Different effects of Ni and Co substitution on the transport properties of BaFe${}_2$As${}_2$

We report resistivity and Hall effect results on Ba(Fe${}_{1-x}$Ni${}_x$)${}_2$As${}_2$ and compare them with those in Ba(Fe${}_{1-x}$Co${}_x$)${}_2$As${}_2$. The Hall constant $R_H$ is negative for all $x$ values from 0.01 to 0.14, which indicates that electron carriers dominate the transport in both the magnetic and the paramagnetic regimes. We analyze the data in the framework of a two-band model. Without any assumption on the number of carriers, we show that the electron resistivity can be estimated with good accuracy in the low-temperature paramagnetic range. Although the phase diagrams of the two families are very similar with respect to the extra electrons added in the system, we find that the transport properties differ in several aspects. First, we evidence that the contribution of holes to the transport is more important for Ni doping than for Co doping. Second, Ni behaves as a stronger scatterer for the electrons, as the increase of residual electron resistivity $\Delta {\rho }_e/x$ is about four times larger for Ni than for Co in the most doped samples.

Figure 1

Temperature dependence of the resistivity for (a) Ba(Fe${}_{1-x}$Ni${}_x$)${}_2$As${}_2$ and (b) Ba(Fe${}_{1-x}$Co${}_x$)${}_2$As${}_2$. Same colors were used as much as possible for respective $\rho (T)$ curves corresponding to equivalent electron doping, that is concentrations of Ni and Co of $2x$ and $x$ respectively. (c) Derivative of the resistivity as a function of the temperature for underdoped Ni samples. (d) Phase diagram indicating $T_{\mathrm{mag}}$ and $T_c$ for the two families. In order to allow comparison between the two families, Co content was divided by 2. Lines are guides for the eye.

Figure 2

(a) The Hall resistivity at different temperatures for $x=0.04$. Continuous lines are linear fits. (b) Temperature dependence of the Hall coefficient $R_H$ obtained for the Ba(Fe${}_{1-x}$Ni${}_x$)${}_2$As${}_2$ family.

Figure 3

Temperature dependence of the Hall number $n_H$ of (a) the Ba(Fe${}_{1-x}$Ni${}_x$)${}_2$As${}_2$ family and (b) the Ba(Fe${}_{1-x}$Co${}_x$)${}_2$As${}_2$ family.

Figure 4

Hall number extrapolated to $T=0$ for Ni and Co-doped samples. Ni concentration was multiplied by a factor of two. The continuous line is a linear fit to the Co data.

Figure 5

Ratio ${\rho }_e/\rho$ for different values of $a=\surd (1-2x/n_H)$. For each sample, $a$ is estimated on the basis of Hall effect measurements.

Figure 6

Temperature dependence of the electron resistivity estimated in a two-band model for (a) Ba(Fe${}_{1-x}$Ni${}_x$)${}_2$As${}_2$ and (b) Ba(Fe${}_{1-x}$Co${}_x$)${}_2$As${}_2$. The error bars are given by the lower and upper limit of ${\rho }_e$ in a two-band model, with no assumption on the number of carriers.

Figure 7

Electron resistivity at $T=0$ in a two-band model for Ni- and Co-doped samples. No assumption is made regarding the number of carriers and its variation with doping. Error bars are given by the extreme values estimated within this analysis. The fact that ${\rho }_e(T=0)$ saturates at large dopings for Ni doping indicates that the scattering rate of electrons increases in this case, whereas it is found almost independent of Co doping (see text).

Figure 8

Minimum value of ${\rho }_h/{\rho }_e$ estimated at 200 K for Ni- and Co-doped systems. Similar curves are obtained at all temperatures in the paramagnetic range.