Thermal stability of nanoparticle size and phase composition of nanostructured Ag2S silver sulfide

Article abstract of DOI:10.1016/j.jallcom.2018.06.351

The thermal stability of phase composition and particle size of nanocrystalline silver sulfide Ag₂S powder has been examined for the first time by in situ high-temperature X-ray diffraction, scanning electron microscopy and Raman spectroscopy methods. It was shown that in oxidizing environment or in the presence of oxidizing impurities, heating of nanostructured silver sulfide is accompanied by its decomposition (dissociation) with isolation of metallic silver. Reduction of silver sulfide particles leads to decreasing the decomposition temperature of Ag₂S. The decomposition of Ag₂S nanopowder with isolation of silver metal as a result of heating in air by high-power laser radiation was confirmed by Raman scattering. In the absence of oxidizing environment, heating of silver sulfide does not lead to its decomposition and nanoparticle growth up to the temperature of ～450 K. The nanoparticle size increases as a result of annealing at a temperature of 450–930 K corresponding to the region of collective recrystallization of silver sulfide nanopowder. For nanostructured silver sulfide, the activation energy Q of collective recrystallization in the temperature range 450–900 K is ～0.12 eV?atom^?1.

Full text of DOI:10.1016/j.jallcom.2018.06.351

Accepted Manuscript
Thermal stability of nanoparticle size and phase composition of nanostructured Ag S
2
silver sulfide
Stanislav I. Sadovnikov, Emma G. Vovkotrub
PII:
S0925-8388(18)32472-1
10.1016/j.jallcom.2018.06.351
JALCOM 46687
DOI:
Reference:
To appear in: Journal of Alloys and Compounds
Received Date: 8 February 2018
Revised Date: 22 June 2018
Accepted Date: 28 June 2018
Please cite this article as: S.I. Sadovnikov, E.G. Vovkotrub, Thermal stability of nanoparticle size and
phase composition of nanostructured Ag S silver sulfide, Journal of Alloys and Compounds (2018), doi:
2
10.1016/j.jallcom.2018.06.351.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Thermal Stability of Nanoparticle Size and Phase Composition of
Nanostructured Ag2S Silver Sulfide
Stanislav I. Sadovnikova* Emma G. Vovkotrubb
,
a Institute of Solid State Chemistry, Ural Branch of the Russian Academy of Sciences, Ekaterinburg 620990, Russia
b Institute of High-Temperature Electrochemistry, Ural Branch of the Russian Academy of Sciences, Ekaterinburg
620137, Russia
ABSTRACT
The thermal stability of phase composition and particle size of nanocrystalline silver sulfide Ag2S
powder has been examined for the first time by in situ high-temperature X-ray diffraction, scanning
electron microscopy and Raman spectroscopy methods. It was shown that in oxidizing environment
or in the presence of oxidizing impurities, heating of nanostructured silver sulfide is accompanied
by its decomposition (dissociation) with isolation of metallic silver. Reduction of silver sulfide par-
ticles leads to decreasing the decomposition temperature of Ag2S. The decomposition of Ag2S na-
nopowder with isolation of silver metal as a result of heating in air by high-power laser radiation
was confirmed by Raman scattering. In the absence of oxidizing environment, heating of silver sul-
fide does not lead to its decomposition and nanoparticle growth up to the temperature of ~450 K.
The nanoparticle size increases as a result of annealing at a temperature of 450-930 K correspond-
ing to the region of collective recrystallization of silver sulfide nanopowder. For nanostructured
silver sulfide, the activation energy Q of collective recrystallization in the temperature range 450-
900 K is ~0.12 eV atom-1.
Key words: Nanostructured silver sulfide, Thermal stability of nanoparticle size, Decomposition,
Silver whiskers
*
Corresponding author. Institute of Solid State Chemistry, Russian Academy of Sciences, Ekaterinburg,
620990, Russia.
E-mail address: sadovnikov@ihim.uran.ru (S.I. Sadovnikov).

ACCEPTED MANUSCRIPT
1. Introduction
Metal sulfide semiconductor nanostructures represent the most promising group of
nanostructured materials with excellent optical and electrical properties for various applications [1-
3]. Among them, much attention is drawn to nanostructured silver sulfide Ag2S [4-6], which is a
semiconductor below ~450 K and a superionic conductor at T > 452 K.
Silver sulfide Ag2S has three polymorphic modifications [6]. The low-temperature mono-
clinic (space group 14 - P21/c) semiconducting phase
~452 K. Argentite -Ag2S has a body-centered cubic (space group 229 - Im3m ) crystal lattice and
exists in the temperature interval 452-859 K. The high-temperature face-centered cubic (space
α-Ag2S (acanthite) exists at temperatures to
β
group № 225 - Fm3m ) phase γ-Ag2S is stable at temperatures from ~860 K to melting tempera-
ture.
Silver sulfide, as one of the most versatile semiconductors, can be used in photoelectro-
chemical cells, photocells, photoconductors and infrared detectors [6-9].
Silicon and germanium containing semiconductors are the basis of modern electronic engi-
neering. The maximum operating temperature of germanium devices does not exceed 323-330 K,
for silicon devices it increases to 400-430 K. Silicon exhibits a better operation stability at high
temperatures. However, with further application of Si and other accepted semiconductors, the im-
provement of the performance and stability of integrated circuits by increasing the operating clock
frequency and the number of transistors is becoming an ever more complicated and expensive task.
The point is that the heat release of transistors increases exponentially as their switching speed in-
creases. Therefore, an important challenge is to determine the regions of thermal stability of semi-
conducting compounds and materials that are potentially suitable for operation at elevated temper-
atures. One of such materials is nanostructured silver sulfide.
Potential application of nanostructured silver sulfide holds much promise in micro- and
nanoelectronics, where Ag2S/Ag heteronanostructures are used in nonvolatile memory devices and
resistive switches. Their action is based on reduction of Ag+ cations of silver sulfide to metallic Ag
atoms, transformation of acanthite
α-Ag2S into argentite β-Ag2S and appearance of a conducting
channel from Ag and argentite -Ag2S [10-15]. Hybrid heteronanostructures Ag2S/Ag exhibit en-
β
hanced efficiency in photocatalysis owing to their band structure peculiarities [16-18].
As a rule, microelectronic devices and photocatalysts, having a nanostructured silver sul-
fide cell, operate in air at elevated temperatures (to ~400-450 K), at which the growth and varia-
tion in the phase composition of Ag2S nanoparticles may take place, leading to the degradation of
2

ACCEPTED MANUSCRIPT
their electrical and optical properties. For example, according to data [19], the fluorescence maxi-
mum of photoluminescence emission spectra changes from 975 to 1175 nm when the size of Ag2S
quantum dots in colloidal solutions increases from 2.4 to 7.0 nm; according to [5, 6], the same
shift of the fluorescence maximum from ~960 to ~1176 nm occurs when the size of Ag2S quantum
dots increases from 2.3 to 15.0 nm. The observed shift is due to the weakening of the quantum
confinement effect (quantum localization) and decreasing the band gap Eg as a result of Ag2S
quantum dot growth. This agrees with the experimental data [20] on the dimensional dependence
of the band gap of Ag2S nanopowders.
The uncontrolled growth of quantum dot and Ag2S nanoparticle dimensions results in the
instability and degradation of properties of devices and compositions with nanostructured silver
sulfide.
When studying the thermal stability of nanostructured silver sulfide, special attention
should be paid to the phase transformation of the semiconductor monoclinic acanthite
α
-Ag2S to
the superionic cubic argentite -Ag2S at a temperature of ~448-450 K. The cubic argentite
β
β
-Ag2S
has a very high mobility of Ag+ ions. This is due to its crystal structure, in which four silver atoms
are statistically distributed in 54 positions 6(b) and 48(j) with probabilities 0.0978 and 0.0711, re-
spectively [6, 21].
The least distance between the Ag1 and Ag1 atoms in the crystal lattice of monoclinic
Ag2S acanthite at 433 K is equal to 0.3353 nm, and the least distance between the Ag1 and Ag2
atoms is in the interval from 0.3087 to 0.3205 nm [22, 23]. In cubic -Ag2S argentite, possible dis-
tances between silver atoms at a similar temperature of 453 K are much smaller. According to [22,
23], in the crystal lattice of cubic -Ag2S argentite at 453 K the least possible distances between
α-
β
β
the Ag1 and Ag1 atoms are equal to 0.2431 nm, between the Ag1 and Ag2 atoms the least distanc-
es lie in the range from 0.0927 to 0.3256 nm, and between the Ag2 and Ag2 atoms the least dis-
tances are from 0.0561 to 0.3002 nm. The covalent radius of Ag atom is ~0.146 nm, and ionic ra-
dius of Ag+ is ~0.126 nm. With that in mind, it is clear that silver atoms in monoclinic acanthite
are at rather large distances from each other and therefore they occupy their crystallographic sites
with probability 1. In cubic β-Ag2S argentite, the possible distances between silver atoms are too
small for 6(b) and 48(j) positions to be occupied by Ag atoms with probability 1. For this reason,
the occupancies of the 6(b) and 48(j) positions by Ag atoms (put it otherwise, the probabilities of
finding Ag atoms in the 6(b) and 48(j) sites) are very small, less than 0.1 [6, 21, 23]. Physically,
this means that 4 silver atoms in the superionic argentite lattice are in continuous motion over 54
possible crystallographic positions. With time, the continuous motion of Ag atoms (ions) can lead
3

ACCEPTED MANUSCRIPT
to Ag2S nanoparticles growth and to a decrease in the thermal stability of devices based on
nanostructured silver sulfide operating at elevated temperature.
There is scarcely any information in the literature on the variation in the size and phase
composition of particles (grains) of nanostructured sulfides with temperature and on their recrys-
tallization. We can mention only the work [24] on recrystallization of cadmium sulfide films, in
which the grain size variation was not discussed, and the study [25] devoted to recrystallization in
lead sulfide nanofilms.
As for nanostructured silver sulfide, the information about its thermal properties is limited
to works [22, 23, 26-28] dealing with the effect of Ag2S particle size on the thermal expansion co-
efficient and heat capacity. Even the recent review article [5] devoted to nanostructured silver sul-
fide and monograph [6] contain no information about the effect of temperature on the phase com-
position and size of silver sulfide nanoparticles. Limited data on the effect of temperature on the
size of Ag2S nanoparticles and the recrystallization of silver sulfide nanopowders are presented in
work [29]. Besides, the phase transformation of acanthite
α-Ag2S into argentite β-Ag2S takes
place in silver sulfide at ~448-450 K. It is not known how this phase transformation affects the re-
crystallization and the silver sulfide nanoparticle growth.
In this connection, in the present work the effect of temperature and phase transformation
“acanthite - argentite” on the particle size and phase composition of nanostructured silver sulfide
was studied experimentally for the first time, and the regions of thermal stability and recrystalliza-
tion of Ag2S nanoparticles were determined.
2. Experimental details
Silver sulfide powders were synthesized by chemical deposition from aqueous solutions of
silver nitrate AgNO3, sodium sulfide Na2S and sodium citrate Na3C6H5O7. The solubility product
Ksp of silver sulfide Ag2S is very small (at 298 K Ksp = 6.3·10-50 [6, 30]) and deposition of silver
sulfide occurs by the following reaction scheme
Na C H O
3
6 5
2AgNO3 + Na2S
7 → Ag2S↓ + 2NaNO3.
(1)
Ag2S powders with different particle size were prepared from the same chemical reagents by
changing their concentration in the solution.
In this work, silver sulfide powders with the average particle size of less than 50 nm are
considered as nanocrystalline, while those with the average particle size of about 200 nm are
called bulk powders. The procedure of synthesis of coarse- and nanocrystalline Ag2S powders is
described elsewhere [31, 32] (see also Supplementary material for experimental details).
4

ACCEPTED MANUSCRIPT
The high-temperature in situ X-ray diffraction (XRD) studies were carried out on a STA-
DI-P (STOE) diffractometer in Debye-Scherrer geometry. The diffractometer was equipped with a
quartz capillary furnace STOE. For better determination accuracy of lattice parameters, measure-
ments were performed in CuKα1-radiation in angle interval 2θ = 20-55° with a step of
∆(2 ) = 0.02° and scanning time of 50 s in point. The quartz capillary was filled by a silver sulfide
θ
powder, the diameter of quartz capillary was 0.3 mm. The diffraction measurements were carried
out with the use of a linear position-sensitive detector while continuous rotation of the capillary at
a temperature from 295 to 773 K with a step of ~25-30 K.
In addition, the high-temperature in situ X-ray measurements of Ag2S nanopowders were
conducted on an X`Pert PRO MPD (Panalytical) diffractometer in Bragg-Brentano geometry. The
diffractometer was furnished with a fast position-sensitive solid-state sector detector PIXCEL and
an Anton Paar HTK-1200 Oven providing heating of samples. The diffraction measurements were
carried out at a temperature from 295 to 398 K in CuKα1,2-radiation in the angle interval 2
67.5° with a step of ∆(2 ) = 0.026° and scanning time of 200 s in point.
The synthesized silver sulfide powders were also examined on a Shimadzu XRD-7000 dif-
fractometer in CuKα1-radiation in the angle interval 2 = 20-95° with a step of ∆(2 ) = 0.02 and
θ = 20-
θ
θ
θ
scanning time of 10 s in point. The determination of the crystal lattice parameters and final struc-
ture refinement of the synthesized silver sulfide powders were performed by means of the X’Pert
HighScore Plus software suite [33].
The nanocrystalline silver sulfide powders were additionally annealed in vacuum 0.013 Pa
at 393, 423, 453 and 493 K for 2 h and heated to the annealing temperature Tann at a rate of not
more than 1 K min-1, as well as in argon atmosphere during heating to 623 K for 2 h at a rate of
1 K min-1.
The average size D of particles (to be more precise, the average size of coherent scattering
regions (CSR)), in the synthesized silver sulfide nanopowders was estimated by the XRD method
from
β
(2
θ
) diffraction reflection broadening using the dependence of the reduced reflection broad-
) = [ (2 )cos ]/ on the scattering vector s = (2sin )/ [34, 35] (see Supplementary
ening β*(2
θ
β
θ
θ
λ
θ λ
material). The diffraction reflections were described by the pseudo-Voigt function. The particle
size was found also from the value of specific surface Ssp, measured by the Brunauer-Emmett-
Teller (BET) method, and from scanning electron microscopy data.
The changes occurring during heating of silver sulfide powders were observed directly us-
ing the scanning electron microscopy (SEM) method on a JEOL-JSM LA 6390 microscope with a
5

ACCEPTED MANUSCRIPT
JED 2300 Energy Dispersive X-ray Analyzer. The elemental composition of Ag2S powders was
obtained on the same microscope by means of EDX analysis.
Nanostructured silver sulfide was examined by transmission electron microscopy (TEM)
method. The high-resolution TEM images were recorded on a JEOL JEM-2010 transmission elec-
tron microscope with 0.14 nm (1.4 Å) lattice resolution.
The Raman spectra were measured on a RENISHAW-1000 (Renishaw plc., UK) micro-
scope-spectrometer with visible continuous radiation in the “green” spectral region at wavelength
532 nm, and at radiation power to 40 mW in the interval 50-4000 cm-1. The measurement condi-
tions were as follows: spectrograph diffraction grating with 1800 groove mm-1; exciting radiation
power on sample was from 0.4 to 40 mW; accumulation (exposure) time of one spectrum was
from 10 to 30 s; and number of scans (number of signal accumulation cycles) was equal from 2 to
10 (see Supplementary material).
3. Results and discussion
The thermal stability of nanostructured silver sulfide was studied on powders with an aver-
age CSR size of about 45 and 85 nm (from XRD data). According to BET data, the particle size in
these powers is ~45 and ~163 nm. Comparison of particle sizes determined by XRD and BET
methods shows that the coarser powder was agglomerated. The TEM image of Ag2S nanoparticle
with a size about 40 nm of synthesized silver sulfide nanopowder is demonstrated in Fig. 1.
3.1. Decomposition of silver sulfide in oxidizing medium
The XRD patterns were recorded in Debye-Scherrer geometry in air at a temperature from
295 to 773 K for Ag2S powder with CSR size equal to 85 nm. Film strip in situ XRD plots are
shown in Fig. 2, and their evolution in standard coordinates “2θ angle - intensity” is presented in
Fig. 3.
Full-profile structure refinement and comparison with data [31, 32] revealed that diffrac-
tion reflections of monoclinic (space group P21/c) acanthite -Ag2S are observed on the XRD
plots during heating up to 448 K (Fig. 2, 3). At further heating from 453 to 623 K, the XRD plots
contain diffraction reflections of cubic (space group Im3m ) argentite -Ag2S. Thus, the transfor-
α
β
mation of acanthite into argentite takes place at ~450 K. At 673 K, along with argentite reflections,
diffraction reflections of cubic (space group Fm3m ) Ag metal appear on the XRD pattern, while
at higher temperatures, 723 and 773 K, the XRD patterns contain only diffraction reflections of
silver. This is due to the following reasons.
6

ACCEPTED MANUSCRIPT
The peculiarity of silver sulfide in comparison with other sulfides is its stability to oxida-
tion because coarse-crystalline (bulk) Ag2S does not interact with oxygen during heating in air to
~770 K. Only as a result of calcinations in air at a temperature of >800 K, coarse-crystalline silver
sulfide with particle size of more than 2-3 ꢀm is decomposed with isolation of silver metal and sul-
fur, which is oxidized to gaseous sulfur dioxide (sulfurous gas):
Ag2S + O2 → 2Ag + SO2 ↑.
(2)
A similar spontaneous isolation of silver in the form of whiskers during heating of Ag2S to 800-
900 K was noted in study [36]. The formation of silver whiskers on the surface of pressed
pellet heated in air up to 570 K was observed in work [37].
α-Ag2S
In oxidizing medium, reduction of silver sulfide particles leads to lowering of its decompo-
sition temperature and to the appearance of metallic silver at lower temperature.
Indeed, heating of the synthesized silver sulfide nanopowder at a rate of 5 K·min-1 and
more in vacuum 0.013 Pa (10-4 mm Hg) or in argon or helium atmosphere under pressure 1.01×105
Pa to ≥623 K is accompanied by the formation of silver metal. Figure 4 shows as an example the
variation in the XRD plot of Ag2S nanopowder with particle size ~45 nm after heating in argon
atmosphere to 623 K at a rate of 10 K·min-1. According to the XRD data, both Ag2S powders con-
tain monoclinic (space group P21/c) acanthite α-Ag2S as basic phase (Fig. 4). The average size D
of the coherent scattering regions of nanopowders was estimated from the broadening of non-
overlapping diffraction reflections (1 1 0), (-1 1 3), (-1 0 4) and (0 3 1). As a result of heating, the
size of powder particles increased from ~45 to ~147 nm, and about 3.5 wt.% silver metal appeared
in the powder.
The reason of isolation of metallic silver during rapid heating of the synthesized silver sul-
fide nanopowder in vacuum or argon is oxidation of some part of sulfur in the sulfide by oxygen or
moisture adsorbed by the nanopowder surface, removal of sulfur in the form of gaseous SO2 and,
as a consequence, liberation of some amount of silver.
Scanning electron microscopy of the synthesized Ag2S nanopowder (Fig. S1, see Supple-
mentary material) and the sample pressed from the same powder and annealed for 2 h in vacuum at
773 K (Fig. 5) confirms the variation in the phase composition as a result of isolation of metallic
silver whiskers. Figure S1 (see Supplementary material) demonstrates a SEM image of the initial
Ag2S nanopowder and the size distribution of nanoparticles in this nanopowder. The average na-
noparticle size is ~45 nm. After heating to 773 K, numerous silver metal whiskers 1-2 ꢀm in diam-
eter and 1 mm and more in length appeared on the surface of the pressed Ag2S nanopowder sample
(Fig. 5a). Energy dispersive X-ray analysis of the surface of compact sintered sample at point
7

ACCEPTED MANUSCRIPT
(001) has shown that the content of silver Ag and sulfur S is ~96.5±0.4 and ~3.5±0.1 wt.%, respec-
tively (Fig. 5b). The content of silver, which is almost 10 wt.% larger than in silver sulfide Ag2S,
confirms the decomposition of silver sulfide. In some places, the whiskers form splices (aggre-
gates) to 40 ꢀm in thickness (Fig. 5c). The isolation of silver metal on the surface of compact sam-
ple sintered at 773 K is observed visually (Fig. 6).
In order to avoid, during temperature variations, the loss of a part of sulfur caused by its
oxidation by adsorbed moisture and oxygen impurities, silver sulfide nanopowders were prelimi-
narily annealed in vacuum 0.013 Pa (10-4 mm Hg). Slow heating in vacuum to the annealing tem-
perature 393 K was performed for 2 h at a rate of 0.5 K·min-1, then the nanopowders were an-
nealed for 2 h at 393 K in vacuum. Heating of annealed silver sulfide nanopowders did not lead to
the isolation of metallic silver. According to differential thermal analysis data, when annealed
Ag2S nanopowders were heated there is no loss of mass of these nanopowders.
Ag2S nanopowder with the average particle size ~45 nm was also studied with the use of
Raman spectroscopy. Raman spectroscopy is a qualitative nondestructive analytical method for the
identification of different silver compounds, whose vibrational frequencies are in the near IR
range. Interfering impurities, such as H2O and CO2, have no effect on the Raman spectroscopy re-
sults.
Raman spectroscopy of Ag2S at the maximal exciting radiation power of 40 mW provided
high intensity, but led simultaneously to photoinduced decomposition of silver sulfide, which man-
ifested itself in the appearance of a very sharp peak at 1437 cm-1 and strong peaks at ~490 and
~1250 cm-1 (Fig. S2b, see Supplementary material). According to [38, 39], the observed strong
peaks at ~490, ~1250 and 1437 cm-1 correspond to vibrations of metallic silver Ag and sulfur ox-
ide compounds that are isolated during photoinduced decomposition of Ag2S in air. Indeed, the
observation of Ag2S powder by means of optical microscope after irradiation and registration of
Raman spectra showed that the sample became black in the irradiation region (Fig. S2a, Supple-
mentary material). Even the registration of spectra at a decreased laser radiation power of 4 mW
(10 % of the maximal power) and with a small number of scans did not allow us to eliminate Ag2S
decomposition completely and to get rid of decomposition-related peaks at ~490, ~1250 and
1437 cm-1.
At the background of strong peaks due to silver sulfide photodecomposition, the intensity
of bands related immediately to Ag2S is small.
The survey Raman spectrum of Ag2S nanopowder recorded at exciting radiation power of
10 mW (25 % of the maximal power) and by only two scans of the sample is displayed in Fig. 7a.
8

ACCEPTED MANUSCRIPT
It is similar to the spectrum of fine silver sulfide powder with monoclinic acanthite structure re-
ported in work [38]. The Raman spectrum of
260 cm-1 is smeared and broadened. For more detailed analysis, Figs. 7b and 7c show the spectra
of -Ag2S nanopowder in the low-frequency region recorded by us in work [40] at 10-fold scan-
α-Ag2S nanopowder in the low-frequency range 210-
α
ning of the sample and radiation power of 10 and 4 mW, respectively.
The spectra (Figs. 7b, 7c) contain a series of bands in the interval 90-260 cm-1. Besides the
bands at 93 and 147 cm-1, related to silver sublattice vibrations [38], other bands are observed,
which are due to Ag–S bonds and symmetric Ag–S–Ag longitudinal modes. This agrees with the
literature data [39, 41, 42], according to which the Raman spectrum of acanthite
terized by intense bands at ~93, ~122, 167-188, 225-253 and ~430 cm-1. According to [41-43], the
peak at ~430 cm-1 corresponds directly to low-temperature monoclinic acanthite
-Ag2S. The
α-Ag2S is charac-
α
presence of a broad band in the region 210-260 cm-1 centered at ~240 cm-1 agrees also with the
data of work [44], in which the peak induced by Ag2S nanoparticles is observed at ~220 cm-1. In
the structure of α-Ag2S acanthite, the triatomic Ag-S-Ag chains are non-linear therefore the Ra-
man spectrum bands are related to the symmetric Ag-S-Ag longitudinal vibrational modes. In
study [44], Ag2S nanoparticles were modeled by a set of (AgS)n clusters with n = 1-9 having the
minimal energy. Model clusters (AgS)n were formed by different number n of Ag-S pair bonds.
The simulation with the use of the Density Functional Theory approximation revealed that the
bands observed in the Raman spectrum of Ag2S nanoparticles in the region of ~250 cm-1 are main-
ly due to the radial breathing modes of (AgS)7 cluster, as well as (AgS)6 and (AgS)8 clusters
(Fig. 8) [44]. According to [43, 45], the fundamental modes that are observed at ~220 and ~430
cm-1 correspond to the first and second order longitudinal optical phonon modes in Ag2S, respec-
tively.
On the whole, Raman scattering confirmed that synthesized silver sulfide nanopowder has
a monoclinic structure of the α-Ag2S acanthite type.
3.2. The thermal stability of Ag2S nanoparticle size
The effect of temperature on the nanoparticle size was examined on preliminarily annealed
silver sulfide nanopowders that did not contain any adsorbed moisture or oxygen impurities.
The in situ XRD patterns of annealed nanocrystalline silver sulfide, recorded at 295, 323,
348, 373 and 398 K in air (Fig. 9), contain an identical set of broadened diffraction reflections cor-
responding to nonstoichiometric monoclinic (space group P21/c) acanthite α-Ag1.93-1.98S [4]. The
average size D of coherent scattering regions in initial powder is equal to 43±6 nm, and it agrees
9

ACCEPTED MANUSCRIPT
with the size of nanoparticles according to the SEM data. Gradual heating of nanopowder to 398 K
did not lead, within the measurement error, to any noticeable narrowing of diffraction reflections,
i. e. it was not accompanied by an increase in the nanoparticle size, although the resolution of dif-
fraction reflections in the region of 2θ < 28° weakened. The average CSR size in acanthite na-
nopowder was estimated is to be 46±6 nm at 398 K.
The silver sulfide nanopowder was additionally annealed in vacuum 0.013 Pa at the anneal-
ing temperature Tann equal to 393, 423, 453 and 493 K. Besides, a sample was pressed from silver
sulfide nanopowder, which was annealed in vacuum at 930 K. Heating to the annealing tempera-
ture Tann was carried out for 2 h. The XRD plots of silver sulfide nanopowder, measured after vac-
uum annealing at 393, 423, 453, 493 and 930 K (Fig. S3, see Supplementary material), contain the
diffraction reflections corresponding to nonstoichiometric monoclinic (space group P21/c) acan-
thite
α
-Ag1.95-1.98S.
With allowance for errors of D size estimation, two-hour annealing at a temperature to
423 K did not lead to any size gain. A weak growth of the CSR size was observed only for the na-
nopowders annealed at 453 and 493 K (Fig. S3, Supplementary material). An appreciable increase
in the CSR size to 260 nm was observed for the sample annealed at 930 K (Fig. S3, Supplementary
material). The value of microstrains, which is ~0.1 % in the initial nanopowder, decreased to
~0.03-0.05 % as a result of annealing.
In the nanopowder annealed in argon atmosphere at 623 K for 2 h, the CSR size of mono-
clinic acanthite α-Ag2S was 147±10 nm (see Fig. 4).
The variation in the Ag2S nanoparticle size vs the temperature of vacuum annealing is
shown in Fig. 10. At Tann < 440 K, the nanoparticle size hardly changes with increasing Tann with
consideration for the measurement errors, therefore the temperature interval 273-443 K can be
treated as the region of thermal stability of Ag2S nanoparticle size. The observed nanoparticle
growth takes place at Tann > 455 K, although even in the temperature range 460-930 K silver sul-
fide particles remain nanosized and do not reach the micrometer dimensions. The temperature
range 460-930 K, in which the particle size begins to grow appreciably and increases 3-6 times,
corresponds to the region of collective recrystallization of silver sulfide nanopowder. Note that the
transition from the region of thermal stability of Ag2S nanoparticle size to the region of collective
recrystallization occurs at a temperature close to the temperature of the phase transformation
“acanthite α-Ag2S – argentite β-Ag2S” equal to ~450 K [23, 26]. The scanning electron microsco-
py of Ag2S samples annealed at 623 and 930 K (Fig. 10, insets on the right) confirmed the Ag2S
particles size found from XRD data.
10

ACCEPTED MANUSCRIPT
The variation in the average particle (grain) size during collective recrystallization in na-
nomaterials is related to its duration t by the general ratio (Dn − D0n ) ~ t , where n = 1-4 [29, 46].
Grain growth (collective recrystallization) takes place as a result of chemical diffusion, when the
concentration gradient is absent, but there is a non-zero gradient of chemical potential. The differ-
ence in the chemical potential between particles of different size is due to the contribution from the
surface energy to the total energy of particle. The smaller is the particle size, the larger is the rela-
tive value of this contribution, therefore, under otherwise equal conditions, small particles possess
excess energy as compared with large particles. Hence it is clear that the thermodynamic reason of
spontaneous particle growth in polycrystalline substance is reduction of the total energy of the sys-
tem in this process. If the particle growth is due only to a decrease in the total surface energy of the
system as a result of diffusion, then the variation in the particle size D in the process of recrystalli-
zation is described by the expression
D2 − D02 = kt
,
(3)
where D0 is the initial particle (grain) size, t is the duration of annealing, and k is a coefficient ex-
ponentially depending on temperature.
However, recrystallization is a more complex process than diffusion and it does not reduce
to the latter. Collective recrystallization is characterized by the activation energy Q. There is no
available information in the literature about the activation energy of collective recrystallization in
sulfides, especially nanocrystalline ones. In works [47, 48] it was shown by the example of urani-
um dioxide, aluminum nitride, titanium boride and other nanostructured substances and materials
that the variation in the grain size D in the process of collective recrystallization is described by
the expression
D3 − D03 = k0t exp(−Q / RT)
where k0 is a normalizing factor. At equal annealing duration t, expression (4) can be written as
D3 − D03 = kt=const exp(−Q / RT)
,
(4)
.
(5)
From Eq. (5) it follows that the larger is the activation energy, the smaller is the particle size, other
factors being equal.
The dependence between the silver sulfide nanoparticle size D and the annealing tempera-
ture, presented in Fig. 10, can be described in the first approximation by function (5). In accord-
ance with the experimental data, the value of D0 was assumed to be 45 nm. Quantitative minimiza-
tion of the D(T) dependence by function (5) with account of D0= 45 nm [29] revealed that the acti-
11

ACCEPTED MANUSCRIPT
vation energy Q of collective recrystallization for nanostructured silver sulfide in the temperature
range 450-900 K is equal to ~33.6 kJ mol-1 (~0.35 eV per Ag2S formula unit or ~0.12 eV atom-1).
Note that the thermal stability of Ag2S nanoparticle size is smaller than that for PbS lead sulfide
nanoparticles [25]. The growth of Ag2S nanoparticles begins at ~460 K, whereas PbS nanoparti-
cles begin to grow at ~700 K. Apparently, the smaller region of the thermal stability of Ag2S na-
noparticle size is due to a smaller melting temperature of silver sulfide (Tmelt for bulk Ag2S and
PbS is 1115 and 1391 K, respectively). Another reason for the smaller thermal stability of Ag2S
nanoparticles can be the presence of a phase transformation in silver sulfide at ~450 K; there are
no polymorphic phase transitions in lead sulfide. It can be assumed that the transition of Ag2S to
the region of collective recrystallization is associated with a sharp spasmodic jump in the mobility
of silver atoms in the transformation of semiconductor acanthite to superionic argentite.
4. Conclusion
Heating of nanocrystalline silver sulfide powder in air or in oxidizing media (moisture and
oxygen adsorbed by the nanopowder surface) leads to its decomposition with isolation of silver
metal. The reduction of silver sulfide particles results in a decrease of the decomposition tempera-
ture and in the appearance of metallic silver at lower temperature.
The Raman studies confirmed the decomposition of Ag2S nanopowder with isolation of
metallic silver as a result of heating by high-power laser radiation.
The monoclinic α-Ag2S acanthite-type structure of the synthesized silver sulfide nanopow-
der was confirmed by Raman scattering. The observation of multiple resonant Raman peaks indi-
cated that the yielded Ag2S nanoparticles have good optical properties and crystallinity.
The absence of oxidizing impurities (adsorbed moisture and oxygen) in silver sulfide na-
nopowder provides a stable phase composition and the absence of decomposition during heating in
vacuum or in protective atmosphere of inert gas.
The elevation of the annealing temperature of nanocrystalline Ag2S powder from 300 K up
to ~450 K leads to an insignificant growth of nanoparticles and to annealing of microstrains, which
allows one to consider this temperature range to be a region of thermal size stability of nanocrys-
talline silver sulfide powder. It means that the electronic devices based on nanostructured silver
sulfide may steadily operate in this temperature interval. The temperature interval from 450 to 900
K, where the particle size increases 3-6 times, corresponds to the region of collective recrystalliza-
tion of silver sulfide nanopowder.
12

ACCEPTED MANUSCRIPT
Acknowledgements
This study was supported by the Russian Science Foundation (grant no. 14-23-00025)
through the Institute of Solid State Chemistry of the Ural Branch of the RAS. Authors are grateful
to Dr. A.P. Tyutyunnik for the help in high-temperature XRD study and to Prof. A.I. Gusev for
useful discussion.
Appendix A. Supplementary data
Supplementary data related to this article can be found at …
13

ACCEPTED MANUSCRIPT
References
[1] L. Argueta-Figueroa, O. Martínez-Alvarez, J. Santos-Cruz, R. Garcia-Contreras, L.S. Acosta-
Torres, J. de la Fuente-Hernández, M.C. Arenas-Arrocena, Mater. Sci. Eng. C. Mater Biolog.
Appl. 76 (2017) 1305-1315.
[2] P. Kulkarni, S.K. Nataraj, R.G. Balakrishna, D.H. Nagaraju, M.V. Reddy, J. Mater. Chem. A 5
(2017) 22040-22094.
[3] Y. Xiao, S.H. Lee, Y.-K. Sun, Advanc. Energy Mater. 7 (2016) paper 1601329 (20 pp).
[4] A. Tang, Yu. Wang, H. Ye H, C. Zhou, C. Yang, X. Li, H. Peng, F. Zhang, Y. Hou, F. Teng,
Nanotechnology 24 (2013) paper 355602 (12 pp).
[5] S.I. Sadovnikov, A.I. Gusev, J. Mater. Chem. A 5 (2017) 17676-17704.
[6] S.I. Sadovnikov, A.A. Rempel, A.I. Gusev, Nanostructured Lead, Cadmium and Silver Sul-
fides: Structure, Nonstoichiometry and Properties, Springer Intern. Publ. AG, Cham - Heidel-
berg, 2018. 317 pp.
[7] W. Zhang, L. Zhang, Z. Hui, X. Zhang, Y. Qian, Sol. State Ionics 130 (2000) 111-114.
[8] N. Du, H. Zhang, H.Z. Sun, D.R. Yang, Mater. Lett. 61 (2007) 235-238.
[9] L.H. Dong, Y. Chu, Y. Liu, J. Colloid. Interf. Sci. 317 (2008) 485-492.
[10] K. Terabe, T. Hasegawa, T. Nakayama, M. Aono, Nature 433 (2005) 47-50.
[11] C.H. Liang, K. Terabe, T. Hasegawa, V. Aono, Nanotechnology 18 (2007) paper 485202
(5pp).
[12] Z. Xu, Y. Bando, W. Wang, X. Bai, D. Golberg, ACS Nano 4 (2010) 2515-2522.
[13] D. Wang, L. Liu, Y. Kim, Z. Huang, D. Pantel, D. Hesse, M. Alexe, Appl. Phys. Lett. 98
(2011) paper 243109 (3 pp).
[14] S.I. Sadovnikov, A.I. Gusev, J. Nanopart. Res. 18 (2016) paper 277 (12 pp).
[15] A.I. Gusev, S.I. Sadovnikov, Mater. Lett. 188 (2017) 351-354.
[16] S.I. Sadovnikov, E.A. Kozlova, E.Yu. Gerasimov, A.A. Rempel, Catal. Commun. 100 (2017)
178-182.
[17] M. Basu, R. Nazir, C. Mahala, P. Fageria, S. Chaudhary, S. Gangopadhyay, S. Pande, Lang-
muir 33 (2017) 3178-3186.
[18] S.I. Sadovnikov, E.A. Kozlova, E.Yu. Gerasimov, A.A. Rempel, A.I. Gusev, Intern. J. Hydro-
gen Energy 42 (2017) 25258-25266.
[19] Y. Zhang, Y. Liu, C. Li, X. Chen, Q. Wang, J. Phys. Chem. C 118 (2014) 4918-4923.
[20] S.I. Sadovnikov, Yu.V. Kuznetsova, A.A. Rempel, Nanostr. Nano-Object. 7 (2016) 81-91.
[21] S.I. Sadovnikov, A.I. Gusev, A.A. Rempel, Phys. Chem. Chem. Phys. 17 (2015) 20495-20501.
14

ACCEPTED MANUSCRIPT
[22] A.I. Gusev, S.I. Sadovnikov, A.V. Chukin, A.A. Rempel, Phys. Solid State 58 (2016) 251-257.
[23] S.I. Sadovnikov, A.I. Gusev, A.V. Chukin, A.A. Rempel, Phys. Chem. Chem. Phys. 18 (2016)
4617-4626.
[24] W. Kahle, H. Berger, Phys. stat. sol.(a) 2 (1970) 717-721.
[25] S.I. Sadovnikov, N.S. Kozhevnikova, A.A. Rempel, Inorg. Mater. 47 (2011) 837-843.
[26] S.I. Sadovnikov, A.I. Gusev, Phys. Solid State 59 (2017) 1887-1894.
[27] S.I. Sadovnikov, A.I. Gusev, J. Thermal Anal. Calorimetry 131 (2018) 1155-1164.
[28] A.I. Gusev, S.I. Sadovnikov, Thermochim. Acta 660 (2018) 1-10.
[29] S.I. Sadovnikov, A.I. Gusev, Phys. Solid State 60 (2018) 1308-1315.
[30] P. Patnaik, Dean’s Analytical Chemistry Handbook, 2nd ed., McGraw-Hill, New York, 2004.
Table 4.2. 1280 pp.
[31] S.I. Sadovnikov, A.I. Gusev, A.A. Rempel, Superlattices Microstruct. 83 (2015) 35-47.
[32] S.I. Sadovnikov, A.I. Gusev, A.A. Rempel, Phys. Chem. Chem. Phys. 17 (2015) 12466-12471.
[33] X’Pert HighScore Plus. Version 2.2e (2.2.5). 2009 PANalytical B. V. Almedo, the Nether-
lands.
[34] A.I. Gusev, A.A. Rempel, Nanocrystalline Materials, Cambridge Intern. Science Publ., Cam-
bridge, 2004. 351 pp.
[35] S. I. Sadovnikov, A. I. Gusev, J. Alloys Comp. 586 (2014) 105-112.
[36] T. Blanton, S. Misture, N. Dontula and S. Zdzieszynski, Powder Diffraction 26 (2011) 110-
118.
[37] J. Corish, C.D. O'Briain, J. Mater. Sci. 6 (1971) 252-259.
[38] I. Martina, R. Wiesinger, D. Jembrih-Simbürger, M. Schreiner, E-Preservation Science: Mo-
rana RTD 9 (2012) 1-8.
[39] J.I. Lee, S.M. Howard, J.J. Kellar, K.N. Han, W. Cross, Metall. Mater. Trans. B 32 (2001) 895-
901.
[40] S.I. Sadovnikov, E.G. Vovkotrub, A.A. Rempel, Doklady Phys. Chem. 480 (2018) 80-82.
[41] A.N. Belov, O.V. Pyatilova, M.I. Vorobiev, Advanc. Nanoparticles 3 (2014) 1-4.
[42] G. Socrates, Infrared and Raman characteristic group frequencies: Tables and charts. 3rd ed.
Wiley, Chichester - New York, 2001. 347 pp.
[43] M. Osada, K. Terabe, C. Liang, T. Hasegawa, In: 214th ECS Meeting Abstracts MA 2008-2
(214th ECS Metting, Honolulu, October 12-17, 2008). Honolulu: ECS, 2008. B10 - Solid State
Ionic Devices 6 - Nano Ionics, 2008. Abstract 1406.
15

ACCEPTED MANUSCRIPT
[44] Y. Delgado-Beleño, M. Cortez-Valadez, C.E. Martinez-Nuñez, R. Britto Hurtado, A.B. Alva-
rez Ramón, O. Rocha-Rocha, H. Arizpe-Chávez, A. Perez-Rodríguez, M. Flores-Acosta, Chem.
Physics 463 (2015) 106-110.
[45] L. Hashmi, P. Sana, M.M. Malik, A.H. Siddiqui, M.S. Qureshi, Nano Hybrides 1 (2012) 23-43.
[46] Thesaurus of nanotechnological terms and nanotechnology-related terms / Ed. S.V. Kal-
yuzhnyi. Fizmatlit, Moscow, 2010. 528 pp. (in Russian)
[47] F.A. Nichols, J. Appl. Phys. 37 (1966) 4599-4602.
[48] Functional Gradient Materials and Surface Layers Prepared by Fine Particles Technology /
Eds. M.-I. Baraton, I.V. Uvarova. Kluwer, Dordrecht, 2001. 317 pp.
16

ACCEPTED MANUSCRIPT
Figure captions for the paper
“Thermal stability of nanoparticle size and phase composition of nanostructured Ag2S
silver sulfide”
by S.I. Sadovnikov and E.G. Vovkotrub
Fig. 1. The TEM image of Ag2S nanoparticle with a size about 40 nm of synthesized nanostruc-
tured silver sulfide powder.
Fig. 2. The film strip plot of in situ high-temperature XRD data for Ag S powder with particle size
2
about 85 nm recorded in Debye-Scherrer geometry on air at a temperature from 298 to 773 K.
Fig. 3. Evolution of XRD patterns of silver sulfide powder with the average particle size of ~85 nm
during heating from 298 to 773 K in air. The yellow lines indicate the position of the main reflec-
tions of monoclinic acanthite (at 298-453 K) and cubic argentite (at 463-623 K). The intensity is
shown in logarithmic scale.
Fig. 4. XRD patterns of synthesized silver sulfide nanopowder with the average nanoparticle size
of ~45 nm and the same powder after heating from room temperature to 623 K in argon with the
heating rate 10 K·min-1.
Fig. 5. The SEM image of the surface of sample pressed from Ag2S nanopowder and annealed for
2 h in vacuum at 773 K: (a) metallic silver whiskers 1-2 ꢀm in diameter and 1 mm and more in
length; (b) elemental EDX analysis of the sample surface at point (001) marked in Fig. 5a; (c)
splices (aggregates) of metallic silver whiskers.
Fig. 6. Silver metal isolated on the surface of sample pressed from Ag2S nanopowder after 2 h an-
nealing in vacuum at 773 K.
Fig. 7. Raman spectra of Ag2S nanopowder recorded at room temperature under different scanning
conditions: (a) the survey spectrum recorded at 2-fold scanning of the sample at radiation power
10 mW, (b) and (c) the spectra in the low-frequency region recorded at 10-fold scanning of the
sample with different radiation power, 10 and 4 mW, respectively.
Fig. 8. Most probable (AgS)n clusters (with n = 7-9) whose radial breathing modes determine pre-
dominately the bands observed in the Ag2S nanoparticles Raman spectrum near 250 cm-1.42
Fig. 9. In situ XRD patterns of nanocrystalline acanthite α-Ag2S during heating from 300 to 398 K
in air. The insets on the right show the average size of the coherent scattering regions estimated
from the broadening of non-overlapping diffraction reflections. The XRD patterns are recorded on
an X`Pert PRO MPD (Panalytical) diffractometer in Bragg-Brentano geometry.
Fig. 10. (a) The effect of the annealing temperature on the size D of nanoparticles of Ag2S powder
(the vertical dash-and-dot line denotes the boundary between the thermal stability region of silver
sulfide nanoparticle size and the region of collective recrystallization). The thermal stability region
in the bottom inset is scaled-up. The insets on the right show the SEM images of Ag2S samples
after vacuum annealing at (b) 623 and (c) 930 K.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Silver sulfide nanopowders are prepared by chemical deposition from aqueous solutions. ● Ther-
mal stability of Ag2S particle size is estimated by in situ high-temperature XRD method. ● Heat-
ing of Ag2S up to the temperature of 450 K does not lead to nanoparticle growth. ● The nanoparti-
cle size increases as a result of recrystallization at a temperature of 450-930 K. ● For nano-Ag2S,
the activation energy of collective recrystallization is 0.12 eV atom-1.
1