Contributions on thermal behaviour and crystal chemistry of anhydrous phosphates. XXXIV [1]: Oxygen equilibrium pressures in ternary systems M/P/O (M = Co, Ni) and heats of formation of anhydrous cobalt(II) and nickel(II) phosphates

Article abstract of DOI:10.1002/zaac.200300131

Oxygen equilibrium pressures have been measured in the temperature range 800°C to 1000°C by coulometric/potentiometric techniques for several equilibrium regions in the ternary systems M/P/O (M = Co, Ni). In both systems oxygen coexistence pressures of th

Full text of DOI:10.1002/zaac.200300131

Contributions on Thermal Behaviour and Crystal Chemistry of Anhydrous Phosphates.
XXXIV [1]
Oxygen Equilibrium Pressures in Ternary Systems M / P / O (M ؍ Co, Ni)
and Heats of Formation of Anhydrous Cobalt(II) and Nickel(II) Phosphates
a
b
a,
Matthias Blum [2], Klaus Teske , and Robert Glaum *
a
Bonn, Institut für Anorganische Chemie, der Rheinischen Friedrich-Wilhelms-Universität
b
Dresden, Institut für Anorganische Chemie der Technischen Universität
Received December 4th, 2002.
Abstract. Oxygen equilibrium pressures have been measured in the
temperature range 800 °C to 1000 °C by coulometric/potentio-
metric techniques for several equilibrium regions in the ternary sys-
tems M / P / O (M ϭ Co, Ni). In both systems oxygen coexistence
pressures of three-phase equilibrium solids phosphide/phosphate
Thermodynamic data from literature for the phosphides of cobalt
and nickel and estimated heat capacities for the anhydrous phos-
phates Co (PO , Co , Ni (PO , Ni and Ni
were used for these calculations. Thus obtained enthalpies and en-
tropies are compared to results from thermodynamic modelling of
observed solid phase equilibria in the ternary systems M / P / O
(M ϭ Co, Ni).
3
4
)
2
2
P
2
O
7
3
4
)
2
2
P
2
O
7
2 4 12
P O
are about 3 to 5 orders of magnitude smaller than p(O
corresponding M / MO system.
Heats of formation ∆ H°298 and standard entropies S°298 for the
2
) above the
s
s
f
nd
rd
phosphates have been obtained from 2 and 3 law evaluation of
the temperature dependence of the oxygen coexistence pressures.
Keywords: Phosphates; Thermodynamic data; Thermochemistry;
Gibbs phase triangles
Beiträge zum thermischen Verhalten und zur Kristallchemie von wasserfreien Phosphaten. XXXIII [1]
Sauerstoffkoexistenzdrucke in den ternären Systemen M / P / O (M ؍ Co, Ni) und
Bildungsenthalpien für wasserfreie Phosphate des zweiwertigen Cobalts und Nickels
Inhaltsübersicht. Mittels coulometrischer/potentiometrischer Mes-
sungen wurden für weite Teile der Dreistoffsysteme M / P / O (M ϭ
Co, Ni) im Temperaturbereich 800 bis 1000 °C die Sauerstoffkoexi-
wasserfreien Phosphate Co
und Ni 12 werden durch Auswertung der Sauerstoffpartial-
druckmessungen nach dem 2. Hauptsatz Bildungsenthalpien
H°298 und Standardentropien S°298 für die Phosphate erhalten.
3 4 2 2 2 7 3 4 2 2 2 7
(PO ) , Co P O , Ni (PO ) , Ni P O
2 4
P O
stenzdrucke bestimmt. Die Werte von p(O
2
) über dreiphasigen
∆
B
Gleichgewichtsbodenkörpern Phosphid/Phosphat liegen in beiden
Dreistoffsystemen 3 bis 5 Größenordnungen unterhalb jener über
den Bodenkörpern Ni / NiO bzw. Co / CoO .
s s s s
Mit thermodynamischen Literaturangaben zu den Phosphiden von
Cobalt und Nickel sowie abgeschätzten Wärmekapazitäten für die
Die Daten werden mit Werten aus einer Auswertung nach dem 3.
Hauptsatz (mit abgeschätzten Entropien für die Phosphate) und
mit den Ergebnissen thermodynamischer Rechnungen zur Model-
lierung der beobachteten Gleichgewichtsräume in den Dreistoffsy-
stemen M / P / O (M ϭ Co, Ni) verglichen.
Introduction
O (PO ) [5], CuP O [6]), the redox behaviour of metal
24
4 24
4
11
oxides towards phosphorus vapour and thermodynamic co-
existence of phosphates and phoshides were of major inter-
est to us. In the meantime phase relations in several ternary
systems M / P / O (M ϭ Ti [7], V [8], Cr, Mo [9], W [10],
Mn [11], Fe [12], Co [13], Ni [14], Cu [15]) have been clari-
fied, mostly by our group. A review of the various phase
diagrams M / P / O has been given recently [16].
In search for new anhydrous phosphates containing tran-
sition metals in comparatively low oxidation states we
started a systematic investigation into the equilibrium
phases in Gibbs phase triangles transition metal / phos-
phorus / oxygen. At the beginning of the study, synthesis of
new phosphates (e.g. V O(PO ) [3], Cr (PO ) [4], Ti31-
2
4
7
4 6
In some cases equilibrium phase relations in systems M /
P / O (M ϭ Ti [7], Mo [9], Mn [11], Co [13]) have been used
*
Prof. Dr. Robert Glaum
for an estimation of heats of formation ∆ H°298 of anhy-
f
Institut für Anorganische Chemie
Universität Bonn
Gerhard-Domagk-Straße 1
D-53121 Bonn
e-mail: rglaum@uni-bonn.de
Fax: 0228 / 73 56 60
drous transition metal phosphates. However, these esti-
mations contain several uncertainties. Experimental infor-
mation on thermodynamic properties of anhydrous tran-
sition metal phosphates is scarce. This lack is due to the
generally low solubility and high thermochemical stability
Z. Anorg. Allg. Chem. 2003, 629, 1709Ϫ1717
DOI: 10.1002/zaac.200300131
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
1709

M. Blum, K. Teske, R. Glaum
of transition metal phosphates, difficult equilibration of re-
action mixtures metal oxide / phosphorus oxide, and com-
plex melting and vaporization behaviour of phosphates.
These characteristics have up to now prevented almost com-
pletely any reliable systematic determination of thermo-
dynamic data for transition metal phosphates. To gather at
least some landmark thermodynamical data we have con-
ducted temperature dependent coulometric / potentiometric
measurements of oxygen equilibrium pressures in the ter-
nary systems M / P / O (M ϭ Co, Ni). These systems were
chosen for a first study for three reasons. Firstly, phase
equilibria between phosphides and phosphates have been
determined particularly accurate in these cases [2, 13]. Sec-
ondly, thermodynamic data for cobalt [13, 17] and nickel
Table 1 Summary of studied three phase regions in the ternary
systems Co / P / O and Ni / P / O and starting materials used for
their preparation.
symbol
equilibrium solidsa)
starting materials/mg
Co II/1
Co II/2
Co III/1 Co
Co, Co
Co, Co
2
2
P, Co
P, Co
3
3
(PO
(PO
4
)
)
2
2
Co
3
Co
3
Co
3
(PO
(PO
(PO
4
4
4
)
2
)
2
)
2
116.2, Co 373.1, P 47.9
116.6, Co 373.2, P 48.0
171.5, Ͱ-Co P O
2 2 7
4
2
P, Co
3
(PO
4
)
2
, Ͱ-Co
2
P
2
O
7
1
88.5, Co 148.3, P 38.9
Co III/2 Co
Co III/3 Co
Co IV/1 Co
2
2
2
P, Co
P, Co
P, CoP, Ͱ-Co
3
3
(PO
(PO
4
)
)
2
2
, Ͱ-Co
, Ͱ-Co
2
2
P
P
2
O
O
7
same as Co III/1
CoO 224.8, P 31.0
4
2
7
2
P
2
O
7
2 2 7
Ͱ-Co P O 116.0, Co 323.2,
P 128.1
Co IV/2 Co
Co IV/3 Co
Co IV/4 Co
2
2
2
P, CoP, Ͱ-Co
P, CoP, Ͱ-Co
P, CoP, Ͱ-Co
2
2
2
4
P
P
P
2
2
2
O
O
O
7
7
7
same as Co IV/1
CoO 170.1, P 53.6
same as Co IV/3
NiO 499.6, P 41.2
NiO 448.0, P 60.8
NiO 166.6, Ni 288.3, P 50.3
same as Ni II/2
Ni I/1
Ni, NiO, Ni
3
(PO
)
2
Ni II/1
Ni II/2
Ni II/3
Ni, Ni
Ni, Ni
Ni, Ni
3
3
3
P, Ni
P, Ni
P, Ni
3
(PO
(PO
(PO
4
)
2
)
2
)
2
[
2, 18] phosphides, that will be needed during evaluation,
b)
b)
3
4
4
turned out to be fairly reliable in modelling chemical vap-
our transport experiments of these phosphides. Last but not
least estimations in these systems let us anticipate oxygen
3
Ni III/1 Ni
Ni III/2 Ni
Ni III/3 Ni
3
3
3
P, Ni
P, Ni
P, Ni
3
3
3
(PO
(PO
(PO
4
4
4
)
)
)
2
2
2
, Ͱ-Ni
, Ͱ-Ni
, Ͱ-Ni
2
2
2
P
2
P
2
P
2
O
7
O
7
O
7
Ni
same as Ni III/1
Ni (PO 52.3, Ͱ-Ni
Ni 66.1
Ͱ-Ni P O 83.7, Ni 383.2, P 71.1
3 4 2
(PO ) 606.4, P 10.6
3
4
)
2
2 2 7
P O 389.9,
Ϫ12
equilibrium pressures of an appropriate magnitude (10
to 10
Ϫ22
Ni IV/1 Ni P, Ni P, Ni P O
atm) to allow their reliable measurement by coulo-
3
2.55
2
2
7
2
2
7
Ni V/1
Ni2.55P, Ni12
P
5
, Ni
2
P
2
O
7
NiO 374.8, Ni 46.4, P 72.6
metric / potentiometric techniques.
Ni VI/1 Ni12
Ni VI/2 Ni12
Ni VI/3 Ni12
Ni VI/4 Ni12
Ni VII/1 Ni12
Ni VII/2 Ni12
P
P
P
P
P
P
5
5
5
5
5
5
, Ͱ-Ni
, Ͱ-Ni
, Ͱ-Ni
, Ͱ-Ni
2
2
2
2
P
P
P
P
2
2
2
2
O
O
O
O
7
7
7
7
, Ni
, Ni
, Ni
, Ni
2
2
2
2
P
P
P
P
4
O
O
O
O
12 NiO 358.8, P 99.2
12 same as Ni VI/1
12 Ni P O12 272.4, Ni 210.1, P 34.6
2 4
12 same as Ni VI/3
NiO 358.4, P 98.4
4
4
4
Experimental
, Ni
, Ni
2
P, Ni
P, Ni
2
P
P
4
O
12
O
12
12
12
2
2
4
same as Ni VII/1
Sample preparation
Ni VII/3 Ni12P , Ni P, Ni P O
Ni VII/4 Ni12
Ni P O 46.7, Ni 365.6, P 88.1
5
5
2
2
4
2
4
12
P
, Ni
2
P, Ni
2
P
4
O
same as Ni VII/3
Measurements of oxygen equilibrium pressures have been conduc-
ted in several three-phase equilibrium regions (Table 1) of the ter-
nary systems Co / P / O and Ni / P / O. As starting materials cobalt
metal (Fluka, >99.8 %), CoO (Merck p. a.), nickel metal (Riedel de
H a¨ n, 99.8 %), and red phosphorus (Fa. Knapsack, 6N “electronic
grade”) have been used.
a)
equilibrium phases at 800 °C (Co / P / O) and 750 °C (Ni / P / O), respec-
tively, identified from Guinier photographs
b)
2 2 7
σ-Ni P O was observed as additional solid phase indicating incomplete
equilibration; these measurements were not used for derivation of heats of
formation.
Olive-green NiO has been synthesized from Ni(NO
Merck p. a.) at 1000 °C under nitrogen, according to a procedure
described in reference [19].
3 2 2
) · 6 H O
(
and elemental phosphorus from the material. After this
procedure complete equilibration and the presence of the
desired phase composition was checked again by Guinier
photographs.
Initial measurement NiII/0 (orientating measurement;
equilibrium region NiII, zeroth measurement; cfg. Table
For synthesis of Co (PO ) , Ͱ-Co P O , Ni (PO ) , Ͱ-
3
4 2
2
2
7
3
4 2
Ni P O , and Ni P O the metal powders were dissolved
2
2
7
2
4
12
in dilute nitric acid. The acidic metal nitrate solution was
mixed with appropriate amounts of (NH ) HPO (Merck p.
4
2
4
1Ϫ3) with an equlibrium solid, that has been synthesized
a.) in aqueous solution. In a beaker the mixture was evapo-
rated to dryness. Then, the residue was transferred into a
silica crucible and heated in air with the temperature rising
slowly from room temperature to 800 °C (600 °C in case of
Ni P O ). Keeping the phosphates several days at the final
without the second heating procedure showed signs of a
poissoning of the Pt-electrodes, probably due to the men-
tioned volatile compounds. Samples equilibrated with two
heating cycles showed no such behaviour.
2
4
12
temperature yielded micro-crystalline samples (Guinier
photographs).
For equilibration appropriate amounts (Table 1) of the
powdered starting materials (Σm ഠ 500 mg) were reacted
Measurements of oxygen coexistence pressures
The p(O ) vs. T functions for the various three-phase equi-
2
for 3Ϫ5 d in sealed, evacuated silica tubes (l ഠ 10 cm, л ഠ
librium regions (Table 1Ϫ3) were measured on powders (ϳ
100Ϫ200 mg) placed in a silica boat within the furnace of
a solid electrolyte device (Fig. 1). A steady stream of argon
3
15 mm, V ഠ 18 cm ) at 750 °C (Ni / P / O) and 800 °C
(Co / P / O), respectively. About 10 mg iodine were added
as mineralizer. After the heat treatment the solid reaction
products were washed with dilute sodium hydroxid, water,
and acetone and dried at 120 °C in air. Eventually, the
products were heated for a second time (1d, same tempera-
ture as before) in sealed, evacuated silica tubes, this time
without mineralizer, to remove all traces of iodine, iodides,
with a p(O ) controlled by cell 1 (“upstream cell”) flowed
2
through the sample furnace. Gases of controlled p(O ) (in-
2
Ϫ13
Ϫ25
vestigated range: 10
to 10
atm) were prepared from
argon / hydrogen (1000 ppm) mixtures by subsequent elec-
trolytic pumping (I of cell 1). The resulting p(O ) were
2
measured by a solid electrolyte potentiometric cell (U of
1
1710
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
zaac.wiley-vch.de
Z. Anorg. Allg. Chem. 2003, 629, 1709Ϫ1717

Contributions on Thermal Behaviour and Crystal Chemistry of Anhydrous Phosphates. XXXIV
Table 2 Oxygen equilibrium pressures (coulometric evaluation) in
the ternary systems Co / P / O and Ni / P / O given as log p(O ) ϭ
A/T ϩ B, average temperature T(av), ∆
Table 3 Heterogenous equilibrium reactions determining the oxy-
gen pressures in the various three phase regions of the ternary sys-
tems Co / P / O and Ni / P / O.
2
R
R T(av)
HT(av) and ∆ S .
Symbol
A, B
T(av)/K
∆
R
H
T(av)
∆
R
S
T(av)
symbol
reaction
kcal/mol
cal/mol·K
Co II
Co III
1/4 Co
4/9 Co
2/7 Co
2/7 Co
1/6 Co
s
ϩ 1/4 Co
3
(PO
4
)
2,s ϭ 1/2 Co
2
P
s
ϩ O2,g
ϩ O2,g
Co II/1a) Ϫ31000(1000), 10(1)
1098
992
142.3
1
b)
47.7
b)
(PO ϭ 2/9 Co
)
P
s
O
ϩ O2,g
ϩ O2,g
7,s ϩ 4/9 Co P
3
2
2
2
4
2
2
2
2 s
c)
c)
a)
30.0
36.4
Co IV
b, c)
P
P
P
2
2
4
O
O
O
7,s ϭ 4/7 CoP
7,s ϭ 4/7 CoP
b)
c)
b)
c)
b)
c)
b)
c)
b)
c)
b)
c)
b)
c)
b)
Co III/1
Co III/2
Co III/3
Co IV/1
Co IV/2
Co IV/3
Co IV/4
Ϫ28040(30), 7.73(3)
Ϫ28160(40), 7.79(3)
Ϫ28710(20), 8.22(2)
Ϫ29030(90), 7.61(8)
Ϫ28930(60), 7.24(5)
Ϫ29460(60), 7.88(5)
Ϫ28400(50), 6.96(5)
128.1
30.6
128.6
30.9
131.1
31.5
132.6
34.8
132.1
36.1
135.6
35.5
129.7
35.0
35.3
Co V
s
c)
b)
1
37.8
Co VI
12,s ϭ 1/6 CoP
ϩ O2,g
2,s ϩ 3/4 Ni
(PO 2,s ϭ 6/11 Ͱ-Ni
7,s ϩ 620/315 Ni
7,s ϩ 400/75 Ni2.55
7,s ϭ 14/49 Ni
12,s ϩ 1/2 Ni12
s
ϩ 1/6 CoP3,s ϩ O2,g
b)
1003
998
35.6
Ni I
2 NiO
1/4Ni
8/11 Ni
s
ϭ 2 Ni
s
c)
1
37.7
Ni II
Ni III
Ni IV
Ni V
Ni VI
3
(PO
4
)
s
ϭ Ni
3
P
s
ϩ O2,g
7 ,s ϩ 4/11 Ni
b)
37.5
3
4
)
2
P
2
O
3
P
s
ϩ O2,g
ϩ O2,g
c)
1
37.7
2/7 Ͱ-Ni
2/7 Ͱ-Ni
2
P
P
2
O
O
3
P
P
s
ϭ 800/315 Ni2.55
P
P
s
b)
1053
1054
1046
1064
34.8
ϭ 124/105 Ni12 5,s ϩ O2,g
12,s ϩ 4/49 Ni12P5,s ϩ O2,g
c)
2
2
s
1
36.7
38/49 Ͱ-Ni
1/6 Ni
2
P
2
O
2
P
4
O
b)
33.1
Ni VII
2
P
4
O
P5,s ϭ 19/6 Ni
2
P
s
ϩ O2,g
c)
1
36.7
b)
35.9
a)
c)
equilibrium solids: Co
not measured
equilibrium solids: CoP, Ͱ-Co
2
P, CoP, Ͱ-Co
2 2 7
P O
1
36.8
b)
c)
b)
31.8
c)
2
P
2
O
/
, Co P O
2 4 12
1
36.6
Ni I/1d)
Ni II/1
Ϫ23022(Ϫ), 7.52(Ϫ)
Ϫ27250(70), 8.43(7)
1109
1066
105.1
124.5
32.3
38.3
37.7
43.1
44.9
52.8
b)
c)
b)
c)
1
23.8
e)
b)
b)
b)
b)
Ni II/2
Ni II/3
Ϫ28240(30), 9.43(3)
Ϫ28760(60), 9.84(5)
Ϫ30800(400), 11.6(4)
1030
1081
1056
129.0
131.4
140.7
e)
b)
b)
Ni III/1
b)
b)
c)
b)
c)
b)
c)
b)
c)
b)
c)
b)
c)
b)
c)
b)
c)
b)
c)
b)
c)
b)
Ni III/2
Ni III/3
Ϫ30400(400), 11.34(4)
Ϫ28630(40), 9.68(3)
1068
1030
138.9
130.8
51.8
b)
44.2
c)
1
25.4
132.2
31.3
129.3
27.7
135.8
32.7
137.4
33.1
127.9
30.4
134.1
31.6
135.5
31.5
133.0
31.8
133.9
31.2
38.9
b)
Ni IV/1
Ni V/1
Ni VI/1
Ϫ28950(40), 9.55(4)
Ϫ28320(60), 8.62(5)
Ϫ29700(100), 9.4(1)
1033
1038
910
43.6
c)
1
44.9
Fig. 1 Schematic representation of the experimental setup with
the solid electrolyte cells. I pumping current; U , U voltages of
1 2
the cells.
b)
39.4
c)
1
37.6
b)
43.1
c)
1
39.7
Ni VI/2f) Ϫ30100(100), 9.73(1)
Ni VI/3g) Ϫ28020(20), 8.11(2)
Ni VII/1 Ϫ29370(50), 8.62(4)
Ni VII/2 Ϫ29670(30), 8.95(3)
Ni VII/3 Ϫ29130(50), 8.37(5)
Ni VII/4 Ϫ29300(50), 8.60()
912
44.4
b)
c)
1
39.6
to adjust the equilibrium, netto oxygen exchange of the
sample with the gas phase is suppressed, and the composition
b)
913
37.0
c)
1
39.6
b)
1036
1031
1053
1052
39.4
of the solid is kept constant as long as p(O ) set in cell 1 is
2
c)
1
36.8
identical to the oxygen co-existence pressure in the reactor.
For systems with sufficiently high equilibration rates
b)
40.9
c)
1
36.8
b)
38.2
(solid/gas) a continous measuring procedure of p(O ) was
2
c)
1
36.6
chosen. In this case the temperature of the reactor was pro-
gram controlled (1200°/h for reaching the start temperature;
b)
39.3
c)
1
36.6
6
1
00°/h for measurement) and the pumping current I of cell
a)
measured stepwise for four temperatures
2
3
b) nd
1
was continously adjusted manually maintaining U ϭ U2
law evaluation of oxygen partial pressure measurement
law evaluation with estimated entropies
measurement of just two control pressures at 1089 K and 1129 K.
c) rd
during the measurement. The development of all exper-
d)
imental parameters (TSample, I, U , U ) was recorded digi-
1
2
e)
f)
2 2 7
σ-Ni P O as fourth phase indicating incomplete equilibration
tally (see Fig. 2). Continous measurement of p(O ) vs. T for
meassured stepwise for six temperatures
only cooling curve measured
2
g)
the system CoII did not give reliable data. Unacceptable
large hysteresis between p(O ) obtained on heating and
2
cooling, respectively, indicated a strong kinetic influence.
Therefore, the continous measurement was abandoned for
cell 1). After equilibration with the sample material, the de-
viation of the oxygen partial pressure from the initial value
is indicated by cell 2 (“downstream cell”). For isostoichi-
a stepwise determination of p(O ) vs. T. This allows suffic-
2
iently long equilibration periods, however, at the cost of a
significantly smaller number of temperatures measured. Ad-
ditionally, measurement NiVI/2 was carried out stepwise,
just to check the equilibration rate, that turned out to be
sufficiently high.
ometric conditions during heating the voltage U of cell 2
2
must be kept equal to U by means of subsequent pumping
1
of oxygen into the argon in cell 1. During the measurements
the temperature of cell 1 and cell 2 was held constantly at
7
50 °C, while the temperature of the sample was varied. For
Regardless of the measurement mode, from the recorded
temperature TSample, I (coulometric evaluation) and U1
(potentiometric evaluation), respectively, the functions
temperature measurement NiCr/Ni thermocouples were
used. On the condition that the rate of oxygen exchange
between the gas phase and the solid phase is sufficiently high
p(O ) vs. T were determined (Tab. 2, Figs. 3 and 4). The
2
Z. Anorg. Allg. Chem. 2003, 629, 1709Ϫ1717
zaac.wiley-vch.de
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
1711

M. Blum, K. Teske, R. Glaum
Fig. 2 Development of experimental parameters during a typical experiment. (Temperature of reactor T, pumping current I (cell 1),
potentials U and U )
1 2
coulometric evaluation takes into account the pumping cur-
For an estimation of the oxygen loss we assumed a poten-
rent I and follows Faradays law combined with the flow
tial difference between U and U of ∆U ϭ 2 mV, which is
1
2
rate of the carrier gas (5 l/h) to calculate p(O ) from the
twice than usual for one hour measurement. This corres-
2
resulting ratio p(H )/p(H O) (eq. 1 and 2 ). An estimated
ponds to a current difference of 100 µA/sec between down-
and upstream cell which is 3.6·10 µA per hour. Thus, the
2
2
0
5
residual moisture of cH ϭ 70 ppm in the carrier gas has
2
O
been allowed for.
amount of electrons is 1.85 µmol, which means an oxygen
conversion of 0.925 µmol during one hour. Taking for ex-
ample the triangle NiVII (see table 3) this means per hour
a conversion of 0.154 µmol (ϳ6.73 µg) nickel-cyclo-tetra-
metaphosphate per hour. Regarding average initial amounts
of 100Ϫ200 mg sample material, this loss is sufficiently
small to exclude the possibility of shifts in sample compo-
sition.
In addition, constancy of sample composition has been
checked after the measurements by X-ray powder diffrac-
tion. In all cases Guinier exposures (IP-technique [21])
showed the same phase composition as before the experi-
ment.
418
p(H
2
)
1000 Ϫ · I
5
ϭ 4
(1)
(2)
18
· I ϩ c 0H O
p(H O)
2
5
0
) ϭ 102·(2.958Ϫ13022/TSample
2
2 2
)Ϫlog(p(H )/p(H O))
p(O
For the potentiometric evaluation (eq. 3) U and the exter-
nal oxygen pressure (p(O2)ext. ϭ 0,209 atm) have been used
2
(Nernst law). The asymmetry potential has been assumed
to be zero. Measuring U and U at different I with deacti-
vated sample furnace showed no detectable difference [20].
1
2
) ϭ 10(4F·U
2
/RTcell)
· p(O
)
ext.
.
(3)
p(O
2
2
Results
In all cases both types of evaluation gave virtually the
same oxygen pressures. In Table 2 and Figs. 3 and 4 data
from coulometric evaluation, which to our experience show
a slightly better accuracy, are given.
In total 24 series of p(O ) measurements for 10 different
equilibrium regions in the Gibbs phase triangles Co / P / O
and Ni / P / O have been carried out (Table 1 to 3). Func-
2
A critical point during coulometric/potentiometric meas-
urements is the constant composition of the equilibrium
solid. Under unfavorable conditions (slight differences be-
tions log p(O ) vs. 1/T are given in Fig. 3 (system Co / P /
O) and Fig. 4 (system Ni / P / O).
2
Control measurement NiI/1 of triangle NiI (Ni / NiO /
s
s
tween p(O ) in the carrier gas and the oxygen coexistence
Ni (PO ) ) at the beginning of our study showed, as ex-
2
3 4 3,s
pressure above the solid in the reactor) a netto loss/uptake
of oxygen by the solid might occur.
pected, no influence of the presence of the phosphate on
the oxygen equilibrium pressure above Ni / NiO . Pressures
s
s
1712
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
zaac.wiley-vch.de
Z. Anorg. Allg. Chem. 2003, 629, 1709Ϫ1717

Contributions on Thermal Behaviour and Crystal Chemistry of Anhydrous Phosphates. XXXIV
2
Fig. 4 log p(O ) vs. 1/T for Ni / P / O.
2
Fig. 3 log p(O ) vs. 1/T for Co / P / O.
same within the accuracy limits of our determination. A
thermodynamic interpretation of this observation will be
discussed (see “Discussion”).
p(O ) for this system observed at 1089 K and 1129 K (Table
2
1
) were in good agreement with values calculated from
tabulated thermodynamic data.
Thermodynamic Evaluation
During the equilibrium measurements CoIV/1Ϫ4
Thermodynamic data
(
Co P / CoP / Co P O ) loss of phosphorus / phosphorus
2 s s 2 2 7,s
oxides was observed (red condensate outside (downstream)
the reactor). However, due to the buffer capacity of the
three-phase solid no change of the equilibrium solids has
been observed. Measurements of system NiVIII (Ni P /
Data for cobalt phosphides [13, 17] and nickel phosphides
[18] have been taken from literature (see also [22]). In case
of the cobalt phosphides the data allow a very good de-
scription of the behaviour of Co P, CoP and CoP in chemi-
2
s
2
3
Ni5P4,s / Ni P O12,s) and higher were not carried out due to
anticipated high phosphorus decomposition pressures
cal transport experiments using iodine [13]. This might be
taken as a hint on the accuracy of the data. Calculated equi-
librium pressures p(P ) and p(P ) above cobalt phosphides
2
4
(Fig. 5).
2
4
In the temperature range 730 to 920 °C oxygen partial
are given in Fig. 5a. Thermodynamic data for the nickel
Ϫ13
Ϫ22
pressures of 10
to 10
atm are observed above the in-
phosphides given in [18] are based on calorimetric (Ni P,
3
vestigated equilibrium solids in the systems Co/P/O and Ni/
P/O. These pressures are two to five orders of magnitude
smaller than oxygen pressures above Co /CoO and Ni /
Ni2.55P) and vaporisation measurements (Ni P to NiP ).
12
5
3
Calculation of phosphorus vapor pressures as a function
of the molar ratio P : Ni leads to slightly higher phosphorus
pressures above Ni /Ni P than above Ni P /Ni2.55Ps. This
s
s
s
NiO , respectively. In general the oxygen equilibrium press-
s
s
3
s
3 s
ure decreases with increasing phosphorus vapour pressure
points to an inconsistency within the data of the two metal
rich nickel phosphides. Apparently, the relative stability of
(
see Fig. 3Ϫ5). However, measured p(O ) for the systems
2
NiII and NiIII and CoII and CoIII, respectively, are the
Ni P and Ni2.55P is not properly reproduced by the data set
3
Z. Anorg. Allg. Chem. 2003, 629, 1709Ϫ1717
zaac.wiley-vch.de
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
1713

M. Blum, K. Teske, R. Glaum
for some time from X-ray crystal structure determination
[23], that this composition most likely belongs to a phase
with composition Ni P . Though we have been able to re-
8
3
produce in our own equilibrium experiments [2] an inter-
mediate phosphide between Ni P and Ni P , the powder
3
12 5
diffraction pattern observed by us for this phase showed
only distant resemblance to the pattern calculated from the
structural data for Ni P [23]. Much better coincidence was
8
3
found with the pattern reported in the powder diffraction
file (JCPDS) for a phase of composition Ni2.55P [24] of yet
unknown crystal structure. For the given reasons we used
the data of Myers and Conti for the nickel phosphides with-
out any changes. Additional remarks on this subject will be
made in section “Discussion”.
Entropies and heat capacities of the phosphates were esti-
mated assuming ∆ S ϭ 0 and ∆ C ϭ 0 (Dulong-Petit,
r
r P
Kopp-Neumann) for the formal formation reactions (eq. 4).
s 4 n m
n MO ϩ m/4 P O10,s ϭ M P O((10m/4) ϩ n),s (M ϭ Co, Ni) (4)
It should be noted, that in case of CoO entropy and
s
heat capacity are significantly higher than expected from
interpolation between the data of CaO and ZnO . This is
s
s
due to additional electronic and atomic disorder in CoOs.
The cobalt phosphates show no such pecularities. We there-
fore assume identical entropy and heat capacity values for
corresponding cobalt(II) and nickel(II) phosphates (Table 4
and 5).
Preliminary heats of formation for cobalt(II) and nickel-
(II) phosphates have been obtained via an estimation of the
enthalpies of reaction for eq. 4. These might be taken as
heats of neutralisation for a solid state acid / base reaction.
A procedure similar to that described in literature [25, 26,
27] has been used, comparing well known enthalpies of re-
action for sodium phosphates (Na PO [28], Na P O [28],
3
4
4
2
7
NaPO3 [28]), and Na SO [28], CoSO [28], and NiSO4
2
4
4
[28]). Very similar data are obtained on the basis of calcium
phosphates, CaSO and the anhydrous transition metal sul-
4
fates.
Evaluation following the 2nd law
From the log(p(O )) vs. 1/T Ϫ plots the heats of reaction
2
∆RHT(av) and entropies of reaction ∆RST(av) were deter-
mined. According to the appropriate decomposition equa-
tions (Table 3) and the thermodynamic data (Table 4) the
heats of formation and entropies of the phosphate were cal-
culated at T(av) as shown in eq. (5) and eq. (6) for the sys-
tem NiVII.
Fig. 5 Vapour pressures log p(P
the binary systems Co / P (a) (T ϭ 800 °C) and Ni / P (b) (T ϭ
50 °C) calculated using the data given in Table 4.
2 4
) and p(P ) vs. composition for
7
1/6 ∆
f
HT(av)(Ni
2
P
4
O
12,s) ϭ 19/6 ∆
f
HT(av)(Ni
2 s
P ) ϩ
∆
f
H
T(av)(O2,g) Ϫ 1/2 ∆
f
H
T(av)(Ni12
P
5,s) ϩ ∆
R
H
T(av) (5)
reported by Myers and Conti [18]. Our own model cal-
culations showed, however, that a small stabilisation (ϳ 1
kcal/mol) in ∆ H° (Ni P ) results in the correct stability
1
/6 ST(av)(Ni
2
P
4
O12,s) ϭ 19/6 ST(av)(Ni
2 s
P ) ϩ
f
298
3 s
sequence of the phosphides. In case of Ni2.55P it is known
ST(av)(O2,g) Ϫ 1/2 ST(av)(Ni12P5,s) ϩ ∆ S
R
T(av) (6)
1714
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
zaac.wiley-vch.de
Z. Anorg. Allg. Chem. 2003, 629, 1709Ϫ1717

Contributions on Thermal Behaviour and Crystal Chemistry of Anhydrous Phosphates. XXXIV
Table 4 Thermodynamic data used in this paper.
compound
∆
f
H°298
S°298
cal/mol·K
A
B
C
Ref.
2
4
kcal/mol
cal/mol·K
cal/mol·K
cal/mol·K
H
H
2,g
0
31.21
45.106
49.005
5.45
53.22
60.6
82.60
96.5
54.68
52.11
66.89
50.24
6.52
7.169
7.16
0.78
2.56
1.00
0.12
0.08
Ϫ0.40
Ϫ
0.0322
Ϫ1.647
Ϫ10.969
Ϫ12.735
Ϫ7.424
Ϫ0.994
Ϫ3.213
Ϫ1.71
[30]
[33]
[30]
[30]
[33]
[33]
[31]
[33]
[32]
[30]
[30]
[30]
2
O
g
Ϫ57.795
0
0
O
P
2,g
a)
red,s
4.051
6.806
10.517
40.699
46.841
35.792
8.675
19.562
9.11
3.559
2.7189
4.4428
22.778
47.5350
77.62
0.1910
0.162
2.86
PO
PO2,g
g
Ϫ2.9
Ϫ71.0
Ϫ440.0
Ϫ677.4
Ϫ719.38
42.68
30.77
1.3
P
P
P
P
P
4
4
4
O
O
O
6,g
10,g
10,s
2,g
4,g
PH3,g
Co
CoO
s
0
7.18
12.65
54.7
45.5
72.9
10.5
16.0
14.3
4.74
4.00
2.04
60.06
58.0
112.04
2.10
5.9
Ϫ
0.4
Ϫ
Ϫ
[30]
[32]
[this work]
[this work]
[this work]
[33,17]
[17]
s
Ϫ57.1
Ϫ613
Ϫ544
Ϫ920
Ϫ28.10
Ϫ39.80
Ϫ51.5
11.54
41.71
30.73
39.1
11.00
16.3
Co
3
Co
2
Co
2
(PO
P
P
4
)
7,s
2,s
2
O
O
4
12,s
Ϫ
CoP
Co
s
P
Ϫ0.36
Ϫ0.54
Ϫ0.27
2
s
CoP3,s
Ni
NiO
21.7
1.4
[13]
s
0
7.141
17.303
54.7
45.5
72.9
26.012
23.431
111.797
19.659
52.424
12.709
16.681
5.995
11.18
41.71
30.73
1.798
2.02
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
[30]
[30]
[this work]
[this work]
[this work]
[18]
[18]
[18]
[18]
[18]
b)
s
Ϫ54.004
Ϫ605
Ϫ533
Ni
3
Ni
2
Ni
2
Ni
3
(PO
4
)
2,s
60.06
58.04
112.04
9.372
7.942
41.70
6.353
14.496
3.790
4.488
P
P
P
2
4
s
O
O
7
12,s
c)
Ϫ915
39.1
Ϫ47,179
Ϫ46.878
Ϫ227.190
Ϫ40.926
Ϫ114.202
Ϫ32.051
Ϫ35.903
20.452
18.069
83.799
14.894
46.666
15.587
20.851
Ni2.55
Ni12
P
s
P
P
P
5,s
Ni
Ni
2
s
5
4,s
NiP2,s
NiP3,s
[18]
[18]
a)
all data for Pred,s as reference state
b)
c)
Thermodynamic data for NiO
s
at 565 K as reference temperature (∆
f
H298 ϭ Ϫ57.5 kcal / mol; S298 ϭ 9.1 cal / · K
see text, section “Thermodynamic Data”
For the use of Ni2.55P instead of Ni
8 3
P
The values obtained were transferred to 298 K according to
Kirchhoff’s principle using the corresponding option in the
computer program CVTRANS [29]. Thermodynamic data
used in this work are summarized in Table 4.
phosphorus activities, namely, triangles CoII, CoIII and
NiII, NiIII. For the adjacent equilibrium regions II and III
very similar oxygen pressures have been observed, probably
due to the oxidation of the same phosphide (Co P and
2
Ni P, respectively). Our measurements show that coulo-
3
metric/potentiometric methods are well suited for the deter-
mination of oxygen coexistence pressures in systems M / P /
O. Apart from equilibrium regions with high vapour press-
ures of phosphorus and/or phosphorus oxides no systematic
problems have been observed during the measurements. A
kinetic hinderance of the oxygen exchange solid / gas has
been found for triangle CoII only.
rd
Evaluation following the 3 law
Partial pressures p(O2)meas. at T(av) for all equilibrium sys-
tems can be taken from the equations given in Table 2.
Using thermodynamic data summarized in Table 4 and the
estimated entropy and heat capacity of the phosphate under
consideration allows the calculation of ∆ H at T(av) and
f
further on ∆ H of the phosphate from these pressures.
The derivation of thermodynamic data from vapour
pressure measurements naturally bears several problems
and uncertainties. This is the more true for systems with
complex composition of the equilibrium solids, and only
limited thermodynamic information available. Thus, for the
present study we had to estimate entropies and heat capacit-
ies for the cobalt and nickel phosphates. Furthermore, due
to the method used, calculated heats of formation for the
phosphates depend on the thermodynamic stabilities of the
coexisting phosphides (see Table 3 and eq. (5)). Every un-
certainty in their thermodynamic data will be reproduced in
the corresponding values for the phosphates. Due to more
f
298
Discussion
Oxygen coexistence pressures determined in the systems
Co / P / O (CoII Ϫ CoIV) and Ni / P / O (NiI Ϫ NiVII),
cfg. Fig. 3 and Fig. 4, provide a quantitative measure of the
redox behaviour of phosphate/phosphide equilibrium
solids. As one would expect p(O ) decreases with increasing
2
phosphorus activity in the solids. The general tendency is
visualized in Figs. 3 and 4. Interestingly, there appears to
be an exception from this experimental trend at very low
Z. Anorg. Allg. Chem. 2003, 629, 1709Ϫ1717
zaac.wiley-vch.de
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
1715

M. Blum, K. Teske, R. Glaum
Table 5 Summary of heats of formation and standard entropies of
nd
cobalt(II) and nickel(II) phosphates. Estimated data, data from 2
law and 3 law evaluation of experimental data a, b)
rd
.
phosphate
symbol
D
f
H°298 (2nd) S°298 (2nd)
f
∆ H°298 (3rd)
kcal / mol
c)
kcal / mol
kcal / mol·K
Co
Co
Co
Co
Co
Co
Co
Co
Co
Co
3
3
3
3
3
2
2
2
2
2
(PO
(PO
(PO
(PO
(PO
4
4
4
4
4
)
)
)
)
)
7
7
7
7
7
2
2
2
2
2
est. [13]
Co II/1
Co III/1
Co III/2
Co III/3
est. [13]
Co IV/1
Co IV/2
Co IV/3
Co IV/4
est. [34]
Ni II/1
Ϫ629.8
Ϫ664.4
54.7
9.5
65.1
65.5
60.1
45.5
52.3
58.3
48.4
62.4
54.65
51.91
Ϫ615.3
a)
a)
b)
Ϫ600.8
Ϫ612.0
a)
a)
b)
Ϫ602.1
Ϫ612.7
a)
a)
b)
Ϫ607.8
Ϫ614.0
P
2
P
2
P
2
P
2
P
2
O
O
O
O
O
Ϫ561.1
Ϫ534.1
Ϫ532.8
Ϫ540.8
Ϫ524.4
Ϫ629
Ϫ541.7
Ϫ546.6
Ϫ544.2
Ϫ542.9
Ni
Ni
3
(PO
(PO
4
)
)
2
3
4
2
Ϫ607.5
Ϫ605.1
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
P
2
P
2
P
2
P
2
P
2
P
2
P
2
P
2
P
2
P
4
P
4
P
4
P
4
P
4
O
O
O
O
O
O
O
O
O
O
O
O
O
O
7
7
7
7
7
7
7
7
7
est. [34]
Ni III/1
Ni III/2
Ni III/3
Ni IV/1
Ni V/1
Ni VI/1
Ni VI/2
Ni VI/3
est. [34]
Ni VII/1
Ni VII/2
Ni VII/3
Ni VII/4
Ϫ550
Ϫ514.9
Ϫ518.2
Ϫ533.1
Ϫ568.1
Ϫ529.6
Ϫ544.8
Ϫ546.8
Ϫ534.7
45.55
67.44
65.58
51.71
29.06
39.50
40.98
39.27
48.44
72.9
57.1
47.7
64.1
57.6
c)
c)
c)
c)
c)
c)
Ϫ539.7d)
Ϫ551.7
Ϫ523.7
e)
e)
f)
Ϫ535.1
e)
e)
f)
Ϫ535.6
e)
e)
f)
Ϫ532.2
12
12
12
12
12
Ϫ957
Ϫ929.9
Ϫ938.0
Ϫ923.4
Ϫ928.6
Ϫ914.6
Ϫ913.2
Ϫ917.0
Ϫ913.5
a)
0
Calculated using ∆
) ϭ 55.4 cal/mol·K from CoIV-meassurements.
Calculated using ∆ 298(av)(Co ) ϭ Ϫ543.8 kcal/mol from Co IV-
third law evaluation and estimated S ( (Co
f 2 2 7
H298(av)(Co P O ) ϭ Ϫ533.0 kcal/mol and S (av)
(
2 2 7
Co P O
b)
f
H
2 7 7
P O
0
2
P
2
O
7
) ϭ 45.5 cal/mol·K and
0
S ((Co
3
(PO
Using ∆
1.91 cal/mol·K from Ni II/1 Ϫ second law evaluation.
4
)
H
2
) ϭ 54.7 cal/mol·K.
c)
0
f
298(Ni (PO
3
4
)
2
) ϭ Ϫ607.5 kcal/mol and S (av)(Ni
3
(PO
4
)
2
) ϭ
5
d)
Using ∆
f 3 4 2
H298(Ni (PO ) ) ϭ Ϫ605.1 kcal/mol from Ni II/1 Ϫ third law
0
evaluation and estimated S (Ni
3
(PO
4
)
2
)
ϭ
54.65 cal/mol·K and
0
S (Ni
2
P
2
O
7
) ϭ 45.55 cal/mol·K.
Calculated using ∆ 298(av)(Ni
Ni 12) ϭ 56.63 cal/mol·K from Ni VII-second law evaluation.
Calculated using ∆ 298(av)(Ni 12) ϭ Ϫ914.5 kcal/mol from Ni VII-
third law evaluation and estimated S (Ni
e)
0
f
H
2
P
4
O
12) ϭ Ϫ930.0 kcal/mol and S (av)
(
2 4
P O
f)
f
H
2 4
P O
0
P
4
O
12) ϭ 72.6 cal/mol·K and
H°298 / mole oxide vs. x(PO2.5). (᭹) 3rd law enthalpies,
∆
f
᭺) estimated enthalpies; dotted lines as a guide for the eye.
2
Fig. 6
(
0
2 2 7
S (Ni P O ) ϭ 45.6 cal/mol·K.
reliable entropies used in their derivation, we recommend
rd
heats of formation obtained via 3 law evalution as being
law evaluation (Table 2 and 5) for the various heterogenous
reactions (Table 3) is not bad. It should also be noted, that
data for the cobalt and the nickel systems fit together nicely,
which appears to be chemically reasonable.
nd
preferable over values from 2 law evaluation. However the
maximum difference between obtained enthalpies from 2
and 3 law evaluation is 14 kcal/mol and generally even
nd
rd
lower. Despite all these difficulties, we believe that thermo-
dynamic data for Co (PO ) , Co P O , Ni (PO ) , Ni P O ,
Heats of formation for cobalt and nickel phosphates de-
rived from the oxygen coexistence pressures are slightly
3
4 2
2
2
7
3
4 2
2
2
7
and Ni P O summarized in Table 4 give an acceptable
lower than estimated values ∆ H°(av). However, the devi-
2
4
12
B
match of observed phase relations and equilibrium vapour
pressures.
ations are not too large, given the uncertainties in the esti-
mations as well as in the experimentally derived values.
∆ H°(Co (PO ) ) has been determined from two phase
First of all, the internal consistency of our measurements
is fairly good (Table 2). Repeated measurement of the same
triangle (e. g. CoIII/1 Ϫ 3) leads to only small variations in
B
3
4 2
triangles (CoII and CoIII), depending on ∆ H°(Co P) and
B
2
(∆ H°(Co P O ), ∆ H°(Co P)), respectively. In both cases
B
2
2
7
B
2
nd
∆RHT(av) and ∆RST(av) (from 2 law). Apart from just a
very similar enthalpies were obtained. ∆ H°(Co P O ) de-
B
2
2
7
few measurements (CoII/1, NiIII/1, NiIII/2) entropies
pends in our evaluation just on ∆ H°(CoP), triangle CoIV.
B
nd
∆RST(av) from 2 law evaluation and estimated reaction
In addition, thermodynamic data given in Table 4 for the
system Co / P / O allow the calculation of all observed equi-
librium regions.
entropies fit together acceptably. In the same way, agree-
ment between heats of reaction derived from 2nd and 3
rd
1716
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
zaac.wiley-vch.de
Z. Anorg. Allg. Chem. 2003, 629, 1709Ϫ1717

Contributions on Thermal Behaviour and Crystal Chemistry of Anhydrous Phosphates. XXXIV
[2] M. Blum, part of planned Ph.D. thesis, Rheinische Friedrich-
∆BH°(Ni (PO ) ) has been determined from phase tri-
3
4 2
Wilhlems-University, Bonn.
3] R. Glaum, R. Gruehn, Z. Kristallogr. 1989, 186, 91.
4] R. Glaum, Z. Kristallogr. 1992, 205, 69.
5] F. Reinauer, R. Glaum, Acta Crystallogr. 1998, B54, 722. F.
Reinauer, R. Glaum, R. Gruehn, Eur. J. Solid Inorg. Chem.
angle NiII, depending on ∆ H°(Ni P). Independently,
B
3
[
[
[
∆BH°(Ni P O ) has been determined from p(O ) in tri-
2
4
12
2
angle NiVII, depending on ∆ H°(Ni P) and ∆ H°(Ni P ).
B
2
B
12 5
Based on the heats of formation of nickel phosphides and
∆BH°(Ni (PO ) ) and ∆ H°(Ni P O ), ∆ H°(Ni P O ) can
3
4 2
B
2
4
12
B
2
2
7
1
994, 31, 779.
6] R. Glaum, M. Weil, D. Özalp, Z. Anorg. Allg. Chem. 1996,
22, 1839.
be derived from p(O ) in triangles NiIII, NiIV, NiV, and
2
[
NiVI. Heats of formation thus obtained for Ni P O are
2
2
7
6
in sufficiently good agreement, especially when considering
[
[
7] R. Glaum, R. Gruehn, Z. Anorg. Allg. Chem. 1990, 580, 78.
8] T. Droß, E. Benser, U. Kaiser, R. Glaum, Z. Anorg. Allg.
Chem. 2003, in preparation.
9] M. Lenz, Ph.D. thesis, Justus-Liebig-University, Gießen, 1995.
10] H. Mathis, R. Glaum, R. Gruehn, Acta Chem. Scand. 1991,
rd
values from 3 law evaluations (Table 5). This gives also
some evidence for the relative accuracy of the thermo-
dynamic data given for the nickel phosphides in literature
18]. In contrast to the system Co / P / O, for Ni / P / O
[
[
[
modelling of the the Gibbs phase triangle does not repro-
duce the observed equilibrium relations in all cases. The
45, 781.
[11] M. Gerk, Ph.D. thesis, Justus-Liebig-University, Gießen, 1996.
[12] C. Gleitzer, Eur. J. Solid Inorg. Chem. 1991, 28, 77.
[13] A. Schmidt, Ph.D. thesis, Justus-Liebig-University, Gießen,
2002. The thesis is available online via: http://bibl7.hrz.uni-
giessen.de/ghtm/2002/uni/d020105.htm
problem with the relative stability of Ni P and Ni2.55P has
3
already been mentioned. Furthermore, slight changes in
∆BH°(Ni P O ) have a large impact on the calculated equil-
2
2
7
ibria between Ni P O and Ni P O on one side and the
2
2
7
2
4
12
[
14] M. Blum, R. Glaum, Z. Anorg. Allg. Chem. 2003, in pre-
nickel phosphides Ni2.55P to Ni P on the other. Due to the
5
4
paration.
complexity of the phase diagram with very narrow equilib-
rium regions and due to the uncertainty of available
thermodynamic data we did not try further to optimize
data available for the system Ni / P / O.
Fig. 6 gives the heats of formation per mole oxide for
cobalt and nickel phosphates, showing the internal consist-
ency of the values.
[
[
15] D. Özalp, Ph.D. thesis, Univ. Gießen, 1993.
16] R. Glaum, Thesis of Habilitation, Justus-Liebig-University,
Gießen, 1999. The thesis is available online via: http://
bibd.uni-giessen.de/ghtm/1999/uni/h990001.htm
17] C. E. Myers, High Temperature Sci. 1974, 6, 309.
18] C. E. Myers, T. J. Conti, J. Electrochem. Soc. 1985, Vol. 32 No.
[
[
2, 455.
[
[
19] V. Propach, D. Reinen, Z. Naturforsch. 1978, 33b, 619.
20] M. Bode, K. Teske, H. Ullmann, GIT-Fachz. Lab. 1994, 38,
CoO
s
s
ϩ 1/4 P
4
O
10,s ϭ 1/2 Co
2 2 7,s
P O
∆
r
H
298 ϭ Ϫ45.6 cal / mol·K (7)
4
95.
[21] K. Maaß, R. Glaum, R. Gruehn, Z. Anorg. Allg. Chem. 2002,
28, 1663.
[
[
NiO
s
ϩ 1/4 P
4
O10,s ϭ 1/2 Ni P O
2 2 7,s
∆
r
H
298 ϭ Ϫ39.7 cal / mol·K (8)
6
CoO
ϩ SO3,s ϭ CoSO4,s
ϩ SO3,s ϭ NiSO4,s
∆
r
H
298 ϭ Ϫ45.9 cal / mol·K (9)
22] M. E. Schlesinger, Chem. Rev. 2002, 102, 4267.
23] O. N. Il’nitskaya, L. G. Aksel’rud, S. I. Mikhalenko, Yu. B.
Kuz’ma, Sov. Phys. Crystallogr. 1987, 32, 26.
NiO
s
∆
r
H298 ϭ Ϫ42.6 cal / mol·K (10)
Enthalpies of neutralisation calculated for the formation of
diphosphates M P O (M ϭ Co, Ni) (eq. (7) and eq. (8))
compare to values obtained for the corresponding anhy-
drous sulfates MSO (eq. (9) and eq. (10)). Enthalpies of
neutralisation ∆ H (LaPO ) ϭ Ϫ86 kcal / mol and
∆rH298(1/2 Ca P O ) ϭ Ϫ77.7 kcal / mol reflect the higher
basicity of La O and CaO in contrast to CoO and NiO.
For the formation of TiPO4 and VPO4 from the only
weakly basic oxides Ti O and V O even lower enthalpies
[24] E. Larson, Arkiv Kemi 1964, 23, 335.
[
[
2
2
7
25] R. Hoppe, Z. Anorg. Allg. Chem. 1958, 296, 104.
26] O. Kubaschewski, C. B. Alcock, P. J. Spencer, “Materials Ther-
4
th
mochemistry”, 6 ed., Pergamon Press, 1993.
r
298
4
[27] H. Schäfer, V. P. Orlovskii, Z. Anorg. Allg. Chem. 1972, 390,
2
2
7
13.
2
3
[
28] O. Knacke, O. Kubaschewski, K. Hesselmann, „Thermochem-
ical Properties of Inorganic Substances“, Springer-Verlag, Hei-
delberg, 1991.
29] O. Trappe, R. Glaum, Computer program CVTRANS for the
calculation of homogenous and heterogenous equilibria solid /
gas, Justus-Liebig-University, Gießen, 1998.
2
3
2
3
of neutralisation ∆ H (MPO ) ഠ Ϫ34 kcal / mol are
found.
r
298
4
[
Acknowledgement. We thank Prof. Dr. H. Ullmann (Inst. f. Anor-
ganischen Chemie, TU Dresden) for allowing us coulometric/
potentiometric measurements in his group. We also thank Mrs.
Carmen Hofmann (Inst. f. Anorganischen Chemie, Uni Gießen) for
her support with the preparative work. This work has been funded
by Fonds der chemischen Industrie and Deutsche Forschungsge-
meinschaft within the program “Reactivity of Solids” (SPP 1010).
[
[
[
[
30] I. Barin, O. Knacke, “Thermochemical Properties of Inorganic
Substances”, Springer Verlag, Berlin, 1973.
31] R. Glaum, M. Walter-Peter, D. Özalp, R. Gruehn, Z. Anorg.
Allg. Chemie 1991, 601, 145.
32] O. Kubaschewski, C. B. Alcock, “Metallurgical Thermochemi-
th
stry”, 5 Edition, Pergamon Press 1979.
33] B. I. Noläng, “FREDACS Ϫ Free Energy Data for chemical
Substances” 1978, Chem. Inst. Univ. Uppsala, Sweden.
References
[
1] Part XXXIII of this series: H. Thauern, R. Glaum, Z. Anorg.
[34] M. Blum, Diploma thesis, Justus-Liebig-University Gießen,
Germany, 1997.
Allg. Chem. 2003, 629, 479.
Z. Anorg. Allg. Chem. 2003, 629, 1709Ϫ1717
zaac.wiley-vch.de
 2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
1717