JOURNAL O F RESEARCH of the National Bureau of Standards- A. Physics and Chemistry Vol. 64A. No. 6, November- December 1960

Heat of Formation of Decaborane Walter H . Johnson, Marthada V. Kilday, and Edward J. Prosen

(August 8, 1960)

The heat of formation of crystall in e decaborane has been deter mined by calorimetric measlIrement of the heat of decompoF ition:

B lOTI j4(c) ...... lOB (am) + 7H2 (g),

MT(25 °C) = S2.8 ± 5.9 kj/ mole, (l9.8 ± 1.4 kcal/mole).

With t he heat of t ransition of crystalline to flmol'phollS boron taken as 0.40 kcal/mole we obtain for t he heat of formation of decaborane : '

M-Tf; ••. I0= - 66.1 kj/ mole, (- J5.8 keal/mole).

1. Introduction bustion of deeaborane. The heat of formation of decaborane [2] was tal<:en as 8 kcal/mole, calculated Combustio n meLhods afford a con venien t and from his results. precise meLhod fo1' determination of the heats of The decomposition of crystalline decaborane into formation of a large number of compounds. In the amorphous boron and gaseous hydrogen proved t o case of decaborane, however, preliminary measure­ be the best m ethod for calorimetric determination of ments of t he heat of combustion of dccaborane llave the heat of formation. The results of this investi­ been found Lo be vcry unreliable. CombusLion ex­ gation are presented here. perimen Ls in an oxyge ll bomb failed to yield resulLs of t he required precision ; the sample detonated in 2. Source and Purity of Materials practically all cases and the combusLion products in chlded varying amounLs of elemenLal bo[,on and The cl ecaborane was obtained from the General higlw r bo],on h:vdricies. There was also some 1111 (,e'1'­ E lectric R esear ch Laboratory through th e courtesy tainty in the amount of dissolved boric acid and the of A. E. Newkirk and A.L. Marshal1. Jt was prepared possibility of t he presence of boron oxides. In ge ll ­ by th e pYl'olysi of diborane and purified by several eral, the eXpel'inlents we're performed with the deca­ v:w uum sublimations, the last two of' which produced bo1'an e sample' in tbe form of a pcllet. I n some materials wi tIt no . measUl'able differ ence in m elting cases, in which the sample was burned in Lh e form point. Prio], to this investigation it was further of loose crystals or of crysLals intcrmixed wi tl) quartz purifi.ed b.v Lwo successive vacuum sublimations and fib ers, t he combustion residue also illcluded deca­ stored in a desiccator over magnesium perchlorate. borane. In other experimen ts a solu tion of deca­ A m ass spectrometric analysis of the helium2 used borane in toluene was burned ; the resulting quan tity showed it to contain less than 0.01 volume percent of carbon dioxide, however, indicated the formation or ail' ; the helium was further treated as described eitber of elemental carbon or of boron carb ide. m t he n ext section of this paper. The quantity of boric acid found in the watel'­ sohlble portion of the combustion products varied from 75 to 95 percent of the theoretical amount cal­ 3. Apparatus culated from the mass of sample. The resulting heat of combustion of decaborane (to crystalline boric The isothermal jacket calorimeter has been de­ oxide and liquid water) was - 2004 kcalfmole. A scribed previously [3]. The calorimetric system reason able estimate of the probable errol' in the beat consisted or the niclcel-plated copper calorimeter can of combustion of decaborane, as determined by the and covel' into which was placed a weighed quantity oxygen bomb process, is about 1 percent. In view of water , the calorimetric reaction vessel, and the of this information, it is obvious that the heat of platinum r esistance thermometer. The apparatus for formation calculated from the heat of combustion is measuring calorimeter temperatures and electrical subject to a relatively large un certainty. energy have been described previously [3, 4, 5]. Liebhafsky [1]1 also determined the heat of com-

2 Tbis analysis was perform ed by V. H . Dibelcr of the Mass Spectrometry 1 Figures in brackcts indicate the literature references at the end of tbis paper. Scction of the Atomic and Radiation P hysics DivisiolJ. 521 A diagr am of th e calorimetric decomposition vessel is given in figure 1. The vessel consisted of a quar tz tube, wound with a 183-ohm nichrome heating coil , surr ounded with a silver r adiation shield, :wd a vacuum jacket. The decaborane sample was placed t u in the Vycor crucible at the lower end o( the quar tz tube. H elium entered at th e top t hrough a coarse capillary immediately above the vessel, fl owed around the crucible and out through t he glass h eli x. A nichrome wire attached to the crucible was led out of th e vessel through the inlet tube and fastened to a circularly ground stopcock in the helium line. By turning the stopcock, the crucible could be raised without interrupting the £l ow of helium or permitting air to enter the vessel. An aluminum cover (no t shown in the diagram) which fi tted over the top of the vessel served to preven t loss of energy by radi­ ation through the opening in the calorimeter can cover. A di agram of t he gas train is shown in figure 2. The helium was purified by passing it successively ovel· titanium, copper , and copper oxide in a Vycor tube which was heated to 600 DC and through absorbers con taining Ascarite, anhydrous magnesium perchlorate, and phosphorus pen toxid e. The exi t gases wer e passed successivel y through a capillary flowmeter, a trap cool ed with liquid nitrogen, a Vycor tube con Laining copper oxide at 600 DC, two water-absorption tubes, E'ac h con taining magnesiu m FIG URE 1. Decomposition ve.ssel.

HELI UM

CAL ORIME TE R

FIGURE 2. Ga s train used in decomposition of decaborane. 522 perchlorate and phosphoru pen toxid e, and a r e­ period. Calorimeter temper atures wer e observed iLt cording gas-volume meter. i-minute in tervals during the en tire 1-houl" "r eac­ The capillary flowmeter ser ved for the approxi­ Lion " peri od. The helium flow was then directed to miLte adjustment of the ra Le of How of helium and by-pa s the decomposition ve el ~tncl Lh e r ecording also to indica te the l"hydride . The hydrogen evolved 20-minute m tin g period. in the decomposition wn. burned to water in the After the experimen t the decomposition vessel copper-o x.ide heater and ciLrried, as vapor, into the \vas r emoved from the calorimeter iLnd di sassembled. weighed absorption tube by th e str eam of dry helium. Al though most of the boron r emained in the crucible, The portion of the water which collected as liquid iL significiLn t amount was transported to the walls of in the bulb b elow the heater was evaporated by the vessel. It was necessary to place a plug of glass warming the bulb sligh tly. The second absorption wool in the helix in order to prevent some of the t ube ser ved only to protect the first from moisture boron from being carried out of the calorimeter by in the atmospher e. The total quan tity of helium the helium and hydrogen. Some small crystals of was m easured by means of the r ecording gas-volume decaborane, apparen tly formed by sublima tion, were meter at the exit end of the gas triLln . The approxi­ found in the upper part of the vessel. These usually mate temperature of the helium which entered the became di slodged a nd fell into the crucible during system dUTin g the experiment was m easured by removal of th e crucible from the vessel. It was means of a thermometer placed in con tact wi th the ther efor e no t possible to collect, quantitatively, all inlet tube of the decomposition vessel. of the boron produced in the decomposition. The quan tity of decaborane decomposed was cal­ 4. Procedure culated from the weigh t of hydrogen, as water, pro­ dllCed by the process. The accuracy of this deter­ The decabol'fLl1 e sample wa.s pl aced in the weighed mination WiLS checked by dissol ving a weighed quan­ Vycor crucible and melted by h eating wi th CiLr e in tity of zin c in dilute sulfuric acid while a slow str eam iL belium atmospher e. When cool, t he crucible plus of h elium was bubbled tlu-ough the solution and sample was weighed, transferred to the decompo­ directed through the gas train. The water vaporized sition vessel, and the vessel assembled. The vessel from the solution WiLS retained in the cold trap ; the was placed in th e calorimeter can, which con tain ed mass of water produced by burning the hydrogen 5,653 g water at 25. 0 °C. The calorimeter assembly agreed within 0.2 percen t wi th the theoretical was lower ed into the water bath, the temperature quan ti ty calculated from the mass of zinc. of which WiLS adjusted to 27.0 °C and maintained Ther e was usually a smiLll quan tity of nonvolatile constant to ± 0.001 °C. brownish materiiLl r em aining at the top of the Ail' WiLS r emo ved from the vessel by evacuation crucible which was assumed to consist of higher and by flushing wi th helium. Thermal equilibrium boron hydrides. The analysis o[ this material WiLS was established after about 20 minutes; the h elium complicated by contamination with decaboran e as was then directed to by-pass the decomposition described iLbove. An attempt was made to r emove vessel iLnd temperatures wer e observed at 2-minute the decaborane by placing the residue in a tan talum intervals during a 20-minu te initial r atin g period. boat and heating slowly to 300 °C in a tream of H elium was then directed through the vessel and helium. The material which r emain ed was weigh ed recording gas-volume meter, and an electric current and then h eated to 800 °C for 1 hour in vacuum. was passed through the heating coil. About 6 The loss in weigh t r esulting from t his treatmen t may minutes wer e required [or the quar tz tube to attain be assumed to be due to r emoval of hydrogen from the decomposition temperature of about 650 °C. the r esidue. The amoun t of hydrogen rem aining in The crucible was then raised very slowly by turning the r esidue was indica ted by this procedure to be the circularly ground stopcock un til hydrogen began approximately 0. 2 percen t by weigh t which corre­ to be evolved as indicated by the in cr ease in the sponds to sli gh tly less than 2 percent of the total reading on the capillary flowmeter. W"h en the evo­ hydrogen in th e original material. lution subsided the crucible was a.~ain raised slightly; The energy equivalent of the calorimeter was de­ this process was continued until the crucible was termined in a separ ate series of experimen ts which lo cated completely within the heated zone of the de·­ wer e performed in the same m anner as the decom­ composition ves el. A period o[ about 6 minutes position experimen ts. In each case the calorimeter was r equired for elevation of the crucible and an ad­ system was made to be n early iden tical wi th thiLt ditional period of 6 to 8 minu tes "fiLS allowed for used in the decomposition experiments excep t that completion 0 f the decomposition process. The the decabor iLne sample was no t presen t. h eating current was th en interrupted and the flow of helium through the vessel was continued for an additional 40 minutes to r e-establish thermal 5 . Results and Calculations equilibrium. M easlU'ements of the curren t through th e heating Th e moleculiLr weigh ts of boron, hydrogen, and coil and the poten tial drop across the coil were made water hiLve been taken iLS 10.82, 2. 016, and 18. 01 6 al ternately at I-minute in tervals elurin g th e heating g/mole, l'especti vely. The following mean h eat 523 capacities in jr C mole for the range of 25 to 27 °C For conversion to the conventional thermochemical were used: calorie, 1 calorie has been taken as equivalent to He(gas) 20.8 [2] 4.1840 joules. The mean of the values given for H 2(gas) 28.8 [2] flH in table 2 corresponds to the process: BlOH14 (c) 219.7 [6] B (amorph) 11. 1 [7]. BlOH14(c) --.? lOB (amorph) + 7H2(g) The results of the electrical calibration experi­ flH(25 °C) = 82.8 ± 5.9 kj /mole, ments are given in table 1, where flRe is the corrected = 19.8 ± 1.4 kcal/mole. temperature rise [4 , 8, 9], E is the quantity of electrical energy introduced, qHe is a correction With the value 0.40 kcal/mole taken for the heat applied to convert the temperatures of the inlet and of transition of crystalline to amorphous boron [2], exit helium to 25 °C, and Es is the heat capacity of there is obtained for the heat of formation of deca­ the standard calorimetric system obtained by the borane: relationship: E + qrre _ E flHf °(25 °C)= - 66.1 kj/mole, flRe - s· =-15.8 kcal/mole. The heat of the transition from amorphous boron to TABLE 1. Results of the electrical calibration experiments the stable crystalline form has not been determined. Some preliminary measurements in this laboratory Expcri- ARc E qH e E , have indicated that it is small. The enthalpy change mcnL for the process: ohm i/o!!", B (c) --.?B (amorph), 1 0. 183984 37350.3 1.0 203014 2 . 161 826 32?48.2 2. 8 203002 3 . 169807 34488. 3 2.0 20311 5 4 .188565 3830 1. 9 2.6 203 137 has been estimated to be 0.40 kcal/mole [1]. No 5 .235845 47888. I 3.5 2030134 uncertainty has been assigned to this arbitrary value. 6 .2513337 52016.6 3. 1 202923 Thore is also some ambiguity in regard to the Mean . . . .. ______. ______... ___ ..• ______203042 thermodynamic state of the amorphous boron Stan ci aI'd deviation of the mean - -. --- --. -._- ± 32. 4 obtained from the decomposition of the hydride. Several types of amorphous boron may exist. In addition, the effect of the particle size is not known. The results of tho decaborano decomposition Tho decomposition of the hydride, however, is experiments are given in table 2. The quantities believed to be a step-wise process and the nature of the resulting amorphous product, obtained at a given qHe, qH 2 , qB, and QB lOH 14 are the corroctions requirod to convert tho actual decomposition process to the temperature, is expected to be independent of the isothermal process at 25°C. Tho change in enthalpy boron-hydrogen ratio of the initial hydride. corresponding to the decomposition process was The heat of formation of decaborane may be com­ determined for each experiment by the relationship: bined with the heats of formation of boric oxide [3] and water [2] to obtain the hoat of combustion:

E + qH e+ qH 2+ qB + qB lOH14- flRc(Es) B IOH I4 (C) + 1l:02(g)--.?5B 20 3 (c) +7H 20(liq), flH (25 °C) mole BIOH14 ~HcO(25 °C)=- 1989.1 ± 4.0 kcal/mole.

Table 2. R esults of the BJOHJ4 decomposition experiments The uncertainties assigned to the values given in this paper have been taken as twice the standard

Exper- flRc E Moles AlI (25 °0) deviation of the means of the calibration and decom­ imcnt B laH " position experiments, combined with reasonable ------. __. estimates of all other known sources of error. ohm kjJrnole 0. 1781 11 3fi330. 0 2.9 - 0.1 -0.4 -0.5 0.001 8321 91. 64 . 180650 36806.8 1. 8 .0 -.4 -.5 .0016701 76. 713 6. Discussion . 178745 3M22. 6 2.2 . 0 -.3 - .5 .0016462 79. 76 .176884 36023. 9 2.2 -. 1 -.3 -.4 .0016111 68. 52 .215651 43948.5 3.6 -. 2 -.4 -.4 .0017796 92.66 . 236109 48154.4 2.8 -. 3 -. 3 -. 4 .00161320 93.38 The heats of formation of diborane and penta­ .213561 43512. 3 1.6 -. 2 -. 4 -.5 . 0017852 84. 53 borane were determined in a similar manner by .238218 484913. 6 2.7 -. 3 -.4 -. 4 . 0017660 7:1.56 . 221232 • 45046.1 3.5 -. 2 -.3 -. 4 . 0016111 84.79 measurement of the heats of decomposition of the vapors [10]. In each of these cases the decomposi­ Mean. __ . ______. ______.. ______. ______. ______82. 84 tion process was accompanied by the evolution of Standard Deviation of the mean .______±2.96 energy; in the case of decaborane, however, the process involves the absorption of energy. In every • There is a correction of -33.0 j/ohm in the energy equivalent of the calorim· eter system used for this experiment. case the quantity of material decomposed was deter- 524 mined by measurement of t he amounL of hydrogen 7. References evolved. This method has the advan tage of limiting the elTor introduced by incomplete decomposition to [11 n . A. Li ebltafsky, 1I npllblished daia. the difference between t he mean energies of the [21 F . ] . Ro sin i, D . D . Wagman, W. H . Evans, S. Levine, a nd 1. Jaffe, Selected values of chemi cal thermody­ broken anel unbroken boron-h.vell'ogen bonds multi­ Jlamic properties, KBS Circ. 500 (U.S. Government plied b y the fl'acLion of unbroken bonels. Printing OIfi ce, Washingion 25, D .C., 1952). There is a possibility LhtLt some of the hydrogen [3] E . J . Prose n, W. lI. Johnson , and F. Y. Pergiel, J. Re­ searcll N BS 62, 43 ( 1959) RP2!)27. may r emain adsorbed on Lhe amorphous boron at the [4] E. J . Prosen, F . W. Maron, a nd F . D. Rossini, J. Re­ decomposition temperature, but this quantity is searc ll NBS 46, 106 (I951 ) RP2181. believed to be mall. Sfl.mples of amorphous boron [5] E. J . Prosen a nd F . D . Ros ini, J . Research NBS 33, 255 prodLLced by pyrolysis of diborane at 600 °C were (1944) RP1607. [6] G. T . Jl'uruka \I'a and R. P . Pa rk, J . Research KBS 55, heated to 850 °C in vacuum with no significant loss 255 (1955) RP2627. of weight. [7] H . L. Johnston, H. N. Hersh, a nd E. C . Kerr, J. Am. Simons, Balis, and Liebhafsky [ll] made an Cilem. Soc. 73, 1112 ( 1951 ) . [8] J . R . Eckman and F . D. Rossini, BS J . Research 3, 597 elemental analysis by heating decaborane in sealed (1929) RPlll. pallfl.dium tubes to 900 °C in vacuum and were [9] E. J. Prosen , Chapter 6 in Experimental thermo­ able to remove 99.99 percent of the available hydro­ chemistry, F . D . Rossini, cd. (Interscience Publishers, gen. In view of these data it appears that any Inc., New York, N .Y., 1956). [10] E. J. Prosen, W. H . Johnson, and F . Y . Pergiel, J. Re­ error introduced by incomplete decomposition Ot' search NBS 61, 247 (1958) RP 290l. by adsorption of hydrogen on the amorphous boron [111 E. L. Simons, E. W. Balis, and H . A. Liebhafsky, Anal. residue is stllfl.ll in comparison with the uncertainty Chern. 25, 635 (1953). assigned to the experimental values. (Paper 64A6- 75)

525