Figure 1

(Color online) (a) Schematic of the Fe spin structure in the ${\text{BaFe}}_2{\text{As}}_2$ which has magnetic Bragg peaks at $Q=\left(1,0,1\right)$, (1,0,3), etc. For our experiment on ${\text{BaFe}}_{1.9}{\text{Ni}}_{0.1}{\text{As}}_2$, we use the same unit cell for easy comparison. (b) Temperature dependence of the spin gap as determined from energy scans [Fig. 3c] and the temperature dependence of the scattering at $Q=\left(1,0,1\right)$ [Fig. 3d]. The solid curve represents the temperature dependence of the BCS gap function. [(c) and (d)] Schematic of the magnetic response and spin gaps at $Q=\left(1,0,0\right)$, and (1,0,1), respectively. Measurements at $Q=\left(1,0,3\right)$ showed similar behavior as those at $Q=\left(1,0,1\right)$.Reuse & Permissions

Figure 2

(Color online) Examples of constant energy scans around the (1,0,3) position for $E=1\text{meV}$ obtained with $E_f=5\text{meV}$ above and below $T_c$ on SPINS. (a) $Q$ scan along the $\left[1,0,L\right]$ direction for $E=1\text{meV}$ at 24 and 2 K. A clear peak centered at (1,0,3) at 24 K disappears at 2 K, indicating the opening of a spin gap. (b) Similar scan along the $\left[H,0,3\right]$ direction showing a peak centered at (1,0,3) that disappears below $T_c$. (c) Using scattering at 2 K as background scattering, we determine the normal-state $L$ modulation of the spin fluctuations by subtracting the 2 K data from 24 K data. It is clear that spin fluctuations are 3D and have similar modulations along the $c$ axis as spin waves. (d) $Q$ scan in the superconducting state through the magnetic-resonance position and above $T_c$ near (1,0,1).Reuse & Permissions

Figure 3

(Color online) (a) $Q$ scans at the $\left[H,0,1\right]$ direction above and below $T_c$ at $\hslash \omega =2\text{meV}$. The data show the opening of a spin gap at 2 meV below $T_c$. (b) $Q$ scans along the $\left[1,0,L\right]$ (signal) and $\left[1.2,0,L\right]$ (background) positions, showing the $L$ modulation of the intensity with maxima at (1,0,1) and (1,0,3). (c) Constant $Q$ scans at $Q=\left(1,0,1\right)$ at various temperatures. The differences between low- and high-temperature data show negative scattering due to the opening of a spin gap. The data suggest a spin-gap value of 1.5 meV at 15 K and 3.0 meV at 2 K. (d) Temperature dependence of the scattering at $Q=\left(1,0,1\right)$ and $E=1\text{meV}$ shows a sudden drop below $18\text{K}\left(=T_c-2\text{K}\right)$ suggesting that the $E=1\text{meV}$ spin gap opens at a temperature slightly below $T_c$.Reuse & Permissions

Figure 4

(Color online) Constant $Q$ scans around the (a) $Q=\left(1,0,0\right)$ and (b) (1,0,1) positions above and below $T_c$, showing the development of the spin gap at low energies, and the enhancement of the magnetic scattering at the resonance energy at each wave vector. Background data are indicated by the open diamonds and dashed curves, and were collected at $Q=\left(1.2,0,0\right)$ and $Q=\left(1.2,0,1\right)$, respectively. [(c) and (d)] ${\chi }^{″}\left(Q,\omega \right)$ above and below $T_c$ obtained by subtracting background and removing the thermal factor (see text). Also shown are values obtained from the constant $E$ scans at various energies above and below $T_c$, which are consistent with the results of constant $Q$ scans, suggesting that the subtracted results are reliable. At both wave vectors there is a clear magnetic intensity gain at the resonance energy of $E=9.0\text{meV}$ at $Q=\left(1,0,0\right)$ and 7 meV at $Q=\left(1,0,1\right)$ and spin gaps of 4.3 and 2.5 meV, respectively. The solid lines show fits using the model described in the text.Reuse & Permissions