Molar Volume and Structure of Solid and Molten Cesium Halides1

Molar Volume and Structure of Solid and Molten Cesium Halides1. J. W. Johnson, P. A. Agron, and M. A. Bredig. J. Am. Chem. Soc. , 1955, 77 (10), pp 27...
0 downloads 0 Views 503KB Size
2734

J. \V.JVI-INSON, E'. Ll.XCRVX A N D hl. &l. Bmmrc; [CONTRIBUTION FROM

THE

Vol. 77

OAKRIDGENATIONAL LABORATORY CHEMISTRY DIVISIOK]

Molar Volume and Structure of Solid and Molten Cesium Halides' BY J. W. JOHNSON, P. A. AGRONAND M. A. BREDIG RECEIVED JANUARY 6, 1955 X-Ray diffraction patterns of solid cesium iodide and bromide taken a t temperatures u p to the melting points gave no evidence of any trysition from a simple cubic to a face-centered cubic structure similar to the transition which occurs in cesium chloride a t 469 . Discussion of the molar volumes of the solid and liquid salts a t the melting points leads to theoconclusi?n that the structural change from 8- to 6-fold coordination, which in the chloride takes place in solid phase a t 469 , occurs in the bromide and iodide a t their melting points as part of the fusion process. The per cent. volume expansions on melting for the bromide and iodide, 26.8 and 28.5, are essentially equal to the sum of the expansions in the chloride a t it? transition and melting points, 17.1 10.0 = 27.1%.

+

Introduction I n the course of investigations concerned with the miscibility of alkali metal halides with alkali a need for knowledge of the volume changes occurring in the melting process for these salts became apparent. A search of the literature revealed that information on this subject is scarce. Lorenz and Herz4 compared the molar volume of the solid salt a t room temperature with the volume of the liquid salt a t the melting point and Stewart5 calculated the ratio of the density of the solid a t room temperature to that of the liquid a t some temperature above the melting point. A compilation of densities, linear expansion coefficients, and lattice constants for the alkali metal halides in the solid and liquid state, as reported in the literature, was made as part of the present study. This information was used to calculate the volume changes for the transition solid-liquid a t the melting point of each salt. The alkali metal halides are naturally divided into two groups; first, those existing in the simple cubic structure a t room temperature, under normal conditions, consisting of cesium chloride, bromide and iodide; and second, the much more numerous remaining members which exist in the face-centered cubic structure and which include cesium fluoride. This discussion will be confined to the first of these groups, while a subsequent paper deals with the second one. A preliminary check of available data for the three cesium halides above indicated that the percentage volume expansion of the bromide and iodide on melting was approximately twice as large as that exhibited by the chloride. Since such a wide variation in the volume expansion of the halides of a given alkali metal was found to be unique, this was examined more thoroughly. Wagner and Lipperts have measured lattice constants of these salts between room temperature and approximately fifty degrees below the melting point. It seemed desirable to extend these measurements to the melting point in an effort to confirm the large volume expansion of cesium bromide and iodide on melting. (1) This paper is based on work performed for the U. S. Atomic Energy Commission. (2) M. A. Bredig, J. W. Johnson and Wm. T. Smith, Jr., THIS J O U R N A L , 7 7 , 307 (1955). (3) M. A. Bredig, H. R Dronstein and Wm. T. Smith, Jr., ibid., 77, 1452 (1955). (4) K. Lorenz and W. Hem, C. anorg. Chcm , 146, 88 (1926). 15) C. W. Stewart, Trans. F a r n d a y Soc., 33, 238 (1037). (I;) G , Woguer a n d I.. Lipprrt, Z . p h y s k . Chem., BS1, 2C3 (1933).

Experimental The cesium bromide and iodide in the form of clear fragments of optical grade salt were ground to a fine powder and mixed with nickel powder a? an internal X-ray diffraction standard. The sample holder of the X-ray diffractometer was a flat plate and incorporated a cylindrical beryllium metal cover which prevented loss of the sample by vaporization while not obstructing X-ray passage. Heating to various temperatures was accomplished by a furnace the core of which was wound with nichrome V wire and had a semi-circular slit for the entrance of the primary beam and emergence of the scattered beams. The primary and scattered beams also passed through a beryllium strip 10 mil thick in the metal envelope surrounding the furnace, and the scattered beam was scanned with a Geiger-Muller counter tube connected to a Brown Electronik recorder. An atmosphere of helium was maintained inside the beryllium can of the sample holder and around the furnace core. Temperatures were measured with a chromel-alumel thermocouple whose junction was placed in the sample holder, and which was directly connected to a Rubicon potentiometer. Compensation for room temperature was made by subtracting the corresponding millivolts from the reading obtained on the potentiometer. The validity of the temperature reading on the plate as a measure of the sample temperature was established by comparing the transition temperature of cesium chloride obtained in this fashion with that reported by Menary, Ubbelohde and Woodward? in their precise determination of the transition with X-ray diffraction. These authors reported the transtion in cesium chloride occurs a t 469" and our measurements indicated a temperature of 470'. These temperatures are considerably higher than 451' Zemczuiny and Rambach,8 and 443' KeitelO obtained from cooling curves, while it is lower than 479" Korranglo obtained from heating curves. Menary and co-workers' state that hysteresis accompanying the transition makes accurate determination of the transition temperatures by thermal methods very difficult. Wagner and Lippert6 reported 445 =k 5" after determining the temperature from a calibration curve of temperature ws. wattage, without stating the melting points of the standards used.

Data and Calculations A comparison of our smoothed data for the lattice constants of cesium bromide and iodide (Tables I and 11) with the experimental data of Wagner and Lippert6 (Fig. 1) shows some discrepancy between the two sets of values, especially in the curvatures a t the higher temperatures. Expressions for CsBr: ab = 4.293 (25-630 ")

and for CsI: at = 4 549 (25419')

+ 1600 X

+ 1767 X

10-4t

+ 1529 X 1 O - V (1)

10-4t

+ 1528 X

10-7t2 (2)

(7) J. W. Menary, A. K. Ubbelohde and I. Woodward, Proc. R o y . Soc (London), A208, 158 (1951). (8) S.Zemczuzny and F. Rambach, 2.anorg. Chcm.. 66, 403 (1910) (9) IT. Reitel. Npues Jaiirb. Mineyo2 Geol., 8378, (1925) (10) E . Kormng, Z. anorp. C h e m . , 91, 101 (1915).

May 20, 1055

MOLAR VOLUME

AND STRUCTURE O F SOLID AND

MOLTENCESIUM HALIDES 2733

relating the lattice constant at in Angstrom units with the temperature in degrees centigrade (eq. 1, 2) were derived from our data and used to construct the curves in Fig. 1. The agreement beTABLE I LATTICE CONSTANTS A N D DIFFERENTIAL LINEAREXPANSION COEFFICIENTS OF CESIUM BROMIDEAS A FUNCTION OF TEMPERATURE Run no.

Temp. t.

oc.

Lattice constant (A,) Obsd. Calcd.

25 4,290" 179 4.328 208 4.310 270 4.352 297 4.361 1 320 4.361 1 380 4.374 3 395 4.380 4 443 4.400 3 475 4,407 1 520 4.410 2 532 4.417 3 546 4.426 4 580 4.436 1 599 4.432 2 625 4.445 1 620 4.451 4 630 4.466 636 ( m p , ) Average of eight measurements.

3 4 3 3

Linear exp. coeff. X 108, ' C .

39.0 49.8 51.8 56.0 57.9 59.4 63.9 64.5 67.0 69.8 72 7 73.4 74.4 76.5 77.8 79.4 79.4 79.7 80.1

4 297 4.327 4.333 4.347 4.355 4,300 4.378 4.380 4.304 4.404 4 418 4 422 4 426 4 438 4 445 4.453 4.454 4.455 4,455

4

4.500c

4

.

4.300

-_-0

4--00 0/O-O 1

4

1

_-- --100

Fig. 1.-Lattice

/

1

-0- Wagner and

Lippert - Present Work I

200 300 TEMPERATURE

400

500

600

(OC.1.

constants of cesium bromide and iodide 3 s a function of temperature.

tween the experimental values and those calculated by eq. 1and 2 is shown in Tables I and 11. The reason for the discrepancy between the measurements of Wagner and Lippert6 and the present results is not clear, but the methods of both temperature and lattice constant determination in their small camera are subject to some question. Differential linear expansion coefficients, p, derived by differentiating the equations above and dividing by the lattice constant, nt 1

dn

at=,xs

(3)

(1

are also listed in Tables I and 11. The value for

TABLE I1 CsI a t 25", 40.4 X 10-6 degrees C.-l is in fair LATTICECOXSTANTS A N D DIFFERENTIAL LINEAREXPANSIONagreement with the average linear expansion coCOEFFICIENTS OF CESIUMIODIDE AS A FUNCTION OF TEMefficient of cesium iodide, 48.6 X l o 4 degrees C.-l PERATURE

Run no.

Temp. 1,

O C .

Lattice constant (A,) Calcd. Obsd.

25 4.553" 4.563 4 83 4 234 4.596 3 251 4.615 2 276 4.617 1 278 4 . 606 4 320 4.619 3 33 1 4.628 3 343 4 . 630 2 354 4 . 630 2 388 4. 637 3 393 4.651 3 398 4.636 3 446 4 . 600 2 402 4.659 3 485 4.672 4 498 4.670 3 512 4.681 1 525 4.680 4 539 4.680 3 572 4.696 1 594 4.710 4 602 4.708 3 607 4.716 4 619 4.713 621 ( m . p . ) = Average of eight measurements.

4.553 4 . 565 4.598 4.603 4.610 4.610 4,621 4.624 4,628 4.631 4.640 4.642 4.635 4.658 4.664 4.671 4.674 4.679 4.684 4.688 4.700 4.708 4.710 4.712 4.716 4.718

Linear exp. coeff,

x

108, o c . - 1

40.4 44.3 54.0 55.0 56.6 56.7 59.4 60.1 60.8 61 5 63.6 63.9 64.4 67.2 68.2 69.6 70.4 71.2 72.0 72.8 74.8 76.1 76.6 76.9 77.6 77.7

between 22 and 36", reported by Rymer and Hambling" in their precision determination of the lattice constant of this salt. Molar volumes of the solid salts were calculated directly from the lattice constants by multiplying the volume of the unit cell with Avogadro's number and dividing by the number of ion pairs per unit cell Simple cubic Face-centered cubic

V = Na3 V = Naa/4

where V is the molar volume in cc. and a is the lattice constant in cm. N was taken as 6.0235 X loz3from the compilation of Dumond and Cohen.12 The molar volumes of crystalline cesium bromide and iodide (Table 111) were calculated from smoothed values of the lattice constants and for cesium chloride from the smoothed values of Menary and co-workers.' The volume of the facecentered structure of cesium chloride was calculated from lattice constant measurements reported by Wagner and Lippert6 up to 530" and a value reported by West13 a t 500'. The volumes of the metastable face-centered modifications a t room temperature were calculated from lattice constants reported by S ~ h u 1 z . l ~The transition temperature in cesium chloride was taken as 469" as reported (11) T. B. R y m e r and P. G. Hambling, Acto C r y s f . ,4 , 565 (1951). (12) J. W. M. Dumond a n d E. R.Cohen, Rrws. M o d . Phys., a l , 651

(1949). (13) C. D. West, Z . K r i s f . ,(A)88,94 (193.1). (14) L. G . Scltulz, J. Chcm. P h y s . , 18, 090 (1950).

J. w.JOHNSON, P. A. AGRON.4ND &I. A. BREDIG

2736

Vol. 77

by Menary, Ubbelohde and WoodwardI7which is not sensibly different from our own value of 470". The melting points used are those listed by the

N.B.S. compilation.'5 The molar volumes of the liquids a t the melting points were calculated from density values reported by Jaeger.'6 Table 111 presents the volume of these salts in TABLE I11 the two modifications and Table I V gives the MOLARVOLUMESOF SOLIDA N D LIQUIDCESIUMHALIDES volume expansion in the melting process in percentage. The data of Tables I11 and I V are sumTemp., CsCl CsBr CSI OC. CN8 CNO CN8 Ch-G CN8 CNG marized in Fig. 2. 25 42.1 50.3" 4 7 . 8 56.9" 5 6 . 8 67.7" Discussion 100 42.6 48.2 57.4 An examination of the data 011 the molar volumes 200 43.2 48.9 58.2 of the solid and liquid salts shows the per cent. 300 44.0 49.8 59.2 volume expansion on melting to be considerably 400 44.8 50.7 60.3 lower for cesium chloride than for the bromide and 469 45.4 53.4 iodide. An extrapolation of the molar volume of 500 53.6 51.7 61.5 the low temperature or simple cubic modification GOO 54.4 52.9 62.9 of cesium chloride up to the melting point would ti21 ( s ) 6 3 . 2 (73.6)" result in a volume expansion on melting of 28.37,, 621 (1) 81.2 in good agreement with that observed for the 636 (s) 5 3 . 4 (62.O)C bromide and iodide, 26.8 and 2S.5%, respectively. 636 (1) 67.7 If the data of Wagner and Lippert6 were used the 6 4 5 ( s ) (47.0)b 54.8 volume changes would be 26.S and 23%, respec645 (1) 60.3 tively, the latter still being considerably larger a Calculated from Schulz data for metastable forms. than the 10% expansion exhibited by cesium Extrapolated from lower temperatures. Extrapolated from per cent. volume change in CsCl transformation and chloride. Schulz data a t room temperature. If the volume expansion (I 7.6%) associated with the transition in cesium chloride from simple cubic TABLE IT to face-centered cubic structure1' is applied to the I'OLUME CHANGES ON MELTIXG FOR C E S I U M HALIDES bromide and iodide and corrected for the increased CsCl CsBr CSI temperature a hvpothetical per cent. expansion CN8 CNG CN8 CNG CN8 CNG on melting of the (metastable) face-centered cubic Mol. vol. solid ( m . p . ) (47.0)' 5 4 . 8 53.4 (02 0 ) G 3 . 2 (i3.G) modifications may be calculated. The calculation Mol. vol. liquid gives 9.2 and 10.3y0expansion, i.e., similar to the (m.p.) 60.3 G0.3 07.7 67.7 81.2 81.2 7% Expansion over value of 10.0% for the chloride. solid (28.3) 10.0 20.8 (9.2) 2 8 . 5 (10.3) The X-ray diffraction patterns of solid cesium Figures in brackets by extrapolation from lower tem- bromide (m.p. 636") and cesium iodide (m.p. 621') peratures. observed up to 2-3" below the melting points gave no indication of a transition in the solid. While the possibility that this occurs in these last few degrees before melting cannot be ruled out, it appears to be unlikely. Thus the conclusion seems inevitable that there is something quite different in the melting process of the bromide and iodide from that in cesiuni chloride. In the chloride the melting consists essentially of the disappearance of the long range order of the crystal, while the short range order is 50.0l I I I I I 1 maintained more or less the same as in the solid, i.e., the coordination of six oppositely charged ions C s Br around a given ion is still present in the liquid.'? 9.2% 3 The formation of a "hole'' even in the first sphere of coordination of a certain fraction of the ions of the melt may or may not have to be assumed to explain the mobility of the liquid. It seems impossible, however, to explain the volume changes exhibited by cesium bromide and iodide on melting by anything but a structural change in the first 60.01 I I I I sphere of co6rdination of all the ions. It is pro3 40.0%

66.0r

',"

J

0 -' > J

u E-

,

------ - - - F.C.C-

50.0

,

4

-----J

cubic

J

0

40'00

128.3%

469'

400

200 300

400

500

600 700

TEMPERATURE ("C,)