Electrical Spin Injection from an Organic-Based Ferrimagnet in a Hybrid Organic-Inorganic Heterostructure

We report the successful extraction of spin-polarized current from the organic-based room temperature ferrimagnetic semiconductor $\mathrm{V}[\mathrm{TCNE}{]}_x$ ($x\sim 2$, TCNE: tetracyanoethylene; $T_C\sim 400\mathrm{K}$, $E_G\sim 0.5\mathrm{eV}$, ${\sigma }_{300\mathrm{K}}\sim {10}^{-2}\mathrm{S}/\mathrm{cm}$) and its subsequent injection into a $\mathrm{GaAs}/\mathrm{AlGaAs}$ light-emitting diode. The spin current tracks the magnetization of $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$, is weakly temperature dependent, and exhibits heavy-hole–light-hole asymmetry. This result has implications for room temperature spintronics and the use of inorganic materials to probe spin physics in organic and molecular systems.

Figure 1

(a) Schematic energy diagram for $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$. (b) Schematic band diagram of the spin LED. The study of the $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}/n\mathrm{\text{−}}\mathrm{AlGaAs}$ interface through which spin-polarized electrons are injected is complicated by the large and nonlinear resistance from bulk $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$, which is represented by the dashed box. (c) The expected polarization signal for HH/LH transitions (right) for a given magnetization of the spin injector (left). Because of quantum selection rules, HH and LH show opposite polarization behavior.

Figure 2

(a) I–V curves for both bare LED (triangles) and $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$ spin-LED (diamonds) devices at $T=60\mathrm{K}$. (b) Typical EL spectrum of a $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$ spin-LED device ($I=0.5\mathrm{mA}$, $V=+18.5\mathrm{V}$) at $T=60\mathrm{K}$. The shaded areas in the spectrum indicate the region of polarization integration over the quantum well heavy-hole (HH) and light-hole (LH) peaks, respectively. Inset: A schematic of the $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$ spin-LED device structure.

Figure 3

(a)–(c) Schematic signal from various sources. (a) Background from a nonmagnetic sample. (b) Background (a) (dots) plus the MCD response from $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$ (dashes). (c) HH and LH polarization from spin injection (long and short dashes, respectively) and backgrounds (dots). (d)–(f) Experimental results corresponding to (a)–(c): EL from a nonmagnetic device, MCD from $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$, and EL from a spin LED, respectively. Solid and open symbols are field sweeping up and down, respectively.

Figure 4

Circular polarization for HH and LH transitions in a $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$ spin-LED device at (a) $T=60\mathrm{K}$, $I=0.5\mathrm{mA}$, and $V=+18.5\mathrm{V}$, (b) $T=60\mathrm{K}$, $I=0.08\mathrm{mA}$, and $V=+15.2\mathrm{V}$, and (c) $T=140\mathrm{K}$, $I=1.5\mathrm{mA}$, and $V=+10.1\mathrm{V}$ (up and down triangles for HH and LH, respectively). The solid line in (a) represents out of plane magnetization of the $\mathrm{V}[\mathrm{TCNE}{]}_{x\sim 2}$ layer. Error bars are determined as in Fig. 3 except for (b), where the long measurement time required for low bias measurements prevents statistically significant averaging. As a result, the error is estimated from the scatter of the LH data.