H E . ~ CI\P.ICITIES T OF MAGKESIUM DIBORIDE .IND TETR.IBORIDR
July 20, 1037
Rossini and co-workersg list values leading to the following heats of formation from the oxides (kcal. /mole) : - 24.7 for strontium metasilicate, - 33.0 for strontium orthosilicate, - 20.7 for barium metasilicate, and -24.6 for barium orthosilicate, based upon older thermochemical work. The present values are considered much superior. I t is of interest to note t h a t the new values become increasingly negative in going from magnesium to barium in both the meta- and ortho-series whereas the solder work indicated a reversal in this trend. HEATS OF
TABLE I'I' FORMATIOX FROM THE OXIDES AT 298.15'K. ( KCAL./MOLE)
Substance
MgSi03" CaSi03* SrSiOa BaSiO3 a
Substance
AH
-
AH
& 0.15
Mg?SiO," -15.12Zk0.21 Ca?Si04(p) -30.19 .23 Ca?Si04(y) -32.7 & =t Sr2SiOl --50.04 & .24 Ba?SiOd -64.48 =t . 2 8 Clinoenstatite. IVollastonite. Forsterite. 8.69 -21.25 -31.24 -:38.03
=t
+
.13 .I6 .li
The heats of formation of the silicates from the oxides may be converted to heats of formation from the elements by means of available data for the oxides. In this connection, the heat of formation of quartz (-210.26 kcal.,/mole) obtained by (9) F. D. Rossini, D. D. Wagman, W. H. Evans, S. Levine and I. Jade. Nati. Bur. Standards Circular 500, Feb. 1, 1952.
[ CONTRIBCTIOS FROM
THE
:3 611
Humphrey and King"' and the recent value of Huber and Holley" for calcium oxide (-151.79 kcal./mole) are adopted. Shomate and Huffman'? and Holley and HuberI3 have obtained values of the heat of formation of magnesium oxide t h a t differ by only 140 cal./mole; the mean, - 113.X kcal. , ' mole, is adopted. Values for strontium oxide ( - 141.1 kcal. /mole) and barium oxide ( - 133.4 kcal./mole) are from NBS Circular 300.9 These values lead to the heats of formation from the elements listed in Table V, which also brings up-toTABLE V HEATS OF FORMATION FROM THE ELEMEXTS AT 298.15"K. ( KCAL./MOLE) AH
Substance
MgSiO, MgSiO: CaSi03 CasSiO,(P) Ca?SiO,(-,,) CasSiOj
-362.7 -512.9 -383.3 -544.0
-546.5 -692.6
Substance
SrSi03 Sr2SiOc BaSinOj Ba?SiaOs BaSiOs Ba?Si04
AH
-382.6
- 542.5 -595.2 -980.0 -:381 . i -541.5
date the values given by Torgeson and Sahama' and King.? (10) G. L. Humphrey and E. G. King, THIS J O U R N A L7 ,4 , 2041 (1952). (11) E. J. Huber, Jr., and C. E. Holley, J r . , J . P h y s . Chem., 60, 498 (1956). 66, , 1627 (12) C. H . Shomate and E. H. Huffman, THIS J O U R N A L (1943). (13) C. E. Holley, Jr., and E. J. Huber, Jr., ibid., 73, 5577 (1931).
BERKELEY 4, CALIF.
DEPARTMENT O F CHEMISTRY, SVRACUSE UNIVERSITY]
Low Temperature Heat Capacities of Magnesium Diboride (MgB2) and Magnesium Tetraboride (MgB4) BY ROBINSON h f . SWIFTAND DAVIDWHITE' RECEIVED FEBRUARY 14, 1957 The heat capacities of magnesium diboride (MgB2) and magnesium tetraboride ( MgB,) were measured in the temperature range 18 t o 305'K. T h e values of heat capacity, entropy, enthalpy and free energy function have been tabulated a t integral values of temperature. The entropy a t 298.16'K. of MgB2 is 8 60 Zk 0.04 c d . deg.-' mole-', t h a t of MgBl is 12.41 Zk 0.06 cal. deg.-' mole-'. T h e heat capacity of these compounds a t the lowest temperatures measured do n i ~ texhibit a t T* relationship characteristic of some substances having a layer structure.
Introduction The existence of magnesium borides with formulas MgBz?v3and MgBr42has been shown by X-ray diffraction studies and chemical analysis. The diboride has a layered structure in which hexagonal nets of boron atoms are separated by layers of magnesium atoms. The structure of the tetraboride is not known. Recent studies have shown that in crystalline substances with layer structures, such as graphite4 and g a l l i ~ m ,the ~ heat capacity a t low temperatures follows a T 2law, rather than the Debye T 3law. However, for CdI4, a compound having a layered structure, it has been found t h a t neither the T 2 nor T 3 relationship was followed (1) Department of Chemistry, Ohio State University, Columbus, 0. (2) V. Russell, R. Hirst, F. A. Kanda and A. J. King, Acta C r y s t . , 6 , 870 (1953). 76, 1431 (1854). ( 3 ) hl. E. Jones and R.E . Marsh, THIS JOURNAL, (4) W. DeSorbo and W. W. Tyler, P h y s . Regs., 8 3 , 878 (1951); J . Chem. P h y s . , '21, 1660 (1953). ( 5 ) W. DeSorbo, ibid., 21, 168 (1953).
a t low temperatures6 The pronounced anisotropy characteristic of the graphite and gallium structures does not prevail in CdI4 which may account for the lack of a T 2relationship in the latter case. The layer structure of !UgB2 is more closely related to that of CdIc than t h a t of graphite from the standpoint of anisotropy. However, in the case of MgB4, if the hexagonal boron network is retained, the interactions between the layers may be sufficiently diminished so as to lead to a T 2dependence in the heat capacity a t relatively low temperatures. Apparatus and Procedure Calorimeter.-The S e r n s t type vacuum calorimeter which was used for the measurements was similar t o t h a t described by Johnston and Kerr7 with only minor changes being made in t h e calorimeter assembly and vacuum system. (6) A. S. Dworkin, D . J . Sasmor and E R . V a n Artsdalen, THIS JOURNAL 77, , 1304 (1955). (7) H L. Johnston and E. C. Kerr, ibid., 73, 4733 (1950).
,.l l i e caliiriineter itself
liutl a caliacit)- o f : i I ) c ~ i i t(io 1111. 11. cylindrical surface \vas wc~und with .I\YG 411, forrrics-inciilated grild-0.1,51'L silver wire. Thii served as both re~ i s t a n c ethertnotneter and heater and it5 resistance :it room temperature was aljout :300 o l i t i i ~ . The winding was covered n-ith giild f(iil cemented witli Ceneral Electric .\tlheiive S I I 7031. . T h e caloriitietcr a ~ i c i i i h l v\v;ic iti.;ertetl rlirectl\. i r t t c , :t Collinh helium cryii>tnt. Electrical Circuits.--.\ IYliite r l i ILIIIIC p~itetitiistiirs calibrated by the S a t i o n d Bure.:iu o f Standard., .i \\7e~inerpotenticitnetcr with :L 10,000 niicrl I\.(ilt r m g c i v a \ used to i n e i t ~ t i r ethc thcrtnr~c~iuple elcctriiitiotix-e forcci. .\ :j(l,llOi):I O i J < i 1 1 1 1 1 \,sjltagc.tiivitler w a I~I ~ L Y I iti tlie energy circuit. Timing.--The tiniiiig rif tlie energ!' iiiput IY;L. t l i i n e by :I c.lutch-i~perntet1St:intIard Elcctric Tiiiiing C ~ ~ n i ~ ~titiling aiiy clock which read tlircctl\- t o 0.01 i e c . Tliii c l i i c k \v:i\ 51 tlri\-en by the arnplified output ( i f a General Radio Ciitnpan~precision fork and \vas :ictu:itctl b y :t iinglc in:i?ter switcli ivliich also contriilleci t h e energ!- input. Temperature Scale.---Tlic tIiertniiilyii,i~iiie tetiipmiturc 1,:ilioratiiry 1va5 c~~t;ihlislietl h y c . ~ i i i ~ ~ ~ : torfi ~ .t i ~ i i a r i t a i i tlirrtiii i c r i ~ i p l cagainit a ~t;iiiti;irdt1ierii111c r ~ u p l e(,litainetl froin the Ohio State l-tiiver>ity Crj-iigetiic Lalxiratory. anti calibrated tlicre agaiiiit :t priniCir>. standard. Si,. 80.' .~h-ciluteteiiiper:iture~are Iirolj'ibly accurate til i\-itliin +iJ.lG3. Tlie unecrtaiiitiei i i i tile ~ i i c ~ i u r cteiiid pcr:iture differeiicc, o f tlic 1ie:it cip;tcit!- ritii; : i w t i l ) greater t h a n zkll.(J05°. Data Computation.-- Tlie c ~ ~ l c i i l ~ i t i\\ere i i i i ~ inatie 1' , iiiitlineti b y Giuuqiie9 anti J o l i n s t i i r i .: Correction5 were ~n:tclr~ t o thc inearureti lieat capacitic~for the energy loqt in the potential divider a n d lead,, heat leaks, c;iliwirneter surface ~~iperlieating during energ!- iiiput , drift.; in I)lock teinper:iturvs. r:idiatiiiti t o thc c;ilorinictcr froin the blocks, I I O ~ ~ I I tiorneter batter>-drifti, i ~ t i dIie sti~ichiiitrictricqu~ititiLie5 of the elenient.;. T h e tnagiie~iuni \IX\ ;t ciitiiiticrcial grade an:ilyzi~ig of bIgB, and larger amiiuuts of unreacted MgB?. T h e latter, mean temperatures. They never exceeded 1057; of the temperature scale or lo", whichever was smaller. togc.ther witli other soluble impurities, was retnovetl by t re,itinent with ~ r n r i ntiilutc liytlriichloric acid and subwComputed thermodynamic functions for these comqiient ~vashitig n.ith wxtvr. C~insider:iblefree Ixjron repounds are given in Tables IV and T:. The ice tnairied r i t h tlie MgB4. S-Ral-diffraction stuclic.5 identified tlie hubstatice> prei- point was taken as 2i3.1A°K. and the results are crit :rnd chemical analysi.: ebtabliihed the amounts in e x 1 1 expressed in calories equal to 1 . 1 SKI absolute o f the tn-o boride prcy~:~rati.I.\ STATE ~ ~ S I V E R S I T Y ]
Heat Capacity and Magnetic Susceptibility of Copper(I1) Tetrammine Sulfate Monohydrate from 1.3 to 24°K. BY J. J. FRITZAND H. L. PINCH XRCEIVEDJASUARY 24, 1957 The heat capacity and magnetic susceptibility of copper tetrainmine sulfate have been investigated lietxvecn 1.3 and 29" K , The results indicate strong interactions between copper ions, leading to a transition, of order higher than tlie second, \vith ~ a t 3 . 0 ° K . The magnetic susceptibility is nearly constant ixtiveeli maximum in the heat capacity of 0.75 cal. ~ n o l e -deg.-' 1 and L I O E ; . , arid thereafter falls gradually; the mnximum valuc of tlie n i d a r susceptibilit>- i, 0.04. I t i i believed t h a t tlic interaction is directly related to tlie coiirdination with ammoni:i.
Introduction The magnetic behavior of crystalline salts of cupric ion varies considerably from salt to salt. The Tutton salt, CuS04.K2S04.GH20, is one of the most ideal of all paramagnetic salts, and has been used in the attainment of temperatures far below 1 OK.' The ordinary hydrated sulfate, CuS04. 5H20, has a magnetic transition? in the vicinity of 1OK., the exact nature of which is not clear. The hydrated chloride, CuC12.2H20, has a lambda-type transition a t about 4.2"K. below which the salt becomes anti-ferromagnetic.B The acetate, Cu(CrHaO?)2.He0, becomes diamagnetic a t about 50 K. 1l;ith the exception of the acetate, the properties of all the salts described above may be explained in terms of the effect of crystalline environment on a substantially free copper ion. The ground state of free cupric ion is ?Dj,,?. At temperatures above the magnetic transitions. the typical copper salts behave almost as ?S states; their magnetic susceptibilities are but slightly higher than the "spin-only" values. Copper tetramniine sulfate differs formally from CuS04.5H20only in the replacement of four of the water molecules by ammonia molecules. In the crystal the basic unit is Cu(TH3)4++,rather than C U ( H ? O ) ~ + +due ; to differences in crystal structure. the less near neighbors of the copper ion are different in the two salts. 9 t the outset of this research, it was expected that the ammine salt would behave in the same general fashion as the hydrate. so that comparison of the two would show the effect of the ammonia ancl in addition shed further light on the behavior of the ordinary hydrate. Experimental The experimental procedures were substantially the same as those employed by Fritz a n d Pinch in the investigation (1) J. Ashmead, Y a t i t r ~ 143, , 853 (193.9). (2) T. H . Geballe and W. P. Giaurjue, THISJ O T R N A L . 7 4 , 3,715 (1952). (3) S Friedberg, P h y s i r o . 18, 714 ( 1 9 5 2 ) ; J. van den Handel, H. I f . Gijsman and A* S f'oulis, ibid, 18,8fi2 (19521. (4) €3 C. Guha. P ~ o r Roy. .COG. ( L o n d o u ) , A206, 333 (19.51).
of rnnndiurn ammonium alum,5 and will not be drscrilictl i n detail. T h e ellipsoidal sample container had a n internal volume of 31.13 ancl contained 32.64 g. of the s a l t . T h e weight of t h e container was 19.5O g. Tlir carbon thernioineter had :t resistance a t room temperature of :ihout 6000 ohms. .At loiv temperatures its sensitivity v a s decidedly higher than t h a t of the thermometer previously rlescriljcd , G but it was also consideralily less stable. The specimen used in the investigation \\-:is prep;irctl l)y addition of C.P. ammonium hydroxide to ;I iolution of C . r . copper sulfate according to the method of \Valton.6 ?'lie crystalline salt was obtained bj. cooling of the resulting hot solution. The crJ-stals used were ground to a po\vtler; the portion selected for use passed througli a 20-171esIiscrcc'ii, but failed to pa53 through a 100-mesh qcrecn. Tlie irrcgular particles were stored for a time under ;I wturatetl solutioii O f the salt and then dried manually. h sample for analysis \vas witlidran.n during tlir filling of the sample tube. Its copper content i i - a s determined electrol>-tically to be 25.927, (theoretical 25.877,); tlie :tninioni:i-copper ratio \vas determined by titration to be 3.92 to I . Measurements of heat cxpacitj- a n d magnetic suscqitihility w r e m a d e in t h e nianner previously tlescril)etl.' For tlic low temperature susceptibility meaSurcniciits, the coil constants ivcrc obtained with the specimen nc:u 7 0 ° K . , :tfter the coil5 1i:td been cooled by liquid Iieliuin or I i ~ ~ d r o g e n . niple a t 70°K., upon tlic coil> i v a i tletere experiment in wliich the spccimen \v;ii cooled from room temperature t o 70'K. T h e xuiccptibilities were thus hasetl upon the room tcmperaturr suiccptihilit>-of the salt. The molar susceptibility a t 290°K. i i I .10 X according to Bliatnage, Lesslieim a n d Klliitiiia.' :in independent check on our specimen by the C;ou!- iiietliotl gave :i result of 1.35 X (For t h e purposc of tile m:rrection this susceptibility is required only to about 211%. 8'
Results The heat capacity of the salt was measured beThe observed measurements tween l ..? and "OK. were corrected for the heat capacity of 19.Xi g. of Pyrex as previously described.j A t the highest temperatures, this represented ','4 of the observed heat capacity. In view of the unusual behavior of the heat capacity between 12 and 20°K.. this region was investigated in three separate sets of experiments; the results of the several sets were consistent. The heat capacity measurements are given J . J F r i t z and H . I< Pinch, THISJ O T - K Y A L , 78, 022.I ( 1 H. F. n'alton, "Inorganic Preparations." Prentice--€I I'ork, h-. Y , 1918. I' 79. 17) S S Bhatnage. 13 1,essheim and If. 1.. Khanna. J Ii>c! C/>civ .So
