Investigation of acoustoionic and acoustoelectronic interaction in fast ionic conductors

Available online at www.sciencedirect.com

Solid State Ionics 179 (2008) 120 – 125 www.elsevier.com/locate/ssi

Investigation of acoustoionic and acoustoelectronic interaction in fast ionic conductors V. Samulionis ⁎, V. Jonkus . Faculty of Physics, Vilnius University, Sauletekio 9/3, LT-10222 Vilnius, Lithuania

Abstract In our contribution we present the acoustoionic interaction study results in piezoelectric mixed ionic-electronic conductors AgI, Ag3SbS3 and Ag3AsS3. The “deformation potential” type of interaction was investigated in Rb4AgI5 crystals, and various polycrystalline and glassy solid electrolytes. The relaxational ultrasonic attenuation peaks and velocity dispersion were observed in temperature dependencies. The attenuation peak height was proportional to the electromechanical coupling constant. The activation energies of ionic conductivity were calculated from ultrasonic data. It was shown, that in DC electric field applied on solid electrolyte the ultrasonic attenuation changes substantially and such behaviour is determined by the increase of electric conductivity in the volume of the crystal. The acoustoelectric current caused by electronic charge carriers was observed in illuminated AgI, Ag3SbS3 and Ag3AsS3 crystals. The sign and mobility of free charge carriers were established from acoustoelectronic voltage measurements. Acoustoelectric current was observed also in these crystals in DC electric field and it was shown that the increase of concentration of free holes is responsible for this phenomenon. The ultrasonic behaviour near superionic phase transitions also is discussed. © 2008 Elsevier B.V. All rights reserved. Keywords: Ionic conductivity; Acoustoelectric interaction; Mixed ionic conductors

1. Introduction The acoustoelectronic interaction in piezoelectric semiconductors is well known [1]. It leads to photosensitive ultrasonic attenuation and acoustoelectric current in photo-sensitive conductors [2]. Acoustoionic interaction was observed in solid electrolytes and manifests itself as ionic conductivity dependent ultrasonic attenuation and velocity dispersion [3,4]. When ultrasonic wave propagates through medium with high concentration of mobile ions i.e. solid electrolyte its amplitude and velocity change because of acoustoionic interaction due to many mechanisms. These mechanisms can be divided into two major classes: (i) piezoelectric interaction in piezoelectric crystals, (ii) interaction caused by elastic wave modulation of chemical potential felt by mobile ions i.e. “deformation poten-

⁎ Corresponding author. E-mail address: [email protected] (V. Samulionis). 0167-2738/$ - see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ssi.2007.12.041

tial” type. The acoustoionic interaction of piezoelectric type was observed mostly in crystalline solid electrolytes such as AgI, Ag3SbS3, Ag3AsS3 [4,5]. This mechanism manifests itself as temperature dependent ultrasonic attenuation peaks and velocity dispersion. The attenuation peak height depends on electromechanical coupling factor of piezoelectric material. The acoustoionic interaction of second class, which can be defined as “deformation potential” type, arises in centrosymetric (nonpiezoelectric) materials such as superionic glasses [6] and also in polycrystalline solid electrolytes [7]. In this case the ultrasonic relaxational attenuation maximum (relaxation strength) depends on the square of deformation potential and also on density of mobile ions. Usually in fast ionic conductors density of mobile ions is high enough and the ultrasonic attenuation peaks and velocity dispersion are easy to reveal in experiment. Ultrasonic attenuation and velocity anomalies at phase transition (PT) in superionic conductors are caused by various mechanisms too. At first there are critical anomalies in the vicinity of phase transition in the rigid lattice showing slowing

V. Samulionis, V. Jonkus / Solid State Ionics 179 (2008) 120–125

Fig. 1. The temperature dependencies of longitudinal ultrasonic attenuation (1,2,3) and velocity (1′,2′, 3′) in pyragyrite crystals at different frequencies.

down of ultrasonic velocity and attenuation peak as in ferroelectrics and ferroelastics. Another type of anomaly can be related to finite ionic conductivity jump at the superionic transition point. At PT it is change of acoustoionic interaction, which was previously described. Also the relaxation time of mobile ions system changes rapidly at the PT. Thus, additional characteristic downward step in velocity and increase of attenuation must arise at the transition to superionic state. Such anomalies have been observed earlier near superionic phase transitions in Ag2HgI4 [8], CsDSO4 [9] and Cu6PS5I [10]. In mixed electronic-ionic conductors the ultrasonic behaviour must reveal properties of semiconductors as well as ionic conductors. Because of high mobility of electronic charge carriers the acoustoelectric (AE) current [11] it is possible to observe experimentally in such crystals. In this case additional information about transport properties could be obtained. Therefore it is of interest to apply ultrasonic method to solid electrolytes in order to understand not only ultrasonic relaxation mechanisms but also conductivity and transport phenomena in fast ionic and mixed conductors. This paper reviews our extensive ultrasonic studies of solid electrolytes. The acoustoionic interaction of piezoelectric and “deformation potential” type was observed and investigated in various ionic materials. It was shown, that in DC electric field applied on solid electrolyte the ultrasonic attenuation changes substantially and such behaviour is determined by the increase of electric conductivity in the volume of the crystal. Acoustoelectric current was observed also in these crystals in DC electric field and it was shown that the increase of free holes concentration is responsible for this phenomenon. The ultrasonic anomalies at superionic phase transitions have been observed and mechanism of such critical phenomena is discussed.

121

electric (AE) investigations, in order to change the direction of acoustic wave propagation, two identical fussed quartz buffers were placed in mechanical system on both parallel faces of the sample. Two lithium niobate transducers attached to buffers could launch an ultrasonic wave into the sample from both sides. Silver conductive paste was painted on the sides of the samples and served as electric contacts. The AE current was measured by measuring the voltage on 50-ohm resistor connected in series with a sample. The AE voltage between the sample contacts was measured using high input resistance voltage follower, a video amplifier with an RF filter (on input ultrasonic frequencies) and an oscilloscope. A DC electric field could also be applied between the same contacts. The length of the samples along the direction of ultrasonic propagation was 3–5 mm. The samples were illuminated by a white light source using an optical wave-guide. A distilled water filter was placed between the light beam and the sample to prevent heating. The temperature stabilization and measurement accuracy was better than 0.03 K. 3. Results and discussion We have measured the temperature dependencies of variation of ultrasonic attenuation and velocity in proustite (Ag3AsS3) and pyrargyrite (Ag3SbS3) crystals in the interval 290–450 K. The typical temperature dependencies of ultrasonic attenuation and velocity are shown in Fig. 1 for pyrargyrite crystal. The piezoactive longitudinal ultrasonic wave along z-axis was chosen. It is seen that the ultrasonic attenuation shows peaks, which shift to higher temperatures when ultrasonic frequency increases. In the same temperature region where attenuation reaches maximum the velocity decreases. Ultrasonic velocity is frequency dependant. It is well known that in this temperature interval ionic conductivity of pyrargyrite crystals increases considerably. Electronic conductivity is very small (sample in dark). Thus we observe the typical relaxation behaviour similar to that which was observed in piezoelectric photo-semiconductors when electronic conductivity was varied by illumination and acoustoelectronic interaction caused such behaviour. In our case we observe acoustoionic interaction of piezoelectric type. As in the semiconductors [1] the

2. Experimental The ultrasonic attenuation and velocity investigations we performed by using the pulse-echo method by automatic computer controlled system described earlier [12]. Silicone oil was the material for making acoustic bonds of longitudinal ultrasonic waves. For shear waves Nonaq grease was used. For acousto-

Fig. 2. The temperature dependencies of ultrasonic shear wave attenuation (1,2) and velocity (1′, 2′) in proustite (1, 1′) and pyragyrite (2, 2′) crystals.

122

V. Samulionis, V. Jonkus / Solid State Ionics 179 (2008) 120–125

Table 1 Parameters of Ag3SbS3 and Ag3AsS3 crystals evaluated from acoustoionic interaction measurements Crystal

Wave mode

K

V, 103 m/s

r0, 10− 4 Sm− 1

ΔE, eV

Ag3SbS3

u33 u22 u32 u31 u33 u22 u32 u31

0.33 0.11 0.17 0.45 0.11 0.09 0.12 0.4

2.6 2.97 1.45 1.65 2.6 3.2 1.3 1.66

4.1 7.4

0.42

1.63 3.48

0.48

AgAsS3

frequency dependant ultrasonic attenuation α = f(ω) and velocity dispersion ΔV/V = f(ω) can be described by following equations: K 2 x2 s ; 2V 1 þ x2 s2

ð1Þ

  DV K2 1 : ¼ V 2 1 þ x 2 s3

ð2Þ

a¼

Here — K is the corresponding tensor component of electromechanical coupling factor, τ-is the Maxwell relaxation time: s¼

ee0 ; r ¼ rion þ rel r

is the total ionic–electronic conductivity. The variation of electric conductivity on temperature can be described by simple equation with activation energy ΔE:   DE r ¼ r0 exp  : ð3Þ kT In this approximation ultrasonic attenuation reaches maximum for given frequency ω at: ωτ = 1. The attenuation peak value is proportional to electromechanical coupling factor: αmax = K2 / 4ω. This relaxation model of acoustoionic interaction

Fig. 3. Temperature dependences of longitudinal ultrasonic attenuation and relative velocity in ceramic samples of Ag8HgS2J6.

Fig. 4. Temperature dependencies of longitudinal ultrasonic attenuation and relative velocity in 0.55AgI–0.45AgPO3 glasses.

describes our results in Fig. 1 quite correctly. We have measured ultrasonic velocity and attenuation and for shear modes along the y-direction and the most typical results are shown in Fig. 2. In this case the peak attenuation values are much higher than for longitudinal mode because of bigger tensor component of electromechanical coupling factor. The curves are shifted to lower temperatures because ionic conductivity in the y-direction is almost two times higher. We measured the temperature dependencies for different directions and different modes and using Eq. 1,2 and 3 calculated electromechanical coupling parameters and conductivity activation energies for proustite and pyrargyrite crystals. The summary of obtained results is shown in Table 1. In materials, which are centrosymetric, piezoelectric effect is absent and there is no acoustoionic interaction of piezoelectric type. But in polycrystalline Ag8 HgS 2 I 6 and 0.55AgI– 0.45AgPO3 glasses in temperature dependencies of longitudinal ultrasonic behaviour we observed large attenuation peaks and velocity dispersion (Figs. 3,4). In these fast ionic conductors the relaxation arises from the elastic wave modulation of the chemical potential felt by mobile ions [13]. The ultrasonic attenuation peak value (i.e. relaxation strength) is proportional

Fig. 5. Temperature dependencies of ultrasonic attenuation, AE current density and AE voltage in proustite single crystals.

V. Samulionis, V. Jonkus / Solid State Ionics 179 (2008) 120–125

Fig. 6. Temperature dependencies of ultrasonic attenuation, AE current density and AE voltage in pyrargyrite single crystals. In order to adjust the scale results are shown in arbitrary units.

to the “deformation potential” constant A2 and concentration of mobile ions no. In the case of single relaxation of ionic system the ultrasonic attenuation and velocity change can be described by equations: a¼

n0 A 2 x2 s ; 3 qV0 kT 1 þ x2 s2

DV ¼

n0 A2 1 : 2 1 þ x2 s2 qV0 kT

ð4Þ ð5Þ

Here ρ — density of material, k — Bolcman constant. But the relaxation peaks are much wider than for single relaxation time therefore the distribution of relaxation time should be used in order to describe ultrasonic behaviour. For more detail description of acoustoionic interaction of deformation potential type we refer to our original paper [14]. Proustite and pyrargyrite crystals are photosensitive mixed electronic-ionic conductors. Therefore acoustoelectric phenomena associated with electronic subsystem should manifest itself also. The acoustoelectric current and voltage is mainly caused by electronic charge carries because they have high mobility. The mobility of ions is less more than six orders. But ionic conductivity, which is high enough, influences acoustoelectric current and reduces it. Also the ultrasonic attenuation caused by mobile ions reduces intensity of ultrasonic wave and also reduces AE current. In case of large attenuation (αL NN 1) the AE current density jae and voltage Uae in mixed ionic electronic conductors could be described by equations: jae ¼

123

Fig. 7. Temperature dependencies of ultrasonic attenuation, AE current density in β-AgI single crystals. In order to adjust the scale results are shown in arbitrary units.

In case of small attenuation (αL NN 1) the AE current density and voltage is written as follows: jae ¼

Ael ael I0 rel ;  V rel þ rion

Uae ¼

Ael ael I0 rel :  V ðrel þ rion Þ2

The temperature dependencies of AE current and voltage are shown in Figs. 5 and 6 in illuminated proustite and pyrargyrite crystals. As one could see the qualitative agreement is quite good. For proustite crystals, the case of small attenuation is valid and AE current follows the attenuation dependencies. For pyrargyrite crystal where attenuation is large the AE current saturates. In both cases the AE voltage decreases more rapidly because of squared dependence on total conductivity. The similar results were found in β-AgI crystals along c-axis. The preparation of crystalline samples for ultrasonic investigations was similar to that of [4]. In this crystal the AE voltage was small at room temperature and increased at low temperatures (Fig. 7). Such behaviour is determined by the increase of photoconductivity of β-AgI crystals at

Ael I0 rel ;  LV rel þ rion

Uae ¼

Ael I0 rel :  LV ðrel þ rion Þ2

Here I0 — intensity of ultrasonic wave, L — length of sample and μel — mobility of electronic charge carriers.

Fig. 8. The DC electric field dependences of ultrasonic attenuation, acoustoelectric current and voltage for longitudinal mode along c-axis in pyrargyrite.

124

V. Samulionis, V. Jonkus / Solid State Ionics 179 (2008) 120–125

Fig. 9. The DC electric field dependencies of ultrasonic attenuation, acoustoelectric current and voltage for longitudinal mode along c-axis in β-AgI crystal.

low temperatures, because photosensitive ultrasonic attenuation increases as well. Very interesting results were obtained when DC electric field was applied on mixed ionic-electronic conductor samples. After applying DC field the ultrasonic attenuation increased in the same way as with temperature increasing. It shows that electric conductivity increases. The AE effect measurements have shown (Figs. 8, 9) that increase of electric conductivity is mainly caused by the increase of positive charge carriers of holes type in the volume of the sample. We carried out extensive investigation of ultrasonic behaviour at superionic phase transitions. In all cases the ultrasonic attenuation and velocity anomalies were found. In Figs. 10 and 11 the most typical results are shown for polycrystalline AgI and Rb4AgI5 samples. The polycrystalline samples were used, because crystals usually crack when exhibiting superionic PT. The samples were obtained by pressing powders into platelets. The structure and phase composition was not checked. As one can see at phase transition there are ultrasonic attenuation and velocity critical anomalies. The main feature is that attenuation in superionic phase is much higher than in normal phase. The superionic phase in these

Fig. 11. Temperature dependencies longitudinal ultrasonic attenuation and velocity in polycrystalline AgI sample.

materials is nonpiezoelectric. Therefore in the superionic phase we observe the acoustoionic interaction of “deformation potential” type (see Eq. 4 and 5). The critical ultrasonic anomalies can be explained by the interaction of ultrasonic wave with order parameter in low temperature phase (Landau–Chalatnikov mechanism [15]) and interaction with order parameter fluctuations in the high temperature phase (fluctuation mechanism [16,17]). 4. Conclusions The acoustoionic interaction of piezoelectric type was observed in Ag3SbS3 and Ag3AsS3 ionic conductors and the electromechanical parameters of these crystals have been evaluated from ultrasonic attenuation and velocity measurements. The acoustoionic interaction of “deformation potential” type was observed polycrystalline, glassy fast ionic conductors and in the high temperature phases of AgI and Rb4AgI5 crystals. We have shown that the DC electric field applied to Ag3SbS3, Ag3AsS3 and AgI crystals induces anomalies of ultrasonic and acoustoelectric properties what is mainly associated to the increase of electronic component of electric conductivity in the volume of material. The temperature dependencies of ultrasonic velocity and attenuation revealed anomalies, which are the indication of the phase transitions in these materials. References

Fig. 10. Temperature dependence of longitudinal ultrasonic attenuation in polycrystalline Ag4RbI5 sample.

[1] A.R. Hutson, J.H. McFee, D.L. White, Phys.Rev.Letters 7 (1961) 237. [2] D.L. White, Journ. Appl. Phys. 33 (1962) 2547. [3] A. Brilingas, N.J. Bucko, J. Grigas, V. Samulionis, J.D. Sheshnitsch, Kristall und Technik 13 (1978) 221. [4] J.H. Page, J.-Y. Prieur, Phys. Rev. Letters 42 (1979) 1684. [5] V.L. Skritskij, V. Samulionis, Liet Fiz. Rink. 24 (1984) 587 (in Russian). [6] G. Carini, M. Cutroni, M. Federico, G. Gali, G. Tripodo, Phys. Rev. B30 (1984) 7219. [7] D.P. Almond, A.R. West, Solid State Ionics 3/4 (1988) 265. [8] V. Samulionis, V.L. Skritskij, A. Kezionis, V. Mikucionis, A. Orliukas, Ju. E. Ermolenko, S.V. Glazunov, Fiz. Tverd. Tela 29 (1987) 2501 (in Russian). [9] R. Mizeris, J. Grigas, V. Samulionis, V. Skritskij, A.I. Baranov, L.A. Shuvalov, Phys. Stat. Solidi (a) 110 (1988) 429. [10] V. Samulionis, J. Banys, Yu. Vysochanskii, Ferroelectrics 336 (2006) 29.

V. Samulionis, V. Jonkus / Solid State Ionics 179 (2008) 120–125 [11] G. Weinreich, T.M. Sanders, H.G. White, Phys. Rev. 114 (1959) 33. [12] V. Samulionis, V. Valevicius, J. Banys, A. Brilingas, J. Phys, IV France 6 (1996) C8–405. [13] J.B. Boyce, B.A. Huberman, Phys. Rep. 52 (1979) 190. [14] R. Vaitkus, A. Kezionis, V. Samulionis, A. Orliukas, V. Skritskij, Solid State Ionics 40/41 (1990) 922.

125

[15] L.D. Landau, I.M. Khalatnikov, Sov.Phys.(Doklady) 96 (1954) 459 (in Russian). [16] A.P. Levaniuk, K.A. Minaeva, B.A. Strukov, Fiz. Tverd. Tela 10 (1968) 2443 (in Russian). [17] V. Valevicius, V. Samulionis, J. Banys, J. Alloys and Compounds 211/212 (1994) 369.