1492
P. 0. 8CHISSEL
AND
0. c. TRULSON
Vol. 66
face area, however, shifts the data to lower separa- ferent adsorption or reaction sites a t two different tion factor values and lower potentials. Since the degrees of surface coverage. porous electrodes can discharge ions a t very low The decrease in separation factor with increasing potentials, the backward electrochemical reaction temperature is due to both the lower hydrogen rate is quite high, and the separation factors are overvoltage and the lower exchange reaction equicorrespondingly lower. I n addition, the internal librium constant at the higher temperature. The surface roughness, which results in a highly cata- thermodynamic equilibrium constant is lowered by lytic surface and unequal potential distribution over 13% as the temperature is increased from 30 to 50'. the true area, increases the rate of the chemical This decrease is close to the total percentage deexchange a t sites where the potential is inoperative, crease in S observed experimentally over the same to the extent that the maximum separation factor temperature range. The overvoltage difference beon the porous electrodes is less than 6 in all in- tween hydrogen and deuterium ion discharge destances. The general shape of the curve of sepa- creases slightly with increasing temperature. The ration factor vs. potential for porous electrodes is shape of the separation factor vs. potential curve the same as for smooth electrodes. The existence of remains the same at increased temperature with a two maxima in the separation factor curves seem t o lower maximum for S, which is shifted to less correspond t o the presence of two energetically dif- negative potentials.
MASS SPECTROMETRIC STUDY OF THE VAPORIZATION OF THE TITANIUM-BORON' SYSTEM^ BY P. 0. SCHISSEL~~ AND 0. C. TRULSON~~ Parma Research Laboratory, U n i o n Carbide Corp., Cleveland, Ohio Received Mmch IS,196.9
A mass spectrometer has been used with Knudsen cells to study the vaporization of the titanium-boron system. The pressures of Ti(g) and B(g) have been determined over several condensed phases and yield A"JsQ8 = 430 kcal./mole and AH02ss = -52 kcal./mole for the heats of vaporization and formation of TiB2, respectively. Thermodynamic functions for TiB(s) have been obtained.
Introduction Brewer and Haraldsen3 have summarized tliermodynamic information on some refractory borides and have made qualitative estimates of heats of formation based on the relative stabilities of boroncontaining compounds a t high temperatures. They mention the experimental difficulties associated with borides and suggest vapor pressure measurements as a promising approach to the evaluation of thermodynamic quantities. Vapor pressure measurements on elemental boron have been made by C h ~ p k aThornj4 ,~ Schissel and TViland Akishin, et aE.,6who find values in agreement for the heat of sublimation near AHo" = 130 kcal./mole. The qualitative observatioiis of Brewer and Haraldsen3 indicate a heat of formation for TiBz(s) near -72 kcal./mole; this value is confirmed by Samsonov.' However, mass spectrometric determinations of partial vapor pressures by Schissel and Williams5showed TiB, to be more volatile than would correspond to AH? = -72 kcal./mole, lead(1) This work was acoomp~iihedwith ARPA suppoit under AOMC Contract DA-30-069-ORD-2787. (2) (a) Union Carbide Corporation, Parma Research Laboratory: (b) Union Carbide Research Institute. (3) L. Brewer and H. Ilaraldsen, J. Electrochem. Soc., 102, 399 (1955). (4) JANAF Interim Thermochemical Tables, The Daw Chemical Co., Midland, Michigan, December 31, 1960. (5) P. 0. Sehissel and W. S.Williams, Bull. Am. Phys. Soc., Ser. 11, 4, No. 3 (1959). (6) P. A. Akishin, 0. T. Xikitin, and L. N. Gorokov, Dokl. Akad. Nauk S S S R , 129, 1075 (1959). (7) G. V. Samsonov, J . Applied Cham. (USSR), 28, 975 (1955).
ing to a value of about -32 kcal./mole instead; the congruent vaporization of the Ti-B system as used by Searcy, et aLls provided bounds for the relative electron impact ionization cross sections for titanium and boron vapor and substantiated the conclusions of Schissel and Williams; and a high temperature calorimetric measurement by Lowell and Williams9 resulted in AHfO = -50 6 kcal./mole. A mass spectrometer was employed in the present work with Knudsen effusion cells where the pressures of titanium and boron vapor were determined from rates of effusion. The vaporization of TiBz was studied under several experimental conditions: (1) TiBz with excess boron added, (2) TiB2, and (3) TiBz with excess titanium added. The results of each set of measurements were used to compute the heat of vaporization of TiB2, a?d current thermodynamic data were used to obtaiii the heat of formation at 298°K.
*
Apparatus.-The mass spectrometer, similar to that developed by Inghram and co-workers,lowas a 60' sector instrument of 12 in. radius, differentially pumped by two 5-l./sec. ion-getter vacuum pumps .I1 Electron bombardment was used to heat the Knudsen cells in an arrangement similar to that of Chupka and 1nghram.lZ A movable shutter interposed between the cell and the electron-impact ionization (8) A. W. Searcy, W. S. WiUiams, and P. 0 Behissel, J . CheTn. Phys., 82, 957 (1960).
(9) C. E. Lowell and W. S. Williams, Rea. Scz. lnstr., in press. (10) M. G. Inghram and R. J. Hayden, National Academy of Soiences and National Research Council, Publication 311 (1984). (11) Varian Associates, 611 Hansen Way, Palo Alto, California. (12) W. A. Chupka and M. G. Inghram, J . Phys. Chem., 59, 100 (1955).
August, 1962
VAPORIZATION OF THE TITANIUM-BORON SYSTEM
source was used to distinguish neutral species from background non-condensable molecules of the same mass. The energy of the ionizing electrons was continuously adjustable from 5 to 150 e.v. for appearance potential measurements. The main detector13 WBS a twenty-stage electron multiplier employing Be-Cu dynodes. Provision was also made t o collect the ion beam directly on an adjustable plate which could be moved in and out of the line of the ion beam The output from each collector was measured with an electromcter and recorded continuously on a strip-chart recorder. Temperature measurements were made with a Leeds and Northrup disappearing-filament optical pyrometer sighted intosmall holes (0.030in. diam., 0.060 in.. deep) in the sides of the Knudsen cells. The pyrometer was calibrated a ainst a standard lamp certified by the National Bureau of h a n d ards to be awurate to i7" a t the highest temperatures employed in t h k study. An auxiliary check was made against a pyrometer calibrated by the National Bureau of Standards and no systematic differences were noted. The calibration error therefore was taken to be &7'. The temperature correction for the sight window was determined before and after all data were taken. The estimated total error wa! &20" for the work with boron added to TiBs and A10 for the remajning m.easurements. Because of the high emissivity of graphite and since the temperatures were measured by Eiighting in holes in the graphite, no emissivity corrections were necessary. Further tests on the temperature measurements showed the above statement to be correct, and further that temperature gradients in the cell were small. In an. auxiliar vacuum system two pyrometers were simultaneously siglted on a tungsten ribbon filament to obtain their relative calibration. The Knudsen cell then was placed a t the same position as the tungsten ribbon with one pyrometer sighted into the temperature measuring hole in the side of the cell and the second pyrometer sighted into the Knudsen effusion hole. At the temperatures used in the experiments for which third law calculations were made in this study no systematic difference could be measured, but a t the very lowest temperatures (1700'K.), attained only during m.easurements pertaining to the second law treatment of the vaporization of boron, a temperature difference was noted but never exceeded 14'. Procedure and Results.-The phase diagram shown by Brewer and I.Iaraldsen3 for the Ti-B-C system indicates the expected mutual stabilities of the various phases. This diagram was used to determine suitable crucible materials for the systems which were studied. For the experimental situations discussed below, the phase syptems are identified by the sample materials initially loaded in the crucible. Subsequent X-ray analyses of the samples after heating showed partial conversion to other phases, particularly in the boron- and titaniumrich cases. The experimental results are discussed in the light of the X-ray observations made a t room temperatures after the runs. TiBs(s) It(s).-X-Ray analysis of an aliquot part of the TiBz and boron powdered samples showed no detectable impurities, and spectroscopic analysis showed 0 .Iyo iron, 0.1% silicon, and smaller amounts of chromium and other metals. The tiamples a-ere run in a B4C crucible (11/a2 in. diam., "/IF, in. long, 1/83 in. wall thickness) which was contained in a graphit,e outer jacket (16/16 in. diam., 1-1/4in. long, '/x in. wall thickness). Two B4C crucibles were made by heating graqhite crucibles in the presence of boron. The first crucible was used to demonstrate the com lete boroniaat'ion of the gmphite; the second was used for the experimental runs. Both crucible and jacket had rectangular orifices 4.5 mm. long and 0.76 mm. wide which provided a ratio of orifice area to sam le area of UA G 0.04 (sample area powcfer area, not the microscopic area effective UA smaller). was loaded, the Knudsen cell was heated SIONI t o the temperature a t which a shutter effect was observeJon Ti48and B". (These isotopic signals were used for all the measurements, whereas signals a t other isotopic positions were used only to confirm the identification of the species.) During the preliminary outgassing period, gaseous titaniixm oxide and boron oxide species were observed, but disappeared completely after several hours of heating. No data were taken until signals obviously characteristic of the oxide species had disappeared and ap-
+
pearance-potential measurements indicated that the B and T i signals resulted solely from the elemental gases. Data were taken on the temperature dependence of the boron signal and were plotted according to the ClausiusClapeyron relation to obtain the heat of sublimation of boron. AHTO. Four exaerimental runs of this tvoe were made.' The results for AHTO in kcal./mole obtainachronologically are: 130.2 i: 4.4 a t 1962"K., 128.4 j , 1.2 at 2085"K., 133.0 =t2.2 a t 2072"K., and 132.6 j , 1.5 at 2121"K., where the errors are the statistical errors determined from least square analyses, and the temperatures are the avera e mid-range temperatures of the individual runs. Tabulate$ data4 were used to reduce the AHTO values to AHoO with the results AH# = 130.8 f 5.1, 129.3 f 2.7, 133.8 f 3.4, and 133.5 f 2.9 kcal./mole. The final error assignments were obtained from the statistical combination of random errors from the least squares analyses and from the instrumental uncertainty in the absolute temperature (fractional error in AWT" = 2(AT/T), as described by Trulson, et aE.14 ' 4 powder pattern X-ray analysis of the sample after the runs showed strong TiBz lines and no boron lines. While the boron lines were weak even at the start, it is possible that all free boron was lost during the course of heating. Nevertheless, the average of the above second law values, 131.9 kcal./mole, is only slightly higher than mass spectrometric determinations using the third law4-+ and is in excellent agreement with recently tabulated data.4 Therefore, computations in this section will utilize the tabulated data for the vapor pressure of boron. The titanium pressure was determined relative to the boron pressure from
where P denotes the partial pressure, I the relative collected ion current, u the cross section for ionization of neutral species, T the temperature, i the isotopic abundance correction factor, and S the electron multiplier conversion factor. The ionization cross sections computed by Otvos and Stevenson15 were used, and the multiplier conversion efficiencies were determined directly for both boron and titanium signals. The relative ion signals were measured a t the same temperature. Electron bombardment energies ranging from 15 to 60 e.v. were used, but always were normalized to 40 e.v. by means of appearance-potential data of relative signal intensity us. electron energy taken when the signals were large. The boron and titanium pressures were used to obtain the equilibrium constant K for the reaction TiBZ(s) + Ti(g) 42B(g). The free energy change was computed (AFTO = -RT In K), and the heat of vaporization was obtained from
where tabulated4 free energy functions for Ti and B were used, and free energy functions for TiBz(s) were assumed equal t o those for Ti(s) 2B(s), which assumption is equivalent to assuming AC, and A S are zero for the solidsolid reaction. Walker, et a1.,16 have shown that AC, 0 up to 700". The heat of formation of TiBz(s) was computed from AHO~~X and the tabulated4heats of sublimation of boron and titanium. Data used to compute the heat of vaporization of TiBz are given in section 1 of Table I. Ten determinations of P T ~ relative to P B were made over a temperature range of approximately 200'K. The table gives the equilibrium constant, free energy change, and heat of vaporization of TiBz(s). The average heat of vaporization at 298'K. is
+
=
(14) 0. C. Trulson, D. E. Hudson, and F. H. Spedding, J. Chem. Phvs., in presa. (15) J. W. Otvos and I). P. Stevenson. J. Ana. Chem. Soc., 7 8 , 546
(1956).
(13) Nuclide Analysis Associates, Box 752, State College, Pennsylvania.
1493
(16) B. E. Walker, C. 61, 1682 (1957).
'r. Ewing, and R. R. Miller, J. Phys. Chem..
P. 0. SCHISSEL
1494
System before heating
B
+ TiB2 in BIC
AND
0. c. TRULSON
Vol. 66
TABLE I CALCULATION OF THE HEATOF VAPORIZATION AND HEATOF FORMATION OF TiB2(s) AFT^ AH^ - AHr’ T (OK.)
1
-log
-log
PB
PTi
2244 2296 2347 2357 2382 2398 2408 2306 2462 2411
(ked./ -log
K
mole)
(kcal./ mole)
(kcal./ mole)
49.8 53.2 54.9 57.5 54.2 55.2 53.5 53.3 53.3 51.8
8.47 8.46 -8.26 8.44 7.97 7.96 7.74 8.40 7.38 7.56
18.92 18.33 17.62 17.69 16.98 16.81 16.50 18.16 15.64 16.29
194.3 192.6 189.2 190.8 185.0 184.4 181.8 191.7 176.2 179.7
427.5 430.9 432.6 435.2 432.0 432.9 431.2 431.1 431.0 429.5
Av. 53.7 =k 0 . 7
431.4 TiBz in C
Ti
+ TiBz in C
2192 2185 2353 2246 2246 2246 2278 2278 2278 2310
6.54 6.62 5.14 6.20 6.22 6 13 5.78 5.73 5.77 5.67
7.19 7.11 6.31 7.12 7.00 7.04 6.83 6.78 6.79 6.83
20.27 20.36 16.60 19.52 19.44 19.30 18.40 18.23 18.33 17.97
203.3 203.5 178.7 200.6 199.8 198.4 191.8 190.1 191.1 189.9
431.4 430.9 422.7 434.1 433.2 431.8 428.3 426.6 427.6 429.7
2271 2269 2269 2307 2063 2107 2272 2276 2266 2340 2327
6.22 6.18 6.11 5.93 7.61 7.04 6.01 5.96 6.61 6.11 6.05
6.02 6.21 6.29 6.21 7.92 7.59 6.14 6.55 5.35 4.96 5.37
18.46 18.56 18.51 18.06 23.15 21.68 18.45 18.47 18.56 17.18 17.47
191.8 192.7 192.1 190.6 218.5 209.0 191.8 192.4 192.5 183.9 186.0
427.7 428.4 427.9 430.1 433.7 428.6 427.8 428.8 427.9 426.7 427.5
53.7 53.2 45.0 56.3 55.5 54.1 50.6 48.9 49.9 52.0 Av. 51.9 f 1.1
429.6 7
8 9 10
50.0 50.7 50.1 52.4 56.0 50.9 50.1 51.0 50.2 49 .O 49.7 428.6
431.4 f 5.9 kcal./mole, and the heat of formation is -53.7 f 5.9 kcal./mole. TiBz in Graphite.-A crucible fabricated from “Kational” TSX graphite was used with a TiB2 powdered sample. The knife-edged circular orifice in the crucible lid was 0.75 In a preliminary mm. in diameter to give U A 5 5 X run the crucible was heated empty t o the highest temperatures anticipated for the TiB2 measurements to demonstrate that no boron background impurity signals were present. The TiBz sample was then loaded with a 5-mg. silver calibration charge and heated to approximately 1100”, where the silver charge was allowed to vaporize completely. This method of calibration has been described by Chupka and Inghram.12 The pressures of titanium and boron were determined from ion intensity data taken a t higher temperatures according to the relation
(3) where I , denotes the relative ion current for species x (titanium or boron), T the cell temperature, and ( v i s ) as described in eq. 1. The constant represents the effective sensitivity constant (the collected charge per effused neutral particle) for the mass spectrometer determined from the silver calibration measurements. The relative intensities of the boron and titanium signals were observed to increase slowly with time a t fixed crucible temperature. Since boron already had the higher partial pressure, this further increase would be paradoxical if titanium and boron only were present. However, equilibration with carbon, while having little effect on the integral
Av. 50.9
+
0.7
free energy for the reaction, TiBZ(s) -+. Ti(g) 2B(g), can change the partial molal free energies of the titanium and boron markedly. I n other terms, the equilibrium constant should remain nearly unchanged when carbon is added while the partial pressures may change significantly. Equilibrium constants and integral free energies of evaporation were calculated by the third law method outlined in eq. 2. Ten determinations of PT, and PB were made over TiBz in graphite and are given in Table I, section 2. These results yield AH%* = 429.6 rt 5.4 kcal./mole for the vaporization of TiBz and &H0298 = -51.9 f: 5.4 kcal./ mole for the heat of formation. Ti TiB2.-Excess titanium was added to the graphite crucible containing the TiBz sample used in the runs described above. The crucible was heated for several hours above the melting point of titanium t o allow complete reaction, and subsequent X-ray analysis of the sample showed TiB2 only, while the inside of the crucible lid showed C and TiC. Additional titanium was added and the heating process was repeated; again the reaction occurred but at a much slower rate. A third loading of titanium was added with a weighed amount of silver. A silver calibration wag performed at 1100’ and was in agreement (-lo’%) with other calibrations. The vapor pressure of solid titanium was determined from eq. 3: a t 1842”K., PT,= 1.0 X atm.; a t 1858”K., P T ~ = 1.6 x 10-6atm.; and a t 1722’K., P T= ~ 1.1 X lo-’ atm. These results are in excellent agreem-ent with tabulated data*and are used t o confirm the ionization cross sections computed by Otvos and StevensonL6for titanium. While Ti(s) cannot be in equilibrium with C, T i c , and TiB2,it is assumed the reaction rate was sufficiently slow after the crucible had been previously exposed to titanium a t much higher temperatures to sustain the pressure
+
August, 1962
VAPORIZATION O F THE TIT.4"KW-BORON
SYSTEM
1495
error assignment comprised the random iiiiccrtainties obtained as the standard deviation of the means compounded with the instrumental uncertainties obtained from the estimated errors in the temperature and vapor pressures. The temperature error has been discussed above; the pressure error was determined from uncertainties in the relati1.c ionization cross sections and in the measured relative multiplier gains for the several species. The internal agreement among the second and third law measurements by Chupka,* the third law measurements by Schissel and WilliamsI5and the second + law data herein presented on the sublimation of boron, shows that the relative cross sections UB/ U A ~predicted by Otvos and StevensonJ5 are in error by less than i50%. Assuming that the + + + tabulated data4 on titanium are correct, the pressure determinations on titanium in this study show the relative cross sections U T ~ / ~to . Abe ~ in error by less than +SO%, and thus the ratio CrB/UTi is known to a t least a factor of two. These observations are consistent with those of Searcy, et c ~ l . for ,~ the free surface vaporization of TiB,. The multiplier response for Ag, B, and Ti was measured in this work with an uiicertainty of *lo%. All + thermodynamic results taken from tabulated data were assumed to be without error. Discussion Summary.-Data from all the experimental Partial pressures of titanium and boron have been poi& given in 'Table I can be used t.o plot the determined over condensed phases of the Ti-B-C equilibrium constant for the vaporization of TiB2. system for several regions of the phase diagram. X least squares analysis of a plot of In K vs. 1/T The first set of measurements was made in the yields AR02287 = 424 f 14 kcal./mole and AHozgs region TiB2-B4C-B, where boron and TiB2powders = 436 kcal./mole, in excellent internal agreement were used in a B4C crucible. The second set of with the third law computations. This result is measurements was made with TiB, powder loaded independlent of any assumptions regarding ioniza- in a graphite crucible, ostensibly along the line tion cross sections. TiB2-C in the phase diagram. The third set of Searches for Ti-B gaseous species were negative; measurements was made with titanium metal ( ~ )less t,han for example, a t 2426OK., P T ~ Bwas added to the TiBz powder in the graphite crucible TiB2 runs. Similar bounds used for the second set of measurements. Results atm. during the B for other simple combinations occurred and, in for the heat of vaporization and formation of TiBz general, extended. sweeps through mass 600 indi- are in excellent agreement, giving an average cated no species to -1% of the B(g) signal. A AHoz9s = 429.8 kcal./mole and an average AHozs8 search for Bz(g) was negative and indicated P B ~ = -52.1 kcal./mole, respectively. ~1"sless t,han 4 X atm. a t 2316'11. over B As discussed by Brewer and Haraldsen3 the obTiB,. This bound is not low enough t,o test the servation that TiB2(s) is stable in graphite bounds tabulat'ed (data. the stability of TiBz relative to T i c . Using the Some impurity signals were observed. In ad- data of Humphreyi7for Tic, one finds AHfo(TiB2) dition to the oxide species which volat,ilized, ele- I -44 kcal./mole, while the data of Fujishiro and mental Fe(g) and Cr(g) impurities were observed Gokcen18 would yield AHfO(TiB2) 5 -31.3 kcal./ and never completely disappeared. During the mole. Brewer and Haraldsen state that the heats final runs where excess t'itaniurn wa,s added, the of formation of titanium borides should be about pressure of Cr(g) was comparable to Ti(g) even -36 kcal./mole of boron. The agreement bet'hough spectroscopic analysis showed the Ti(s) to tween this result and that of Samsonov7 for TiB, cont'ain no Cr. It is presumed that chromium apparently is fortuitous, since Samsonov used the borides in the TiB2a t tra,ce levels were converted t80 value -66 kcal./mole for the heat of formation of titanium borides and excess chromium. The rather BgC in disagreement with the value used by Brewer high impurity level of Cr on the inside of the lid and Haraldsen and with recently tabulated values. 4 corroborates this assumpt'ion. The present results for TiB2 do not agree with Some unexplained signals occurred a t the mass AHfO = -72 kcal./mole. 32 through 37 positions which appeared by every The measurements of Schissel and Williamsi test t'o be legitimat,e neutral species leaving the yield AHfO % --32 kcal./mole for TiB,. Thew TiB, in B4C. It is mass spectrometric measurements were made with cell during the work with B believed they are not related to the Ti-B system a mixture of boron and Ti& powders contained in a but are possibly ga,seousB-C species. (17) G . L. Humphrey, J . Am. Ciiem Soc., 75, 2261 (1951). In all the previous resuhs, the net experimental (18) S. Fujishiro and N. A. Gokoen, J . Phvs. Chem., 66, 161 (1961).
over solid titanium t.hough t,he pressure of pureJitanium was not sustained over the more reactive liquid. rhiv assumption is corroborated by the observation that even after the second loading of titanium, a plot of In P T ~ vs. 1 / T taken before the 1:itaniumwas melted yielded a heat of sublimat,ion of titanium only 7 kcal. larger than given in tlhe JAKAF tables. Below the melting point of titanium, the boron signal could not be observed, but a t higher temperatures both signals wwe measured. At the higher temperatures the reaction of titanium again became evident., since the vapor pressure (which wa9 determined after each of the three successive titanium loadings) of liquid titanium could not be maintained, although the pressure exceeded that to be expected over T i c C. The premure data are given in section 3 of Table I, and unlike the previoua results, were not obtained a t a fixed position in the phase diagram. As excess titanium was added, difysrent phases formed, and X-ray analyses showed Tic C .fTiB2 aft8erthe first heating and T i c TiB TiB2 after the find heating. The effect (orthorhombic) of the additional titanium manifests itself in a 20-fold increase in tibanium pressure a t 2300°K., as inspection of the values of in section 3 of Table I will show, and corresponds to the formation of TiB. Titanium diboride was always present and each set of pressures gives consistent values for its heat of vaporization. Results for eleven sets of d a h give an average heat of vaporization for TiBz of 428.6 f 5.3 kcal./mole and a heat of formation of -50.9 f 5.3 kcal./mole. The pressures from run No. 10 of Table I, taken in the TiC-TiB-TiBz region of the phase diagram, yield for the reaction TiB(s) -+ Ti(g) B(g), log K ~ w ,= -11.07, AIi'0n3a~= 118.5 kcal./mole, and = 279 kcal./ mole. The: associated heat of formation of TiB(s) a t 298'K. is --34 kcal./mole.
+
+
+
RICHARD K. WOLFORD AND ROGER G. BATES
1496
TiBz crucible which 11-as held in a tungsten outer jacket. The boron pressure determinations are in agreement with the data of this report although the titanium pressure was anomalously high. The discrepancy in the titanium pressure is believed to be due to titanium vapor escaping from the region between the TiBz cell and the tungsten jacket, possibly enhanced by reaction with the tungsten, thus leading to erroneous heats of vaporization and formation of TiB2. In a recent experiment with a high temperature calorimeter, Lox-ell and Williams!’ obtained AHfO = -50 f 5 kcal./mole for TiB2,in agreement with the result of the present> investigation. The question of vaporization coefficients has not been systematically investigated, but the values of u for the t x o cells were different by one decade and resulted in a statistically insignificant difference in the TiBz values. The observation that TiB is found with TiB2 after samples of the appropriate composition range are heated does not preclude the possibility that TiB is stable over a limited temperature range only. However, for TiR(s) to be unstable a t low tempera-
Vol. 66
-
tures relative to Ti(s) and TiBz(s), the heat of the reaction Ti(s) $- TiBz(s) BTiB(s) must be positive.lg The value found in this research is -16 kcal./mole. TiB may be unstable a t some temperatures relative to decomposition to TiBz plus a titanium rich phase, however. If, as suggested by Hansen and AnderkolZ0TiB decomposes a t temperatures above about 2330°K., by the reacTi2B(s) TiB2(s), a few of these tion 3TiB(s) measurements may have been made for TizB and TiBz mixtures. Hoiyever, AF must be small for the decompcsition reaction near the transition point, and no significant error is introduced by the tsscmption that TiB was the phase present with Ti%,. Acknowledgment.-The authors wish to thank Dr. R. Goton for preparing the B& crucibles and C. E. Lowell for the X-ray analyses. Several helpful discussions with Prof. A. W. Searcy are greatly appreciated,
-
+
(19) A. W. Searcy in “Proceedings of Smond International High Temperature Symposium,” Asilomar, California (Oct. 1959), MoGrawHill Book Co , Inc., New York, N. Y . , 1960. (20) M. Hansen and K. Anderko. “Constitution of Binary Alloys,” MoGrawv-Hill Rook Co., Ino., New York, N. Y., 1958.
KISETICS OF THE HYDROLYSIS OF ACETAL I N 9-METHYLPROPIOXAMIDE-WATER AND TU’,P6-DInlEa’I3CUL~O~~~A~~I~ WATER SOLVESTS AT 20, 25, 30, ASD 40’ BY RICHARD Ez. WOLFORD AND ROGER G. BATES Solution Chemistry Section, National Bureau of Standards, Washington 25, D. C. Received ,March IY, l g S 8
Rate constants for the acid-catalyzed hydrolysis of acetal in the binary solvent mixtures K-methglpropionamide-water and iY,N-dimethylformamide-~~ater have been obtained as a function of solvent composition and temperature. There is a marked decrease in the second-order rate constant when either amide is added to the aqueous solvent. The energy of activation remains essentially constant in the temperature range 20 to 30” but manifests an unexplained complex behavior in the 30 to 40’ range. The decrease in rate when the aqueous solvent is enriched with either amide appears to follow the Winstein Y solvent parameter but shows no simple relationehip with the chanqe of dielectric constant of the bulk sclvent. The various factors which mav influence the rate of the reaction as the solvent composition i s altered are discussed. Qualitatively it seems likely that the abrupt drop in rate of the acid-catalyzed reaction reflects to a large degree a decreased proton energy level resulting from the addition of relatively basic amide species.
Introduction The effect of a change in composition of a binary solvent mixture on the acid-catalyzed hydrolysis of acetal has been studied at8 different temperatures and acid concentrations. The reaction
+
CH~CH(OC~H~)Z HfzO* CHSCHO
+ 2CzHsOH
was chosen for several reasons: it has been studied many times in aqueous solution,1 , 2 the hydrolysis reaction does not affect the acidity or ionic strength of the solution, the uncatalyzed hydrolysis is negligible, and the reaction appears to be catalyzed only by hydrogen ion. The non-aqueous constituents of the aqueous
binary solvent mixtures were N-methylpropionamide and N,X-dimethylformamide. The dielectric constants of these compounds (172.2 and 36.71, indicate a high degree of respect,ively, a t solvent polarity. The study of rate processes in amide-water mixtures may be expected to throw some added light on the structure of the solvent and the nature of specific solvation or “solvent sorting” around a charged ion. The kinetics of a few reactions already have been investignted*J using the pure amides and their mixtures as solvents. Studies dealing wit.h the basicities of the amides have been published16g7 while proton magnetic resonance techniques ha,ve
26, 59 (1929).
(3) C. R. Leader and J. 1’. Gormlev, ibid.. 7 5 , 5731 (1951). (4) S. D. Ross and hI. ill. Lahes. ihid., 79, 4155 (1957). ( 5 ) L. R . D e m o n , J. E. Berger, and E. C. Eokstrom, J . Phys. Chem., 65, 986 (1961). (6) R . S. L. Martin and I. H. Reeoe, J . Chem. Roc.,4697 (1960).
(2) L. K. J. Tong and A. R. Olson, J . A m . Chem. Soc.,66, 1704 (1943).
(7) A. R. Goldfarb, A. Mele, and N. Gutstein. J. Am. Chem. SOC., 77, 6194 (1955).
(1) J. N. Br@nstedand W. F.K. Wynne-Jones. Trans.Faraday h e . ,
