RICHARD I?. PORTER A N D SURESH K. GUPTA
280
hydroxyl groups with water is antagonistic to iceberg formation. Unfortunately, no data are available for N,N-dimethylglycine itself. AcknowZedyment.-The authors thank Miss Anna Straker for the calculations in this work. They are
also grateful to the Wellcome Trust for a travel grant t o S. P. D. which made this collaboration possible. Some of the equipment used in 1,ondon was provided by the Central Research Fund of the University of London.
Further Observations of the Stabilities and Reactivities
of Gaseous Boroxines'a
by Richard F. Porterlb and Suresh K. Gupta Department of Chemistry, Cornell University, rthaca, New York
(Received August 9, 1965)
Gaseous B303H3(boroxine) produced in a high temperature reaction of HzO(g) with elemental boron is thermodynamically unstable with respect to BzH6(g)and RZO3(s) a t room temperature. Molecules of B303H3 produced a t high temperatures exhibit kinetic stability a t low temperature. Condensation of gas to solid boroxine provides a kinetic path to BJIa and boric oxide. hfolar free energy and enthalpy relationships for R303H3(g), B303H3(s),and '/zBzHe(g) BzOa(s) have been established. Heats of formation of solid and gaseous boroxine are -301.7 f 2.5 and -291.0 f 2.0 kcal./mole, respectively, a t 298'K. Gaseous boroxine reacts with HCl at high temperatures to produce mono-, di-, and trichloroboroxine. Low pressure reactions of BCla with liquid BzO3 produce B3OsCl,(g) and B404Cl,(g). For the reactions BC1&) Rz03(1) = Ba03C13(g)and 4/3BC13(g) ~ ~30.8 ~ f 5.0 4/,B2O3(1) = B404C14(g),A H o I W o= o ~16.6 f 2.5 kcal./mole and A H o l l z = kcal./mole, respectively. Heats of formation a t 298OK. of gaseous mono-, di-, and trichloroboroxine are -314.5 f 4.0, -342.2 f 4.0, and -378.8 f 8.5 kcal./mole, respecis -494.0 f 6.0 kcal./mole a t 1125OK. tively. The heat of formation of B4o4Cl4(g)
+
+
Introduction Gaseous boroxine has been observed as the major product generated in a high temperature reaction of Hz with R-B2Oo mixtures.2 The thermal stability of &03H:{ was demonstrated in studies of the reaction
+ + B203
3/~Hdg) B
=
R303H3(g)
At a temperature of about 1400'K. the equilibrium constant for the reaction has been found to be of the order of unity. Indirect evidence of the stability of B303H3(g)is also noted by observation of the reaction The Journal of Phygical Chemistry
+
product isolated in a trap a t low temperatures. Investigation of the behavior of the solid has led t o some perplexing problems. On warming t o room temperature, the solid decomposes to yield diborane and a solid residue. These effects indicated that a more thorough study of the thermodynamic behavior of boroxine would be of value. A t present there is little or no information on the chemical behavior of this substance. (a) Work supported by the Advanced Research Projects Agency; (h) Alfred P. Sloan Fellow. (2) W . P. Sholette and R. F. Porter, J . Phys. Chem., 6 7 , 177 (1963). (1)
28 1
STABILITIES nxr, REACTIVITIES OF GASEOUS ROROXINES
The present study, therefore, was undertaken for the purpose of examining the condensation-evaporation bchavior of boroxine and investigating the chemical properties of the gaseous molecule. During the course of this study, several derivatives of boroxine have been ohserved and thermodynamic data for a number of these have hccn obtained.
Experimental Ijoroxine was produced by the reaction described previously. In these experiments I&O(g) was passed over a heated sample of elemental boron mixed with boric oxide powder. In the reaction with H,O(g) additional boron oxide is produced along with Hg(g). The exit of the furnace tube was connected by means of a U-t.ube and stopcocks to the inlet valve of a mass sp~ctrorneter.~With this assembly it was possible to ohserve reaction products while experimental conditions were altered. The assembly is illustrated in Fig. 1 . Water vapor maintained a t room temperature was introduced through a glass frit into the reaction zone which was usually held a t a temperature of about 1325'K. The water source was then cut off after a few minutes arid the reaction products were examined mass spectrometrically. I t is important to note that the reaction products, as they are observed, are essentially a t room temperature. The initial hydrogen pressure as observed with H n +was very high and the abundance of B303H3as noted by the intensity of B303112+was two to three orders of magnitude lower than H2+. The 13303H2+/Hz+ratio increased slowly with time. This was attributed to the difference in rates of diffusion from the hot reaction zone. A small proportion of 13,04H, (hydroxyboroxine) was also observed but in general the ratio of B3O4H2+/B3O3Hz+ was less than 0.05. With fresh samples of B-B2O3 mixtures in the reaction chamber, reaction with water was always complete; traces of unreacted water were observed only after several series of measurements with one sample. Diborane also was not observed in the
early stages of the reaction but was observed after several quantities of boroxine had been generated from one reactant mixture. The trap located between the reaction zone and the mass spectrometer was used to condense R30aH3(g). This tube also served as a means of introducing solids that undergo reaction with gaseous B303H3generated in the hot zone. A series of experiments was also undertaken to examine the reactions of gaseous boroxine with other reactive gases. The results presented here relate primarily to reactions with HCl(g). The experimental arrangement is similar to that used previously. This consisted of a Knudsen-type oven mounted to a gas inlet system. Product and urireacted gases leaving the cell are ionized by electron bombardment and the positive ion mass spectrum is ~ b s e r v e d . ~For these experiments, effusion cells were constructed of compacted boron nitride. Several small chips of BN were also placed in the cell. The mounting stem and cover for the cell were made of molybdenum. A molybdenum cylinder surrounding the cell was used as a conducting shield for electron-bombardment heating. The assembly is illustrated in Fig. 2. S o t shown in the figure are the radiation shields that enclose the oven. Temperatures were recorded with a platinum, platinum-rhodium thermocouple. The advantage of BX cells was that they provided a large
Mo Covet
/J
To
Moss Spect. V ~ c o r
Resistonce furnace Figure 1. High ternpersture assembly for preparation of gaseous B303II3.
Figure 2 . Effusion oven for mas$ spectrometric studies of reactions of gaseous borosine. (3)
R. E'. Porter and R. C . Shoonmaker, J . Phiis. ('hem., 6 2 , 234 (1958).
reactant surface in which boron is essentially at unit N2(g)4). activity (note that 13K decomposes to B(s) Decomposition or reduction of the BK to produce pU'p(g) became a limitation for good vacuum conditions a t temperatures above about 1300'K. Condensation Behavior of B&yH:l(g). I'rcvious studies of the Condensation behavior of 13aOaHj(g)were concerned with trapping of solid only a t liquid nitrogen temperatures. RIore detailrd information of the temperature-pressure relationship has now been obtained. In these experiments a source of gaseous B303H3was produced as described previously except that the U-tube was immersed in a series of "slush A baths" a t temperatures between 25 and -85'. series of coolants was prepared from ethanol-Dry Ice, chloroform-Dry Ice, and acetone-Dry Ice mixtures. For these experiments the best conditions for observation were during the initial runs where the only other product was noncondensable Hz(g). Data were taken by observing the intensity of & 0 3 H 2 + (the major ion fragment of as the temperature of the coolant
+
'
O
o
0
d
was changed. The temperature was first, lowered and then gradually raised to check rcproducibility. A series of ion currctit--tenipt:rat~iremeasurcmerit,s is given in l'ig. 3. X decrease of the 13,O,Hz+ peak was rarely observed for t,rap temperatures above - 40'. Below this tcmpcrature condensation behavior appears to be normal. It should be emphasized that the measuremcnt corresponds to what must be takttri as a limiting condensation pressure. ;Is we note later the evaporation behavior is much more complex. I.'rom rough pressure calibrations, we estimate that the pressure of 13:~O:lHsin the system was gcricrally between 0.1 and 0.5 torr for most of these expcriments. This prcssure range would also correspond roughly to the vapor pressure for the highest tcmperat.ure at which condensation was observed. The extrapolation of the intensity- -temperature curve to. higher temperature indicates that a very much higher pressure of B,03F13 is possible a t room temperature in a I'yrex system. Evaporation Uehavior of Solid R30alis. For cxamination of solid-vapor reactions of boroxine, it was necessary to isolate solid samples in an external system similar to that indicated in Fig. 1. The experimental conditions were identical with those employed for study of the condensation reaction except that the product, gases mere cont,iriually pumped through the system. Samples of solid were condensed in a U-tubc which could be easily joined to the inlet system to the mass spectromctcr. ]:or a reaction temperature of about 1300'K. with inlet water vapor at room temperature, it was possible to isolate about 100 mg. of solid in a period of 1.5 to 2 hr. The solid sample was held a t liquid nitrogen tcmperature during transfer to the mass spectrometer inlet. Samples then were warmed to about -80" and evaporation was followed by monitoring the mass spectrum. It was generally observed that H2He(g) was present Tho B2IIs(g) in the U-tube when warmed to -85'. pressure was found to be independent of temperature until the tube had reached a temperature of about, 0'. .4 rapid decomposition was then observed and the B2&+ rose abruptly. The high pressure of l37H6 in the trap a t -85' must result from condensation of DZHOformed by dissociation of solid condensed on warmer sections of the system above the trap during preparation. This may be due to the small amount of hydroxyhoroxine which should have a higher condensation temperature than boroxine. T o avoid this problem, samples
I / T (1000) Figure 3.
Mass spectrometric d a t a illustrating condcrisiition behavior of boroxine gas: 0 , d a t a are for B308113: A, d a t a are for R303H3.
The .Journal of Physical Chemicrtry
(4)
(a) 1'. Svhissel and W. Williattls, BILII.Am. f'h?/s. Soc.. 4 , 139 J. I,. Margrave,
(1959): (b) L. 1-1. Dreger, V. \'. Dadape, and J . I'hys. Cht-m., 6 6 , 1556 (1962).
STABILITIES AND REACTIVITIES OF GASEOUS BOROXIKES
283
-
20.01 224
.e 2.00 0
.J
I
I
I
I
I
2.9
2.8
I
3.0
f X 1000 Figure 5 . Temperature dependence data for evaporation products of solid boroxine.
above about 80' this ratio was nonreproducible and usually tended to decrease with an increase in temperature (Fig. 5 ) . I n some experiments after diborane had been pumped from the U-tube, it was found that reheating the residue yielded additional B&. This effect indicates that the solid residue behaves thermodynamically like a solid solution of boroxine in boric oxide although the structure of this material is still ~ n k o w n . ~ Thermodynamics os Condensation-Evaporation Reactions. Formally, we may consider the reactions Condensation: B303H3(g) = B303H8(s) Evaporation: B303H&)
=
'/zBzH6(g)
Figure 4. Mass spectrometric data illustrating evaporation behavior of boroxine solid to diborane.
(1)
+ Bz03(s)
(2)
It,will also be convenient t o consider the reaction were subsequently prepared by holding the trap at -85' while Bz& was removed by pumping. Evaporation behavior of these samples is illustrated by the data in Fig. 4 where we plot the intensity of BzHs+ us. temperature. These experimental data were obtained by successively replacing a cooler trap by a warmelr trap. The point corresponding to the threshold for decomposition was found to depend to some extent on the warming rate. A sample held for several hours in a closed U-tube a t -25' was also found to undergo partial decomposition. At temperatures between -85 and -40°, B303H3(g) was not detectable. The partial pressure of B303H3(g)for samples a t 25' was about 0.01 that of B2Ha(g). Thus it is evident, that the condensation-evaporation processes are not reversible. The relative change in intensities of BzH6+ and B303&+ was also studied as a function of temperature (see Fig. 4). This was achieved by immersing the U-tube in an oil bath which could be heated to temperatures of about 1 5 0 O . The ratio B303H2+/ BzHs+ increased very slightly but for temperatures
B303H3(g)
=
'/2B2&(g)
+ &03(S)
(3)
From the temperature dependence curve for condensation of boroxine, we obtain a corresponding heat of condensation. A series of six runs gives an average value of AHI' = -10.7 f 0.5 kcal./mole. Experiments with deuterated boroxine gave the same value within the experimental error quoted (see Fig. 3). This value is comparable in magnitude to heats of sublimation of similar molecules.6 During the course of the experiment liquefication was never observed either on condensation or during evaporation. We have evaluated the heat for reaction 3 by the third-law procedure in which AF3, computed from equi( 5 ) S. K. Gupta and R. F. Porter, J . Phys. Chem.. 67, 1286 (1963).
There is n o intention to imply here that. boroxine rings are ne
