This work combines new and earlier obtained results on electron hole and oxygen-ion transport in the La2NiO4-based solid solutions. The effect of lanthanum substitution with Ca/Sr and nickel with Fe, Mn, Co or Cu on transport properties of La2−xAxNi1−yMeyO4+δ was analyzed and discussed at different substitution levels. Besides the changes in concentration and mobility of electron holes induced by the doping with cations of different nature, the partial transformation of Ni3+ from low-spin to high-spin state was shown to have a profound effect on transport properties of these materials leading to a notable decrease in mobility of electron holes, especially in the strontium-rich oxides. The obtained results suggested that the size factor was the main driving force behind the observed transformation of Ni3+. The oxygen-ion transport in La2−xAxNi1−yMeyO4+δ was characterized by significant surface exchange limitations, which can be reduced only at relatively high concentrations of strontium and iron, and should be taken into account while evaluating the ionic conductivity by means of oxygen permeation or the modified Hebb-Wagner polarization method.
The La2NiO4-based oxides are considered as promising cathode materials for intermediate-temperature solid oxide fuel cells (SOFCs) and oxygen separation membranes due to relatively high mixed ionic/electronic conductivity, with advantages in dimensional and chemical stability over the prevalent Co-based perovskites , , . It is known that La- and/or Ni-site doping in La2NiO4+δ can drastically change its functional properties such as oxygen content, thermal expansion, p-type conductivity, oxygen-ion transport, surface exchange kinetics, etc. , , , , , , .
In particular, the electronic p-type conductivity of La2NiO4+δ (≈80 S/cm at 600°C in air ) increases after doping with calcium or strontium due to a partial oxidation of Ni2+ to Ni3+, which is necessary to compensate the effective negative charge of Ca/Sr on La-site . According to refs. , ,  the increase in conduction is observed only at x≤0.3 in La2−xCaxNiO4+δ and x≤0.75–1.2 in La2−xSrxNiO4+δ, respectively. The analysis of p-type conduction in La2−xSrxNiO4+δ revealed that the mobility of electron holes remained almost constant in the range of 0≤x≤0.4 , indicating a predominant role of Ni2+/Ni3+ oxidation when the conductivity increases at x≤0.4. Moreover, complex analysis of oxygen non-stoichiometry and transport properties for La1.9Ca0.1NiO4+δ and La1.9Sr0.1NiO4+δ revealed lower mobility and higher concentration of electron holes in the former . The substitution of Ni in La2NiO4+δ with other 3d-metals, such as iron, cobalt or copper deteriorates the hole transport , , , . In general, B-site doping leads to a decrease in concentration and mobility of electron holes, the latter is usually addressed to the elongation of Me–O distances and hole trapping , .
The calcium/strontium substitution for lanthanum in La2NiO4+δ lowered oxygen tracer diffusion coefficient and ionic conductivity by one order of magnitude already at x=0.1 , , which was explained by a reduction in concentration of oxygen interstitials  and their mobility . According to ref.  oxygen diffusion in La2−xSrxNiO4+δ reaches a minimum at x=0.6 followed by a dramatic increase in approximately 4 orders of magnitude at x=0.8. The strontium doping has little effect on surface exchange kinetics when x≤0.2 , , while a noticeable decrease in surface exchange coefficients at higher x can be observed . The B-site doping with higher-valence cations (Fe3+, Co3+) was found to suppress anion migration despite the increase in interstitial oxygen concentration , , . At the same time, such doping was shown to have a positive effect on surface exchange kinetics , . For example, the surface exchange coefficient for La2Ni0.9Co0.1O4+δ was higher by one order of magnitude than that for La2NiO4+δ at 650°C .
The highly anisotropic p-type conduction and oxygen diffusion (by means of interstitial oxygen at δ>0 or oxygen vacancies at δ<0) in the La2NiO4-based solid solutions occurs predominantly along the ab plane , . The oxygen-ion transport in the La2NiO4-based solid solutions can be described in terms of the intersticialcy mechanism and/or the vacancy migration mechanism depending on the main migrating species , . However, the electron hole transport mechanism and electronic state in the La2NiO4-based oxides are still under discussion , , , , , . At T<500°C the p-type conduction in La2NiO4+δ and its derivatives with low doping level was described by a small polaron hopping mechanism , , . However, at T>500°C the observed decrease in conductivity was ascribed whether to the loss of lattice oxygen (followed by a reduction in the electron hole concentration) , ,  or to the change of hopping mechanism to a metal-like band conduction , .
Previously, we proposed a model, describing the temperature dependence of the Seebeck coefficient in La2NiO4+δ and its derivatives, which was based on the small polaron hopping mechanism and included equilibrium between the low-spin and high-spin Ni3+ cations . We obtained satisfactory fits for La2NiO4+δ and all studied compositions in the La2−xAxNi1−yMeyO4+δ (A=Ca, Sr; Me=Fe, Cu) systems , . Introduction of dopants was found to induce partial transformation of Ni3+ from low-spin to high-spin state when temperature increased, significantly reducing the mobility of electron holes. Furthermore, another study on La1.2Sr0.8Ni0.9Fe0.1O4+δ showed that electron holes localized on Ni2+ forming Ni3+ in low-spin state can be considered as quasi-delocalized, which behavior can be interpreted in terms of band conduction . Despite the mixed electronic state in La1.2Sr0.8Ni0.9Fe0.1O4+δ the small polaron hopping mechanism was shown to be predominant , . In this work, we continue the discussion on the electron hole and oxygen-ion transport in the La2NiO4-based oxides with different substitution levels. To complete the data on the earlier studied La2−xAxNi1−yMeyO4+δ (A=Ca, Sr; Me=Fe, Cu) solid solutions we also present the results for individual oxide phases in the La2−xSrxNiO4+δ, La2−xSrxNi1−yCoyO4+δ, La2−xSrxNi1−yMnyO4+δ and La2−x(Ca/Sr)xNi1−yFeyO4+δ systems.
La2−xSrxNiO4+δ (x=0, 0.2, 0.3), La1.5Sr0.5Ni0.9Co0.2O4+δ, La1.5Sr0.5Ni0.7Fe0.3O4+δ, La2−xSrxNi0.9Mn0.1O4+δ (x=0.7, 0.8) and La1.5Ca0.5Ni1−yFeyO4+δ (y=0.3, 0.33) were synthesized by a citrate-nitrate technique followed by calcination of the as-prepared powders at 1373 K in air as reported in , . At the final step the powders were uniaxially pressed into rectangular-shaped bars at a pressure of 20 bar and sintered at 1623 K for 20 h in air. The X-ray powder diffraction (XRPD) showed that all synthesized samples were obtained as single-phase, which was in agreement with the phase diagrams reported earlier in , , . As an example, the XRPD pattern of La1.5Sr0.5Ni0.7Fe0.3O4+δ refined by the Rietveld method is presented in Fig. 1. The refined XRPD patterns for other La2−xSrxNi1−yMeyO4+δ oxides can be found in , , .
Oxygen non-stoichiometry of the powder samples was studied by thermogravimetric analysis (TGA) using a Netzsch STA 409 PC instrument within the 303–1373 K temperature range in air as described in . The absolute values of oxygen content were obtained by a direct reduction of the samples in H2 (20 ml/s) gas flow at 1200°C for 10 h. Total conductivity and the Seebeck coefficient were measured simultaneously using the standard 4-probe DC technique in the temperature range of 298–1273 K in air (the details can be found in refs. , ). Oxygen-ion conductivity was determined by the modified Hebb-Wagner polarization technique in the temperature range of 973–1123 K for La1.5Sr0.5Ni0.7Fe0.3O4+δ and 873–1273 K for La1.5Ca0.5Ni0.7Fe0.3O4+δ with a step of 50 K in air. The technique details were reported in .
Results and discussion
The temperature dependencies of oxygen non-stoichiometry, total conductivity and the Seebeck coefficient for some La2NiO4-based solid solutions are shown in Fig. 2. The obtained temperature dependencies are typical for the La2NiO4-based materials and had been discussed elsewhere , . As can be seen from Fig. 2a the replacement of iron by cobalt in La1.5Sr0.5Ni0.8Me0.2O4+δ leads to a decrease in oxygen excess and, as a consequence, the concentration of electron holes according to the electroneutrality condition , . The latter can be written as follows using the Kröger-Vink notation :
where p or stands for the concentration of electron holes localized on Ni2+ forming Ni3+ cations.
Taking into account the oxygen content and electroneutrality condition (eq. 1) one can suggest that an average oxidation state of cobalt cations is expected to be mixed (Co2+/Co3+) in La1.5Sr0.5Ni0.8Co0.2O4+δ compared with that of iron cations which is shown to be close to +3 in La2−xSrxNi1−yFeyO4+δ .
It should be stressed that the contribution of ionic conductivity to the total conductivity of the La2NiO4-based compounds is less than 0.01% , , . Thus, the observed values of total conductivity mainly represent the electron hole transport. The studied samples possessed semiconducting behavior in the whole temperature range studied, contrary to the undoped La2NiO4+δ (Fig. 2b). As a result, at intermediate and high temperatures the studied samples showed comparable or higher conductivity values than that for La2NiO4+δ. The lnσT=f(1/T) plots indicated thermally activated conduction, implying the small polaron hopping mechanism in the oxides.
The temperature dependencies of the Seebeck coefficient (Fig. 2c) were described by the model reported in , . In this model the observed Seebeck coefficient was interpreted as a sum of two constituents:
where tLS, tHS and SLS, SHS denote the transference numbers and the partial Seebeck coefficients for electron holes localized on Ni2+ forming low-spin (LS) and high-spin (HS) states of Ni3+, respectively.
Assuming that the small polaron hopping mechanism was predominant in these materials, the partial Seebeck coefficients were defined by the Heikes formula  taking into account spin degeneracy of nickel cations as it was shown in , . In order to determine the concentration of electron holes (Ni3+) at a given temperature by eq. 1, the oxidation state of iron in La1.5Ca0.5Ni0.7Fe0.3O4+δ and La1.5Ca0.5Ni0.67Fe0.33O4+δ was considered to be +3 in whole temperature range studied , . Similarly, following the discussion in , the oxidation state of manganese in La1.3Sr0.7Ni0.9Mn0.1O4+δ and La1.2Sr0.8Ni0.9Mn0.1O4+δ was assumed to be +4 at all studied temperatures. Since both, iron and manganese, behave as electron hole traps in these materials, their contribution to the total Seebeck coefficients was neglected. The fitting results are shown in Fig. 2c by the solid lines. One can observe good agreement between the calculated curves and the experimental data in the studied temperature range.
Similar to nickel cations, the oxidation state of cobalt in La1.5Sr0.5Ni0.8Co0.2O4+δ was suggested to be mixed between +2 and +3 in the 298–1273 K range. This was taken into account by using (eq. 1) as an additional fitting parameter. The fitting result was represented in Fig. 2c by the solid line. The average concentration of the Co2+ and Co3+ cations in the temperature range of 298–1273 K was found to be 0.08 and 0.12, respectively. In earlier works it was reported that all Co2+ and Co3+ cations in La2−xSrxCoO4+δ were considered to exist in HS state for 0.4≤x≤0.6, and in intermediate (LS+HS) or LS state for Co3+ cations for x>0.6 , . Taking into account the aforementioned results all cobalt cations in La1.5Sr0.5Ni0.8Co0.2O4+δ are expected to be in HS state. Moreover, cobalt behaves as an electron hole trap in the La2NiO4-based materials , . Bearing this in mind, the contribution of cobalt cations to the total Seebeck coefficient (eq. 2) can be omitted. The final fitting parameters for the mention above samples will be shown later in the work in comparison with that for the studied earlier La2NiO4-based oxides.
La2−xAxNiO4+δ (A=Ca, Sr)
In this and following sections we compare and discuss the characteristics of electron hole transport in the La2NiO4-based materials as a function of dopant content. Let’s start with the La2−xAxNiO4+δ (A==Ca, Sr) system, where x≤0.4. Figure 3a–c show the total conductivity (σ), total concentration of electron holes (p), activation energy of the electron hole transport (Ea), enthalpy of formation of high-spin Ni3+ from low-spin Ni3+ cations (ΔH) and the fraction of high-spin Ni3+ cations (χHS). Figure 3d–f represent the heat of transfer (QLS and QHS=ΔH+QLS), small polaron hopping energy (WLS and WHS) and mobility (μLS and μHS) for electron holes localized on Ni2+ cations forming Ni3+ in low- and high-spin state, respectively. The concentration of electron holes can be calculated by the electroneutrality condition (eq. (1)) using the data on oxygen non-stoichiometry from refs. , , , . The Ea values were calculated from the corresponding lnσT=f(1/T) dependencies in the 370–570 K temperature range to exclude the effect of charge carrier concentration changes on the conductivity at high temperatures. The fraction of high-spin Ni3+ cations and the parameters of electron hole transport were obtained from the fitting results for the S=f(T) and μT=f(1/T) dependencies, as had been reported in , .
The total conductivity of La2−xAxNiO4+δ exponentially increases with x in the range of 0≤x≤0.4 (Fig. 3a). The higher σ values for La2−xSrxNiO4+δ in O2 atmosphere (PO2=1 bar) are explained by the increase in oxygen excess, which leads to an increase in the electron hole concentration (eq. 1) , , . As expected, La2−xSrxNiO4+δ shows slightly lower conductivity values than La2−xCaxNiO4+δ in air. Indeed, La2−xCaxNiO4+δ possessed higher oxygen excess (and concentration of electron holes) compared to La2−xSrxNiO4+δ at the same conditions .
The activation energies approach maximum in the range of 0.15≤x≤0.25, which is especially noticeably for the strontium doped system (Fig. 3b). In general, the activation energy can be expressed as follows , :
where Q and WH denote the heat of transfer (or the energy required to generate the carrier from the ideal state) and the polaron hopping energy, respectively. These parameters can be obtained from the experimental data using the following equations :
where A and B are the temperature independent constants, R and F are the universal gas constant and Faraday constant, respectively.
In our case the situation is more complicated: after an electron hole localized on Ni2+, low-spin or high-spin state Ni3+ cation can be formed. Assuming that low-spin and high-spin states coexist in the studied oxides both, Q and WH (from eq. 3), consist of partial contributions from QLS, QHS (Fig. 2d) and WLS, WHS (Fig. 3e), respectively. As can be seen from Fig. 3e the WLS and WHS parameters are significantly higher than 0, indicating that the small polaron hopping mechanism is predominant , . Figure 3d,e clearly show that calcium/strontium doping decreases QLS and QHS and, thus, facilitates formation of polarons. Therefore, the activation energy in La2−xAxNiO4+δ (0≤x≤0.4) is mainly defined by the hopping energy, namely the WLS values, since their contribution is significantly higher than that for WHS due to low concentration of high-spin Ni3+ (Fig. 3e).
As can be seen from Fig. 3c the lanthanum substitution with calcium or strontium promotes transformation of Ni3+ cations from low- to high-spin state. The doping gradually decreases the enthalpy of formation of high-spin Ni3+, particularly in La2−xSrxNiO4+δ. As a result, almost half of Ni3+ cations are in high-spin state in La1.6Sr0.4NiO4+δ at elevated temperatures (Fig. 3c). The transformation drastically reduces the mobility of electron holes (Fig. 3f); the μLS values are higher by almost one order of magnitude than μHS. One can also observe that La2−xSrxNiO4+δ show higher μLS values compared to La2−xCaxNiO4+δ. It should be noted that mobility is almost independent from x in the given range of dopant content. These findings confirm that charge carrier concentration (rather than mobility) plays a major role in electron hole transport when lanthanum is replaced by Ca and/or Sr in La2−xAxNiO4+δ at low substitution levels.
La1.6Ca0.4Ni1−yMeyO4+δ (Me=Fe, Cu)
In this section we continue the discussion on the data reported in ref. , representing the results as a function of dopant content (Fig. 4). As can be seen from Fig. 4a, the substitution of iron or copper for nickel in La1.6Ca0.4NiO4+δ leads to a decrease in total conductivity. For La1.6Ca0.4Ni1−yFeyO4+δ the conductivity noticeably decreased already at y=0.1, while for La1.6Ca0.4Ni1−yCuyO4+δ it remained almost constant, with a minor decrease in the range of 0.2≤y≤0.4. In both systems the concentration of electron holes is gradually reduced with y (Fig. 4a). However, it is still comparable or higher than that in La2−xAxNiO4+δ. It should be noted that contrary to iron, nickel, manganese and cobalt, the copper cations are mainly in +2 oxidation state in the La2NiO4-based materials .
Figure 4b shows that the activation energy of electron hole conductivity in La1.6Ca0.4Ni1−yFeyO4+δ linearly increased with y in the range of 0≤y≤0.2 as a result of the rapid increase in the QLS and WLS energies (Fig. 4d and e) and growing contribution of QHS and WHS (Fig. 4c). On contrary, the Ea values for La1.6Ca0.4Ni1−yCuyO4+δ were almost independent from y. One can observe that in La1.6Ca0.4Ni1−yCuyO4+δ the QLS and WLS values change insignificantly in the studied range of y and WHS increase is partially compensated by the QHS reduction.
The B-site doping of La1.6Ca0.4NiO4+δ with copper or iron decreases the ΔH values (Fig. 4c), thus, promoting the formation of high-spin Ni3+ cations. The comparison of effective ionic radii for A-site cations: lanthanum calcium strontium and B-site cations: nickel iron and copper let us assume that the size factor is the main driving force for the partial transformation of Ni3+ from low- to high-spin state in the La2NiO4-based materials, at least at low and intermediate doping levels .
The mobility of electron holes represented by low-spin Ni3+ cations is almost 2 orders higher than that represented by high-spin Ni3+ (see Fig. 4f). The μLS values remained unchanged in the 0≤y≤0.1 range for both iron- and copper-doped systems. At y>0.1 the μLS values for La1.6Ca0.4Ni1−yFeyO4+δ dropped almost twice, while for La1.6Ca0.4Ni1−yCuyO4+δ only a small increase had been observed.
The results reveal that the decrease in electron hole concentration is a major factor of decrease in conductivity with y in La1.6Ca0.4Ni1−yCuyO4+δ. For the iron-doped system both factors, electron hole concentration and mobility, were shown to have a significant impact on the decrease in conductivity with y.
La1.5A0.5Ni1−yMeyO4+δ (A=Ca, Sr; Me=Fe, Co)
Similar to La1.6Ca0.4Ni1−yFeyO4+δ, the total conductivity in the strontium-doped system La1.5Sr0.5Ni1−yFeyO4+δ, decreases with iron doping (Fig. 5a). The complete replacement of strontium by calcium, or iron by cobalt in La1.5Sr0.5Ni1−yFeyO4+δ results in a slight increase in σ values. La1.5Sr0.5Ni0.8Co0.2O4+δ possesses similar concentration of mobile electron holes compared to that for La1.5Sr0.5Ni0.8Fe0.2O4+δ. Interestingly that partial substitution of nickel with iron, in both, strontium and calcium doped systems, resulted in small difference in the concentration of electron holes (Fig. 5a).
Figure 5b shows exponential increase in the activation energy with y in La1.5Sr0.5Ni1−yFeyO4+δ. The substitution of strontium for calcium had a small effect on the Ea values. On contrary, the replacement of iron by cobalt has doubled the activation energy. The possible reason of such increase is relatively high values of WLS, WHS and QHS energies for La1.5Sr0.5Ni0.8Co0.2O4+δ.
It is interesting to note that iron doping in La1.5Sr0.5Ni1−yFeyO4+δ and La1.5Ca0.5Ni1−yFeyO4+δ increases enthalpy of formation of high-spin Ni3+ cations, which is opposite to the results obtained for La1.6Ca0.4Ni1−yMeyO4+δ (Me=Fe, Cu). Figure 5c shows that iron doping suppresses the formation of high-spin Ni3+ cations, at least in La1.5Sr0.5Ni1−yFeyO4+δ.
The mobility of electron holes that form low-spin Ni3+ cations (μLS) is gradually reduced with y in La1.5Sr0.5Ni1−yFeyO4+δ. Furthermore, it is seen from Fig. 5f that slightly higher conductivity values for La1.5Ca0.5Ni1−yFeyO4+δ can be explained by the increased mobility of electron holes. This result is opposite to that observed for La2−xAxNiO4+δ (A=Ca, Sr), where an increase in the concentration of electron holes was shown to be the main factor for conductivity growth with substitution of calcium for strontium. Finally, it should be noted that the μLS values for La1.5Sr0.5Ni0.8Co0.2O4+δ are higher than that for La1.5Sr0.5Ni0.8Fe0.2O4+δ. At the same time, the partial transformation of Ni3+ from low-spin to high-spin state in both, La1.5Sr0.5Ni0.8Fe0.2O4+δ and La1.5Sr0.5Ni0.8Co0.2O4+δ drops the mobility of electron holes (forming high-spin Ni3+) down to ~0.003–0.005 cm2 V−1 s−1 (Fig. 5f), although the concentration of high-spin Ni3+ is noticeably smaller in the latter, which agrees well with the assumption that the size factor is one of the main reasons for the LS to HS state transformation of Ni3+ at elevated temperatures.
La2−xSrxNi0.9Me0.1O4+δ (Me=Fe, Mn)
Figure 6a shows that total conductivity and the concentration of electron holes in La2−xSrxNi0.9Fe0.1O4+δ continue to increase with x at higher strontium content (x≥0.5). The La2−xSrxNi0.9Mn0.1O4+δ samples possesses lower σ and p values than that in La2−xSrxNi0.9Fe0.1O4+δ at the same strontium content. Moreover, the conductivity of La2−xSrxNi0.9Mn0.1O4+δ noticeably decreases at x>0.8. According to the results reported in , the observed decrease can be related to a significant growth in Q and W energies, which is reflected in the drastic increase in the activation energy as can be seen from Fig. 6b. For La2−xSrxNi0.9Fe0.1O4+δ the Ea value decreases with x, showing only minor growth at x>0.8 (Fig. 5b).
Contrary to La2−xAxNiO4+δ (x≤0.4), strontium-rich compositions La2−xSrxNi0.9Fe0.1O4+δ showed an increase in the QLS/QHS and WLS/WHS values with strontium doping (Fig. 6d and e). It should be noted that the QHS and WHS energies are significantly higher than QLS and WLS even at x>0.5, and their contribution to the activation energy increases with x as the fraction of high-spin Ni3+ continues to rise up to 0.5–0.6 at x=0.8 in La2−xSrxNi0.9Me0.1O4+δ (Fig. 6c). As can be seen from Fig. 6f the mobility of electron holes μLS gradually increases with strontium doping and partially compensates the increasing contribution of low μHS values.
Similarly to the La2−xSrxNiO4+δ (x≤0.4), the increase in concentration of electron holes with x in La2−xSrxNi0.9Fe0.1O4+δ remains the main factor for conductivity growth in the range of 0.5≤x≤0.8. However, considering the obtained results, one could expect a maximum on the σ=f(x) dependence at x>0.8 followed by a reduction of the σ values since the high-spin state of Ni3+ cations becomes predominant, decreasing the mobility of electron holes. Indeed, the decrease in conductivity at x>0.8 in La2−xSrxNiO4+δ was reported in numerous works , , .
Oxygen-ion transport in La1.5A0.5Ni1−yFeyO4+δ (A==Ca, Sr)
The oxygen-ion transport in La2−xSrxNi1−yFeyO4+δ was previously studied by means of oxygen permeation and the modified Hebb-Wagner polarization technique, which were referred to as Method 1 and Method 2, respectively . Briefly, the bulk oxygen-ion conductivity (σO) was shown to increase with y in La1.5Sr0.5Ni1−yFeyO4+δ despite the rise in activation energy, which was ascribed to the incorporation of additional interstitial oxygen ions (see Fig. 7a). The results also indicated a decrease in the ionic conductivity with strontium doping in the range of 0.5≤x≤0.8. The combined data on La2Ni0.9Fe0.1O4+δ  and La2−xSrxNi0.9Fe0.1O4+δ  presented in Fig. 7b clearly demonstrate a drop in the σO values with the increase in strontium content. Furthermore, the ionic conductivity of La1.5Sr0.5Ni0.6Fe0.4O4+δ calculated from the oxygen permeation flux was shown to be in good agreement with that obtained by the polarization technique .
Another important issue to consider while studying the oxygen-ion transport in the La2NiO4-based materials is surface exchange limitations , , . Figure 7a and b show critical membrane thickness (dc) as a function of dopant content. The critical membrane thickness can be determined by using the oxygen permeation data , , and interpreted as a thickness value, which corresponds to the change of rate-limiting step from surface exchange (for thinner samples) to bulk diffusion (for thicker samples). As can be seen from Fig. 7a and b strontium and iron doping leads to a fast increase in the dc values, thus, showing the increasing role of surface exchange in the oxygen permeation flux through the La2−xSrxNi1−yFeyO4+δ membranes. Further substitution of iron for nickel (y>0.3) decreases the critical membrane thickness down to zero indicating the elimination of surface exchange limitations in La1.5Sr0.5Ni0.6Fe0.4O4+δ at elevated temperatures (Fig. 7a). For La2−xSrxNi0.9Fe0.1O4+δ the dc values continued to increase in the range of 0.5≤x≤0.8 at T=1223 K. However, at T=1173 K the critical membrane thickness decreased almost fivefold in the same range of x and returned to ~0.5 mm, which was close to that for La2Ni0.9Fe0.1O4+δ. In general, the results indicate that the partial substitution of lanthanum and nickel with strontium and iron respectively can improve the surface exchange kinetics only at relatively high concentrations of dopants. At this stage it is difficult to interpret such dependencies unambiguously. While the key for understanding of this phenomena could be the microstructure and defect chemistry on the surface of samples , , further research is necessary in this field.
Figure 7c demonstrates the reciprocal temperature dependencies of oxygen-ion conductivity in La1.5A0.5Ni1−yFeyO4+δ (A=Ca, Sr) obtained by oxygen permeation and the polarization technique. The results for La1.5Sr0.5Ni0.7Fe0.3O4+δ show the importance of the surface exchange limitations: the σO values calculated directly from oxygen permeation flux (closed square symbols) are one order in magnitude lower than that recalculated ones (as described in ) taking into account the critical membrane thickness (open square symbols). Moreover, the ionic conductivity in La1.5Sr0.5Ni0.7Fe0.3O4+δ measured by the polarization technique correlates well with the former calculated results on oxygen ion conductivity indicating that redox processes inducing surface limitations on the La1.5Sr0.5Ni0.7Fe0.3O4+δ/microelectrode interface may still occur and have to be taken into account. Indeed, the results obtained both from oxygen permeation flux measurements and the polarization method are in good agreement for La1.5Sr0.5Ni0.6Fe0.4O4+δ – the composition with no surface exchange limitations (see Fig. 7c). It also should be noted that although the ionic conductivity of La1.5Sr0.5Ni0.7Fe0.3O4+δ is slightly lower than that for La1.5Ca0.5Ni0.7Fe0.3O4+δ at the same conditions, the difference between them lies within the experimental error.
In this work we combined the results on the electronic and ionic conductivity in the La2NiO4-based materials and discussed the effect of A-site (A=Ca, Sr) and B-site (Me=Fe, Cu, Co, Mn) doping on transport properties at different substitution levels. In general the electron hole conduction of La2−xAxNi1−yFeyO4+δ (y=0, 0.1) increases with calcium/strontium doping in the whole range of x studied (0≤x≤0.4 for Ca and 0≤x≤1 for Sr), owing it mainly to the increase in concentration of electron holes. The mobility of electron holes localized on Ni2+ forming low-spin Ni3+ cations, μLS, was a few orders in magnitude higher than that for electron holes forming high-spin Ni3+, μHS, at all substitution levels. The μLS and μHS changed insignificantly in the range of 0≤x≤0.4 in both, Ca- and Sr-doped system, while at higher strontium content, the increase in μLS was partially compensated by the growing contribution of low μHS values as the fraction of high-spin Ni3+ increased with x and exceeded 0.5 at x=0.8 in La2−xSrxNi0.9Fe0.1O4+δ.
The substitution of iron, copper, cobalt or manganese for nickel in La2−xAxNi1−yMeyO4+δ tends to hinder the electron hole conductivity at all studied y. While Fe3+ and Mn4+ cations behaved as electron hole traps, decreasing the concentration of electron holes and their mobility in La2−xAxNi1−yMeyO4+δ, the Cu2+ cations deteriorated the total conductivity only by reducing the amount of electron holes as their mobility was generally unchanged in the range of 0≤y≤0.3 at the same external conditions. Contrary to Fe3+, Mn4+ and Cu2+, the cobalt cations were considered to be in the mixed Co2+/Co3+ oxidation state in La2−xAxNi1−yMeyO4+δ. The concentration of Co3+ in La1.5Sr0.5Ni0.8Co0.2O4+δ was shown to be approximately 60% of all cobalt cations. La1.5Sr0.5Ni0.8Co0.2O4+δ showed only slightly higher electron hole conductivity compared with that for La1.5Sr0.5Ni0.8Fe0.2O4+δ.
The oxygen-ion conductivity in La2−xSrxNi1−yFeyO4+δ exponentially reduced with x in the range of 0≤x≤0.8 (at y=0.1), noticeably increasing with iron doping up to y=0.4 (at x=0.5). The obtained data revealed that elimination of surface exchange limitations can be expected only at relatively high concentrations of strontium and iron. The difference between the σO values in La1.5Ca0.5Ni0.7Fe0.3O4+δ and La1.5Sr0.5Ni0.7Fe0.3O4+δ was shown to be within the experimental errors. However, in order to obtain the final values of ionic conductivity for these oxides, the surface exchange limitations should be taken into account, which requires further development of the polarization technique.
A collection of invited papers based on presentations at the 16th International IUPAC Conference on High Temperature Chemistry (HTMC-XVI), held in Ekaterinburg, Russia, July 2–6, 2018.
This work was supported in parts by the Ministry of Education and Science of Russian Federation (State Task 4.2288.2017) and by Act 211 Government of the Russian Federation, agreement 02.A03.21.0006.
 V. V. Kharton, E. V. Tsipis, E. N. Naumovich, A. Thursfield, M. V. Patrakeev, V. A. Kolotygin, J. C. Waerenborgh, I. S. Metcalfe. J. Solid State Chem. 181, 1425 (2008).10.1016/j.jssc.2008.03.019Search in Google Scholar
 F. A. Kröger. The Chemistry of Imperfect Crystals, Vol. 2: 2nd Revised ed., North-Holland Publishing Company, Netherlands (1974).Search in Google Scholar
©2019 IUPAC & De Gruyter. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. For more information, please visit: http://creativecommons.org/licenses/by-nc-nd/4.0/