Colossal electric conductivity in Ag-defect Ag5Pb 2O6

Article abstract of DOI:10.1002/pssa.200306738

The Bystroem-Evers compound Ag₅Pb₂O₆ was annealed at 500-540 K under a flow of electric current which results in a textured structure and anisotropic colossal electric conductivity (>10 ⁹ Ω^-1 cm^-1) between 210 and 525 K. The related physical properties are primarily governed by dissociation of Ag from c-axis channels and lattice strains, which in turn depend upon electric current.

Full text of DOI:10.1002/pssa.200306738

phys. stat. sol. (a) 201, No. 3, 544–549 (2004) / DOI 10.1002/pssa.200306738
*
Colossal electric conductivity in Ag-defect Ag Pb O
5
2
6
*
*, 1
1
2
3
D. Djurek , Z. Medunić , M. Paljević , and A. Tonejc
1
A. Volta Applied Ceramics (AVAC), Kesten brijeg 5. Remete, 10000 Zagreb, Croatia
Ruđer Bošković Institute, P. O. Box 180, 10000 Zagreb, Croatia
Physics Department, Faculty of Science, University of Zagreb, PO Box 331, 10000 Zagreb, Croatia
2
3
Received 3 February 2003, revised 15 September 2003, accepted 10 November 2003
Published online 9 February 2004
PACS 61.66.Fn, 72.80.Jc, 74.10.+v
The Byström–Evers compound Ag Pb O was annealed at 500–540 K under a flow of electric current
5
2
6
9
–1
–1
which results in a textured structure and anisotropic colossal electric conductivity (>10 Ω cm ) be-
tween 210 and 525 K. The related physical properties are primarily governed by dissociation of Ag from
c-axis channels and lattice strains, which in turn depend upon electric current.
©
2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1
Introduction
In previous papers [1, 2] the Pb–Ag–C–O system has been reported as a possible carrier of superconduc-
tivity (SC) which extends up to room temperature (RT). More detailed analysis of data leads to the con-
clusion that the Byström–Evers (BE) compound Ag Pb O [3] is responsible for the properties reported
5
2
6
in [1], as well as for properties appearing in an alternative phase discussed in the present paper.
Originally Ag Pb O was prepared in the trigonal form by hydrothermal synthesis [3], while Jansen and
5
2
6
co-workers [4–6] prepared the BE compound from PbO and Ag O in a solid-state reaction proposing the
2
2
hexagonal structure shown in Fig. 1a. Ag Pb O consists of channels stretched along the hexagonal c-axis
5
2
6
(Fig. 1b) and filled with two Ag(1) atoms per unit formula. In order to differentiate between the two types
of Ag cations in the BE compound, a more suitable unit formula, Ag [Ag Pb O ], may be proposed.
2
3
2
6
2
Results and discussion
Preparation of the BE phase Ag [Ag Pb O ] was carried out with PbO and Ag O mixed in the atomic
2
3
2
6
2
2
proportion Ag/Pb = 5/2 using a magnetic stirrer and benzene. The PbO contained traces of H O visible
2
2
by the infrared (IR) technique but undetectable by conventional thermogravimetric (TG) analysis with
sensitivity higher than 0.02 mol%.
The mixed powder was then fused for 100 h at 620 K and a pressure of 200 bar O . TG decomposition
2
in air to Pb O + Ag at 723 K and subsequent decomposition of Pb O to PbO at 793 K (Fig. 2, inset)
3
4
3
4
revealed the unit formula Ag [Ag Pb O ]. X-ray diffraction angles and relative intensities correspond to
2
3
2
6
those cited by Jansen and co-workers.
Heating of the BE phase in air, nitrogen or in vacuo results in transformations at 550 and 582 K, as is
shown by differential thermal analysis (DTA) (Fig. 2) recorded from a pellet of weakly compacted pow-
der (p ~ 30 bar). An annealing at 523 K in air for 48 h results in a partial release of elementary silver,
which is confirmed by X-ray analysis and by visual identification of small silver grains. No traces of
*
See also M. Stutzmann and S. Hildebrandt, Editorial Note, phys. stat. sol. (a) 201, 543 (2004).
Corresponding author: e-mail: danijel.djurek@inet.hr
*
*
©
2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

phys. stat. sol. (a) 201, No. 3 (2004) / www.pss-a.com
545
b)
a)
Fig. 1 a) Unit cell of Ag [Ag Pb O ]; b) c-axis projection of the unit cell of Ag [Ag Pb O ].
2
3
2
6
2
3
2
6
Pb O were recorded which excludes the possibility that silver evolves from the decomposition of the BE
3
4
compound. When a higher uniaxial stress (about 1500 bar) is applied to the BE powder silver is removed
from the compound up to β > 0.30, with the basic hexagonal structure preserved.
The four-probe electric resistance was measured in a conventional way using 100 µm gold wire and
3
silver paint. The pellet (5 × 5 × 0.4 mm ) was prepared by compression (p ~ 22 bar) of the BE powder
and is shown in the inset of Fig. 3. Heating was performed in air under a dc current of 100 mA with
current contacts connected to A–B and voltage contacts to C–D. The pellet was annealed at 523 K for
2
4 h and subsequently cooled to RT. The reheating is followed by a decrease of electric resistance at
–
7
TD = 345 K to less than 10 Ω (Fig. 3), which is the ultimate resolution of our dc resistance measure-
ment. This colossal electric conductivity (CEC) state persists up to 525 K when a partial deterioration
appears which manifests as a lower temperature (450 K) of recovery of the CEC state by cooling to RT.
Fig. 2 Differential thermal analysis (DTA) of Ag [Ag Pb O ] recorded in 1 bar N . Inset shows the
2
3
2
6
2
thermogravimetric decomposition curve of Ag [Ag Pb O ] recorded in air.
2
3
2
6
©
2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

546
D. Djurek et al.: Colossal electric conductivity in Ag-defect Ag Pb O
5 2 6
Fig. 3 Temperature dependence of the electric
resistance of Ag-defect sample Ag2–β⋅[Ag Pb O ]
3
2
6
heated from RT (1) up to 525 K (2), and then
cooled to 100 K (3). Inset shows specimen ar-
rangement with electric contacts glued with silver
paint.
Based upon repeated preparations the main characteristic of the CEC state is the downturn temperature TD
which exclusively depends on strains. After cooling to RT and pulling of the holder with sample in free air
the CEC state was manifested, under flowing current, over the next three months, independent of the inter-
change of electric and voltage contacts. In parallel tests an interchange of current contacts to A–D and volt-
age contacts to B–C revealed normal electric resistance of 0.12 Ω which increased up to 0.55 Ω after 3
months of exposure to air, and the cited anisotropy also appeared after interchange of B–C (current) and
A–D (voltage) contacts. It should be borne in mind that deterioration of the CEC state at 525 K has nothing
in common with the classic transition from superconducting to normal state but it should be looked upon as
a result of the appearance of another high-temperature structural phase. Cooling below RT results in a re-
versible deterioration of the CEC state near 210 K which is also the lowest T recorded by the four-probe
D
resistance measurement. By heating to RT and switching off the current the pellet undergoes, in four days,
destruction by self-pulverisation. It should be noted that small Ag grains were visible between the current
contacts and this might be included in the scheme of mutual dependence of the CEC state, Ag dissociation
and driven electric current strength. It should be pointed out that an anisotropic textured state was also in-
duced by annealing under flowing ac currents (230 cycles per second) supplied from a lock-in amplifier.
Part of the pellet considered in Fig. 3 was mounted in a microwave cavity (9.3 GHz). The temperature
dependence of the inverse 2Q factor, which scales microwave resistance, is shown in Fig. 4. The sample
–1
was first cooled to 90 K and than heated at a rate ~1 K min . Decrease of the microwave resistance is
clearly visible at T = 210 K, and is in fair agreement with our four-probe resistance data, as shown in
D
Fig. 3. By repeated cooling–heating cycles conversions to normal and CEC states took place, as shown in
Fig. 4. Application of a magnetic field up to 8 T revealed an absence of field dependence of the micro-
wave resistance in the CEC state.
An investigation of microwave absorption in modulated magnetic field (MAMMA) pointed out an
absence of characteristic absorption hysteresis which contradicts the results reported in [2], and calls for
attention to be paid to at least two possible modifications of BE samples exhibiting independently classi-
cal superconductivity and the CEC state.
Anisotropic electric properties induced by electric currents at elevated temperatures suggested an
experimental set-up consisting of a metal tube filled with BE powder. We used a copper tube with outer
(
OD) and inner diameters (ID) of 4 and 2 mm, respectively. The tube was filled to 18% of theoretical
–
3
density of the BE phase (8.92 g cm ) and then extruded to OD = 1.5 mm and ID = 0.62 mm. An empty
tube of similar size connected in series with the filled one was used as a reference, and electric contacts
were realised by screw-tight fittings. Voltage drops at temperature T of the filled and empty tube are
denoted as V (T) and V (T), respectively, whereas the values at RT are denoted as V (0) and V (0). A
S
R
S
R
dimensionless coefficient K = V (T)V (0)/V (T)V (0) was plotted against temperature as shown in
S
R
R
S
Fig. 5. A decrease of the resistance of the BE-filled tube on heating from RT is evident (Fig. 5, curve a),
©
2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

phys. stat. sol. (a) 201, No. 3 (2004) / www.pss-a.com
547
Fig. 4 Temperature dependence of the inverse
microwave 2Q factor of the sample of Fig. 3
heated from 90 K (1) up to 330 K, then again
cooled to 90 K and reheated to 220 K (2).
Fig. 5 Temperature dependence of the dimen-
sionless K factor (see text for explanation); (a) cop-
per tube (OD = 1.5 mm, ID = 0.62 mm) filled with
the BE powder; (b) empty copper tube of similar
size. Resistance measurement current was I = 0.5 A.
Fig. 6 Temperature dependence of the dimensionless K
factor (see text for explanation); (a) copper tube (OD =
0.5 mm, ID = 0.18 mm) filled with the BE powder; (b)
full cross-section copper wire of similar size. Voltage
contacts were realised with silver paint. Resistance meas-
urement current was I = 0.5 A.
©
2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

548
D. Djurek et al.: Colossal electric conductivity in Ag-defect Ag Pb O
5 2 6
Fig. 7 Temperature dependence of the electric resist-
ance (I = 0.5 A); (a) copper tube (OD = 1.5 mm,
ID = 0.62 mm) filled with the insulator Pb O ; (b)
3
4
tube of similar size (in series with the former) filled
with BE powder.
which shows a good correspondence with the decrease of resistance at T = 345 K shown in Fig. 3. A net
D
resistance along the wire is a result of interruptions in the powdered core as a result of the low filling
factor. The filled tube was then replaced by another empty tube of similar size and the corresponding
coefficient K was plotted against T (Fig. 5, curve b).
Figure 6 (curve a) shows the temperature dependence of the dimensionless factor K for the tube filled
with BE powder as above, but with OD = 0.5 mm and ID = 0.18 mm, and Fig. 6 (curve b) for the case
when the filled tube was replaced by the full cross-section copper wire. Note that voltage contacts were
realised using silver paint.
In the next experiment a BE-filled tube was heated under a 0.5 A dc current up to 510 K and
then quenched in cold water. Reheating was again performed under 0.5 A and Fig. 7 shows the tempera-
ture dependence of the electric resistance of extruded copper tubes: curve a when filled with the insulator
Pb O and curve b when filled with BE. An increase of the resistance of the BE-filled tube at 568 K
3
4
to that of the tube filled with Pb O is evident. Cooling re-established the CEC state at 548 K. At RT
3
4
the resistance was nearly 0.6 times that before heating, but was partly recovered over time to the
initial value. A rather strong noise visible on the voltage contacts usually precedes a complete loss of
the resistance above RT in the copper tube filled with BE powder. The electric resistivity calculated
from the cross-section and length of the BE core at 523 K is nearly 3 times smaller than that of
silver, although with regard to the low filling factor such a calculation is an overestimate of the
resistivity of the BE phase. In addition, tubes equipped with a concentric copper stud 1 mm in dia-
meter and filled with the BE phase in the space between the stud and the inner tube wall show an
absence of CEC when the voltage is measured on the stud and current contacts are attached on the
outer tube. This means that the CEC is confined to a thin layer of the BE phase on the inner wall, i.e.
conductivity is induced by a tangential interaction of the BE phase and electric current flowing in the
copper.
3
Conclusions
It may be summarised that: (1) the BE phase undergoes transformations at temperatures of 550 and
82 K, not reported in previous publications; (2) annealing at 523 K results in a partial release of silver
5
from c-axis channels; (3) dissociation at 523 K under flowing current results in a textured structure with
high anisotropy of electric resistance; (4) there is a close interdependence of strain, dissociation level of
silver and electric current; (5) an absence of magnetic field dependence of microwave resistance data, as
well as an absence of MAMMA effect indicate an alternative conclusion to those reported in [2]. Based
upon recent experience both effects observed in the BE phase critically depend on strains and related
silver defects. To summarise the listed observations, both effects may be recognised, in some instances,
as two forms of the same, as yet unknown, phenomenon.
Efficient control of the strain and dissociation is a decisive factor in further research which will shed
more light on the physical and chemical properties of the BE compound.
©
2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

phys. stat. sol. (a) 201, No. 3 (2004) / www.pss-a.com
549
Note added in the proof
The wire sample (OD = 0.5 mm) from Fig. 6a extruded to OD = 0.20 and ID = 0.075 mm exhibits a
temperature independent dimensionless factor K = 0.682 ± 0.05 between 77 and 703 K when BE phase
decomposes and K = 1, that is, the slopes of the temperature dependent resistances of the empty and BE
filled tubes are identical, which in turn means that parts of Cu tube are short circuited by BE phase ex-
hibiting CEC between 77 and 703 K.
Acknowledgments The authors are indebted to Dr. M. Prester of the Institute of Physics in Zagreb for an inde-
pendent evaluation of Fig. 6, and to Professor A. Dulčić of the University of Zagreb for the microwave resistance
measurements. The ELKA facility performed the extrusion of the copper tubes.
References
[
[
[
[
[
[
1] D. Djurek, Z. Medunić, M. Paljević, and A. Tonejc, Physica C 341–348, 723 (2000).
2] D. Djurek, Z. Medunić, M. Paljević, and A. Tonejc, Physica C 351, 78 (2001).
3] A. Byström and L. Evers, Acta Chem. Scand. 4, 613 (1950).
4] M. Jansen, M. Bortz, and K. Heidebrecht, J. Less-Common Met. 161, 17 (1990).
5] M. Bortz, M. Jansen, H. Hohl, and E. Bucher, J. Solid State Chem. 103, 447 (1993).
6] M. Jansen, M. Bortz, and K. Heidebrecht, Z. Kristallogr. 147 (1989).
©
2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim