Comment on "Colossal Ionic Conductivity at Interfaces of Epitaxial ZrO₂:Y₂O₃/SrTiO₃ Heterostructures"

García-Barriocanal et al. (1) fabricated yttria-stabilized zirconia (YSZ)/strontium titanate (SrTiO₃ or STO) epitaxial heterostructures on STO substrates, in which YSZ layers with thickness t ranging from 1 to 62 nm were sandwiched between two 10-nm-thick STO layers. They claimed a huge ionic conductivity enhancement in the heterostructures, and the enhancement was attributed to the high oxygen vacancy concentration and the high mobility at the YSZ/STO interfaces. Their major experimental evidence is that the dc conductance was orders of magnitude lower than the ac conductance. However, it should be pointed out that their dc conductance is not purely electronic.

A typical cell investigated by García-Barriocanal et al. (1) is Ag/STO_10nm-YSZ_1nm-STO_10nm-STO_Substrate/Ag, in which STO is an important part. Nominally undoped STO is almost always doped with acceptor-type impurities. Therefore, it is a mixed conductor of oxygen vacancies and holes. Owing to the higher mobility of holes, nominally undoped STO shows p-type conductivity (2–4). According to the defect chemistry of acceptor-doped STO (2, 3), the p-type conductivity dominates the overall conductivity when the acceptor concentration is low. Therefore, one has to consider the p-type conductivity of the 10-nm-thick STO layers and the STO substrate.

The electronic (for example, p-type) partial conductivity can be experimentally deduced by means of the Hebb-Wagner polarization (5, 6). A key point of the Hebb-Wagner polarization is that an electrode should block the ionic species (e.g., oxygen vacancies). Under this condition, the electrical conduction takes place only via the electronic species (e.g., holes) at the steady state. The solubility of oxygen in solid silver is relatively high, and the bulk diffusion of oxygen in Ag is also remarkable (7–9). Therefore, the Ag electrode is not ionically blocking, which is evidenced by the relatively small electrode resistance shown in Fig. 1. In addition, the oxygen diffusion in STO is much too sluggish in the temperature range of 80 to 260°C [the temperature range in figure 3 in (1)]. Estimated from the chemical diffusion coefficient of oxygen in STO (10), it takes about 10⁸ s for oxygen to diffuse a distance of only 1 mm at 170°C. Therefore, the dc conduction of the cell Ag/STO_10nm-YSZ_1nm-STO_10nm-STO_Substrate/Ag can never reach the steady state within a reasonable time period. In view of the fact that the Ag electrode is not ionically blocking, and that the steady state cannot be reached within a reasonable time period, one can conclude that the dc conductance of the cell Ag/STO_10nm-YSZ_1nm-STO_10nm-STO_Substrate/Ag consists of the conributions of oxygen vacancies in STO and YSZ and holes in STO, that is, it is not purely electronic.

View larger version (24K): [in this window] [in a new window]   Fig. 1. Impedance spectrum of STO substrate measured at 220°C in air.

As shown in figure S2 in (1), the conductance of the cell Ag/STO_10nm-YSZ_1nm-STO_10nm-STO_Substrate/Ag is larger than that of bulk YSZ only by a factor of ~10. Figure 3 in (1) shows that the conductivity becomes almost eight orders of magnitude higher than the conductivity of bulk YSZ, but this is because only the 1-nm-thick YSZ layer was taken into consideration while calculating the conductivity. It is noted that the activation energy (~0.66 eV) of the STO substrate conductivity is quite close to the activation energy of the heterostructure conductivity (~0.64 eV) (see Fig. 2). Therefore, the contribution of the STO substrate to the heterostructure conductivity may not be negligible, especially at higher temperatures. When the 10-nm-thick STO layers and the 0.5- or 1-mm-thick STO substrate are taken into consideration, the calculated conductivity immediately decreases by six orders of magnitude, which is about three orders of magnitude higher than the conductivity of the STO substrate (see Fig. 2). A previous study (11) on BaTiO₃ demonstrated that the p-type conductivity of BaTiO₃ ceramic with an average grain size of ~35 nm is about two orders of magnitude higher than that of the microcrystalline counterpart, which was ascribed to a greatly reduced oxidation enthalpy in nanocrystalline BaTiO₃. For the same reason, the p-type conductivity of the 10-nm-thick STO layers is expected to be higher, which can account for the conductivity difference shown in Fig. 2.

View larger version (13K): [in this window] [in a new window]   Fig. 2. Temperature dependence of the conductivities of STO substrate and Ag/STO10nm-YSZ1nm-STO10nm-STOSubstrate/Ag heterostructure. The conductivity of the heterostructure is replotted from figure 3 in (1).

To corroborate that the measured conductivity is really ionic, one has to check the oxygen partial pressure dependence of the heterostructure conductivity and do the Hebb-Wagner polarization experiments under well-controlled conditions.

www.sciencemag.org/cgi/content/full/324/5926/465-a/DC1

SOM Text

Figs. S1 and S2

References

The editors suggest the following Related Resources on Science sites:

In Science Magazine

TECHNICAL COMMENTS
Response to Comment on "Colossal Ionic Conductivity at Interfaces of Epitaxial ZrO₂:Y₂O₃/SrTiO₃ Heterostructures"

J. García-Barriocanal, A. Rivera-Calzada, M. Varela, Z. Sefrioui, E. Iborra, C. Leon, S. J. Pennycook, and J. Santamaría (24 April 2009)
Science 324 (5926), 465-b. [DOI: 10.1126/science.1169018]
 |  Abstract »  |  Full Text »  |  PDF »

REPORTS
Colossal Ionic Conductivity at Interfaces of Epitaxial ZrO₂:Y₂O₃/SrTiO₃ Heterostructures

J. Garcia-Barriocanal, A. Rivera-Calzada, M. Varela, Z. Sefrioui, E. Iborra, C. Leon, S. J. Pennycook, and J. Santamaria (1 August 2008)
Science 321 (5889), 676. [DOI: 10.1126/science.1156393]
 |  Abstract »  |  Full Text »  |  PDF »  |  Supporting Online Material »

THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES:

ADVERTISEMENT