Solid state synthesis of binary metal chalcogenides

Article abstract of DOI:10.1039/b101645h

Solid state reaction of metal halides MX_n (n = 1 or 2) with stoichiometric amounts of sodium chalcogenide (Na₂S₂ or Na₂E where E = S, Se or Te) at 300 °C for 48 h in evacuated ampoules affords a range of transition and main-group metal chalcogenides: ME₂ (M = Fe or Co, E = S or Te); M_(1-x)E (M = Fe or Co, E = S); Ag₂E (E = S, Se or Te), Ni_(1-xE (E = S, Se or Te); NiS₂, MnS, FeSe, SnSe and SnTe along with co-formed salt. Washing of the highly sintered, fused product mixture with water resulted in isolation of crystalline binary chalcogenides, typically of a single phase in good yield (90%). The products were characterised by X-ray powder diffraction, scanning electron microscopy (SEM), energy dispersive X-ray analysis (EDXA) and infrared spectroscopy.

Full text of DOI:10.1039/b101645h

View Article Online / Journal Homepage / Table of Contents for this issue
Solid state synthesis of binary metal chalcogenides
Graham A. Shaw, Daniel E. Morrison and Ivan P. Parkin*
Department of Chemistry, Christopher Ingold Laboratories, University College London,
20 Gordon Street, London, UK WC1H 0AJ
Received 20th February 2001, Accepted 1st May 2001
First published as an Advance Article on the web 25th May 2001
Solid state reaction of metal halides MXn (n = 1 or 2) with stoichiometric amounts of sodium chalcogenide (Na2S2
or Na2E where E = S, Se or Te) at 300 ЊC for 48 h in evacuated ampoules affords a range of transition and main-
group metal chalcogenides: ME2 (M = Fe or Co, E = S or Te); M(1 Ϫ x)E (M = Fe or Co, E = S); Ag2E (E = S, Se or
Te), Ni(1 Ϫ x)E (E = S, Se or Te); NiS2, MnS, FeSe, SnSe and SnTe along with co-formed salt. Washing of the highly
sintered, fused product mixture with water resulted in isolation of crystalline binary chalcogenides, typically of a
single phase in good yield (90%). The products were characterised by X-ray powder diffraction, scanning electron
microscopy (SEM), energy dispersive X-ray analysis (EDXA) and i nfrared spectroscopy.
Introduction
Binary transition and main group metal chalcogenides, of the
formulae ME or ME2 (E = S, Se or Te), have varied tech-
nological uses. The most extensive use of metal chalcogenides
stems from their wide range of semi-conducting properties.1
Group 12 chalcogenides are classed as direct wide band-gap
semiconductors, whilst tin() sulfide has a band gap (1.3 eV)
between that of silicon (1.12 eV) and gallium arsenide (1.43
eV).
Many procedures exist for the synthesis of bulk metal chal-
cogenides. These include the decomposition of organometallic
precursors,2 electrodeposition from solution3 and elemental
combination at elevated temperatures.4 Hydrogen sulfide and
selenide have widely been employed in the synthesis of metal
5–8
Fig. 1 XRD pattern of washed silver selenide synthesized by SSM
chalcogenides
through reactions with aqueous metal cat-
reactions heated to 300 ЊC for 48 h.
ions,6 metal oxides, and sol–gel processing.8 Most procedures
to form metal chalcogenides require multiple steps with long
processing times, carefully tailored precursors and/or the use of
highly toxic reagents.9
flame, and that the use of low-oxidation state metal halide
precursors gives a measure of stoichiometric control in the
synthesis of crystalline mono- and di-chalcogenides.
One synthetic approach that circumvents these problems is
the metathetical solid state reaction of metal halides with alkali
metal chalcogenides.10 Once initiated, these intimately mixed
reagents undergo a rapid, self-propagating reaction to produce
highly sintered metal chalcogenides and a co-produced salt
such as LiCl. The reactions proceed by passage of a hot yellow-
orange synthesis wave that progresses through the material as a
solid flame. Kaner and co-workers have carried out some
detailed mechanistic studies into the synthesis of ME2 (M = Mo
or W; E = S, Se or Te) from filament initiated solid state meta-
thesis (SSM) reactions involving MoCl5 and WCl6.11 They show
that the reaction temperature is limited by the melting point of
the co-produced salt. Product crystallinity and stoichiometry
were optimised by the use of high oxidation state metal fluor-
ides or chlorides. Consequent work resulted in the formation of
crystalline NiS (vaesite) from the reaction of Na2S with
K2[NiF6].12 We have shown, through a wide range of survey
work, that metathesis techniques can also be applied to the
synthesis of both binary main-group, lanthanide and actinide
chalcogenides, as well as to ternary Group I–lanthanide
sulfides.9,10,13
Results and discussion
Synthesis and characterisation
The solid state metathesis reactions of metal halides and
sodium chalcogenides (Na2E, E = S, Se or Te, and Na2S2) in
sealed evacuated ampoules at 300 ЊC for 48 h are described by
eqns. (1)–(11) (Table 1). In terms of monochalcogenides, the
reactions that afforded a single crystalline phase of desired
stoichiometry are described by (1)–(4). Fig. 1 shows the XRD
2AgF ϩ Na2E → Ag2E ϩ 2NaF (E = S, Se or Te) (1)
SnBr2 ϩ Na2E → SnE ϩ 2NaBr (E = Se or Te) (2)
MnF2 ϩ Na2S → MnS ϩ 2NaF
FeBr2 ϩ Na2Se → FeSe ϩ2NaBr
(3)
(4)
pattern obtained for silver selenide (Ag2Se) formed from the
SSM reaction after removal of the co-produced salt by washing
with water.
In addition, a series of reactions (5), (6) afforded a single
crystalline phase of non-stoichiometric metal chalcogenides.
In this paper we extend the scope of the solid state metathesis
procedure to the synthesis of binary chalcogenides of man-
ganese, iron, cobalt, nickel, silver and tin. We also show that
solid state metathesis reactions can proceed without a solid
1872
J. Chem. Soc., Dalton Trans., 2001, 1872–1875
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b101645h

View Article Online
Table 1 XRD data of the products from the SSM reaction of metal halides with Na2E and Na2S2 at 300 ЊC in sealed glass ampoules for 48 h
Initial
Product
colourc
Washed and
Lattice values/Å
( 0.01)
Lit.15 lattice
values/Å
a
e
Reagents
productsb
annealed materiald
2AgF ϩ Na2S
2AgF ϩ Na2Se
NaF
NaF
Black
Black
Ag2S (acanthite)
Ag2Se (naumannite)
a = c = 4.46
a = 4.33
b = 7.06
c = 7.76
a = 8.09
b = 4.47
c = 8.96
a = c = 4.46
a = 4.33
b = 7.06
c = 7.76
a = 8.09
b = 4.47
c = 8.96
2AgF ϩ Na2Te
NaF
Black
Ag2Te (hessite)
2MnF ϩ Na2S
2MnF ϩ Na2S2
FeCl2 ϩ Na2S
NaF
NaF
NaCl
Black
Black
Black
MnS (alabandite)
MnS (alabandite)
Fe1 Ϫ xS (x = 0.04)
a = c = 5.24
a = c = 5.24
a = 3.43
c = 3.79
a = c = 5.41
a = 3.77
c = 5.52
a = 3.85
b = 5.34
c = 6.26
a = 3.38
c = 5.15
a = c = 5.24
a = c = 5.24
a = 3.43
c = 3.79
a = c = 5.41
a = 3.77
c = 5.52
a = 3.85
b = 5.34
c = 6.26
a = 3.38
c = 5.15
FeCl2 ϩ Na2S2
FeCl2 ϩ Na2Se
NaCl
NaCl
Black
Black
FeS2 (pyrite)
FeSe
FeCl2 ϩ Na2Te
NaCl
Black
FeTe2
CoCl2 ϩ Na2S
NaCl
Black
Co1 Ϫ xS (x = 0.05)
CoCl2 ϩ Na2S2
CoCl2 ϩ Na2Te
NaCl
NaCl
Black
Black
CoS2
CoTe2
a = c = 5.54
a = 3.88
b = 5.30
c = 6.30
a = c = 5.54
a = 3.88
b = 5.30
c = 6.30
NiCl2 ϩ Na2S
NiCl2 ϩ Na2S2
NiCl2 ϩ Na2Se
NaCl
NaCl
NaCl
Black
Black
Black
Ni1ϪxS (x = 0.02)
NiS2ϩx (x = 0.02)
Ni1ϪxSe (x = 0.06)
a = c = 3.42
a = c = 5.62
a = 3.66
c = 5.33
a = 3.78
c = 6.06
a = 4.30
b = 11.20
c = 3.99
a = c = 3.42
a = c = 5.62
a = 3.66
c = 5.33
a = 3.78
c = 6.06
a = 4.30
b = 11.18
c = 3.98
NiCl2 ϩ Na2Te
SnBr2 ϩ Na2S
NaCl
NaBr
Black
Black
Ni1ϪxTe (x = 0.45)
SnS
a Reagents shown with molar ratios. b Phases characterised by X-ray powder diffraction prior to work-up of product. c Colour of isolated phases (the
solid material remaining after reaction product work-up). d Phases characterised by X-ray powder diffraction after annealing at 250–300 ЊC for 48 h.
Values of x assessed by Vegard’s law. e Unit cell parameters were determined for major phase binary material only.
This phase was not observed in Kaner’s reactions employing
K2[NiF6] where NiS was formed.12
The reactions that afforded a single phase of crystalline metal
dichalcogenides are described in eqns. (7) and (8). Fig. 2 shows
MX2 ϩ Na2S2 → MS2 ϩ 2NaX
(MX2 = FeBr2, CoCl2 or NiCl2)
(7)
(8)
MX2 ϩ Na2Te → 0.5MTe2 ϩ 0.5M ϩ 2NaX
(MX2 = FeBr2 or CoCl2)
the XRD pattern of the initial reaction product from the reac-
tion of Na2S2 with NiCl2, showing both the formation of co-
produced salt (NaCl) and a slightly sulfur deficient phase of
nickel disulfide, NiS(2 ϩ x). The telluride reactions, (8), showed
that elemental metal was produced in the process. The ele-
mental metal could not be seen in the X-ray powder diffraction
pattern probably as it was amorphous. It was, however, readily
seen in the back-scattered electron micrographs from SEM and
by EDAX (energy dispersive X-ray analysis) which showed
regions of phase segregated metal and metal ditelluride.
Fig. 2 XRD of unwashed nickel disulfide (NiS2.02) synthesized by
SSM reactions heated to 300 ЊC for 48 h. Stick pattern is for
stoichiometric NiS2.
NiCl2 ϩ Na2E → Ni(1 Ϫ x)E ϩ 2NaCl (E = S, Se or Te) (5)
MX2 ϩ Na2S → M(1 Ϫ x)S ϩ 2NaCl
(MX2 = FeBr2 or CoCl2)
(6)
Eqns. (9)–(11) show those reactions that did not produce
single phase metal chalcogenide products after trituration.
The stoichiometries were derived from application of Vegard’s
law. The results are shown in Table 1, and represent an average
stoichiometric value derived from the four most intense peaks
in the range 20 ≤ 2θ ≤ 70Њ. Application of Vegard’s law to the
XRD spectrum of the Ni(1 Ϫ x)S product showed it to be a
slightly metal deficient phase (with an empirical formula of
NiS1.02).14 By changing the sulfide precursor to Na2S2, eqn. (7),
the vaesite structure was obtained, NiS2. However, upon close
examination of the XRD spectrum (Fig. 2) a shift in peak
position to lower 2θ values is observed, signifying a slight
increase in size of the unit cell. This observation is consistent
ZnCl2 ϩ Na2S2 → ZnS* ϩ 2NaCl ϩ S
(9)
* Two distinct phases of zinc sulfide were characterised by
XRD (sphalerite and α-wurtzite-2H)
SnBr2 ϩ Na2S → SnS ϩ Sn2S3* ϩ Sn3S4* ϩ 2NaBr (10)
(* These are minor phases less than 10%)
with the formation of a slightly metal deficient phase NiS(2 ϩ x)
.
2CrCl3 ϩ 3Na2S → NaCl ϩ ‘amorphous material’ (11)
J. Chem. Soc., Dalton Trans., 2001, 1872–1875
1873

View Article Online
The yield of metal chalcogenide obtained for each reaction as
described in eqns. (1)–(10) was typically greater than 90% . All
products were black, and after washing with water, showed
characteristic IR absorptions for binary metal chalcogenides
is likely that the intimate mix of reagents allows neighbouring
grains to react via close contact. This could be triggered by
decomposition of the sodium chalcogenide. This gives rise to an
observed dark wave front that progresses through the material.
Hence, propagation is accomplished without the massive
evolution of heat and resultant thermal flash associated with
conventional SSM reactions. It should be noted that the dark
wave is an exothermic phenomenon, just insufficiently exo-
thermic to promote a solid flame.
with broad bands between 600 and 300 cmϪ1
.
All reactions of metal halides with sodium chalcogenides at
300 ЊC for 48 h gave products of sufficient crystallinity to be
characterised by XRD. The metal chalcogenide phases pro-
duced correlated well with literature values (cell parameters
within 0.02 Å, Table 1).15 The XRD pattern of each product, as
formed, also showed the presence of crystalline sodium halide.
This was easily removed by trituration with distilled water
without decomposition of the crystalline metal chalcogenide.
In each case, analysis of the washed product by SEM showed it
to be a highly sintered, fused mass. EDXA analyses of the
washed material confirmed the presence, for the majority of
cases, of homogeneous products, whose elemental ratio was in
good agreement with that observed by the XRD pattern (to
detection limits). The notable exception to this was reaction
(10) that formed tin sulfide. Evaporation of the water washings
left colourless co-produced salt, which was confirmed by X-ray
powder diffraction.
The product from the SSM reaction of SnBr2 with Na2S
(annealed at 300 ЊC) showed appreciable differences in atomic
distribution by EDXA. X-Ray powder diffraction studies
showed the presence of two tin sulfide phases (SnS and Sn2S3).
Increasing the annealing temperature to 500–550 ЊC produced
predominantly SnS with a 2–5% trace of Sn2S3. This is con-
sistent with the formation of tin sulfide thin films from the
APCVD of SnCl4 with H2S,16 where Sn2S3 predominates at low
temperatures and SnS predominates at 525 ЊC and above.
The reaction of CrCl3 with Na2S, (11) shows the formation
of a co-produced salt suggesting that a metathesis reaction has
occurred. However, the chromium product was amorphous to
X-rays, even after annealing at 300 ЊC for 48 h. EDXA studies
of this X-ray amorphous material, both before and after wash-
ing, showed variable elemental composition of Cr and S over
many surface spots. This suggests either the existence of differ-
ent chromium sulfide phases or impurities resulting from
oxygen contamination. The origin of the impurities is unknown,
although contamination from the reaction vessel has been
observed in SSM reactions.10,13 The Cr–S system has six discrete
phases in the composition range Cr0.95S to Cr1.5S, each of which
may also be encountered in a non-stoichiometric form,17 so
definitive elucidation proved problematic.
Kinetic products and reaction mechanism
The relatively low temperatures at which these SSM reactions
proceed allow some degree of stoichiometric control in the
reaction simply by varying the sulfur content of the precursor.
Thus it is possible to form single phases of both M(1 Ϫ x)S and
MS2 (M = Fe, Co or Ni) by using Na2S and Na2S2 respectively.
In addition, the low reaction temperature may allow the form-
ation of kinetic products, and so the formation of mixed phases
of tin sulfide.
Solid state metathetical reactions are believed to proceed via
reductive recombination or ionic metathesis pathways.11,19 Ionic
metathesis involves the exchange of ions within a salt melt.
Reductive recombination involves reduction to the elements
followed by recombination to form the metal chalcogenide. The
reactions resulting in metal monochalcogenides, (1)–(4), are
likely to proceeded via partial reduction of the reagents to the
elements, followed by their recombination to form the product,
eqns. (12), (13). Moreover, the lack of a solid flame raises
MX2(s) ϩ Na2E(s) → [M ϩ E ϩ 2NaX] →
ME(s) ϩ 2NaX(s) (12)
MX2(s) ϩ Na2S2(s) → [M ϩ 2S ϩ 2NaX] →
MS2(s) ϩ 2NaX(s) (13)
serious doubts as to the free mobility of ions within the reaction
mixture required for efficient ‘ionic exchange’. The nickel
arsenide structure adopted by many of the metal products
infers a high degree of covalency (close approach of dipositive
metal atoms, 2.60–2.68 Å for FeS, CoS and FeS). This taken in
conjunction with the covalent nature of the metal precursors,
and the common crystal structure,20 infers that an ionic inter-
mediate is unlikely.
In addition, elemental metal was produced in the synthesis
of the metal ditellurides, eqn. (8), indicating reductive recom-
bination.
By considering the comparative heats of formation of M–Te,
M–M and Te–Te, it is apparent that for the low temperatures at
which these SSM reactions proceeded the most favorable
product is MTe2. Increasing the reaction temperature would
favour the reverse reaction, and so the synthesis of MTe.
Whilst it was not possible to synthesize ZnS2 from the reac-
tion of ZnCl2 with Na2S2, the resultant metathesis afforded
both cubic and hexagonal structures of ZnS with the formation
of free sulfur, eqn. (9). Interestingly, previous studies have
shown that the reaction of ZnCl2 with Na2S results only in the
more thermodynamically stable hexagonal phase,9 suggesting
that the reactions studied here may follow a less energetic route.
Role of chalcogenide precursor
Conclusion
Previous studies9 have shown that metathetical reactions
between solid alkali metal chalcogenide and metal halide occur
very rapidly, with a large evolution of heat. In virtually all cases
this gives rise to a characteristic thermal flash or solid flame,
with temperatures in excess of 1000 ЊC being reported.9 For
solid-state metathesis reactions it has been suggested that
propagation will only occur once the heat of reaction is
sufficient to melt the co-produced salt. In so doing, the reaction
can proceed through successive layers and hence propagate
through the solid. In the reactions observed here, however, a
dark wave passes through the reagents (rather than a solid
flame) once the surroundings reach approximately 200 ЊC,
irrespective of which metal halides are used. In addition, the
calculated heats of reaction (200–250 kJ molϪ1) are insuffi-
ciently exothermic to support a solid flame synthesis wave.18 It
Solid state metathesis reactions of low oxidation state metal
halides with sodium chalcogenides at 300 ЊC for 48 h offer a
convenient route to the synthesis of a range of crystalline
transition metal chalcogenides in high yield (> 90%). In the
majority of cases a single stoichiometric phase is produced.
Most conventional syntheses of metal chalcogenides require
either multiple steps with long processing times, the existence of
specific precursors or the use of highly toxic reagents. These
problems are circumvented by the SSM reactions of metal
halides with alkali metal chalcogenides. Once initiated, these
intimately mixed reagents undergo reaction to produce highly
sintered metal chalcogenides and a co-produced salt. The
reaction is not self-propagating in the sense that no solid flame
is observed, however a dark wave was observed promoting
transformation of products into reactants. The alkali metal
1874
J. Chem. Soc., Dalton Trans., 2001, 1872–1875

View Article Online
halide co-produced salt is easily removed by trituration, afford-
ing pure metal chalcogenide. In comparison to conventional
elemental combination reactions, SSM occurs at significantly
lower temperatures with shorter reaction times and fewer
preparative steps. We have shown that metal halides in their
common low oxidation states can be used in conjunction with
facile heating (300 ЊC for 48 h) to afford a selection of crystal-
line binary metal mono- and di-chalcogenides with, for the
most part, preservation of the oxidation state of the metal.
With the use of alkali metal chalcogenide reagents prepared
from a room temperature synthesis, it is possible to achieve a
low temperature SSM reaction, thereby allowing the possibility
of kinetic product formation and amorphous materials. Clari-
fication of reaction mechanisms is not, however, forthcoming,
although in these reactions the reductive recombination route is
favoured.
Reaction of sodium chalcogenide with metal halide
The same general reaction scale and procedure was adopted for
all these reactions, as exemplified here for FeS2.
Disodium disulfide (100 mg, 0.91 mmol ) and iron()
bromide (196 mg, 0.91 mmol ) were ground together inside a
glove box, using an agate pestle and mortar, until intimately
mixed. The mixture was added to an ampoule, which was then
sealed under vacuum and heated slowly (ca. 20 ЊC minϪ1) in a
tube furnace. For each reaction, a colour change throughout
the mixture was observed once the temperature reached 170–
220 ЊC, resulting in the formation of a fused black product. A
synthesis wave was not observed. The product was then
annealed for 48 h at 300 ЊC, after which it was triturated with
3 × 20 cm3 of distilled water, before being dried under vacuum.
The resultant black powder was analysed by XRD, SEM/
EDXA and IR spectroscopy. XRD analysis was also carried
out on the pre-washed reaction product.
Experimental
All reagents used were of 99.9% purity, or better, purchased
from Aldrich Chemical Company and used without further
purification. Ammonia was purchased from BOC and used
without drying. Na2E (E = S, Se or Te) and Na2S2 were
prepared by the direct combination of stoichiometric quantities
of sodium metal and elemental chalcogen in liquid ammonia.
These reactions were carried out at room temperature in Teflon-
in-glass, Youngs-type, pressure vessels, using Schlenk tech-
niques. Once prepared, the sodium chalcogenide was used
immediately in the SSM reactions without annealing or
exposure to atmospheric conditions. All manipulations were
carried out in a dinitrogen filled glove box. Reactions involving
the synthesis of the sodium chalcogenide precursors were
carried out (using Schlenk techniques) in thick walled (3–4
mm), Teflon-in-glass, sealed, Youngs-type Schlenk tubes (sealed
by a large rota-flow tap) which were surrounded by safety
netting. The solid state metathesis reactions were performed in
sealed, evacuated ampoules using a Lenton Thermal Designs
tube furnace. X-Ray powder diffraction patterns were deter-
mined on a Siemens D5000 transmission powder diffractometer
using germanium monochromated Cu-Kαl radiation (λ =
1.5405 Å). They were indexed using either TREOR or
METRIC-LS programs14 (lattice parameters matched to within
0.02 Å with literature15). The SEM/EDXA measurements were
made on a Jeol JSM820 microscope, equipped with a Kevex
Quantum Delta 4 detector and a Hitachi SEM S-570 camera.
The electron beam was focused (1 µm spot at surface) with an
excitation energy of 20 keV. Electron-probe analyses were
conducted on a Jeol EMA instrument, using polished samples,
and compared to those of metal and chalcogen standards.
FT-IR spectra were recorded using a Nicolet 205 FT-IR
spectrometer, as pressed KBr discs.
References
1 A. Lopez and A. Ortiz, Semicond. Sci. Technol., 1994, 9,
2130.
2 R. Nomura, K. Konishi, S. Futenma and H. Matsuda, Appl.
Organomet. Chem., 1990, 4, 607.
3 S. Massaccesi, S. Sanchez and J. Vedel, J. Electroanal Chem., 1996,
412, 95.
4 R. Coustal, J. Chim. Phys., 1931, 31, 277.
5 J. Livage, M. Henry and C. Sanchez, Prog. Solid State Chem., 1998,
8, 2453.
6 H. C. Metcalf, J. E. Williams, J. F. Caskta, Modern Chemistry, Holt,
Reinhart, Winston, New York, 1982, p. 54.
7 M. Sato, G. Adachi and J. Shiokawa, Mater. Res. Bull., 1984, 19,
1215.
8 M. A. Sriram and P. N. Kumta, J. Mater. Chem., 1998, 8,
2453; M. A. Sriram and P. N. Kumta, J. Mater. Chem., 1998,
8, 2441; M. A. Sriram and P. N. Kumta, J. Mater. Chem., 1994, 77,
1381.
9 I. P. Parkin, Chem. Soc. Rev., 1996, 199.
10 I. P. Parkin and A. T. Rowley, Polyhedron, 1993, 3, 689.
11 P. R. Bonneau, R. K. Shibao and R. B. Kaner, Inorg. Chem., 1990,
29, 2511; P. R. Bonneau, R. F. Jarvis and R. B. Kaner, Nature
(London), 1991, 349, 510.
12 P. R. Bonneau, R. K. Shibao and R. B. Kaner, J. Am. Chem. Soc.,
1990, 29, 2511.
13 J. P. Cotter, J. C. Fitzmaurice and I. P. Parkin, J. Mater. Chem., 1994,
4, 1603.
14 P. F. Warner, L. Eriksson and M. Westdahl, J. Appl. Crystallogr.,
1985, 18, 360.
15 The United Kingdom Chemical Database Service, D. A. Fletcher,
R. F. McMeeking and D. Parkin, J. Chem. Inf. Comput. Sci., 1996,
36, 746.
16 L. S. Price, I. P. Parkin, A. M. E. Hardy and R. J. H. Clark, Chem.
Mater., 1999, 11, 1792.
17 N. N. Greenwood, E. A. Earnshaw, Chemistry of the Elements,
Pergamon Press, Oxford, 1990, p. 804.
18 J. F. Crider, Ceram. Eng. Sci. Proc., 1982, 3, 519.
19 P. R. Bonneau, R. F. Jarvis and R. B. Kaner, Inorg. Chem., 1992, 31,
2127; P. R. Bonneau and R. B. Kaner, J. Am. Chem. Soc., 1993, 32,
6084.
CAUTION: liquid ammonia generates a pressure of about
7 atm at room temperature. Care should be exercised in its use
in thick walled glass vessels. All reactions should be conducted
behind a safety screen with blast proof netting around the
reaction vessel.
20 N. N. Greenwood, E. A. Earnshaw, Chemistry of the Elements, 2nd
edn., Pergamon Press, Oxford, 1994, pp. 765 and 766.
J. Chem. Soc., Dalton Trans., 2001, 1872–1875
1875