American Mineralogist, Volume 67, pages 470482, 1982

Heat capacitiesand entropiesof Mg2SiOa,Mn2SiOa,and Co2SiOa

between5 and 380 K
Rrcnnno A. RonE. Bnuce S. HpvrrNcwey

U.S. Geological Survey

Reston, Virginia 22092

eNo Hurvrrutro Terer

Research Institute for lron, Steel and other Metals
Tohoku University, Sendai, Japan

Abstract

The heat capacities of synthetic single crystats of Mg2Sio4 (forsterite), Mn2Sioa

(tephroite), and co2Sioa (cobalt olivine), were measuredbetween 5 and 3g0 K using an
adiabatically shieldedcalorimeter. Mg2SiOais diamagnetic,and its heat capacity follows a
normal sigmoidal curve at low temperatures. co2sioa shows a single sharp )r-type
transition at 49.85+0102K associatedwith the antiferromagneticordering of the magnetic
moments of the Co2+ ions into a collinear spin arrangementbelow 49.8 K. In contrast to
co2sioa, Mn2Sioahas two transitionsin ci, a sharpL-typetransitionat 47.3g10.05K and
a smaller "shoulder" in Ci centered near 12 K. The upper transition correspondsto the
paramagnetic(disordered) to collinear antiferromagneticordering of the Mn2+ moments,
whereas the shoulder near 12 K correspondsto the changefrom the collinear to a canted
spin structure. Our calorimetric values for the antiferromagnetic-paramagnetictransition
temperature(N6el Temperature)are in excellentagreementwith those obtained by powder
magnetic susceptibility measurements, 49t2 K and 5015 K for co2Sioa and Mn2Sioa
respectively.
The thermal Debye temperature, 0$, of Mg2Sioacalculatedfrom our c$ measurements
between6.3 and 13.8K is 768+15 K and agreeswell with the elastic vaiue ofi of 758 K
based on the mean sound velocity calculated from the room temperature elastic stiffness
constants (ci:) of Graham and Barsch.
At 298.15K (25"c) the molar heat capacitiesare 118.6,128.7 , and 133.4J/(mol . K) and
the molar entropiesare 94.11-r0.10,155.910.4and,142.6-+0.2 Jl(mol . K) respectivelyfor
Mg2SiO4,Mn2SiOa,and Co2SiOa.

Introduction olivine) occurs very rarely in nature and Co2SiOa

does not occur as a discrete mineral. However.
The olivine group minerals are one of the more studies by Ross and others (1954)indicate that Co
important constituents of igneous rocks and are and Ca are common, though minor components in
believed to be a dominant phasein the upper mantle many olivines, and fCazSiO+ is important in the
ofthe earth. They are also significantas the precur- phase chemistry of Portland cement clinker.
sor minerals from which certain commercial talc The crystal structure of the olivine form of
and asbestos deposits have formed. Accurate val- Mg2SiOahas been refined by Hazen (1976)and that
ues for their thermodynamic properties over a wide of Co2SiOaby Morimoto et al. (1974).Fajino et al.
temperature range are thus necessaryfor the quan- (1981) have refined the structures of synthetic for-
titative solution of both geochemicaland geophysi- sterite and tephroite preparedby one of us (H.T.).
cal problems. The olivine structure is based on a slightly dis-
The olivines have the general formula M2(SiO4) torted hexagonal close-packedarray of oxygen at-
where M is Mg, Fe, Mn, Ni, Ca, and Co. They oms in which one-eighth of the tetrahedral intersti-
crystallize in the orthorhombic space group pbnm ces are filled by silicon and one-half of the
(62) with 4 formulas per unit cell. yCa2SiOa(calcio- octahedral interstices are occupied by the divalent
0003-0G)v82l050G0470$02.00 470
 ROBIE ET AL.: ENTROPIES OF MgzSiOa,,Mn2SiOa, and Co2SiOa
cation M. The divalent cations occupy two crystal-
lographically distinct sites normally called M(1) and
M(2). The M(1) site has the point symmetry 1
whereas M(2) has point symmetry m.
Powder neutron diffraction studiesby Nomura e/
al. (1964)indicate that in CozSiO+the spins associ-
ated with the Co2* ions are collinearly ordered
parallel to [001] at temperaturesbelow 49 K. Similar
studies by Santoro et al. (1966)show that tephroite
has a canted antiferromagneticspin structure below
13 K, and has the same collinearly ordered spin
structure as CozSiO+between 13 and 47.4 K.
Previous studies of the low-temperature heat
capacities of Mg2SiOawere made by Kelley (1943)
over the temperature range 53.2-295.0K. King and
Weller (reported by Mah, 1960) measured C[ of
MnzSiOr between52.3 and 296.1K. Orr (1953)has
measuredthe heat content of MgzSiO+between 298
and 1800 K and Christiansen (reported by Mah,
1960) has determined lft-Iffx up to 1795 K for
crystals and liquid Mn2SiOa.We are unawareof any
previous measurementson the heat capacity or heat
content of cobalt olivine.
Materials
Mg2SiOa
The forsterite sample was a single crystal boule
grown by the Czochralski method using an iridium
crucible. The starting materials were 99.9 percent
purity MgO and SiO2which were blended in the 2: 1 ing the intermittent heating method under quasi-
molar proportions and sintered in a resistancefur-
nace for 6 hours at 1300'C. The crystal was pulled
from the molten MgzSiO+(:1900'C) at 3 to 10 mm/
hour with a rotation rate of 20 to 30 revolutions/min Table l. Chemical analysesof MgzSiOr, Mn2SiOaand Co2SiOa
under a nitrogen atmosphere (flow rate 1-3 liter/
min). Details of the apparatus and growth proce-
dures are given by Takei and Kobayashi (1974)who
also provide chemical analyses,unit cell parameters
and densities of several other boules grown under
the same conditions. The boule initially was 2.2 cm
in diameter with a length of 5.6 cm. In order to fit
into our calorimeter it was sliced into quarters, with
a diamond saw, parallel to the long axis.
Electron microprobe analyses of our sample by
L. B. Wiggins(U.S. GeologicalSurvey)are givenin
columns 2 and 3 of Table 1. The tabulated values
are the averages of 7 and 6 separate analyses 4
respectively.
Mn2SiOa
The tephroite sample was also a single crystal
boule, initially 1.7cm maximumdiameterby 5.7 cm
471
in length. The starting materials were 99.9 percent
MnCO3 and SiOz reagentsmixed in the 2:1 molar
ratio and fired at 1000'C under a nitrogen atmo-
spherefor severalhours. The boule was grown from
the melt (-1350"C) in an iridium crucible under a
protective atmosphere of nitrogen plus hydrogen
(500:1 by volume)using the Czochralskitechnique.
The pulling rate was 3 to 8 mm/hour with a rotation
rate of 10 to 15 revolutions per minute. Additional
details of the growth procedure and chemical analy-
ses, X-ray cell parameters, and optical absorption
spectra for crystals grown under the same condi-
tions with the same apparatus are given by Takei
(1976). Microprobe analyses of the heat capacity
sample are listed in columns 5 and 6 of Table 1 and
are the averages of 8 and 3 separate analyses
respectively.
Co2SiO4
The cobalt orthosilicate samplewas composedof
fragments of 3 separate boules. The fragments
weighed between 3 and 10grams. The crystals were
grown by the floating zone method using the appa-
ratus and techniques described by Takei (1978)'
Electron probe analysesof the cobalt orthosilicate
sampleare listed in columns8 and 9 of Table l.
Experimental measurements
The heat capacity measurementswere made us-
adiabatic conditions using the cryostat describedby
single crystals used for heat capacity measurements.
456789
123
Si02 42.71 42.11 42.26 2 9 . 7 5 2 9 . 3 0 2 e . 4 1 ? . 8 . 6 2 2 7 . 3 2 7 . 4
t4s0 51.29 58.31 57.05
l'1n0 70.25 69.A6 70,44
7t.38 7t.8
C o 0
100.00100.42 99.30 100.00 s9.16 q9.84 100.00 cs.l 99.5
I Mg2Si04
t'4icropnobaen a l y s i s h y L . B l , l i g g i n s ,U ' s . G - " o l o g i c a l
Survey
l,4n,5i0,
5,6 Microproba e n a l y s t s h y L . B - l , l i g q i n s ,U . s . G e o l o g i c a l
Survey,Reston,VA
C o2 5 i0 4
7
l l i c r o p r o b ea n a l y s i s b y L . B . l i i g g i n s . t J . S . G e o l o g i c a l
S u f v e y ,R e s t o n ,V A
472 ROBIE ET AL.: ENTROPIES OF MgzSiOq Mn2SiOa, and Co2SiOa

Robie and Hemingway (1972); together with the Table 2. Experimental molar heat capacity measurementson
calorimeter and semiautomaticdata acquisition sys- synthetic single crystals of MguSiOa(forsterite).
tem describedby Robie et al. (1976).Temperatures
were determined with a Minco model S 1059-l r e n p' renp. renp. Ilea t
cap6clty
platinum resistance thermometer (Ro : 100.02 ""1;:i., ",1::1.,
K J/(s'K) K J/(c.K) K J/(s.K)
ohms) which had been calibrated by the Tempera-
ture MeasurementsSection of the National Bureau Series 1 Serles 5 serle6 8
of Standards on the International Practical Tem-
299.45 0.8451 I33.54 0.3741 270.34 0.7885
perature Scaleof 1968(IPTS-68)between 13.8and 304.80 0.8552 139.17 0.3967 2 76 . 2 0 0.8005
310.61 0.8651 t44.73 0.4189 282.08 0.8r24
505 K, and the provisional temperature scale used 3r5.37 0.875r I 50.28 0.4405 2 8 7. 7 t 0.8234
0 ,8839
in this laboratory between 4.2 and 13.8 K as dis- 322.09
327.77 0.8935
155.65
1 6 1, 0 0
0.4609
0.4808
2 9 3. 4 6
2 9 9. t 4
0 . 8 34 3
0.8444
cussed by Robie, et al. (1978).The sampleswere 333.56 0.9024 166.55 0.5008
t72.15 0.5207 Series 9
sealedin the calorimeter under pure helium gas at a serles 2 r77.63 0.5395
7.66 0.0000r8
pressureof 5 kPa, (approximately4 x l0-5 mol of 335,44 0.9053 Serles 6 9.t4 0.000r49
He). 341.31
346.97
0.9134
0.9224 r 75 . 3 1 0 . 5 3r 5
I0.10 0.000215
Il.ll 0.000305
The formula weights used were basedon the 1975 180.79 0.5499 t2 ,35 0.000401
Serles 3 186.41 0.5585 r3,82 0.000578
values for the atomic weights and are 140.694, I92 .02 0.5862 L 5. 4 7 0.000830
35I.06 0,9259 L97.73 0.6039 L 7. 2 2 0.001004
201.960, and 209.950 g mol-r for Mg2SiOa, 355.20 0.9324 2 O 3. I + 6 O. 6 2 1 1 19.I8 0.001698
Mn2SiOa, and Co2SiOarespectively. The sample 350.95 0.9407 209.08 0.6375 21.46 0.002463
3 6 7 .I 3 0.9494 23,7| 0.003402
weights, corrected for buoyancy, used in our mea- 373.75 0.9s75 Serles 7 25.85 0.OO4524
380.3r 0,9654
surementswere 46.537,41.200 and 46.012grams for 2 0 7, 8 6 0.5338 Series 1O
Serles 4
Mg2SiOa,Mn2SiOa,and Co2SiOarespectively. 2L3.42
2r9.rl
0.6495
0.6655 6.23 0.000037
For the measurementson Mg2SiOr the calorime- 5 2. 8 7 0 . 0 4 9I 5 224.92 0 . 6 8r I 1.56 0.000086
58.s3 0.06511 2 3 0. 6 2 0 . 6 9 6I 8.44 0.000r17
ter contributed 88 percent of the total measured 63.54 0.08r14 236.40 0.7107
68.82 0.09976 242.25 0.7249 Serles 11
heat capacity at 10 K, 60 percent at 100 K and 43 74,21 0,I203 24A.Ot 0.7389
percent at 300 K and above. For Mn2SiO4 the 79.63 0.r42I 253.84 0.7527 26,O4 0.004621
85.16 0,L652 259.74 O,7662 28.58 0.006425
calorimeter contributed approximately 53 percent 90.77 0.1892 2 6 5. 5 6 0 . 7 78 3 31,43 0.008845
9 6 . 4I O.2t37 34.70 0,O1244
to the measuredheat capacity between 60 and 380 r0r,99 0.2380 38.25 0.0r732
107.50 0,2622 41.85
K, dropping to 28 percent at 25 and 5 percent at l0 112.83 0.2854 45.88
0.02330
0.03133
K. For the measurementson Co2SiOothe calorime- II8.02 0.3080 50.12 0.04158
123.10 0.3298
ter was 60 percent at l0 K, 38 percent at 50 K and t28.25 0.35r6
averaged 55 percent between 50 and 380 K. The
corrections for deviations from true adiabatic con-
ditions were less than 0.2 percent above 30 K and
less than 0.1 percent between35 and 250 K except in their chronological order of measurement.The
for the measurementsmade with small temperature heat capacities of Mg2SiOaand Co2SiOaare shown
increments (A?<0.5 K) in the neighborhoodof the in Figure I and those of Mn2SiOain Figure 2.
antiferromagnetic transitions. For Mn2SiOa and CozSiOr we made replicate
Mg2SiOa is diamagnetic and its heat capacity measurements in the region of the C! anomalies
follows a normal sigmoidal curve at low tempera- using temperature increments of 0.3-0.18 K in
tures. Co2SiOaand Mn2SiOaare both paramagnetic order to delineatethe exact form ofthe heat capaci-
and as was anticipated show a sharp \-type transi- ty in the transition regions. These are shown in
tion in Q, arising from the antiferromagneticorder- Figures 3 and 4 for Co2SiOaand Mn2SiO4respec-
ing of the Co2+ and Mn2* spins at low tempera- tively.
tures. In addition, MnzSiOr exhibits a second,
rounded, maxima (shoulder)in C! near 12 K which, Thermodynamic properties of Mg2SiOa,Mn2SiOa,
on the basis of the neutron diffraction studies of and Co2SiOa
Santoro et al. (1966), we believe is related to the Our experimental heat capacitieswere graphical-
changefrom a collinearly ordered antiferromagnetic ly extrapolated to 0 K using a plot of Cf,lT versus P
spin arrangement to a canted arrangement of the as shown in Figure 5.
spins with decreasingtemperature. Above 30 K the heat capacity data were analyti-
Our experimental results are listed in Tables 2-4 cally smoothed using an orthogonal polynomial
 ROBIE ET AL.: ENTROPIES OF MgzSiOt,, Mn2SiOa, and CozSiOq 473

Table 3. Experimental molar heat capacity measurementson Table 4. Experimental molar heat capacity measurementson
synthetic single crystals of MnzSiOa(tephroite). single crystals of CozSiO+(olivine structure).

Ternp Heat Teqp H e at Hea! Ie'np Heat

Capaci ty C a p a c it y Capaci ty C a P a c ti Y

mol.K m o l. K J/(uol.K) K J/(npl.K)

S"ri es I Series X S e r ie s I S e r i e sV I I
303.53 l2C.B 219.a] 108.4 309.15 135.8 194.71 102.6
30c.63 I 3l .l 22_5.53 110.2 3 t5 . 0 0 I 3 7 .l 200.38 104.7
3 15 . 3 6 132.2- 2 3 1. 4 2 ]l 2.0 32A.82 I38.3 206.01 I06.8
3 2 1. 0 5 133.3 ?-37,32 113.7 325.09 I 39.5 2 1 1. 7 6 I 0 8 .B
326.69 I 34.4 243.24 |5.3 332.55 I 40.7 217.41 110.7
338.48 I4l.S 223.21 112,5
Se.ies 1l Series \I
326-43 134.4 ?-49.12_ ll 5.4 S e r ie s I I Sere
i s V lI I
332.34 135.7 254.96 I I 5.8 344.32 143.1 228.82 1I 4 . 5
333.57 136.8 350.22 144.2 234.50 115.2
344.19 137.6 Series XI 356.19 145.3 24A.26 I I8.0
35n,94 I 38.7 2 5 0 - lI 111,? 362.24 1,15.4 246.01 I I9.7
35/.16 139.7 ?56.q3 lls.B 368.3/ 141.4
3 6 3 .4 3 I 4 0 .7 ? 6 1. 9 7 I 2D.3 3 14 . 5 7 I48.5 S e r i e sI X
i69.5€' I 4l .7 2-67,92 1?l.8 380,8C I49,5 251.20 121.2
315.1\4 142.6 27i.8A 123.3 251.18 123.0
3Bl.Cq I 43.',1 2.79.85 124.1 S e . i e sI I I 263.04 1?4.6
2 8 5 . 7a 12 , 6 . O 50.12 32.17 264.85 125.2
S e r r e sI I I 2 9 1. 6 e )2t.5 51.02 2/.61 2 74 . 5 9 121 .7
5).12 2e.83 5 2 .I 0 26.52 ?n0.?-5 129,2
52.q9 3q.qq SeriesXIII 53.17 2-6.21 285.90 I 30.6
60.04 i2.6tl 4.65 1.q97 54.23 26.30 291.50 132,0
5.51 2.625 55.25 26.51 2 9 7. 0 6 I 3 3 .I
S , o nei s I V 6.50 3. t02 5 3 . l3 2 1. 6 5
34.69 L50 4.601 62.97 3A.21 S e . ie s x
70.06 3 7 .l t q.5? 5.430 6 7. 9 0 3 3 .3 9 1.53 ,^235
74.97 ln.?1 9.68 6.800 12.92 36.77 9.57 .OB?5
10.70 1.149 1I . 9 7 40.27 10.61 .)295
S - . r ie s V 11.70 1.1e6 83.05 43.83 I1.64 ,1902
]N )2 42 t4 12.89 7.2q1 BB.2) 41,37 12.89 .2933
rq3.42 45.4. 14.2) 7.5iJ 14,38 .4B3 l
3,1.67 4rl.5rl t5.65 1. 9 ? 6 S e r ie s I V 15.13 .i989
9 4. ? 2 5t .q6 l1.2) 8 . 6 13 38.63 4 1. 6 7 18.06 1.302
18.91 1.466 94.10 51.34 20.01 2.423
Si:ries VI 20.86 I q.55 99.36 51.81 22.41 3.088
. 3 1i 4 45.541 2?-.q\ I I .89 104.60 58.16 24.94 4, 5 5 6
,8,7q 4 9 ,5 9 2\.24 I 3.41 I09.95 6 l. 5 0 2 t. 4 5 6.359
q3,96 51.q2 ?-1.15 I 5.23 I r5.54 51.89 30.08 8.621
q!.6r 5 5 -I t 30.7q 11.74 121.25 68,26 32.91 I I .58
34.25 2i,q4 t26.8t 71.45 35.95 I 5.39
S:ri es VII 38.10 25.?6 39.41 24.51
I n0.39 55.56 tr2.?3 30 4q S e r ci s V 43.23 28.55
I 06.t6 5q.97 45,55 43,5q* )32.a2 t4.51
il 1.53 62.Oti I38.29 17.64 S e r ei s X I
117.19 65,24 S-ries XIV 143.32 91.43 5.15 .0081
122,o2 6 P.'3 1 42.72 il.62 149.40 83.15 6.16 .A1a?
12 _ ' \ . . ? 7 ). i 2 43.61 i3.?C I55.02 85.i15 7.51 .0283
I 3 4 .n 2 74 , 1 i 44.3r) 14.71 150.tt3 88.53 8.46 .0513
I lq.3t t5.t5 4 5 , 0q i6.23 166.66 91.15 s.65 .oBBn
I41.73 1 9. 3 1 45.51 35.38 10.73 .1380
I 5 0 - l7 9t .91 45.B5 3 A . l9 S e r i e sV I
t 55.6/ 84.33 46.14 3q.35 lt2_.37 93.62 S e . i e sX I I
4 6, 4 2 3 q .9 2 117.95 95.911 16.05 36,48
5.ries VIIi 46.6^ 4l , n0 183,51 98.3C 41.69 43.40
, 5 1. 2 9 86.94 46.31 '11,60 t89.35 10r).6 48.03 4 5 . 71
t61.11 q9.r2 4 1. a n 42.15 18.32 47.?1
| 72 . 9 8 9t.77 41.18 4i.5^ 48.54 43.92
lla.t\1 9 4 . l7 11,39 4F,,O6 48.16 50.84
I 84.83 95.44 41.68 tl .92 4i1.97 5 3 .I 5
I 90.75 q8.64 44.n4 30.71 49.11 55.80
4q.42 29.58 49,i7 59.31
Seri.s IX 48.31 29.42 49.55 64,05
195.52 I C0.7 49,73 64.04*
2 A 2. 3 2 102.1 49.95 39.4/
20ll,lq 104.7
?14.a3 106.6

previous (unpublished) data on the Calorimetry

curve fitting routine (except in the transition re-
gions). The analytically and graphically smoothed
parts were joined smoothly and corrected for curva-
ture. Our measurementsfor Mg2SiOahave an aver-
age deviation from the smoothed curve of 10.06
percent between 25 and 380 K and maximum devi-
ations of 0.5 percent.
From replicate measurementson all three sam-
ples in the range 5 to 10 K together with our
Conference Copper sample (Osborne, Flotow, and
Schreiner, 1967)between 4.5 and 20 K, we believe
that our smoothedCi values are accurateto approx-
imately 15 percent between 5 and 15 K, increasing
in accuracy to 10.1 percent for temperaturesabove
20 K except in the immediate region of T1qwhere
the small temperatureincrementsreducedthe accu-
racy to 11.0 percent.
Smooth values for the thermodynamic functions
C;,S$-.Sfi,Gft-m)ff,and(Gt-I13)/TforMg2SiOa,
 ROBIE ET AL.: ENTROPIES OF MgzSiOa, Mn2SiOa, and Co2SiOa

150

140 Nf
t40 JXX
r
^f
.{
130 ^t
Ja?"
130
"A-
't20 ry
F
.l r20
F

lr0 . T 'n0
J-
Y ?.
r00 d .
X
,s 100
,f
= tr o
rr =eo
z tr z- 8 0
X

>. t >.
U
o ,t l
tr
d
U
n u60
F
o
U
T 4
I
C
-{
T
f
SA
.P ! M92SiO4 USGS
) o Mg25iOo
l0
T ( K E L T E Yr 9 4 3 ) x MnzSiOr
l0 (TEPHROTTE
a Co2SiOo x'
o
r00 150 350 400
TEMPERATURE, IN KETVINS
Fig. 1. Experimental molar heat capacity of synthetic single T E M P E R A T U R EI ,N K E I - V I N S

crystal Mg2SiO4(forsterite), formula weight 140.694g mol-r and Fig. 2. Experimental molar heat capacitiesof synthetic single
synthetic singlecrystal Co2SiOaformula weight 209.950g mol 1. crystal Mn2SiO4(tephroite), formula weight : 201.960g mol-r.
Earlier measurementsof Kelley (1943) on powdered Mg2SiOa
between 53.2 and 295.0 K are shown for comparison.

Mn2SiOa,and Co2SiO4are listed at integral tem-

peraturesin Tables 5, 6, and 7 up to 380 K.
At 298.15 K the entropy changes, Sies-Sfi, for
Mg2SiOa, Mn2SiOa, and CozSiO+94.11-{-0.10, 55

155.9-f 0.4, and 142.6-10.2Ji(mol.K) respectively. 50

Y
The zero point entropy, Sfi,is presumablyzero for
each of these three olivines. Accordingly, the S|-Sfi o
=
values are equal to ,Si and are the correct valuesfor
z 10
use in thermodynamic calculations.
F
35
L)
Mg2SiOa 4
30
Kelley (1943) measured the heat capacity of
Mg2SiOaprepared from "a mixture of oxides con-
U

u
-
25
t' \"/

taining an 0.82Voexcessof magnesiumoxide over 20

the theoretical 2:l molar ratio was rotated in the I
chamber of a porcelain ball mill for nine hours to
assure intimate mixing. The mixture was moist- lo
ened, tamped into a nickel cartridge, dried at 100', ,( a Co2SiOa
5 (Co-OtlVlNE)
and finally heated in a tube furnace at 1150-+20'for 3
d
four days under a vacuum of l0-4 mm. After this 0
20 30 40 50
treatment, the product contained 3.67ouncombined TEMPERATURE,
IN KEIVINS
magnesium oxide. It was reground and similarly Fig. 3. Molar heat capacity of Co2SiOain the region of the
heatedfor five additionaldavs at 1180-+10o. Analv- paramagnetic-antiferromagnetictransition, TN = 49.7610.02K.
 ROBIE ET AL.: ENTROPIES OF MezSiO+, Mn2SiOa, and Co2SiOa 475

50 Table 5. Molar thermodynamic properties of MgzSiO+

(forsterite). Formula weight = 140.694g mol t.
I
TEI]I'. ilEAT f,I:TROPY ENTIIALPY CIBIS TNERGY
CAPACITY FUItCTIOit FLrltCTI0li

* T C* rsi-sir rri-Hirrr -<ci-rirzr

v35
il XDLIII: J/(tro1'K)

o
:^^
II
)ai
5
IO
15
0.0032
0.0284
0.106
c.00I0
0.008e
0.0327
0.0007
0.0066
o.o249
0.o003
0.0022
0.0078
2A a.275 0.0834 0.0637 0.0197
25 0.570 0.1736 0.1_129 a.04a1
z 30 1.059 0.3169 a.2429 0,0740
Fas 35 1.80a 0.533 0,41C 0.I23
40 2.P43 C.839 0.646 0.193
U 45 4.r71. r.247 0.96I 0.286
50 5.7A7 1.76A L361 0.407
X
l) zv
60 9.804 3.163 2 . 4 1 9 0.744
70 ra .69 5 .032 3 . 8 1 4 r .217
U 80 2A.20 7.347 5 . 5 1 3 1.835
rt5 90 26.12 10.07 7 . 4 7 1 2.595
100 32.26 13.13 9 . 6 4 2 3.492
110 38.44 16.50 4 .520
120 44.55 20.1I t4.44
l0 130 50.51 23.9r 16.99 6.923
I40 s6.27 27.86 r9.59 P ,. 2 7 6
fo'"" 150 61.81 31..94 22.22 9.7r7

160 6 7. L 2 36.10 24.P6 r1.24

x Mn2SiO, I70 12.r9 40.32 21.50 t2.e2
l8c 77.02 44.58 30.12 14.47
(TEPHROTTE)
190 P1.60 48.87 32.7t r6.t7
010203040506070 2AO 85.95 53.L7 35.26 17.9r
210 90.07 57.46 37.71 r9.69
T E M P E R A T U R EI ,N K E L V I N S 224 93.98 6l .74 4 A- 2 4 2I .50
230 97.70 66.00 42.66 23.35
Fig. 4. Molar heat capacity of MnzSiO+in the region of the low 24 A 101.3 10.24 45.03 25-2r
250 104,6 ]1.44 47.34 27.rC
temperaturespin orderingtransition,47.38+0.05K.
26A 107.P 1P.(,1 49.61 2t.00
27r) 110.9 82.73 51.82 30.91
2Po 113.7 86.?,2 53.qP 32.P4
2 -ea 116.4 90,85 56.09 34-71
T E M P E R A T U RI N
E .K E T V I N S 300 119.1 94.85 5B.14 3C.70
310 r2t.6 s8.79 60.15 38.64
510t5 320 t24,A 1C2.7 62.11 4C.58
0r0 42.52
33C 126.2 106.5 64.01
128.3 1I0.3 65.87 44.L6
350 13c.2 114.1 67.68 4(,.44

3 60 132.1 II7.8 69.45 48.33

3 7A 134.1 l2l.A 11.r1 50.26
3 {'C 135.{1 125.0 52.1{r
o08 2 73.r5 1lI.t ?4-O2 5 2. 5 r 3r .52
2 fr8.15 I18.6 94 lI 51.11 36.34

007

o
E
vvv
sis now gave98.6Vomagnesiumorthosilicate,O.SVI
uncombined magnesiumoxide, no free silica, and
- o05 no magnesiummetasilicate".His Ci results(53.2to
F 295.0 K) differ frorn our measurementsby a maxi-
oq
U
004 mum of 0.5 percent over the common temperature
range.Our valuefor Stes,94.11!.10Ji(mol'K)is 1.1
percent less than that obtained by Kelley (1943).
Mg, SiOo
Almost all of this difference arisesfrom the empiri-
SERIES9 ^
cal extrapolation of his data from 50 to 0 K. Over
S E R I E SI O T
that part of the temperature range for which he
Co2 SiOo a actually measuredCi his value for,SisrS3o and ours
agreeto 0.3 percent.
0 r00 200 300
We have also combinedour heat capacityresults
T E M P E R A T U R IEN2 K 2 for the range 300-380 K with the heat content data
Fig. 5. Cp/T versus 72 extrapolation of heat capacity data on of Orr (1953)between398.1and 1807.6K to gener-
MgzSiOr and Co2SiO4 to 0 K. ate values for the molar heat capacity at higher
476 ROBIE ET AL.: ENTROPIES OF MgzSiOt, Mn2SiOa, and Co2SiOa

Table 6. Molar thermodynamic properties of Mn2SiOa Table 7. Molar thermodynamic properties of Co2SiOa.Formula
(tephroite). Formula weight : 2019ffi g mol-r. weight : 209.950g mol-r.
TE}1P. IIEAT N}'TIIOPY f,llTltlLPY clxts Et LilicY i.IIIP. II'AT EI'TRoPY utiTl ALtY cIDES llirRCl
CAPACITY FLrtiCTI0li FUI:CTl0l. CAPACITY FUIICTION }'UNCTIOI]

t a, ( s r - s o) i l : _ . - r r; r^ ) / T
I
-(c--ti^\/i
I u_
T C- <si-sil tr;i-rrilrr -<ci-rri>rr
TELVII] .i/(Do1'k) ruLvlli

5 2.2rr 1.267 0.80e 0.459 o,0050 0.00r4 0.001o 0.0004

l0 6.893 4.263 2.112 r.55r l0 0.10r 0.023 0.0186 0.0045
t5 7.760 7.217 4.244 2.973 o.574 0.130 0.rc6 c .024
20 10.08 9.773 5.384 4.349 20 1.975 0.458 o.372 c.086
25 13.21 r2 .3t 6 . (,27 5 .682 25 4.(tO1 1.I59 0.935 o.224
30 17.05 15.04 8.010 7.011 30 8.555 2.326 1.856 0.470
35 2r.16 l8,01 9.643 4.367 J J 14.13 4.039 3.I88 0.851
40 27.11 2t .29 r1.51 9.714 40 2t.71 6.393 5 .003 1,390
45 36.0e 24.99 t3.74 11.25 33.18 9.553 1.44A 2.113
50 29.9r 28.85 16.02 12.83 50 3 8. 5 0 14.22 11.15 3.069
26.41 16.60 12.60 4.20
55 30.71 3I.71 r7.29 14.42
60 32.L.) 34.46 18.48 15.98 60 28.10 19.19 13.84 5.355
70 37.29 39.80 20.80 19.00 70 34.75 24.O5 r6.38 1.677
80 43,29 45.17 23.23 21.94 80 4r.64 29.r4 19.10 10.04
90 49.42 50.62 25.80 24.A2 90 48.59 34.45 2t.9,o t2.45
100 55.4r 56.14 2t,.47 27.6A 100 55.27 39.92 24.99 r4.93
110 61.58 45.48 28.03 17.45
rlo 6r.23 61.70 3r. r8 30.52 120 67.53 3I.OB 20.o2
120 66.80 33.92 33.35 130 73.t4 56.13 34.10 2 2. 6 3
130 72.A8 72.82 36.65 36.r7 140 78.44 62.34 37.08 25.26
140 77,C7 74.35 39.36 38.98 150 83.46 67.93 40.01 27.92
t50 81,81 83.83 42.O4 4r.79
160 86.30 a9.25 44.66 44.59 160 88.19 73.47 42.87 30.60
170 90.56 04,61 47.24 47,37 92,(,5 78.95 45.61 33.2B
170
1P0 94.57 99.91 49.76 50.15 100 96 ,86 ? , 4. 3 6 48.40 35.91
la0 98.36 105.12 52.22 52.90 190 r00,8 e.9.1r 51.05 38,66
200 101.9 110.3 54.62 55,64 200 tott.6 94.98 53.64 4t .34
210 108.1 100.2 56.15 44.02
2r0 105.3 115.3 56.95 56.36 220 111.5 105.3 58.59 46.69
224 I08.5 120,3 59.22 61.07 230 114.8 110.3 60.96 49.34
234 111.6 t25.2 6r.43 63.75 244 117.9 115.3 63.27 5I.99
2LA 114.5 I30.0 53.58 66.41 250 r20.9 120.1 65.52 5 4. 6 2
250 IL? .2 134.7 65.67 69.05
260 119.8 139.4 67.11 7I .66 260 123.8 124.9 67.71 57.23
270 122.3 143.9 69.68 74.25 270 t26.5 129.7 69,83 5 9. 8 2
28C t24.7 r4a.4 7|.6t 76 . 8 2 280 r29.1 r34.3 71.9L 6 2. 4 0
29A t27 .O 152.8 73.tt8 79.37 290 131.6 138.9 73.92 64.96
300 129.r r57.2 75.30 81.89 300 133.9 143.4 75.88 6 7, 5 0
310 136.1 t47.8 77.79 70,02
310 t3r.2 t6r.5 77 . O 7 84.39 320 t3a.2 t52.2 79.64 72.52
320 r 33 .1 r65.7 78.79 86.86 330 140.2 156.4 tr1.45 75.O0
330 135.1 169.8 PO.46 89.31 340 r42.2 I60.7 83.20 77.45
340 t36.9 173.8 82.10 -qt.74 350 r44.1 I64.8 84.92 79.89
350 138.6 t77.8 83.69 94.14
360 140.2 ItI.8 85.24 96.52 360 146.0 I6P.9 66,59 82.3r
370 141.P I85.6 86.74 9e.e8 370 147.7 r72.9 84.22 84.70
3EO 143.0 I89.4 88.21 ICr.21 380 t49.3 t76.9 89.80 87.07
273.t5 123.1 t45.4 70.30 15.07 273,r5 1 2 7. 4 l3r.l 70.49 6 0. 6 4
298.r5 128.7 156.4 74.96 81.43 2 9 8 .t 5 L33.4 r42.6 67.O3

temperatures; value is however based on an estimated value for

: - AH.or,, of MgO in the molten 2PbO'B2O3used to
Cp 87.36+ 0.08717r 3.699x1067-2
- dissolve Mg2SiOa and SiO2. This estimated vahre
+ 843.610'5 2.237x10-sT1
for A.EI".6 of MgO in the 2PbO.B2O3melt has the
(298-1800K). This equation fitted the 23 datapoints opposite sign from that measuredat a later time by
with an average deviation of 0.2 percent and joins Navrotsky and Coons (1976)at a slightly different
the low temperature Cfi curve smoothly. temperature and we therefore did not include this
The enthalpy change for the reaction result in our analysis.Charlu et al. (1975)obtained
-62.6-1 1.1 kJ at 970 K and Kiseleva et al. (1979)
2MgO + SiO2: Mg2SiOa (1)
obtained -58.211.4 kJ for Llff"./tfor this reaction
has been measured by Torgeson and Sahama both by molten salt solution calorimetry. Kiseleva
(1948),and King et al. (1967)by HF(aq) solution et al. (1979)value corrected to 298 K using the heat
calorimetry. These authors' results were corrected capacity data of Robie et al. (1979)yields A,ffies :
by Hemingway and Robie (1977)for SiO2 particle -56.5 kJ, and is in excellent agreementwith that
size effect. Navrotsky (1971)obtained -62.8!4.2 adopted by Robie et al. (1979,p. 364), LIffes :
kJ for LIfios (1) by molten salt calorimetry. Her -56.7-r.7 kJ/mol which we believe is the best
 ROBIE ET AL.: ENTROPIES OF MgzSiO+ MnzSiOa, and CozSiOt 477

Table 8. Thermodynamic properties of Mg2SiOa at high

G = 261.28 - 1.378x
1,0-zT
temp€rature and Alf and AGe for the reaction 2MgO + SiO2 : - 2217.7T-o'5
(Zgg-t0oor)
MgzSiO+.
for Mn2SiOa.
Jeffes et al. (1954)obtained LIfrsE: -49.2+2.9
GI IJBS
TEMI. (HT-Hze8)/T si -(cT-r12e.)/T ci ENTHALev FRIE
ENERGY
kJ for the reaction
K J/rnol.K J/mol.K J/nol.K J/nol.K kJlocl
2MnO + SiO2 eMn2SiO4
298.15 0.00 94,il
1 .10
94.11
1 .10
118.60 -56.69
I .69
-56.i? (2)
manganosite o,-quartz tephroite
400 32.92 13 t . 9 7 99.0s 137.71 -56.73 -55.20
500 55.01 163.92 108.91 148.28 -56.84 -56.05
600 tl .2t I 9 r, 6 5 I20.44 155.1t -57.12 -55.87 by HF(aq) solution calorimetry (at 25'C). Heming-
/00 83.73 216.12 13 2 _ . 3 9 I 6l .75 -57.53 -55.62
300 93.81 238.06 144.25 166.83 -58.06 -55.33 way and Robie (1977) adjusted this result to
.-47.1-*'3.0
900 1 0 2 , 17 251.97 I 55.80 171.25 -59.00 -54.91 kJ to accountfor the very fine particle
I 000 I 09.28 276.22 166.94 175.14 -58.71 -54.46
ll00 I I 5.43 293.08 1 7 7. 6 5 178.s6 -58.28 -54.07 size of the quartz used in the AI1"o1measurements.
I 200 I 2 0 .B r 3 0 87 .4 -51,74
Navrotsky and Coons Gnq obtained L,Iffeg :
187.93 t8t.54
r 300 125.59 323.38 197.79 lrJ4,09 -57.10 - 53 . 4I
I 400
I 500
129.84
I 33.66
3 3 7 .I 0
350.01
2 0 1. 2 6
216.35
186.)2
187.93
-56.39
-55.52
- 5 3 .I 3
-52.93
-48.5-'-1.4 kJ for this reaction using molten
I 600 r37.10 362.18 225.08 189.2? -54.84 -52,8 1
I 700 I 4 0 .I 9 3 73 . 6 8 233,49 I 90.09 -54.05 -52.F,6 2PbO'B2O3as the solvent. Schwerdtfegerand Muan
I 800 14 2 . 9 8 384.56 2 4 1, 5 9 I9C.54 -53.2c -52,62
(1966)give -43.5-r 1.7 kJ for the Gibbs free energy
change, A6f,, for this reaction at 1423K based on
measurementsof the equilibrium gas ratios of CO/
CO2 coexisting with phases in the "FeO"-MnO-
value. Combining this value with the enthalpies of SiO2 system. Biggers and Muan (1967) obtained
formation of periclase and quartz from Robie et a/., ACfrszt : -42.3!3.4 kJ from equilibriummeasure-
we obtain Ll4,zgs: -2170.4+ 1.4 kJ/mol. Using the ments of the exchangereaction
entropies of the elements adopted by Robie er a/.
and the entropy offorsterite obtained in the present CoSiO3 + 1/2Mn2SiO4e MnSiOr
work we get AGi,2es: -2051.0-r1.4 kJ/mol. + 1/2 Co2SiOa (3)
The heat capacity equation for forsterite was combined with the AGF values for CozSiO+from
combined with the value of ffi,2es adopted above Lebedev et al. (1962)and Aukrust and Muan (1963).
and the high temperature thermodynamicfunctions Using A6f, from Schwerdtfeger and Muan, and
for periclase and quartz tabulated by Robie et al. Biggers and Muan, and our heat capacity (and
(1979)to obtain the values of A/lt and AGi listed in entropy) data together with thosefor MnO and SiOz
Table 8. Our values for AGi differ significantlyfrom tabulated by Robie et al. (1979)we can also obtain a
those obtained by the e.m.f. measurementsof Rog value for Llfrslfor reaction (2) as follows: from the
et al. (1974)between 1373and 1573K. heat capacity data, we obtain ASi.r+zrand from the
relation
MnzSiOa LIi":Ad+TAS"
King and Weller's measurements on Mn2SiO4 we get ffi1ay and using Kirchoff's relation calcu-
between 52.3 and 296.1 K (Mah, 1960)were made late A1{.2es: -53.7t2.1 kJ from Schwerdtfeger
on "crystalline manganese orthosilicate (teph- and Muan and -53.5-+4.4 kJ, from Biggers and
roite)" with no further description of the prepara- Muan respectively. Schreiber (1963) studied the
tion of the material. Their measured values are reaction
systematically larger than ours, being about 2 per-
2MnCOr + SiO2 = Mn2SiO4+ 2CO2 (4)
cent greater at 50 K and decreasingto +0.7 percent
at 300 K. Their entropy value at 298K includes an between 738 and 813 K. From his equation for
empirical extrapolation below 51.0 K of 24.9 Jl logfco,as a function of temperature and our entropy
(mol'K) and was apparently made without any data, and using unpublishedmeasurementsof Robie
knowledge of the two low-temperature magnetic and Hemingway on the entropy of MnCOr, we get
transitions. In the same way as described for AG?oo: -44.2 kJ/mol for the formation of Mn2SiOa
Mg2SiOawe combined our low temperatureQ, data from the oxides, and -49.3 kJ/mol for AI4m (2).
with the heat content measurementsof Christensen The agreementbetween the 5 independentvalues
(Mah, 1960)to generate is only fair and we therefore simply adopted a
478 ROBIE ET AL.: ENTROPIES OF MgzSiOq, Mn2SiOa, and Co2SiOa

rounded averagefor ffi,2es (2) of -50.8-r2.5 kJl Table 9. Thermodynamic properties of MnzSiO+ at higher
mol. We combined this value with ancillary data temperatures and A-El"and Ad for the reaction 2MnO + SiOr :
Mn2SiOa.
from Robie et al. (1979)to obtain -1732.0t 2.9 and
-1631.5-f3.0 kJ/mol for AI/i,2e3and AGF,zrsrespec- G IB B S
TEIIP. (tlf-il)9s)/r st -(Gt-H2eB)/r cD E N T H A L P YF R E t E N E R G T
tively. In Table 9 we list the high temperature
J/mol'K J/mol.K J/mol'K J/,iol'K kJ/nol kJ/nol
thermodynamic properties for tephroite together
with AIIt and AGg for reaction (2). 0.00 155.90 155.9C 128.50 -50.84 -49.35
r0.40 t0.40 10,15 !2.50 !2.60

400 35.00 19 6 . 17 l6l.l7 1 4 4 . 8 4- 5 1 . 1 3 -48.79

500 58.06 229.61 l7l,61 1 5 5 . 1 7- 5 1 . 3 2 -48.19
CozSiot 600 14.88 254.64 133,75 r62.43 -51.54 -47.54
700 87.79 284.10 196,31 167,11 -5r.90 -46.86
iJ00 98.05 306.78 208.13 111.80 -52,47 -45.09
There are apparently no high temperature heat
900 106.42 327.21i 220.73 r74.9r -53.55 -45,24
capacity data for Co2SiOa.At 298.15 K AS' is I 000 I I 3.40 345.75 232.35 177.33 -53.55 -44.30
-4.8+0.4 J/K for reaction(5) basedon our entropy
il00 I I9.30 362.16 243.46 I 79.21 -53,53 -43.37
I 200 124,36 378.41 251.05 180.68-53.78 -42.42
for Co2SiOa and those tabulated by Robie and I 300 128.14 392.9? 264,18 181.81-54.05 -4t.47
182.67-54.46 -40.49
others (1979)for CoO and SiO2.In order to use the r 400
I 500
132.56
135.92
405,43
419.06
2 13 . 8 7
2 8 3 .I 4 183.30-55.00 -39.46
I 500 138.90 ta3.t4 -55.12 -38.41
high temperature equilibrium data we have estimat- r 620 I 39.45
430.90
435.21
292.()0
295,15 183.31-55,74 -41.09
ed the heat capacity of Co2SiOabetween 400 and
1500K by using the sum ofthe heat capacitiesof2
moles of CoO and 1 of SiO2,joining theseestimates
smoothly with our measured values for Co2SiOa
ZCo + 2CO2:ZCoO + ZCO (7)
between 300 and 380 K, making proper allowance;
that is, by smoothing thru the temperatureregion of using data from Robie et al. (1979)for SiO2and the
anomalousheat capacitiescausedby the a-B transi- JANAF Thermochemical Tables for CoO to obtain
tion in qwrtz. Our estimated heat capacities for AG+ (1083-1573K) for reaction(6). We did not use
Co2SiOaare representedby the equation the equation given by Lebedev et al. for reaction (5)
becausein deriving it from their experimental data,
q: 3 4 3 . 9- 0 . 1 0 0 6 ?+ 3 . 8 1 4 xl | - s P
- 31757-o's(289-1500K) they had used values for AGi,1 for CO taken from
Richardson and Jeffes (194q which are incorrect.
This equation was used together with our measured Aukrust and Muan (1963)obtainedtheir AGfr values
entropy at298.15 K to calculate Si and lli-Ifrgsfor for (5) by combining their equilibrium constantsfor
use in a thirdlaw treatment of the high temperature (6) with their own measurementsfor (7). A@ for
equilibrium studieson Co2SiOa. reaction (5) from the CO reduction equilibrium
Navrotsky (1971) and Navrotsky er al. (1979) studies are in reasonableagreementbut are -3 KJ
obtained LH\as -21.6-12.1, and A!s6 : more negative than the AG" obtained from the emf
-23.0-f 0.8 kJ respectively,for the reaction measurements.A third-law treatment of these data
leads to LIffss for reaction (5) of -23.0-1 1.5 kJ.
2CoO + SiO2: Co2SiOa (5)
Combining our values with ffi,2es for CoO and
by solution calorimetry using 2PbO.B2O3as the SiO2 from rANAF (1971)or Robie et al. (1979),we
solvent. The Gibbs energy changefor this reaction obtain AGitsa : -1305.8-13.0kJ/mol and A.Ef,2es :
has been determined by Lebedev et al. (1963) -14ll.6-12.9 kJ/mol for Co2SiOa(from the ele-
between 1073and 1423K,and by Aukrust and Muan ments).
(1963) (1273-1573K) by CO reduction equilibria, At 975 K, the mean temperature of the molten
and by Kozlowska-Rog and Rog (1979) by solid salt calorimetric determination of Alf for reaction
state emf studies. (5), our treatment yields -27.9 kJ. This is approxi-
We have used the experimental data of Lebedev mately 5.6 kJ more negativethan the value given by
et al. (1962)and of Aukrust and Muan (1963)for the Navrotsky et al. (1979).
equilibrium
Magnetic entropiesand heat capacitiesof Mn2SiOa
Co2SiOa+ zCO : ZCo + SiO2 + ZCO2 (6) and Co2SiOa
over the temperature range 1083to 1573K. To the In order to separatethe magneticcontributions to
experimental A@ data of Lebedev et al. for (5) we the heat capacity from those arising from the lattice
added AGfr for the reaction vibrations we have adopted the corresponding
 ROBIE ET AL.: ENTROPIES OF MezSiO',, MnzSiOe, and CozSiOq 479

temperature at which the quanity (Si - 2Rln6) for

Mn2SiOahas the samevalue as ,S'(Ca2SiOa) at the
temperature ?, see Figure 6. This ratio was linearly
extrapolated to 0 K and used to estimate the lattice
./_.
/l rl entropies for MnzSiOa.This estimatedlattice entro-
,/ py was then subtracted from our smoothed values
( 5 6 o r g6 o - 2 R l n 2 i- r l
for the total entropy of Mn2SiOa,listed in Table 5,
o ,
E to obtain the entropy associatedwith the ordering
, of the magnetic moments of the Mn2+ ions. The
F ,/ Scars oo resultant magnetic entropy is shown in Figure 7.
,iJ
/. Our measurements for Co2SiOawere treated in a
I similar fashion. For Co2+ the spin quantum number
^Sis 3/2. However, at low temperature Coz* fre-
quently behaves as though S were 1/2, that is the
effectivespin S' : ll2 (e.9., Carlin and van Duyne-
veldt, 1977, p. g:71) and accordingly for our
o r:0"."",",","1
corresponding statesestimation of the lattice entro-
i" k"t'i:: py we used the quantity (,St - 2Rln2) for CozSiO+.
Fig. 6. Plot of the entropy of lCa2SiOa and the quantity
Scorsioo- 2Rln2.
We therefore expect the magnetic entropy of
CozSiO+to approach2Rln2 for ? D Trqand this is in
fact what we observed.
From Figure 7 we see that for Mn2SiOa the
states model suggested by Stout and Catalano magnetic entropy is only about 75 percent of 2Rln6
(1955).We used both our Ci data for Mg2SiOaand (29.79
J/mol.K) at the N6el temperature,and that of
those of King (1957)for yCa2SiO4 with the corre- CozSiOris only about 71 percentof 2Rln2 (11.53J/
sponding states model to estimate the lattice entro-
pies of MnzSiOaand Co2SiOa.r
We proceededas follows: first, from a large scale
plot of our smoothed values for the molar heat
capacity of Mg2SiOaand those of King (1957)for y
Ca2SiOa,which cover only the range 52 to 295 K,
we calculated the ratio TIT' where ?' is the tempera-
ture at which Ci of 7-Ca2SiOais the sameas C[ for
Mg2SiOaat the temperature ?. The ratio TIT' was
then plotted as a function of temperatureand extro-
polated to zero kelvin. From the extrapolated val- 20
ues of TIT' and our measured value of Ci (I) of Y
Mg2SiOawe generated values of Ci for yCa2SiOa o
E
for temperatures between 5 and 52 K. These esti-
mated values for Ci of yCa2SiO4 wer€ combined
with King's measuredvalues above 52 K to calcu-
late ^Sfrfor yCa2SiO+ at 5 K intervals. Next we o;
o
plotted Si for yCa2SiOa and the quantity (S1,"." -
2R1n6) for Mn2SiOa where 2R1n6 is the limiting
value for the magneticentropy at high temperature,
and derived the ratio TIT' wherc again T' is the

rAlthough neither Mg2SiOaor yCa2SiOaare ideal models for

calculating the lattice heat capacity or entropiesof the transition
metal olivines, we use them becauseZn2SiOa,which would have 0 1.0 2.0
been a more satisfactory model, does not crystallize in the
T/T N6"t
olivine structure---€venup to 150 kilobars (15 MPa) pressure,
(Syono et al. l97l). Fig. 7. Magnetic entropies of CozSiOaand Mn2SiOa.
4E0 ROBIE ET AL.: ENTROPIES OF MezSiOa,,Mn2SiOa, and Co2SiOa

mol'K) at fN. This indicates that appreciableshort Table 11. Magnetic heat capacity of Co2SiOa(cobalt olivine) in
range order is retained by the magnetic spin sys- the neighborhoodof the N€el temperatwe, 49.76K.
tems at temperatureswell above Zryand in fact the
total expected magnetic entropies are not fully Temp C o( t o t a l )
excited until temperaturesof the order of twice the
K e lv i n s J/(mol K) ,r?#t'*r ,r?ffiY'*l
N6el temperature have been reached.
39.41 20.67 8 . 85 1 l. 8 2
The lattice heat capacities of Mn2SiOa and 43.28 2 8 .5 5 1 1 .3 8 L7.17
CozSiOa were derived from our estimated lattice 4 6 .0 5 36.48 13.22 23.26
47.69 43.40 14.31 29.09
entropies utilizing the thermodynamic relation 4 8 .0 8 45.7| 14.57 3 1 .l 4
48.32 47.21 14.73 32.48
Cl.t : (dsht/d7) 4 8 .5 4 48.92 14.88 3 4 .0 4
48.76 50.84 15.04 3s.80
Values for C61 so obtained were subtracted from 48.97 53.15 1 5 .1 8 3 7. 9 7
o'rr measured total molar heat capacities (Figs. 4 49.17 55.80 _ tc . J J 40.47
49.37 5 9 .3 1 t5.47 4 3. 8 4
and 5) to obtain the magnetic heat capacities of 4 9 .5 5 64.05 15.60 48.45
Mn2SiOaand Co2SiOalisted in Tables 10 and 11. 49.95 39.47 1 5. 8 8 23.59
50.12 32.77 16.0r 16.76
51.02 2 7. 6 7 t6.67 11.00
The Debye temperatureof Mg2SiOa 52.10 26.62 t7.46 9 .1 6
5 3 .l 7 26.21 18.26
The Debye temperature, 0o is the most useful 54.23 26.30 1 9 .l 0 7.20
single parameter for characterizing the thermal 55.25 26.5r 19.78 6.73
properties of a solid. The Debye temperature may
be calculated from low temperature heat capacity
measurementsfrom equation (8)

6: Ie2ls)qrraRf lcufit3 (s)

heat capacity results for T < 15 K, Table l, we
where q is the number of atoms in the "molecule" obtain6:7681t5K.
and R is the gas constant (8.3143J/(mol.K)). This The Debye temperature is also related to the
relation is strictly rigorous only for temperaturesof mean velocity of sound in a solid by relation (9)
the order of T < fu/50, Blackman(1955).From our
6: (trlt<)tfqNp/4zMlr/3v- (9)
where h is Planck's constant (6.62618 x 10-34
J'sec), k is Boltzman's constant (1.38066x 10-23
Table 10. Magnetic heat capacity of Mn2SiOa(tephroite) in the
J'K-l), q is the numberof atomsin the "molecule",
neighborhoodof the N6el temperature,47.38 K.
N is Avogadro's number (6.02209x 1023mol-r), p
is the density in g ' cf,-3, and M the formula weight
*liT?,, 37[;:i'l] J/?;rt.K)
,,?ffiy.,r in grams. Robie and Edwards (1966) have shown
how the mean sound velocity, vm, may be calculat-
34.25 20.94 5.68 15.26 ed for a crystal of any symmetry from the single
38.10 25.26 17.58 crystal elastic constantsand the density by means
42.?3 30.49 9.94 20.55
42.72 3t.62 10.27 21.35 of numerical integration with a computer.
43.61 33.20 10.81 22.39 We have calculated the Debye temperature of
44.39 34.7I 11.29 23.42
45.09 7A t1 11.74 24.49
synthetic forsterite from the room temperatureelas-
45.57 36.88 72.04 24.84 tic stiffness constants (c;i) of Graham and Barsch
4 5 .B 5 3 8 .1 9 12.15 26.04
46.14 ?O ?E
(1969),and of Sumino et al. (1977)using a modified
1?.40 26.95
46.42 39.92 t2.58 tt .54
(nonrneN-IV) version of the computer program
46.64 41.00 L2.73
4 6. 8 1
28.27 mentioned by Robie and Edwards (1966). These
41.60 12.84 28.76
47.00 42.75 12.96 29.79 calculationsyield v- : 5.553and 5.532km/secfrom
4 7. t 8 +J.50 13.08 30.48 the data of Graham and Barsch and from Sumino et
47.68 41.92 13.41 2 8 .5 1
4 8 .0 4 30.71 13.65 l7.06 a/. respectively. The corresponding ff298; are
48.42 2 9 .5 8 1? 01
15.67 757.9 and 755.9 K. Sumino et al. (1977)also mea-
48.81 29.0? 14.17 1 4 . 85
51.72 29.83 16.14 13.69
sured the elastic constants at 83 K from which we
52.99 30.99 1 7. 0 4 LJ.v) obtainedv- : 5.604km/secand 6s (83) : 766.6K.
From results of Sumino et al. we estimate6tOt :
 ROBIE ET AL.: ENTROPIES OF MezSiOe, Mn2SiOa, and CozSiOa
768.6 K which is in excellent agreementwith our
calorimetric value.
'At the lowest temperaturesof our measurements,
the heat capacity of MnzSiO+ and Co2SiOa are
composed of a lattice (Debye) term plus a contribu-
tion arising from the spin waves (magnons)(seefor
example, Gopal, 1966).Both of these terms vary as
T3 and accordingly we cannot evaluate their Debye
temperatures (6) from our measurements. One
may however make a first order estimate of the
coefficient of the spin wave term as follows: from
the room temperature elastic constant values of
Sumino (1979)for Mn2SiOaand Co2SiO4one calcu-
lates 6 of 533 and 548 K respectively.Assuming
these (elastic) Debye temperatures to be indepen-
dent of temperature, the lattice heat capacity may
be calculated from (10)
Cy: |215 rraqn7dltdo (10)
for T ( enl50. This calculated lattice heat capacity
is then substracted from the measured (total) heat
capacity to obtain that arising from the spin waves.
At 10 K for Co2SiOathe spin wave heat capacity is
approximately 18 percent of the measured molar
q.
Acknowledgments
This work was in part supported by the U.S. Department of
Energy through the Los Alamos Scientific Laboratory, contract
X69-9915F-3.We thank our U.S. GeologicalSurvey colleagues,
H.T. Haselton, Jr., and Susan Werner Kieffer for their many
useful suggestionsfor improving the manuscript.
References
Aukrust, E., and Muan, A. (1963)The stabilitiesof CoO.Al2O3,
CoO.Cr2O3,and 2CoO.SiO2.Journal of the American Ceramic
Society, 46, 35E.
Biggers, J.V., and Muan, A. (1967)Activity-+omposition rela-
tions in orthosilicate and metasilicate solid solutions in the
system MnO-CoO-SiO2. Journal of the American Ceramic
Society, 50, 230-235.
Blackman, M. (1955)The specific heat of solids. In S. Flugge,
Ed., Encyclopedia of Physics, p. 325-382. Springer-Verlag,
Berlin.
Carlin, R.L., and van Duyneveldt, A.J. (1977)Magnetic proper-
ties of transition metal compounds. Springer-Verlag, New
York, New York.
Charlu, T.V., Newton, R.C., and Kleppa,O.J. (1975)Enthalpies
of formation at 970 K of compounds in the system MgO-
Al2O3-SiO2 from high temperaturesolution calorimetry. Geo-
chimica et CosmochimicaActa. 39.l4E7-1497.
Fujino, K., Saski,S., Takeuchi,Y., and Sadanaga, R. (1981)X-
ray determination of electron distributions in forsterite, faya-
Iite and tephroite. Acta Crystallographica,B37, 513-518.
Gopal, E.S.R. (1966)Specificheatsat low temperatures.Plenum
Press. New York.
4El
Graham, E.K., and Barsch, G.R. (1969)Elastic constantsof
single crystal forsterite as a function of temperature and
pressure. Journal of GeophysicalResearch,74, 5949-5960.
Hazen, R.M. (1976)Etrect of temperature and pressure on the
crystal structure of forsterite. American Mineralogist, 61,
t2E0-1293.
Hemingway, B.S. and Robie, R.A. (1977)Enthalpiesof forma-
tion of low albite (NaAlSi3O8), gibbsite (AI(OH):), and
NaAlOz; revised values for AH'r.zss and AG"s.2esof some
aluminosilicate minerals. Journal of Research of The U.S.
Geological Survey, 5, 413429.
Jeffes,J.H., Richiirdson,F.D., and Pearson,J. (1954)The heat
of formation of manganousorthosilicate and manganoussul-
fide. Transactions of the Faraday Society, 50,364-370.
Kelley, K.K. (1943) Specific heats at low ternperatures of
magnesiumorthosilicate, and magnesiummetasilicate.Journal
of the American Chemical Society, 65,339-341.
Kiseleva,I.A., Ogorodova,L.P., and Topor, N.P. (1979)Deter-
mination of the heat of dissolution of magnesiumoxide by a
high-temperaturecalorimetric method. Vestnik Moskovskogo
Universiteta Seriia 2, Khimia, 20, 243247. [Chemical Ab-
stracts, 9l, 97622a1
King, E.G. (1957) Low temperature heat capacities and entro-
pies at 29E.15 K of some crystalline silicates containing
calcium. Journal of the American Chemical Society, 79,5437-
5438.
King, E.G., Barany,R. and Weller,W.W. (1967)Thermodynam-
ic properties of forsterite and serpentine. U.S. Bureau of
Mines Report of Investigations, 6962.
Kozlowska-Rog, A., and Rog, G. (1979) Thermodynamics of
zinc and cobalt silicates. Polish Journal of Chemistry, 53,
2083-2086.
Lebedev, 8.G., Levitskii, V.A., and Burtstev, V.A. (1963)
Equilibrium in reduction of cobalt orthosilicate by carbon
monoxide. Russian Journal of Physical Chemistry, 36, 460-
461.
Mah, A.D. (1960)Thermodynamic properties of manganeseand
its compounds. U.S. Bureau of mines Report of Investiga-
tions, 5600.
Morimoto, N., Tokonami, M., Watanabe,M., and Koto, K.
(1974) Crystal structure of three polymorphs of Co2SiOa.
American Mineralogist, 59, 475485.
Narvrotsky, A. (1971) Thermodynamics of formation of the
silicatesand germanatesof somedivalent transition metalsand
of magnesium. Journal of Inorganic and Nuclear Chemistry,
33, 4033-4050.
Navrotsky, A., and Coons, W.E. (1976)Thermochemistryof
some pyroxenes and related compounds.Geochimicaet Cos-
mochimicaActa. 40. 128l-1288.
Navrotsky, A., Pintchovski,F.S., and Akimoto, S. (1979)Calo-
rimetric study of the stability of high pressure phasesin the
systems CoO-SiOr and "FeO"-SiO2 and calculation of the
phasediagramsin MO-SiOz systems.Physicsof the Earth and
Planetary Interiors, 19, 275-292.
Nomura, S., Santoro,R.P., Fang,J., and Newnham,R.E. (1964)
Antiferromagnetism in cobalt orthosilicate. Journal of Physics
and Chemistry of Solids, 25, %l-925.
Orr, R.L. (1953)High temperature heat contents of magnesium
orthosilicate and ferrous orthosilicate. Journal of the Ameri-
can Chemical Society, 75,528-529.
Osborne,D.W., Flotow, H.E., and Schreiner,F. (1967)Calibra-
tion and use ofgermanium resistancethermometersfor precise
4E2 ROBIE ET AL.: ENTROPIES OF MgzSiOa,,Mn2SiOa, and Co2SiOa
heat capacity measurementsfrom I to 25'K, high purity
copper for inter-laboratory heat capacity comparisons. Re-
view of ScientificInstruments.38. 159-168.
Richardson,F.D. and Jeffes,J.H.E. (1949)The thermodynamic
background of iron and steel making processes.I. The blast
furnace. Journal ofthe Iron and Steel Institute, 163,397420.
Robie, R.A., and Edwards, J.L. (1966)Some Debye tempera-
tures frorn single-crystal elastic constant data. Journal of
Applied Physics, 37, 2659-2663.
Robie, R.A. and Hemingway, B.S. (1972)Calorimetersfor heat
of solution and low-temperatureheat capacity measurements.
U.S. Geological Survey ProfessionalPaper 755.
Robie, R.A., Hemingway,B.S., and Fisher,J.R. (1979)Thermo-
dynamic properties of minerals and related substances at
298.15 K and 1 bar (105 pascals)pressure and at higher
temperatures. U.S. Geological Survey Bulletin 1452.
Robie, R.A., Hemingway,B.S., and Wilson, W.H. (1976)The
heat capacities of Calorimetry Conferencecopper and of musco-
vite KAl2(AlSi3)Oro(OH)2,pyrophyllite AlzSi+Oro(OH)2,and
illite K3(AlTMgXSir4Al2)O4o(OH)s between 15 and 374 K and
their standard entropies at 298.15K. Joumal of Researchof
the U.S. GeologicalSurvey, 4,631-644.
Rog, G., Langanke, 8., Borcherdt, G., and Schmalzried,H.
(1974) Determination of the standard Gibbs free energies of
formation of the silicatesof cobalt, magnesium,and strontium
by e.m.f. measurements.Journal of Chemical Thermodynam-
ics.6.lll3-1119.
Ross, C.S., Foster, M.D., and Meyers, A.T. (1954)Origin of
dunites and of olivine-rich inclusions in basaltic rocks. Ameri-
can Mineralogist, 39, 693J37 .
Santoro,R.P., Newnham,R.E., and Nomura, S. (1966)Magnet-
ic properties of Mn2SiOaand Fe2SiOa.Journal of Physics and
Chemistry of Solids,27,655-666.
Schreiber, E. (1963)The equilibrium boundary for the reaction
2MnCO3 + SiO2? Mn2SiOa+ 2CO2. Ph.D. Thesis, State
University of New York, Alfred.
Schwerdtfeger,K., and Muan, A. (1966)Activities in olivine and
pyroxenoid solid solutions of the system Fe-Mn-Si-O at
ll50'C. Transactions of the Metallurgical Society of AIME,
236,201-211.
Stout, J.W., and Catalano, E. (1955)Heat capacity of zinc
fluoride from ll to 300'K. Thermodynamic functions of zinc
fluoride. Entropy and heat capacity associatedwith the anti-
ferromagnetic ordering of manganousfluoride, ferrous fluo-
ride, cobaltous fluoride and nickelous fluoride. Journal of
Chemical Physics, 23, 2013-2022.
Sumino, Y. (1979) The elastic constants of MnzSiO+,Fe2SiOa
and Co2SiOa, and the elastic properties of olivine group
minerals at high temperature.Journal ofPhysics ofthe Earth,
27,209-238.
Sumino, Y., Nishizawa,O., Goto, T., Ohno, I. and Ozima, M.
(1977) Temperature variation of elastic constants of single
crystal forstenite between 190"and 400"C. Journal of Physics
of the Earth. 25. 377-392.
Syono,Yasuhiko,Akimoto, Syun-iti,and Matsui,Yoshito(1971)
High pressuretransformationin zinc silicates.Journal of Solid
State Chemistry, 3, 369-380.
Takei, Humihiko (1976) Czochralski growth of MnzSiO+(teph-
roite) single singlecrystal and its properties.Journal of Crystal
Growth. 34.125-131.
Takei, Humihiko (197E) Growth of fayalite (FezSiOa single
crystals by the floating zone method. Journal of Crystral
Growth. 43.46f468.
Takei, Humihiko and Kobayashi, Takaaki (1974) Growth and
properties of Mg2SiOa single crystals. Journal of Crystal
Growth. 23.l2l-124.
Torgeson,D.R., and Sahama,T.G. (1948)A hydrofluoricacid
solution calorimeter and the determination of the heats of
formation of Mg2SiOa,MgSiO3, and CaSiOr. Journal of the
American Chemical Society, 70, 2156-2160.
Manuscript received, July 20, 1981;
acceptedfor publication, January 20, 1982.