Thermodynamic properties of Ag3AuSe2 from 300 to 500 K by a solid state galvanic cell

Journal of Alloys and Compounds 583 (2014) 176–179

Contents lists available at ScienceDirect

Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jalcom

Thermodynamic properties of Ag3AuSe2 from 300 to 500 K by a solid state galvanic cell Dawei Feng ⇑, Pekka Taskinen Department of Materials Science and Engineering, Aalto University, FI-00076 Aalto, Finland

a r t i c l e

i n f o

a b s t r a c t

Article history: Received 16 July 2013 Received in revised form 23 August 2013 Accepted 26 August 2013 Available online 5 September 2013

The numerical values of the standard thermodynamic functions of Ag3AuSe2 (fischesserite) in the Ag–Au– Se system were determined by the electromotive force (EMF) method in a solid state galvanic cell with RbAg4I5 as the solid electrolyte below 500 K. Ag3AuSe2 was synthesized from pure elements in stoichiometric composition by evacuated ampoules made of quartz glass. On the basis of EMF vs. temperature experimental data, the analytical equations were obtained, from which the temperature dependent relationships of the Gibbs energy in the relevant reactions and the standard thermodynamic functions of Ag3AuSe2 within the temperature range of 300–500 K were calculated. The results show that there is possible a new polymorphic transformation around 350 K. Ó 2013 Elsevier B.V. All rights reserved.

Keywords: Silver chalcogenides Electromotive force (EMF) method Experimental thermodynamics

1. Introduction Large magneto-resistance has been found in silver chalcogenides [1], with the promising application as magnetic field sensors [2]. The Ag3AuSe2 is one of the ternary silver gold chalcogenides and exists in nature as fischesserite discovered by Johan et al. [3]. By electron microprobe and physical characterization, they deduced fischesserite is the selenium analogue of petzite, Ag3AuTe2. More recently, the structural and physical properties for fischesserite has been investigated by Bindi et al. [4] from the De Lamar mine, Owyhee County, Idaho, USA. They confirmed that fischesserite is topologically identical to petzite, apart from the slight deviations from the size of the anions. The electronic and ionic conduction of the pseudo-binary system Ag2xAuxSe (0 6 x 6 0.5) was first investigated by Wiegers [5] through differential thermal analysis (DTA) and Guinier-Lenne High Temperature X-ray Camera. Two ordered compounds a-Ag2Se and a-Ag3AuSe2 were found at low temperatures, while he proposed a solid solution with the b-Ag2Se existing over the whole composition range at higher temperatures. Sakai et al. [6] and Wagner et al. [7] measured the ternary silver gold chalcogenides by Mössbauer spectroscopy and suggested that the gold in the compound should be considered as monovalent. Fang et al. [8] made band structure calculation for the low temperature modification of the silver chalcogenide a-Ag3AuSe2 and the results show that a-Ag3AuSe2 is a semiconductor with an energy gap of about 0.2 eV. Tavernier et al. [9] investigated the Ag3AuSe2 by DTA and found the phase transition temperature to be 543 K. Later, Smit et al. [10] ⇑ Corresponding author. Tel.: +358 50 4602751; fax: +358 9470 22798. E-mail addresses: (P. Taskinen).

dawei.feng@aalto.fi

(D.

Feng),

pekka.taskinen@aalto.fi

0925-8388/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jallcom.2013.08.170

studied by X-ray diffraction (XRD) and DTA and found the phase transition temperature from the low temperature a-Ag3AuSe2 (fischesserite) to high temperature b-Ag3AuSe2 (body centered cubic) to be 540 K, and they found the melting point of Ag3AuSe2 to be around 1000 K. Prince [11] reviewed the Ag–Au–Se system in 1988. However, the thermodynamic properties of Ag3AuSe2 at the low temperatures remain poorly studied. Recently, Osadchii et al. [12] investigated the Ag–Au–Se system below 405 K. Nevertheless, enlarged temperature range can prove the results and may bring us better understanding of the thermodynamic behavior on Ag3AuSe2, owning to the fact that RbAg4I5 melts until 500 K [13]. The objectives of the present study are to determine and evaluate the standard thermodynamic properties of Ag3AuSe2 at low temperature. DTA measurement has been extensively used in thermodynamic investigations, but the relatively fast cooling or heating rate makes it difficult to reach thermodynamic equilibria. The DSC is more sensitive but also shares the same problem. The EMF method, however, operates at constant temperature. Thus it takes enough time to reach equilibria, and EMF values can be observed online to be constant corresponding to the operation temperature. Therefore, the EMF method where Gibbs energy can be directly obtained has been applied in this study.

2. Methods 2.1. Materials Ag3AuSe2 was prepared by a direct synthesis from elements in stoichiometric composition using evacuated quartz glass ampoules. Gold powder (99.9%, Alfa Aesar, Germany, 200 mesh), silver (99.99%, Alfa Aesar, Germany) and selenium powder (99.999%, Koch-Light Laboratories Ltd., UK) were used in synthesis of the intermetallic compounds.

177

D. Feng, P. Taskinen / Journal of Alloys and Compounds 583 (2014) 176–179

Se

Ag3AuSe2

AuSe

Ag2Se

Ag

Au

Fig. 1. Phase relations in the Ag–Au–Se system at 298–500 K with the working electrode shown in the symbol I. The synthesis was carried out in a vertical resistance tube furnace. Firstly sealed ampoule was heated up from room temperature to 673 K and kept for 3–4 days to ensure reaction of elements; afterwards, the temperature was raised to 1273 K and then annealing continued at 673 K for 1–2 days. Finally the ampoule was cooled down slowly to room temperature in the furnace over one week and unsealed. The compounds were characterized to be fischesserite phase by Panalytical X’Pert Pro MRD (Philips, Netherland) using Co Ka radiation and examined to be homogenous by SEM (LEO 1450), and EDS (Oxford instruments, UK). The synthesized Ag3AuSe2 was ground in an agate mortar with Ag2Se and Au in a mole ratio 1:0.75:0.325 and then pressed under pressure of 0.1 GPa thus to obtain a pellet of 20 mm in diameter and 2 mm in thickness. Ag2Se was synthesized as in our previous experiment [14]. The ternary equilibrium relations of the Ag–Au–Se system in the cell are shown in Fig. 1. The solid electrolyte, RbAg4I5, was synthesized according to the method described by Owens and Argue [13]. Weighing of 0.8 mol fraction of silver iodide from Alfa Aesar (99.9% in purity, Germany) and 0.2 mol fraction of rubidium iodide from Alfa Aesar (99.8% in purity, Germany) was followed by mixing. The mixture was sealed in a glass tube under vacuum, and heated at 220 °C for 2 h. Then after being cooled down and maintained at 160 °C for 15 h, the solid state electrolyte RbAg4I5 was obtained. In addition, platinum (99.9%, Johnson-Matthey, UK) was used as lead wire in the experiment.

attached permanently to a quick-coupling head (10). The platinum resistance thermometer (7) and the platinum lead (8) come out through insulated holes at the quick-coupling head (10). All these holes are sealed with epoxy resin to make a gas tight assembly. The quick-coupling heads are provided with side tubes (11) to maintain an inert gas flow through the cell. All the above parts are located inside a Lenton tube furnace (11). During the EMF-measurements, temperatures on each end of the galvanic cell were measured using two PT100 sensors (platinum resistance thermometers, PRT). The PRTs, with tolerance class B 1/10 DIN, i.e. ±0.01 K variation at 0 °C, were calibrated in a mixture of ice and water at 273.15 K. The obtained resistance values above 100 O (R0 = 100.026 O and R0 = 100.03 O, respectively) were inserted to a program based on a LabVIEW software code from National Instruments (US). The platinum wires for the EMF-measurements and the lead wires from the PT100 sensors for temperature measurements on each end were connected to a KEITHLEY6517B electrometer and two KEITHLEY-2000 multi-meters (KEITHLEY, US), respectively. Input impedance of the electrometer for the EMF-measurements was 2  1014 X, which allows the cell reaction to be reversible. The measured EMF-values and temperatures were simultaneously transferred to a computer through IEE-488GPIB cable and KEITHLEY-KUSB-488A USB-to-GPIB interface adapter, and the readings were recorded by software storing two measured values every 5 s. The accuracy of temperature and EMF readings were 0.0001 K and 0.001 mV, respectively. Gas flow of argon (99.999% in purity, AGA, Finland) to the EMF cell was purified before introduction to the cell, by passing though an auxiliary furnace with titanium wire at 973 K for removing the oxygen traces. Owing to these improvements, the accuracy of EMF measurements and its stability as a function of time are very good, allowing very long measuring campaigns for each experimental cell. Most measurements were performed by heating and cooling the cell in steps at 10 K. To reach a steady state EMF reading, it took from few hours up to 2 weeks. The equilibria were considered to be reached when the EMF-values were constant, or their variations were not significant (<0.01 mV within 10 h). Temperature differences between the two electrodes of the EMF cell were controlled to be less than 1 K, by manually adjusting the horizontal position of the galvanic cells and observing the real-time temperature readings over the cell owning to high accuracy PRTs. The uncertainties of temperature and EMF were established to be ±0.5 K and ±0.01 mV, respectively. Thus, the possible thermoelectric effect generated to the cell EMF by the temperature gradients is regarded to be negligible. Functionality of the experimental electrochemical system was tested by measuring the EMF of the symmetric cell Ag|Ag+|Ag, which theoretically should not result in any measurable EMF or electric potential difference. The equilibriums in this study were considered reproducible when the EMF readings in heating and cooling coincided. 2.3. EMF cell The measurements were performed on the solid state galvanic cell when silver selenide and gold is formed from solid pure silver and Ag3AuSe2. The virtual reaction of the electrochemical cell is

2.2. Apparatus The experimental setup for the e.m.f. measurements consists essentially of the following module: (I) constant temperature profile; (II) cell assembly; (III) gas purification system; (V) measurement system for temperature and voltage; (VI) data recording system and (VII) bubble system for gas tight. The details are briefly shown in Fig. 2. The reference electrode (1), electrolyte (2) and the working electrode (3) form a ‘sandwich’ that rests on an alumina crucible (4). The alumina crucible (4) with a central hole to pass the platinum lead (8) from the reference electrode (1) could be placed in the center of an alumina tube (5) as a holder for the galvanic cell. A thinner alumina crucible is placed on top of the working electrode (3) as a pressure block (6). On the pressure block (6), stands a moveable platinum resistance thermometer (7) inside an alumina sheath (13). The loading springs (9) exert enough pressure to ensure good electrical contact between the electrolyte and the electrodes. The platinum lead (8) comes out through a small hole near the flat end of the central alumina tube (5). The open end of the alumina tube (5) is

AgðfccÞ þ Ag3 AuSe2 ¼ 2Ag2 Se þ AuðfccÞ; The EMF of the cell

ðÞPtjAgjRbAg4 I5 jAg3 AuSe2 ; Ag2 Se; Auj PtðþÞ;

1

10 9

3

ðBÞ

was measured in a temperature range of 298–500 K, under the ambient atmospheric pressure. The half cell reactions are

RE : Ag  e ¼ Ag

þ

ðCÞ

WE : Agþ þ e þ Ag3 AuSe2 ¼ 2Ag2 Se þ Au

ðDÞ

The Gibbs energy change of the cell reaction, except for the work of volume expansion, is related to the reversible EMF of reaction (A) by the Nernst equation

5

12 8

ðAÞ

2

6

4

11

13

7

Fig. 2. Schematic illustration of set up, (1) reference electrode (RE), (2) electrolyte, (3) working electrode (WE), (4) alumina crucible, (5) alumina tube, (6) pressure block, (7) moveable platinum resistance thermometer, (8) platinum lead, (9) loading springs, (10) quick-coupling head, (11) inert gas inlet/out, (12) Lenton tube furnace and (13) alumina sheath.

178

D. Feng, P. Taskinen / Journal of Alloys and Compounds 583 (2014) 176–179

Dr G ¼ zEF

ð1Þ

Dr G1 ; J=mol ¼ 8798:757ð175:892Þ  11:192ð0:579ÞT=K; ð298 < T=K < 350Þ

ð7Þ

Dr G2 ; J=mol ¼ 3440:945ð588:559Þ  26:533ð1:544ÞT=K; ð350 < T=K < 407Þ

ð8Þ

ð2Þ

Dr G3 ; J=mol ¼ 7144:91ð426:657Þ  52:488ð0:965ÞT=K; ð408 < T=K < 488Þ

ð9Þ

ð3Þ

Thereby the standard Gibbs energy of reaction of the virtual cell reaction at 298.15 K and 1 bar can be obtained from Eq. (7), as at low temperature Au, Au2Se and Ag3AuSe2 does not have any solubility and they are pure substances.

where E is the electromotive force produced by the cell, F is the Faraday constant (96 485 C  mol1) and z is the number of elementary changes transferred, z = 1. The other fundamental thermodynamic properties of Ag3AuSe2 can be derived from Gibbs energy and its functional relationships with temperature in isobaric conditions:

Dr S ¼ zF



dE dT

 P

    dE T Dr H ¼ zF E  dT P

The lower temperature limit was determined by the rate of reaction and corresponded to the minimum temperature of obtaining stable and reproducible EMF values in the cell. Cooling and heating cycle with the same electrode compositions were measured in this study to confirm the reproducibility of the obtained data.

3. Result and discussion The observed EMF values as a function of temperature in this study are shown in Fig. 3 where temperature is the mean value of both the electrodes and E is the final stable value at equilibria. The only well established phase transformation for Ag2Se occurs at around 407 K as shown in previous study [14]. However, there is apparent slope change at around 350 K, thus three ranges are formed. On basis of the knowledge of the phases involved in the reaction and the linear relationship E = a + b  T, the temperature ranges were determined by the linear analysis with the least square fitting. The analytical equations were obtained as:

E1 ; mV ¼ 91:193ð1:823Þ þ 0:116ð0:006ÞT=K; ð298 < T=K < 350Þ;

R2 ¼ 0:979

ð4Þ

Dr G0 ðAÞ ¼ 12:136ð0:348Þ kJ K1 mol

R ¼ 0:952

Dr S2 ; J K

1

ð408 < T=K < 488Þ;

R ¼ 0:993

¼ 11:192ð0:579Þ; ð298 < T=K < 350Þ

ð11Þ

mol

1

Dr S3 ; J K

1

¼ 26:533ð1:544Þ; ð350 < T=K < 407Þ

ð12Þ

mol

1

¼ 52:488ð0:965Þ; ð408 < T=K < 488Þ

ð13Þ

Dr H1 ; J=mol ¼ 8798:757ð175:892Þ; ð298 < T=K < 350Þ

ð14Þ

Dr H2 ; J=mol ¼ 3440:945ð588:559Þ; ð350 < T=K < 407Þ

ð15Þ

Dr H3 ; J=mol ¼ 7144:907ð426:657Þ; ð408 < T=K < 488Þ

ð16Þ

The standard Gibbs energy and enthalpy of formation and standard entropy of Ag3AuSe2 at 298.15 K and 1 bar were calculated from the experimental data as following: 



Df G ða  Ag3 AuSe;2 Þ ¼ 2f G ða  Ag2 SeÞ  Dr G 



ð17Þ 









ð18Þ

ð5Þ

Df H ða  Ag3 AuSe2 Þ ¼ 2f H ða  Ag2 SeÞ  Dr H

ð6Þ

where the entropies for the elements Ag and Au equal to 42.55 and 47.4884 kJ mol1 [15], respectively. The Gibbs energy, entropy and enthalpy of formation of Ag2Se were measured in our previous paper [14] as:

E3 ; mV ¼ 74:052ð4:422Þ þ 0:544ð0:010ÞT=K; 2

1

Dr S1 ; J K1 mol

S ða  Ag3 AuSe2 Þ ¼ 2S ða  Ag2 SeÞ þ S ðAuÞ  Dr S  S ðAgÞ

E2 ; mV ¼ 35:663ð6:100Þ þ 0:275ð0:016ÞT=K; ð350 < T=K < 407Þ;

ð10Þ

As the EMF of the cell was measured as a function of temperature, the molar entropies and enthalpies for the cell reactions corresponding to the two polymorphic forms of Ag2Se and Ag3AuSe2 can be derived by Eqs. (2) and (3), as following:



2

1

The obtained Gibbs energy changes of reaction (A), according to Eq. (1), representing the two polymorphic forms of Ag3AuSe2 over their stability rages in these measurements, as well as over the polymorphic transformation of Ag2Se, are given by equations:



1

Df GaAg2 Se ¼ ð49:24  0:46Þ kJ mol 

ð19Þ

ð20Þ

1

SaAg2 Se ¼ ð154:60  0:22Þ J  K1 mol

ð21Þ



1

Df HaAg2 Se ¼ ð40:87  0:58Þ kJ  K1 mol

ð22Þ

Thereby we can obtain the standard Gibbs energy and enthalpy of formation and entropy for Ag3AuSe2 at 298.15 K from Eqs. (17)– (19) as: 

1

ð23Þ

1

ð24Þ

Df GaAg3 AuSe2 ¼ ð86:344  1:268Þ kJ mol 

SaAg3 AuSe2 ¼ ð302:946  1:019Þ J K1 mol 

Df HaAg3 AuSe2 ¼ ð72:941  1:336Þ kJ mol

Fig. 3. Temperature dependence of the EMF obtained in this experiment.

1

ð25Þ

Fig. 4 shows a comparison of the obtained EMF values with the observations of Osadchii et al. [12]. The EMF values show an excellent agreement with those obtained by Osadchii et al. [12] from 400 K to 453.1 K (the upper temperature limit for the cell in their work). Nevertheless, the high temperature data were not used for the calculation of thermodynamic functions in their study. Obviously, the change in the EMF vs. T slope around 400 K results from the phase transition of Ag2Se [14]. By solving the Eqs. (5) and (6), the phase transition temperature has been determined to be 407.86 K, which is very close to the phase transformation temperature of Ag2Se, 407.7 K from our previous work [14].

D. Feng, P. Taskinen / Journal of Alloys and Compounds 583 (2014) 176–179

179

4. Conclusion

Fig. 4. Variation of obtained EMF functions compared with the results from Osadchii et al. [12].

Table 1 Standard thermodynamic properties of Ag3AuSe2 at 298.15 K. Compound



Df G ; kJ mol

1

a-Ag3AuSe2 86.451 ± 0.32 a-Ag3AuSe2 86.344 ± 1.268



S ; J K1 mol

1



Df H ; kJ mol

1

290.80 ± 1.26 77.172 ± 0.611 302.946 ± 1.019 72.941 ± 1.336

Reference [12] This work

Non-negligible discrepancies, however, can be found at lower temperatures. Our data show a minor deviation from observations obtained by Osadchii and Echmaeva [12] by 3–4 mV from room temperature to 350 K. This could not be easily attributed to the sluggish reaction at low temperatures, since there is apparent slope change by carefully analysis at 350 K for the stable EMF values. The already known phase transformation from the low temperature a-Ag3AuSe2 (fischesserite) to high temperature b-Ag3AuSe2 (body centered cubic) was determined by Tavernier et al. [9] and Smit et al. [10] to be around 540 K by DTA. Based on the current results, we consider the possibility of an unknown phase transition of Ag3AuSe2 at low temperatures, according to the slopes of the experimental EMF vs. temperature plot. Solving the Eqs. (4) and (5), we determined the new polymorphic transition temperature of Ag3AuSe2 to be at 349.3 K, which shall be the new phase transition of Ag3AuSe2 other than the reported a-b transformation at 543 K by Smit et al. [10]. Here we call the phase a0 -Ag3AuSe2 to distinguish it from both the a and b phases. Thus, the enthalpy of the transformation of Ag3AuSe2, based on this work, is given as following:

Dtrs Hða  a0 Ag3 AuSe2 Þ; kJ=mol ¼ 5:358ð0:764Þ; ðT ¼ 349 KÞ

ð26Þ

The obtained values of the standard thermodynamic functions and their comparison with the data found by the other authors have been collected to Table 1.

In this study, new experimental data on the thermodynamic stability of Ag3AuSe2 (fischesserite) in equilibrium with Ag2Se, have been obtained over an enlarged temperature range of 300– 500 K by the electrochemical EMF method in a solid state galvanic cell, with RbAg4I5 as the solid electrolyte. The results obtained in this study essentially confirm the data provided by Osadchii et al. at high temperatures, above 400 K [12]. However, discrepancies are found at the low temperature range in this study compared with previous data. The current experimental data suggest that a phase transformation of Ag3AuSe2 has been deduced at 349.3 K. The temperature was obtained by solving the experimental EMF functions and taking into account the a–b solid state phase transformation of Ag2Se at 407.7 K. The enthalpy of the phase transformation based on the EMF vs. temperature plot is 5.538 ± 0.764 kJ  mol–1. The slope change of the temperature dependence gives thermodynamic evidence for a phase transformation. Further investigation, such as XRD characterization of the phase, will be carried out in the near future. The advanced temperature measurement arrangement of the electrochemical cell makes it possible to distinguish the polymorphic transformations of Ag3AuSe2 at low temperatures within a narrow temperature range. Acknowledgements The authors are grateful to Markus Aspiala for helping the set up, Marko Järvenpää for SEM and EDS analyze and Juha Larismaa for the XRD instruction. This work was carried out as a sub task of ISS project of Elemet Program (Fimecc Oy), supported financially also by Boliden Harjavalta, Boliden Kokkola, Norilsk Nickel Finland Oy and Outotec Finland Oy. References [1] R. Xu, A. Husmann, T. Rosenbaum, M. Saboungi, J. Enderby, P. Littlewood, Nature 390 (1997) 57–60. [2] A. Husmann, J. Betts, G. Boebinger, A. Migliori, T. Rosenbaum, M. Saboungi, Nature 417 (2002) 421–424. [3] Z. Johan, M. Kvacek, P. Picot, R. Pierrot, Bull. De La Soc. Francaise Miner. Et De Cristallogr. 94 (1971) 381–384. [4] L. Bindi, C. Cipriani, Can. Mineral. 42 (2004) 1733–1737. [5] G. Wiegers, J. Less Common Metals 48 (1976) 269–283. [6] H. Sakai, M. Ando, S. Ichiba, Y. Maeda, Chem. Lett. (1991) 223–226. [7] F. Wagner, J. Sawicki, J. Friedl, J. Mandarino, D. Harris, Can. Mineral. 30 (1992) 327–333. [8] C.M. Fang, R.A. de Groot, G.A. Wiegers, J. Phys. Chem. Solids 63 (2002) 457–464. [9] B.H. Tavernier, J. Vervecken, P. Messien, M. Baiwir, Zeitschrift für Anorganische und Allgemeine Chemie 356 (1967) 77–88. [10] T.J.M. Smit, E. Venema, J. Wiersma, G.A. Wiegers, J. Solid State Chem. 2 (1970) 309–312. [11] A. Prince, Silver–gold–selenium, in: G. Petzow (Ed.), Ternary Alloys: A Comprehensive Compendium of Evaluated Constitutional Data and Phase Diagrams. 1. Ag–Al–Au to Ag–Cu–P, VCH, Weinheim, 1988. [12] E.G. Osadchii, E.A. Echmaeva, Am. Mineral. 92 (2007) 640–647. [13] B.B. Owens, G.R. Argue, Science 157 (1967) 308–310. [14] D. Feng, P. Taskinen, F. Tesfaye, Solid State Ionics 231 (2013) 1–4. [15] A.T. Dinsdale, Calphad 15 (1991) 317–425.