fluorine compounds studied in this laboratory [7 , 8]. various tem peratures for the reacti on, CIF:1(g) ~ The determination of the heat of formation of chlorine CIF(g) + F2(g). Von Wartenberg and Riteri s [3] and trifluoride by direct combination of the elements is Schmitz and Schumacher [4] measured the heat of the complicated by the formation of mixed halides. As an reaction, CIFAg) + 3NaCI(c)~ 3NaF(c) + 2Cb(g). These alternative, the reaction of chlorine trifluoride with studies have been evaluated repeatedly in different hydrogen followed by solution of the products to form reviews for the purpose of selecting a "best" value for an aqueous mixture of hydrofluoric and hydrochloric i1HJ (C I F~). The values currently given are -38.0 r5] acids was selected for this study. The heat of formation of chlorine trifluoride, the *Research sponson::d by the Air Force Office of Scientific Researc h, Office of Aerospace Research, Unit ed Slales Air Force. under AFOSR Cont ract No. ISSA-69-{)OO 1. enthalpy of reaction (1), is derived from reactions (2), uPresent address: ChcmisII'y Departme nt , York Coll ege of the Cit y Uni versity of New (3), (4), and (5). (Reaction 1 = Reaction (3 + 4 - 5 - 2).) York, N.Y. 11 365. I Figures ill brackets indi cate the lit erature references at the e nd of thi s paper. The heats of the ClF3 - H2-H20, and CI2-H2= HF(aq), 769 reactions, (2) and (3), were measured directly in this TABLE 1. Compositions of the chlorine and chlorine study. The heat of the F 2 -H2 -H20 reaction (4) was trifluoride samples studied earlier [8], and a necessary dilution energy, the enthalpy of reaction (5), obtained from the literature [5] . Sample The calorimeter was calibrated with the oxygen­ Constituent CIF 3 hydrogen reaction (6)_ (gas phase) 1/2Cb(g) + 3/2F2(g)~CIF 3(g) (1) Weight percent CIF 3 99.82 CIF3(g) + 2HAg) + 1 OOHzO(l) ~ CI 2 99.93 [HCl . 3HF . 100H20](l) (2) F2 O2 0.15 0.02 N2 .01 .05 1/2Cb(g) + 1/2H2(g) + [3HF . 100H20](l) ~ Ar * * [HCl· 3HF . 100H20](l) (3) CO 2 0.01 CF. 3/2F2(g) + 3/2HAg)+ 1 5 0H20(l)~ [3HF· 150H20](l) (4) C2F. * C3 F. C.F. [3HF . 100HzO](l) + 50HzO(l) ~ [3HF . 150H2 0](l) (5) S02F, SiF4 (6) - Not detected. *Traccs detected. 1/2Cb(g) + 1/2H2(g) + 100H2 0(l) ~ [HCl· 100H20](l) (7) atm at 25 °C. The principal impurities in the chlorine trifluoride sample were hydrogen fluoride, oxygen, Reaction (7) is closely related to reaction (3). Because nitrogen, and carbon dioxide. In the process of transfer accurate data for reaction (7) are available in the litera­ to the weighable container the gas was passed over ture, a measurement of the enthalpy gives a way of activated sodium fluoride for removal of the hydrogen checking the overall validity of the procedures used fluoride. in this study. After removal of the hydrogen fluoride, the remaining impurities are more volatile than the chlorine tri- - 2. Experimental Apparatus and Procedures fluoride. It is possible that the sample becomes purer in chlorine trifluoride as a series of experiments is 2.1. The Samples completed. To check this point, duplicate analyses were performed on the sample at different times. The Commercially available samples of hydrogen, results showed the sample to contain 0.47 and 0.45 oxygen, chlorine, and chlorine trifluoride were us~d. mole percent impurities, and give no evidence of such The oxygen and hydrogen samples were of high punty a fractionation. and are the same grades used for earlier flame calori­ For a quantitative analysis of the sample, as with the metric studies [8]. The mass fraction of O2 in the chlorine, the chlorine trifluoride was reacted with oxygen sample was 0.99987 and the mole fraction of H2 mercury and the residual gas was analyzed by mass in the hydrogen sample was 0.999. The hydrogen was spectrometry. Chlorine trifluoride reacts vigorously used directly from the source cylinder. Each of the with mercury. For these analyses the mercury and con­ oxidizer gases was used from small weighable con­ tainer were conditioned with chlorine trifluoride and tainers with a volume of about 250 cm3 [7, 8]. The then reevacuated prior to filling to the final pressure of containers were chilled with solid carbon dioxide to 1 atm. The results of the analyses are given in table 1. condense the chlorine and chlorine trifluoride into Limitations of the absorption method have been them. discussed [8]. Possible impurities that would also Chlorine - The chlorine sample was a high purity react with the mercury under the conditions of the grade and was not subjected to any additional purifica­ test are C}z, CIF, and CI-O-Fcompounds. Oxygen tion. The chlorine was transferred to the weighable liberated by reactions of impurities might affect the container and was contained as a liquid' under its vapor observed amount of free oxygen. The possible errors pressure of 5.8 atm at 25°C. The purity was checked due to lack of information about such impurities are using the mercury-absorption method [8, 9, 10], and taken into account in the discussion of errors. subsequent analysis of the residual gas by mass spectrometry. In carrying out the usual procedure some ·2.2. The Reaction Vessel and Flow System of the sample from the weighable container was trans­ ferred to an evacuated Pyrex glass bulb containing The reaction vessel is illustrated in figure 1. It con­ mercury. The bulb was immersed in hot water to sists of a heat exchanger, combustion chamber, two initiate the reaction. The composition of the chlorine solution vessels, various connecting tubes; and smaller as found by this analysis is given in table 1. parts. See section 2.3 for a brief description of its use. Chlorine trifluoride - The chlorine trifluoride was It is similar to the one used for the study of oxygen contained as a liquid under its vapor pressure of 1.47 difluoride which has already been described [8]. The 770 a certain extent, leaving a soot-like deposit on the walls of the combustion chamber. A spectrochemical analysis of this deposit revealed it to be platinum. The overall design of the fl ow syste m is similar to that used earlier.