Preparation and crystal structures of Ni(NH3)2Cl2 and of two modifications of Ni(NH3)2Br2 and Ni(NH3)2I2

Article abstract of DOI:10.1006/jssc.2000.8666

Diammine nickel(II) halides, Ni(NH₃)₂X₂ (X = Cl, Br, I), were prepared by decomposition of the corresponding hexaammines at 120°C in dynamical vacuum. Their crystal structures are of the Cd(NH₃)₂Cl₂ type ('β-type', space group Cmmm, Z = 2): for Ni(NH₃)₂Cl₂, a = 8.019 A, b = 8.013 A, c = 3.661 A; for β-Ni(NH₃)₂Br₂, a = 8.273 A, b = 8.297 A, c = 3.851 A. Both were obtained by Rietveld refinement of diffractometer data with standard deviations < 0.001 A. For β-Ni(NH₃)₂I₂, a = b = 8.753(3) A, c = 4.127(1) A, obtained by Guinier film data. In the case of the bromide and the iodide, annealing at 300°C leads to an irreversible structural rearrangement. A new modification is formed which is of the Mg(NH₃)₂Br₂ type ('α-type', space group Pbam, Z = 2): for α-Ni(NH₃)₂Br₂, a = 5.865 A, b = 11.723 A, c = 3.856 A, obtained by Rietveld refinement of diffractometer data with standard deviations < 0.001; for α-Ni(NH₃)₂I₂, 2a = b = 12.359(3) A, c = 4.126(1) A, obtained by Guinier film data. Both types of structures contain infinite chains of edge-sharing octahedra (∞)/¹[NiX(4/2)(NH₃)₂]. These run parallel to the c-axes of the unit cells of the corresponding structure types and are arranged in different ways relative to one another. The decomposition of hexaammine nickel(II) halides to the corresponding diammines is analyzed in terms of structural analogies. (C) 2000 Academic Press.

Full text of DOI:10.1006/jssc.2000.8666

KANNAN BALA VVC
DISC USED NO !
JSSC 8666
DATE. 2...9..}..3...}..2..0..0...0..........
Journal of Solid State Chemistry 152, 381}387 (2000)
doi:10.1006/jssc.2000.8666, available online at http://www.idealibrary.com on
Preparation and Crystal Structures of Ni(NH3)2Cl2 and
of Two Modifications of Ni(NH3)2Br2 and Ni(NH3)2I2
A. Leineweberꢀ and H. Jacobs
Lehrstuhl Anorganische Chemie I, Fachbereich Chemie der Universita( t, 44221 Dortmund, Germany
Received December 15, 1999; accepted February 24, 2000
formations (e.g., in hexaammine metal(II) halides (4)). With
Diammine nickel(II) halides, Ni(NH3)2X2 (X ؍ Cl, Br, I),
were prepared by decomposition of the corresponding hexaam-
mines at 1203C in dynamical vacuum. Their crystal structures
are of the Cd(NH3)2Cl2 type (99b-type::, space group Cmmm,
Z ؍ 2): for Ni(NH3)2Cl2, a ؍ 8.019 A, b ؍ 8.013 A, c ؍ 3.661 A;
for b-Ni(NH3)2Br2, a ؍ 8.273 A, b ؍ 8.297 A, c ؍ 3.851 A. Both
were obtained by Rietveld re5nement of di4ractometer data with
standard deviations < 0.001 A. For b-Ni(NH3)2I2, a ؍ b ؍
8.753(3) A, c ؍ 4.127(1) A, obtained by Guinier 5lm data. In the
case of the bromide and the iodide, annealing at 3003C leads to
an irreversible structural rearrangement. A new modi5cation is
formed which is of the Mg(NH3)2Br2 type (99a-type::, space
group Pbam, Z ؍ 2): for a-Ni(NH3)2Br2, a ؍ 5.865 A, b ؍
11.723 A, c ؍ 3.856 A, obtained by Rietveld re5nement of dif-
fractometer data with standard deviations < 0.001; for
a-Ni(NH3)2I2, 2a ؍ b ؍ 12.359(3) A, c ؍ 4.126(1) A, obtained
by Guinier 5lm data. Both types of structures contain in5nite
respect to the last point, the role of hydrogen bonds in
ammines was examined (e.g., (5, 6)). Additionally, ammines
may be used as precursors for crystal growth of nitrides
in supercritical ammonia (e.g., Cu N (7), Cu PdN (8), or
ꢂ
ꢂ
g-Mn N (9)).
ꢂ ꢃ
At the beginning of the 20th century thermodynamic
studies were carried out on several systems M(NH ) X in
ꢂ L K
order to determine the di!erent values of n and the corre-
sponding NH pressures p(n, ¹). Many results of these
ꢂ
works are reported in (10).
For the ammines of M"Mg, Mn, Fe, Co, Ni (with
m"2) and X"Cl, Br, I strong analogies were found. The
di!erent systems show similarities in solvation numbers of
n46 with n"1, 2, 6 for X"Cl, Br and n"2, 6 for X"I
(10). Ammines with n'6 are only stable below ambient
temperatures (10) and are not of interest here.
1
chains of edge-sharing octahedra [NiX4/2(NH3)2]. These run
ꢁ
The structures of the hexaammines M(NH ) X were
parallel to the c-axes of the unit cells of the corresponding
ꢂ ꢄ ꢃ
thoroughly examined by X-ray (11}13) and neutron di!rac-
structure types and are arranged in di4erent ways relative to one
another. The decomposition of hexaammine nickel(II) halides to
the corresponding diammines is analyzed in terms of structural
analogies. ( 2000 Academic Press
ꢃ>
tion methods (14). They are built up by [M(NH ) ] and
ꢂ ꢄ
ꢃ
\
X
ions arranged with the motif of the cubic CaF type (at
ambient temperatures Fm3m with disordered H atoms
Key Words: diammines of metal(II) halides; ammoniates of (12}14)). The hexaammines can in many cases be crystal-
nickel halides; polymorphism of metal diammine halides; dy-
namical disorder of ammonia ligands; Rietveld re5nement;
structural analogue:s between ammoniates.
lized from aqueous ammonia solutions. This is not possible
for the ammines with a lower n. This causes di$culties for
the preparation of well-crystallized materials.
We had started structural investigations of ammines of
lower NH content to examine the diammine halides of
INTRODUCTION
ꢂ
magnesium, Mg(NH ) X (X"Cl, Br, I) (15). In this study
ꢂ ꢃ ꢃ
Ammines of simple salts M(NH ) X are of interest for we have found the Cd(NH ) Cl structure type (Cmmm) (16)
ꢂ L
K
ꢂ ꢃ ꢃ
several reasons. They can serve as model systems for a step- for the chloride and established the then unprecedented
wise solvation of salts. For example, this was systematically Mg(NH ) Br structure type (Pbam) for the bromide
ꢂ ꢃ
ꢃ
examined on the crystal structures of a series of ammines and iodide. Both structure types contain parallel chains
ꢀ
Al(NH ) Cl with n"1, 2, 3, 5, 6 (1}3). Furthermore, there
[MX (NH ) ] of edge-sharing octahedra, which are ar-
ꢂ L
ꢂ
ꢁ
ꢅꢆꢃ
ꢂ ꢃ
is fundamental interest in the rotational dynamics of the ranged in di!erent ways relative to one another (Fig. 1). For
NH molecules "xed by coordinative bonds to metal the bromide and the iodide single crystals were grown by
centers. Often this is accompanied by order}disorder trans- reaction of magnesium with the corresponding ammonium
ꢂ
halide.
Now, we have extended our studies to the systems
ꢀPresent address: MPI fur metallforschung, seestra{e 92, D-70174
Stuttgart, Germany.
NiX /NH (X"Cl, Br, I). These ammines have already
ꢃ
ꢂ
381
0022-4596/00 $35.00
Copyright ( 2000 by Academic Press
All rights of reproduction in any form reserved.

JSSC=8666=TSK=VVC=BG
382
LEINEWEBER AND JACOBS
FIG. 1. Schematic representation of the Cd(NH ) Cl and the Mg(NH ) Br type structures in a (001) projection. Chains ꢀ [MX (NH ) ] run
ꢂ ꢃ
ꢃ
ꢂ ꢃ
ꢃ
ꢁ
ꢅꢆꢃ
ꢂ ꢃ
parallel to the c-axis. The subdivision into square units shows the relation to a CsCl type arrangement of X and N. &&Double CsCl cells'' are indicated as
units, in which one ꢀ [MX (NH ) ] unit is located.
ꢁ
ꢅꢆꢃ
ꢂ ꢃ
been reported in (17, 18) including determinations of the
UV spectroscopic investigations of diammine nickel(II)
NH absorption isotherms.
chloride and bromide con"rm octahedral coordination of
ꢂ
X-ray powder di!raction data on the diammine chloride Ni (e.g., (25)). IR spectroscopy was performed on the
are given in (19). A tetragonal primitive unit cell (a"5.60 A, chloride (5) and hydrogen bonding was analyzed.
c"3.70 A) is reported and &&similarity'' of the crystal struc-
We will now report optimized methods of preparation
ture of Ni(NH ) Cl to the Cd(NH ) Cl type (16) is sug- and elucidation of the structures of one modi"cation of
ꢂ ꢃ
ꢃ
ꢂ ꢃ ꢃ
gested. Further support for this assumption comes from the diammine nickel(II) chloride and of two modi"cations each
fact that Ni(H O) Cl (20, 21) forms a slightly distorted of diammine nickel(II) bromide and iodide.
ꢃ
ꢃ ꢃ
(due to hydrogen bonding) variant of the Cd(NH ) Cl
ꢂ ꢃ
ꢃ
structure type. This was already pointed out in (22). How-
ever, no rigorous structural analysis of the crystal structure
of Ni(NH ) Cl has been reported until now.
PREPARATION
All diammines are sensitive to moist air and were handled
under inert conditions (26). Furthermore, the hexaammines
ꢂ ꢃ
ꢃ
In (23) the structures of Ni(NH ) Cl and Ni(NH ) Br
ꢂ ꢃ ꢃ ꢂ ꢃ
ꢃ
were reported to be probably isotypic and powder patterns show slow exchange of NH by H O indicated by a change
ꢂ
ꢃ
were indexed with a tetragonal primitive unit cell (a" of color from deep blue to green. This is prevented by
5.66 A, c"3.66 A for the chloride; a"5.86 A, c"3.85 A keeping the materials in closed containers. The sensitivity
for the bromide).
to moist air is most pronounced for the chloride. Its equilib-
ꢀ
For Ni(NH ) I again chains of [NiI (NH ) ] were rium NH pressure is the highest (17, 18).
ꢂ ꢃ ꢃ ꢅꢆꢃ ꢂ ꢃ
suggested to constitute the structure (24). This was deduced
ꢁ
ꢂ
The method described for the growth of single crystals of
from a small magnetic transition entropy (antiferro- diammine magnesium halides in (15) was not successful for
magnetic}paramagnetic) which supports a low dimen- the nickel compounds; nickel powder does not react with
sionality of magnetic interactions. Powder di!raction pat- NH X (X"Cl, Br, I) at 3503C in autoclaves.
ꢅ
terns were indexed with an orthorhombic unit cell
Therefore, "rst Ni(NH ) X (X"Cl, Br) was prepared by
ꢂ ꢄ ꢃ
(a"12.38 A, b"13.17 A, and c"8.28 A). However, the introducing NH gas (99.999%, Messer-Griesheim) into
ꢂ
proposed structural elements were not related to the cell a saturated aqueous solution of NiCl ) 6H O (purity '97%,
ꢃ
ꢃ
parameters.
Merck) or NiBr (99%, Aldrich) along with some NH Cl
ꢃ
ꢅ

JSSC=8666=TSK=VVC=BG
PREPARATION OF DIAMMINE NICKEL(II) HALIDES
383
('99.8%, Merck) or NH Br ('99%, Aldrich). The hexa- 3003C showed the same re#ections. However, the
ꢅ
ammines crystallize from these solutions on slow cooling.
thermal treatment leads to a signi"cant narrowing of the
Ni(NH ) I was prepared from a 1 : 2 (molar ratio) aque- re#ections.
ꢂ ꢄ ꢃ
ous solution of NiSO )6 H O (99%, Aldrich) and NH I
A powder di!raction pattern was taken from the an-
('99.8%, Merck) by introduction of NH gas. The crystal- nealed material. It is similar to the tetragonally indexed
ꢅ
ꢃ
ꢅ
ꢂ
line product precipitates readily from the solution.
pattern reported in (19). A C-centered orthorhombic
The hexaammine materials were gathered on a glass "lter Cd(NH ) Cl structure type (16) was used as a starting
ꢂ ꢃ
ꢃ
crucible. It was "rst washed with an aqueous solution of model for the Rietveld re"nement (27). For this model the
NH /NH Cl (each 0.5 molar) and then with EtOH unit cell dimensions are roughly 2ꢀꢆꢃ;2ꢀꢆꢃ;1 larger than
ꢂ
ꢅ
saturated with gaseous ammonia.
the cell proposed in (19). The increased cell size does not
Second, hexaammine nickel(II) chloride was placed into require new re#ections as a result of the C-centering but
a glass tube attached to dynamic vacuum (p(0.1 Torr). allows a splitting of the positions of certain re#ections. This
The tube was slowly heated by an oil bath, "nally reaching was in fact observed in the case of Mg(NH ) Cl for well-
ꢂ ꢃ
ꢃ
1203C. The material changed its color from the initial blue resolved Guinier patterns (15), but not for Ni(NH ) Cl .
ꢂ ꢃ
The Rietveld re"nement readily converged. A preferred ori-
ꢃ
to light green.
If the temperature is increased too fast, the evolution of entation according to the [001] direction as an ellipsoid
NH gas can occur vigorously. A glass fritted disc prevented (March} Dollase model (28, 29)) had to be considered. This
ꢂ
ꢀ
the contamination of the vacuum line with the "ne powder direction corresponds to that of the [NiCl (NH ) ]
ꢁ
ꢅꢆꢃ
ꢂ ꢃ
and helped to keep the decomposition reaction under con- chains.
trol. The "nal product was a light green very "ne powder. It
The scattering power of hydrogen is signi"cant in this
was annealed in a steel autoclave for 2 days at 3003C in compound. A re"nement without taking H into account
order to improve crystallinity. The autoclave (<"7 ml) leads to negative thermal parameters of N and to worse
was "lled and closed under an argon atmosphere at ambient residuals. Indeed, the contribution of H is due to integral
pressures. At elevated temperatures a small amount of scattering of the whole NH molecule whereas a superstruc-
ꢂ
the ammine decomposes to build up its equilibrium NH
ture due to a speci"c orientational ordering of NH would
ꢂ
ꢂ
pressure (about 600 Torr at 3003C (17)) to prevent further cause only very weak new re#ections. Most probably, the
decomposition.
The procedure for the preparation of Ni(NH ) X
NH molecules are rotationally disordered at ambient temper-
atures, as is known for the hexaammines (12}14). This is in
ꢂ
ꢂ ꢃ ꢃ
(X"Br, I) was the same as that for the diammine chloride: accordance with the evaluation of "rst neutron di!raction data
The decomposition reaction with ¹ "1203C gave b- on Co(ND ) Cl (b-type modi"cation) and Mg(ND ) Br
ꢀ!ꢁ
ꢂ ꢃ
ꢃ
ꢂ ꢃ ꢃ
Ni(NH ) X . Annealing of these products at 3003C as de- (30). In most of the above-mentioned cases the hydrogen
ꢂ ꢃ ꢃ
scribed for the diammine chloride led, for both the bromide density distribution can approximately be described by four
and the iodide, to an irreversible transformation into a- maxima pointing to the four halogen atoms next to the NH
ꢂ
Ni(NH ) X as revealed by powder di!raction patterns. group. We introduced the H positions for all our re"ne-
ꢂ ꢃ ꢃ
The bromide is yellowish green, the iodide orange brown. ments. They were modeled by soft constraints to form a
The colors appeared brighter after annealing.
square with edges of 1.1 A and a distance N}H of 0.85 A,
Annealing experiments (t"24 h) on b-Ni(NH ) Br as each site with an occupancy of 3/4 and a thermal displace-
ꢂ ꢃ
ꢃ
a function of temperature (1753C(¹(3003C, intervals of ment parameter ;(H)"2;(N).
253C) were performed in glass ampoules under an argon
Details of data collection and Rietveld re"nement for the
atmosphere.
data of Ni(NH ) Cl are given in Table 1, the "nal struc-
tural parameters in Table 2, and selected atomic distances
ꢂ ꢃ
ꢃ
X-RAY INVESTIGATIONS AND RIETVELD REFINEMENT and angles in Table 3. The di!raction data, the "tted pro"le,
and the di!erence curve are displayed in Fig. 2.
Guinier patterns were taken from all products on an
Figure 3 shows a zoom of the range 103(2h(453 of the
FR552 camera (CuKa radiation, Enraf-Nonius, Delft, The di!raction data for b- and a-Ni(NH ) Br , i.e., before and
ꢀ
ꢂ ꢃ
ꢃ
Netherlands). Si was used as an internal standard.
after annealing. The thermal treatment leads to a consider-
X-ray powder di!ractograms were taken from #at sam- able narrowing of the re#ection pro"les and to additional
ples (thickness +1 mm) covered by Kapton foil (d+ re#ections. Slow cooling after the annealing shows that the
50 lm, type 200 HN, August Krempel Sohne GmbH & Co., transition is monotropic.
Enzweihingen, Germany) on a Siemens D500 powder dif-
In analogy to Ni(NH ) Cl the powder di!ractograms of
ꢂ ꢃ
ꢃ
fractometer (CuKa radiation, Bragg-Brentano geometry, the b-modi"cation of Ni(NH ) Br can be re"ned on the
ꢂ ꢃ
basis of a Cd(NH ) Cl type model whereas the patterns of
ꢂ ꢃ
Samples of Ni(NH ) Cl after thermal decomposition of the a-modi"cation "t to an Mg(NH ) Br type structure. H-
ꢂ ꢃ ꢂ ꢃ
the hexaammine at 1203C as well as after annealing at atoms were introduced "xed to N as mentioned above. For
ꢃ
graphite as secondary monochromator).
ꢃ
ꢃ
ꢃ

JSSC=8666=TSK=VVC=BG
384
LEINEWEBER AND JACOBS
TABLE 1
Details of X-Ray Powder Di4ractometry and Subsequent Rietveld Re5nement for Diammine Nickel(II) Halides
Formula
Formula weight
Di!ractometer
Ni(NH ) Cl
ꢂ ꢃ
b-Ni(NH ) Br
252.6
Siemens D500, Bragg Brentano
a-Ni(NH ) Br
ꢃ
ꢃ
ꢂ ꢃ
ꢃ
ꢂ ꢃ
163.7
252.6
Radiation
CuKa (Graphite secondary monochromator)
2h range
2h step width
Time per step
10}1203
0.013
5 s
Space group
Cmmm (No. 65)
Pbam (No. 55)
Cell parameters@ a in A
8.019
8.013
3.661
235.2, 2
2.31
8.273
8.297
3.851
264.3, 2
3.17
4
8#8 "xed background
5
!0.057(2)
0.874(2)
306
5.865
11.723
3.856
265.1, 2
3.16
b in A
c in A
Volume < in Aꢂ, Z
Calculated density in g/cmꢂ
Pro"le parameters
Background parameters
Structural parameters
Zero point in 3
4
4
8#7 "xed background
5
!0.0247(6)
0.940(2)
248
6#8 "xed background
7
0.0313(6)
*
472
Preferred orientation parameter r according to (28, 29)
No. of re#ections
wR
8.0
8.9
7.1
9.7
8.5
8.9
.
R(Fꢃ) (Fꢃ5p(Fꢃ))
? Simpson's rule integration of Pseudovoigt function with correction for asymmetry (34, 35).
@ Standard deviations resulting from Rietveld re"nements are smaller than 0.001 A.
the b-modi"cation we again had to consider preferred ori-
The di!raction patterns of the iodides are similar to those
entation as described above for Ni(NH ) Cl again using the of the bromides. We did not perform extended powder
ꢂ ꢃ
direction [001] as ellipsoid axis. For a-Ni(NH ) Br an di!ractometer studies. We only determined the unit cell
ꢂ ꢃ
analogous attempt does not improve the residuals signi"- parameters from Guinier patterns. For both modi"cations
ꢃ
ꢃ
cantly. pseudo-tetragonal axes were used for a re"nement of lattice
Details of the Rietveld re"nements are given in Table 1. parameters resulting in a "12.359(3) A and c"4.126(1) A for
ꢄ
The "nal structural parameters are given in Tables 2 and 4. a-Ni(NH ) I (a "a /2, b "a ) as well as a " 8.735(3) A
ꢂ ꢃ ꢃ
ꢃ
ꢄ
ꢃ
ꢄ
ꢄ
In Figs. 4 and 5 we present the "nal pro"le "ts of both and c"4.127(1) A for b-Ni(NH ) I (a "b "a ).
ꢂ ꢃ ꢃ
Samples of b-Ni(NH ) Br annealed for 24 h at certain
ꢂ ꢃ
ꢃ
ꢃ
ꢄ
modi"cations of Ni(NH ) Br . Selected atomic distances
ꢂ ꢃ
ꢃ
ꢃ
are given in Table 3.
temperatures were analyzed by Guinier "lm methods. For
temperatures of ¹42003C no changes of the b-phase pat-
terns were observed. At 2253C a pronounced narrowing of
the re#ections of the b-phase occurred, while a new broad
re#ection at 2h"16.53 appeared. This has to be assigned to
the a-phase which has been formed partially. With increas-
ing temperatures, this re#ection becomes more intense.
At 3003C the transformation seems to be complete: the
re#ection at 2h"16.53 is now as sharp as the others. The
patterns do not change by prolonged annealing which indi-
cates a complete transformation of the b- into the a-phase.
TABLE 2
Re5ned Positional and Thermal Parameters of Ni(NH3)2Cl2 and
b-Ni(NH3)2Br2 (Cmmm)
Wycko!
100
Atom
site
x
y
z
f
; /Aꢃ
-!!
*ꢂꢃ
Ni(NH ) Cl
ꢂ ꢃ
ꢃ
Ni
Cl
N
H?
2a
4h
4i
0
0
0
1/2
0
1
1
1
1.13(7)
0.7(1)
0.91(6)
2 ;(N)
0.2148(2)
0
0.069
0
0.2608(4)
0.303
16r
0.150
3/4
TABLE 3
Selected Distances and Angles
b-Ni(NH ) Br
ꢂ ꢃ
ꢃ
Ni
Br
N
H?
2a
4h
4i
0
0
0
1/2
0
1
1
1
3.7(1)
1.81(6)
4.2(2)
Distances/Angles
Ni(NH ) Cl
b-Ni(NH ) Br
a-Ni(NH ) Br
ꢃ
ꢂ ꢃ
ꢃ
ꢂ ꢃ
ꢃ
ꢂ ꢃ
0.2271(2)
0
0.067
0
0.2520(8)
0.294
4;Ni}X/A
2;Ni}N/A
Bridging X}Ni}X/3
2.513(1)
2.089(3)
86.51(1)
2.690(1)
2.091(6)
88.59(4)
2.656(1)
2.044(5)
86.89(2)
16r
0.143
3/4
2 ;(N)
? Constraints: see text.

JSSC=8666=TSK=VVC=BG
PREPARATION OF DIAMMINE NICKEL(II) HALIDES
385
TABLE 4
Re5ned Positional and Thermal Parameters of a-Ni(NH3)2Br2
(Pbam)
Wycko!
100
Atom
site
x
y
z
f
; /Aꢃ
-!! *ꢂꢃ
Ni
Br
N
2a
4h
4g
8i
0
0
0
1/2
0
0.143
0.143
1
1
1
1.89(6)
1.25(3)
2.9(1)
0.2810(1)
0.2479(9)
0.355
0.38925(7)
0.1226(5)
0.109
H(1)?
3/4 2 ;(N)
3/4 2 ;(N)
H(2)?
8i
0.225
0.177
? Constraints: see text.
However, it is di$cult to prove the absence of the
b-Ni(NH ) Br from considering only the positions of re#ec-
ꢂ ꢃ
ꢃ
FIG. 3. Comparison of the 2h"10}453 range of the powder di!rac-
tion patterns of a- (- - -) and b-Ni(NH ) Br (*).
tions. All positions of re#ections of the a-phase include also
those of the b-phase. However, a-Ni(NH ) Br produced
ꢂ ꢃ
ꢃ
ꢂ ꢃ
ꢃ
from the b-material by annealing for 2 days at 3003C shows
no discrepancies in the Rietveld re"nement (see above). This
may indicate the absence of residual b-Ni(NH ) Br .
A comparison of the di!raction patterns of the two modi-
"cations of Ni(NH ) Br (Fig. 3) indicates that the a-phase
ꢂ ꢃ
ꢃ
ꢂ ꢃ
ꢃ
DISCUSSION
may be a superstructure of the b-modi"cation. Figure 6
shows a &&tree'' of space groups according to (31, 32). The
The crystal structures of the diammine nickel(II) halides space groups of the Cd(NH ) Cl and Mg(NH ) Br types
ꢂ ꢃ
ꢃ
ꢂ ꢃ
ꢃ
annealed at 3003C are isotypic to those reported previously of structures are derived from a CsCl type. Figure 6
for the corresponding diammine magnesium halides (15): displays the successive group}subgroup relations. The two
the chloride crystallizes in the Cd(NH ) Cl type (16) and structures are located in di!erent branches of the tree. The
ꢂ ꢃ
ꢃ
the bromide as well as the iodide in the Mg(NH ) Br type a-structure is not a superstructure of b-form. The similarity
ꢂ ꢃ
(15). This tendency seems to hold for the corresponding of the di!raction patterns is mainly caused by the unit cell
ꢃ
compounds of Mn, Fe, and Co, too (30).
dimensions. These are in both cases close to tetragonal.
However, the diammine nickel(II) bromide and iodide
The transition of b-Ni(NH ) Br into a-Ni(NH ) Br
ꢂ ꢃ
ꢃ
ꢂ ꢃ
ꢃ
form a metastable b-modi"cation of the Cd(NH ) Cl type seems to be sluggish or of very low energy. We failed to
ꢂ ꢃ
after decomposition of the corresponding hexaammine monitor the phase transformation by DSC methods
ꢃ
nickel(II) halides at 1203C.
(Pyris 1, Perkin Elmer) with samples in sealed Al crucibles.
FIG. 2. Powder di!raction pattern and results of Rietveld re"nement
FIG. 4. Powder di!raction pattern and results of Rietveld re"nement
of Ni(NH ) Cl : #, observed pro"les points; *, calculated pro"le; bot- of b-Ni(NH ) Br : #, Observed pro"les points; *, calculated pro"le;
ꢂ ꢃ ꢂ ꢃ
tom, di!erence curve, markers indicate re#ection positions. bottom, di!erence curve, markers indicate re#ection positions.
ꢃ
ꢃ

JSSC=8666=TSK=VVC=BG
386
LEINEWEBER AND JACOBS
FIG. 5. Powder di!raction pattern and results of Rietveld re"nement
of a-Ni(NH ) Br : #, observed pro"les points; *, calculated pro"le;
ꢂ ꢃ
ꢃ
bottom, di!erence curve, markers indicate re#ection positions.
In order to analyze structural aspects of the observed
phase transformation Ni(NH ) X }4NH Pb-Ni(NH ) X
ꢂ ꢄ ꢃ
ꢂ
ꢂ ꢃ ꢃ
Pa-Ni(NH ) X (the last step is observed only for X"Br,
ꢂ ꢃ ꢃ
I) one has to consider the presence of an approximate or
exact cubic primitive array of X in all the relevant crystal
FIG. 6. Tree of group}subgroup relationship. Space groups that lead
to the Cd(NH ) Cl and the Mg(NH ) Br type starting from Pm3m of an
structures. The hexaammines form an anti-CaF -like (or
ꢃ
ꢂ ꢄ
ꢂ ꢃ
ꢃ
ꢂ ꢃ
ꢃ
ꢃ>
anti-K PtCl type) arrangement with [Ni(NH ) ]
tions occupying half of the cubic interstices formed by X.
ca-
undistorted CsCl type structure are given. NH and X play the role of Cs
ꢃ
ꢄ
ꢂ
and Cl. The size of the unit cell within the a,b-plane is given with respect to
The removal of two-thirds of the NH molecules co-
ordinating Ni leads to four free coordination sites. These are
the parent CsCl cell. The c-axis is the same for all relevant space groups. k2
and t2 denote klassengleich and translationengleich relations of index 2 be-
tween the corresponding groups. This "gure shows that the Cd(NH ) Cl
ꢂ
ꢃ>
saturated for the linear Ni(NH )
a square of X. This is reached in such a way that one NH
units by going into
ꢂ ꢃ
ꢃ
ꢂ ꢃ
and the Mg(NH ) Br type structures are not derived from one another in
ꢂ ꢃ
ꢃ
the sense of a group}subgroup relationship.
ꢂ
molecule is located in each cubic interstitial site. This leads
to the image of a CsCl type structure formed by X and NH
ꢂ
groups. Each metal atom connects two NH molecules of
A comparison of the interatomic distances for a- and
ꢂ
two CsCl cells to double units (Fig. 1). The a- and the b-Ni(NH ) Br shows di!erences. However, standard devi-
ꢂ ꢃ
ꢃ
b-modi"cation are di!erent variants of such a structure, in ations of Rietveld re"nements are generally underestimated
which all the double cells are located perpendicular to one (33). An especially severe overlap of broad re#ections and
ꢀ
crystallographic axis leading the [NiX (NH ) ] chains. poor signal-to-noise ratio in the b-Ni(NH ) Br di!raction
ꢁ
ꢅꢆꢃ
ꢂ ꢃ
ꢂ ꢃ
ꢃ
ꢃ>
A disordered arrangement of Hg(NH ) units in a cubic data enforce a critical evaluation of the resulting structural
ꢂ ꢃ
array of Cl is known for Hg(NH ) Cl (16) with a cubic parameters.
ꢂ ꢃ
ꢃ
primitive unit cell of the dimension of an CsCl cell and with
a statistical distribution of Hg.
The re"ned preferred orientation parameters for
Ni(NH ) Cl and b-Ni(NH ) Br (see Table 1) indicate
ꢂ ꢃ
ꢃ
ꢂ ꢃ
ꢃ
The initial formation of the b-modi"cation by removing a plate-like shape of the crystals which is contrary to what is
ꢃ>
ꢂ ꢄ ꢃ
NH at low temperatures from Ni(NH )
indicates that expected from the chain structure. However, this may ori-
ꢂ
ꢂ ꢄ
the decomposition of the Ni(NH ) X crystals is kinetic- ginate from the speci"c conditions for the growth of the
ally controlled. The transformation b-Ni(NH ) X Pa- crystallites from the Ni(NH ) X crystals.
ꢂ ꢃ ꢃ
Ni(NH ) X may be understood as a rearrangement of
ꢂ ꢃ ꢃ
ꢂ ꢄ ꢃ
ꢃ>
Ni(NH ) units maintaining the array of X. However, the
ACKNOWLEDGMENTS
ꢂ ꢃ
transition can also be accomplished by a rotation of half
ꢀ
of the [NiX (NH ) ] chains by 903. Further investiga-
We thank the Fonds der Chemischen Industrie, the Deutsche Forschun-
gsgemeinschaft, and the Bundesministerium fur Bildung, Wissenschaft,
Forschung und Technologie (JA4DOR) for "nancial support of this work.
ꢁ
ꢅꢆꢃ
ꢂ ꢃ
tions are needed to analyze the mechanism of this trans-
formation.

JSSC=8666=TSK=VVC=BG
PREPARATION OF DIAMMINE NICKEL(II) HALIDES
387
REFERENCES
16. H. MacGillavry and J. M. Bijvoet, Z. Kristallogr. 94, 249 (1936).
17. W. Biltz and B. Fetkenheuer, Z. Anorg. Allg. Chem. 83, 163 (1913).
18. W. Biltz and B. Fetkenheuer, Z. Anorg. Allg. Chem. 89, 134 (1914).
19. R.-G. Lucas, Ph.D. thesis, Universitat Bonn, 1954.
20. B. K. Vainshtein, Zh. Fiz. Khim. 26, 1774 (1952).
1. H. Jacobs and B. Nocker, Z. Anorg. Allg. Chem. 614, 25 (1992).
2. H. Jacobs and B. Nocker, Z. Anorg. Allg. Chem. 619, 73 (1993).
3. B. Nocker, Ph.D. thesis, Universitat Dortmund, 1991.
4. L. Asch, G. K. Senoy, J. M. Friedt, J. P. Adlo!, and R. Kleinberger,
J. Chem. Phys. 62, 2335 (1975).
5. R. E{mann and Ch. Mockenhaupt, Spectrochim. Acta A 52, 1897
(1996).
6. R. E{mann, J. Mol. Struct. 351, 91 (1995).
7. H. Jacobs and U. Zachwieja, J. ¸ess-Common Met. 161, 175 (1990).
8. H. Jacobs and U. Zachwieja, J. ¸ess-Common Met. 170, 185 (1991).
9. H. Jacobs and C. Stuve, J. ¸ess-Common Metals 96, 323 (1984).
10. J. Jander, &&Anorganische und allgemeine Chemie in #ussigem Ammo-
niak,'' Band I, Teilband 1 der Reihe Chemie in nichtwa{rigen
ionisierenden Losungsmitteln, Friedr. Vieweg & Sohn, Braunschweig
(1966).
21. B. Morosin, Acta Crystallogr. 23, 630 (1967).
22. M. A. Porai-Koshits, Russ. J. Inorg. Chem. (Engl. Transl.) 4, 332 (1959).
23. C. Mockenhaupt and R. E{mann, 8. Vortragstagung der Fachgruppe
Festkorperchemie der GDCh Darmstadt (book of abstracts) (1996),
PIII 32.
24. R. D. Worswick, J. C. Cowell, and L. A. K. Staveley, J. Chem. Soc.,
Faraday ¹rans. 1590 (1974).
25. R. L. Chiang and R. S. Drago, Inorg. Chem. 10, 453 (1971).
26. H. Jacobs and D. Schmidt, Curr. ¹op. Mater. Sci. 8, 379 (1982).
27. A. C. Larson and R. B. von Dreele, General Structure Determination
System (GSAS), Los Alamos National Laboratory, Los Alamos, 1995.
28. A. March, Z. Kristallogr. 81, 285 (1932).
29. W. A. Dollase, J. Appl. Crystallogr. 19, 267 (1986).
30. A. Leineweber, M. W. Friedriszik, H. Jacobs, R. E{mann, G. Bottger,
F. Fauth, and P. Fischer, Z. Kristallogr. Suppl. 16, 46 (1999).
31. H. Barnighausen, MATCH 9, 139 (1980).
32. Th. Hahn, &&International Tables for Crystallography, Vol. A, Space
Group Symmetry,'' Kluwer Academic Publishers, Dordrecht, 1995.
33. R. A. Young, &&The Rietveld Method,'' Oxford Univ. Press, Oxford,
1993.
11. I. Olovson, Acta Crystallogr. 18, 889 (1965).
12. H. Jacobs, J. Bock, and C. Stuve, J. ¸ess-Common Met. 134, 207
(1987).
13. R. E{mann, G. Kreiner, A. Niemann, D. Rechenbach, A. Schmieding,
T. Sichla, U. Zachwieja, and H. Jacobs, Z. Anorg. Allg. Chem. 622, 1161
(1996).
14. P. Schiebel, A. Hoser, W. Prandl, G. Heger, W. Paulus, and P.
Schweiss, J. Phys.: Condens. Matter 6, 10989 (1994).
15. A. Leineweber, M. W. Friedriszik, and H. Jacobs, J. Solid State Chem.
147, 229 (1999).
34. C. J. Howard, J. Appl. Crystallogr. 15, 615 (1982).
35. P. Thomson and J. B. Hastings, J. Appl. Crystallogr. 20, 79 (1987).