sign in sign up
<<
Home , Beryllium oxide, Magnesium oxide

JO URNAL OF RESEARCH of the National Bureau of Standards- A. Physics and Chemistry Vol. 67A, No.4, July- August 1963 Thermodynamic Properties of and Oxide from 298 to 1,200 oK 1

Andrew C. Victor2 and Thomas B. Douglas

(February 26, 1963)

As a step in developin g new standards of hi gh-temperature h eat capacity and in deter­ mi ning accurate thermodynamic data fOl" simple s ubstances, t he enthalpy (heat content) relative to 273 oK, of hi gh purity fu sed , MgO, and of sintered , BeO, was measured up to 1,173 ole A B unsen ice calorimeter and the d rop m ethod were used. The two samples of BeO measured ha d s urface-to-volume ratios differing by a factor of 15 or 20, yet agreed with each other closely enough to preclude appreciable error attributable to the considerable surface a rea. The enthalpies found for MgO arc several percent highe r than most previously reported values. The values arc represented within t heir uncertainty (estimated to average ± 0.25 % ) by t he foll owing empirical equations 3 (cal mole- I at l' OK )

MgO : T-f ~- T-f;73 15= 10.74091'+ 1.2177(1 0- 3) 7'2 - 2.3183(10- 7) 1'3 + 2.26151 (10 5) 1'-'-3847.94. BeO: Tl ~- II;73 1 5= 11.10841'+ 7.1245(10- 4) 1'2 + 8.40705(105) 7'-1 - 5.31245(107) T-2- 5453.21. Value'S of rnthalpy, IwaL capacit.v, rntropy, a nd G ibbs free-energy function a rc tabulated from 298. 15 to 1, 200 oK.

( 1. Introduction Canada. Spectrochemical and spectrographic analyses at the Nation al Bureau or " Standards A previous paper [1] 4 has described the need for indicated that the sample contil.in ed 99 .90 weight heat-capacity standards at temperatures above the per cent MgO if the detected metallic impurities are range o/" the presen t a-aluminum oxide (corundum) assumed to be present as their highest sLable standard. Enthalpy measurements on (table 1) . dioxide have been presen ted [1] as a first step in the in vestigation of new materials for this purpose. TABLE 1. I mpurities in the samples The chemical stability and high melting poin ts Element a BeO M gO of MgO and B eO (3 ,000 and 2,800 oK , respectively sample 1 [2]) recommend these materials for consid eration ------1------as possible heat-capacity standards. Although both A g ______lVeight % Weight % compounds have lower meltin g points than thorium Al ______('J < 0. 001 Be ______0. 007 . 004 dioxide (3 ,300 oK [2]), the lower sensitivity of their Ca ______

1 The measurements on M gO were su pporteci by th e Wright Air Development respectively). These two samples, whose densities Divisioll..J... Air Research and Dc velopl11clli ComJl1 and, United States Air Force, Wright-L'atterson Air Force J3ase, Ohio. were about 72 and 50 percent of the single-crystal 2 Present acidress: U .S. Naval Orci nance T est Station, China L ake, Calif. (X-ray) value, will hereafter be referred to il.S BeO a US ing the defin ed t hermochemi cal calorie = 4.1840 joules . • F igures in brackets indicate the literature referell ees at tbe end of tbis p aper. samples 1 and 2, respectively. Spectrochem ical 325 analyses of both samples at the BUTeau indicated T ABLE 2. R elative enthalpy of magnesium oxde " t hat they contained 99 .96 percent BeO by weigh t (table 1). In a petrographic examination, sample 1 was found to consist of approximately isometric F urn a.ce I n dividual Mean Mean tempera- enth alpy observed Calc. eq observecl- particles 25f.J. on an edge. BeO sample 2 was ob­ turc, T ]11CaSllrc- enthalpy ( I) ca.lc. served to be composed of needlelike particles esti­ lncnts b ---- ",J mated to average 10}l in length and 1f.J.2 in cross \ oJ{ cal mole- 1 cal mole-I cal molrl ca l mole-1 section. The surface-to-volume ratio of such a 373. 15_. ____ { 923.5 922.7 } 023. 1 923.6 - 0.5 particle is the same as that for a cube with an edge ISS0.3 473.1 5_. ___ ~1 { 1960.7 ] 960. 1 of 1. 5f.J.. The sintered samples of BeO used in the 1961. 1 } + 0.6 3060. 4 573.15 __ __._ { 3050.2 3059.2 enthalpy measurem ents each consisted of two cylin­ 3058. 0 } + 0.0 4194.8 ders 2 cm long and 1 cm in diameter. 4203.6 4197.5 4109. 3 - 1.8 The samples were sealed in containers of annealed 4194.0 } "'I 5375.5 pme preparatory to making enthalpy measm e­ "WI773 .1 5_. ____ 5371. a } 5374. 4 5369.6 + 4.8 ments [1] . BeO sample 2 lost weight during the 5376.7 873.15 __ ___ ~ { 6560.6 - 4. 1 first attempts at sealing it in its container. Further 6558.2 } 6550. 4 6563.5 973.15 ___ __ ~ { 7777. 6 - 0.9 study of the weigh t following successive h eat treat­ 7773.6 } 77 75.6 7776. 5 lO73.15 ._. __ { 9007. 5 ment and exposure to the room atmosphere showed 9000.1 } 9008.3 9005.3 + 3.0 ) t hat at least 0.3 per cent of the original sample mass 1173.l5 .. __ ~ { lO245.4 10246.0 10247. 1 - 1.1 was lost on heating to about 1,100 oK, but was lO246.6 J regain ed by the sample after cooling in a desiccator a Mol wt= 40.311 g. an d then standing in room ail' for 30 min. T ests on b Sample lnass=JO.7578 g. six specimens of thi s lower- sample or BeO all showed the same hygroscopic behavior. Simil ar tests mado with BeO sample 1 showed no detectablo mass change. When BeO sample 2 was fina ll)~ TABLE 3. Relative enthalpy of beryllium oxide a sealed in its container its mass was the lowest attain­ able by the heat treatmen t mentioned above. It is I ndividual enthalpy M ean possible, however, that the sample was still co n­ F urnace tOlll- lllcasurCHl.cn ts observed Mean taminated by a small amount of water. perature, l' enthalpy Calc. eq (2) obser ve cl - (sa lilple I) calc. Sampl e 1 h Sampl e 2 c 3. Enthalpy Measurements 0J< cal mole- 1 cal mole-1 cal mole- 1 cal mole I caZmole- 1 323. ] 5 ~~ ___ ~ __ ~ __ { 303.6 305.6 T 11 e "chop" method and cal orimoter employed 303.9 305. 5 } 303.7 303.7 +0.0 665.5 in the enthalpy measuremen ts have been described 373.15 ~ ~ _~ __ ~_ ~ __ 665.8 } 662.2 662.6 -0.4 in detail ill a previous publication [3]. In brief, t11 e L ~~ ~~~ ~ ~ 666.6 473. 1 5~~ ___ ~ ___ __ { 1503. 1 1490.3 -0.0 method used was as follows . The sample, sealed 1500.4 1497.8 } 1501. 8 1501. 8 2457.4 in it silver container, was suspended in a silver-core 573.15 ______2452.6 2453.7 2452. 7 +1.0 furnace until it Imd time to come to a constant L ~ ~~~~ ~ ~ - 2453. 3 } 673.l5 __ ~ __ ~ __ ~ __ { 3481. 6 34.79.8 .! known temperatme. It was then dropped (wilh 3477. 0 3481. 0 } 3479.3 3478.0 + 0.4 I 4558. 0 4562.2 } almost free fall) into the Bunsen ice calorimeter, 773 .l 5 _~~~~ ____ ~~ { 4555 . 1 4557.8 4559.6 -2. 2 which measured the heat evolved b y the sample 4560.3 ~ -~~~~~~~~~- 873. 1 5_~~~ __ ~.~~_ { 5670.8 5684.0 plus container in cooling to 273 .15 Ole In order to 5684. I 5684.2 } 5682.0 5682.4 -0.4 6843.4 6841. 4 973.15_ ~~_~_~_~ __ { 6840.7 6839.5 account for the enthalpy of the silver con tainer 6838. I 6843.3 } + 1. 2 { 8025.9 8026.7 and the heat lost during the drop , a similar experi­ 1073.1 5~ ~ ~ ___ . ~~_ 8022.2 8026.3 8025.6 + 0.7 mon t was made with an identical empty container .--~~~~~ - ~- 8024.2 f 9242.4 at the same furnace temperature. The difference ______. { 9232.6 0236.5 9237.2 - 0.7 1173.15 ~~ ~ 92:J7. G between the two values of heat is a llleasme of tIl e 9233. G enthalpy change of the sample between 27 3. 15 OK ~~~~~ ~!~~~U and the furnace temperature. a Mol wt=25.012 g. .( b M ass of sample 1=0.7171 g. I The temperatme of the central portion of the , Mass of sample 2= G.3H7 g. furnace was measured by a strain-free platinwn resistance thermometer (ice-point l"esistance, about '" 24 0111ns) up to and including 873 OK, and by an <:: I annealed Pt/P t-l0 percent Rh thermocouple at The measured heat values obtained in individual higher temperatmes.5 The thennOlneter and ther­ runs on an em.pty con tainer that were used in the mocouple were frequently intercompared during the calculations of this paper are recorded in a previous heat measurelnents, were calibrated shortly after­ publication [1]. The second colum.n of table 2 wards, and were concordant with the good consistency gives the fully corrected measmed enthalpy values, of these two instruments over a period of more than in defined tbermochemical calories per mole, for the 5 years. magnesium oxide. In table 3, the second and third columns contain the corresponding information for 'Temperatures are on the International Temperature Scale of 1948 as medi fled in 1054 [4J. BeO samples 1 and 2. The individual enthalpy 326 Y~l l ucs li sted in these two tables were obtained by TABLE 4. Thermodynamic properties of magnesium oxide subtl'tLc g the emp ty contain er valu es (usin g the CO mca n ftt each temperature) from the observed l' II~- II;o K S~-S;oK F;·-H;oK en t11alpies for sample plus con tain er and then con­ • - --']'- ve rting to molar vftlues. Correction s had b een OJ( cal! mole-d eg cal/mole cal/mole-deo cal/mole-dey appli ed in the usual manner rl]. The enthalpy 298. 15 8.906 J234.6 6.439 2.298 y correction for the impurities in the MgO ft lTlounted to 300 8.939 J251.1 6.494 2.324 I 320 9.261 1433.2 7.082 2.603 less than 0.02 percen t of the entllttlpy of the sample. 340 9.532 1621. 2 7.651 2.883 I For BeO an impurity correction of less th ft n 0.01 360 9.782 18H.4 8.204 3. 16'J percent of the sample enthalpy was used. Tbese 380 10.000 2012.3 8.738 3.443 I 400 10.190 2214. 2 9.256 3.721 corrections were made assuming that the foreign 420 10.359 24 19. 7 9.758 3.996 I 440 10.510 2628.4 10.243 4.269 (- chemical elements were present as the highest stable 460 10. 645 2840.0 10. iJ3 4. 539 oxides and that their enthalpies contributed ftCldi­ 480 10.768 3054.2 11. 169 4.806 ~ tively to the total observed enthalpy. The cal­ 500 10. 880 32iO.6 11 .611 5.070 550 11. 122 3820.9 12.660 5.712 culated values of enthalpy in tables 2 and :3 are 600 11.324 4382.2 13.636 6.332 smoothed values arrived at as described in section 4. 650 11. 495 4952.8 14.549 6.930 700 11. 643 5531. 3 15,40i 7.505 i50 11. 774 6 U6.8 16.215 8.059 4. Smoothed Thermodynamic Functions 800 11.891 6708.5 16.9i8 8.593 850 11. 996 730.).7 17. i02 9.108 900 12. 090 i907.8 18.391 9.604 The mean en tktlpy valu es for MgO and Jor BeO 950 12.1i6 8514.6 19.047 10.084 sample 1 wcrc used to derive eqs (1) and (2) (in 1000 12.255 91 25. 4 19.673 10. 548 1050 12. 326 9739.9 20.273 10. 997 cttl mole- I ftt T OK ) by the lll etllOd oJ lcast squares. 1100 12.391 10358 20.848 11. 432 TIl e rcsults for BcO sample 2 were ignored, for the 1150 12.451 10979 21. 400 11. 853 reasons discussed in sec Lion 5. 1200 12.505 11 603 21. 931 12.262

M gO: H;-H;7:J l.5 = 10.7409 T + ] .2 177(1 0- 3) T 2 7 5 -2.3183(10- ) '['3+ 2.26151(10 ) T - I- 3847.94. (1) TAB I , l ~ 5. Thermodynamic properties of beryllium oxide BeO: U ;'-EG73 15= 11.1084T+ 7.1245(10- 01 ) T 2 + 8.40705(105) T - I- 5.31245(lOi) T -2- 5453 .21. (2) 110 0 so () _ ( F;-~;9S . 150K) l' CO T-l-JZ9B.150K ~T-S2QB.150K . -S!l9S.1S0K EnLhD lpy valu es calculated fr om cqs (1) rmd (2) aro Labuhtted and cOJllpared with experim entaJ values oJ( cal/mole-deo cal/mole cal/mole-deo cat/mole-dey 298. 15 6.102 0 0 0 in Lhe last two column s of t~tb l ('s 2 and 3. T lL bles 300 6. 146 11. 3 0.038 0.000 4 a nd 5, which g-ive the COJllmon thermody namic 320 6.597 138.8 .449 .015 340 i.012 2i5.0 . 862 .053 proportics of 111 agnesiulH oxide and boryllium oxide, 360 7.393 419.1 1. 273 .109 ,,' C1'O obLained by a four-point JHlmeric,d inLognltion, 380 7.746 5iO.5 1. 683 . 181 11 t 10-clcg intervals, of hea t-cn p data [5, 6], are given relative to 0 OK. Becauso no low-temperature thermodynamic data have been published for macrocrystalline BeO, the S,Lmo 50 Discussion r thcrmodynamic properties of this substance m'e given relative to 298.15 OIL The statements made in an earlier pttper [1] con­ On the basis or previous work with the BUllsen ice cerning the usual precision of enthalpy measure­ calorimeter, the authors bclicve that the uncertainty ments made with the Bunsen ice Cttlorimeter apply of the thermodynamic properties in tables 4 and 5 again in the present paper. Tho doviation oJ eq corresponds to 0.4 percent in the . (1) from the mean obser ved enthalpios of MgO Changes in values of the thermodynamic properties averages 0.04 percent or 1.9 caI/mole. The corres­ introduced by smooth-joining to low-temperature ponding numbers for eq (2) and BeO sample 1 are dftta are within the cxperim cntal uncertainty. 0.02 percent or 0.7 cal/mole. The m aximum clevia- 327 tion of eq (1) from the mean observed enthalpies are in general many times the estimated uncertain­ is 0.09 percent, while that of eq (2) is 0.06 percent. ties of the authors' values, a fact which is in line with Beryllium oxide is at present difficult to obtain the belief that the latter are the most accurate in the form of particles large enough to be sure that available for this substance. the total surface area is too small to affect the heat In figure 2 the present BeO measurements are I capacity appreciably. For this reason two samples compared with those of earlier workers [6, 11,12, 13]. of widely different surface-to-volume ratios were Of particular importance is the recent work by "'1 measured in this investigation in an effort to test Rodigina and Gomel'skii on a sintered specimen, this possible complication, which has been discussed which shows a reproducibility comparable with that by one of the authors in a recent publication [7]. of the present measurements and agreement with eq Columns 2 and 3 of table 3 show that sample 2 (2) which is generally better than 0.5 percent of the consistently exhibited slightly higher relative­ enthalpy relative to 273.15 oIL I enthalpy values at 323 oK and 373 oK than did Magnesium oxide is probably the more favorable ~ sample 1, though there is no consistent difference of the two materials as a high-temperature heat­ in sign at the higher temperatures. In contrast capacity standard, since it is available in large crys­ < to sample 1, sample 2 had a surface-to-volume tals of reasonably high purity. The agreement ratio 15 to 20 times as large, and also may have between BeO samples 1 and 2 at and above 473 oK retained a very small amount of water (see sec. 2). and the unusually good agreement with one other Both factors would tend to increase the measured worker [12] on a specimen similar to sample 1 support heat capacity somewhat, and may have contributed the serious consideration of sintered beryllium oxide to the differences between the two samples noted as a heat-capacity standard, subject to the conditions up to 373 oIL However, since these differences do that the samples used do not have too small a particle not exceed 1 percent, it appears unlikely that the size and are not contaminated by adsorbed water. relatively small surface area of sample 1 could in A phase change in BeO has reportedly been this temperature range have contributed more than observed at approximately 2,220 oK [14]. Such a a few hundredths of 1 percent to its heat capacity. transition would adversely affect the use of beryllium All the smoothed results for beryllium oxide in this oxide as a heat-capacity standard at higher tempera­ paper are based on sample 1. tures. Measurements on this material up to 1,800 oK I Figure 1 affords a comparison of the present meas­ or higher are planned at the National Bureau of < urements on MgO with previously reported values Standards to extend the temperature range of avail­ [8, 9, 10] in the form of percentage deviation of indi­ able accurate values. At the present time the vidual unsmoothed measurements of (IiT -H;73.l5 OK) heat-capacity discrepancy between reference [13] and from eq (1). The deviations of other workers' the present work is greater than 2.5 percent at values from the measurements reported in this paper 1,200 oK.

1.0 ,---,-----,---,---'---,--1--,---.,.----.,.----,----,

0 .5 o

00 ~ o,o r----~~.r--~~~~&~.~--·~---r~·-----.~-~.r---~~~-~~ ~ 0 ·.... · N 0 '7 0 -0.5 o i~ ~ ~ -1.0 If) a 6~ f= (5 - 1.5 >~ w, W 0 o ~ -2.0 -.J 0: & f':! W Z gj -25 ~.Q . it' o ~ -3.0 x w o -3.5 o - ~47% 0 1 1 1 -4.0 1 1 & L___ ~ __J I __~ I ___~L_ __L_ __ L__~ 300 400 500 600 700 800 900 1000 1100 1200 TEMPERATURE , oK J FIGU RE 1. Comparison of the enthalpy, relative to 278 oK, of magnesium oxide obtained fron1 eq (1 ) with individual val1.es obtained in va1'ious investigations. c; (Some of the observed points h ave been di splaced horizontall y by small amounts in order to avoid the confusion of overlapping.) -, NBS smoothed, eq (1); . , N BS observed (present p aper); A , Arthur [8]; 0, M agn us [9]; 0 , I Wilkes [10]. 328

------2.5 ,------,------,----,------,-----,----,-----I --'---I --~1--,-----[--

2.0 , (

x -2.0

-2.5 L--=-":-.,---_--:-':--=--_~_::_---=-':_:_-----'---L---.l---L---..l-----.J 300 400 500 600 700 800 900 1000 1100 1200 TEMPERATURE , oK

FIGURE 2. Compm'ison of the enthalpy, Telative to 273 oK , of beTylliwn oxide obtained j1-om eq (2) with individual values obseTved in vaTious investigations.

(Some of the observed pOints h ave been displaced horizontally by s1l1all amounts in order to avoid t.he co nfusion of overlapping.) -, NBS smoothed, eq (2); . N BS observed (presen t paper) , sample I ("high density"); o, N B Sob­ served (present p a per), sarnplc 2 (" low density"); x, Magnus and Darn llJ]; 6., Rodigin a and Oo mcl'skii l I2]; 0 , K ancl yba et a1. (13] ; "', }' urukawa and Reilly [6] .

The authors acknowledge help from certain mem­ [7] A. C. Victor, J. Chem. Phys. 36, 2812 (1962). [8] J. S. Arthur, J. Appl. Phys. 21, 732 (1950) . bers of the Bureau: F. Knudsen prepared the samples [9] A. Magnus, Phys. Ztschr. He, 5 (1914). of beryllium oxide, and B. F. Scribner and his associ­ [10] G . B. Wilkes, J. Am. Ceram. Soc. 15, 72 (1932). ates analyzed all the samples spectrochemically. [11] A. Magnus and H. Daln, Ann. Ph~ r s ik , [4]81, 407 (1926). [12] E. N . Rodigina and K. Z. Gomel'skii, Zhur. Fiz. Khim. 35, 1828 (1961). 6. References [13] V. V. Kandyba, P. B. Kantor, R. M. Krasovitskaya, and E. N. Fomiehev, Doklady AI,ad. Nauk S.S.S.R. 131, [1] A. C. Victor and T. B. Douglas, J . Rescarch NBS (Phys. 566 (1960). and Chem .) No.2, G5A, 105 (1961). [14] S. B. Austerman, High-temperature phase transition in [2] F. D . Rossini , D. D . Wagman, W. H. Evans, S. Levine, beryllium oxide, paper presented before the American and I. Jaffe, Selected values of ehcmical t hermody­ Physical Society, N ew York, N.Y. (Jan. 24, 1962). namic properties, NBS Circ. 500 (U.S. Government Printing Office, Washington, D.C., 1052). h [3] G. T. Furukawa, T. B. Douglas, R. E. McCoskey, and D. C. Ginnings, J . Research NBS 57, 67 (1956) I ~ RP2694. [4] H. F. Stimson, Am. J. Phys. 23, 614 (1955). [5] T. H . K. Barron, W. T. Berg, and J . A. Morrison, Proc. Roy. Soc. (London) 250A, 70 (1959). [6] G. T. Furukawa and M. L. R eilly, private communica­ tion. (Paper 67 A4-221) \,. > ;;, f r ~ .

329