Is there really spin transport in Alq$_{3}$ spin-valves?

There have been reports of GMR and extremely long spin relaxation in Alq$_{3}$- based spin valves.\footnote{Z. H. Xiong et al. Nature, 427, 821 (2004); S. Pramanik et al. Nature Nanotech. 2, 216 (2007).} However, it has also been suggested that direct tunneling through locally-thin regions of the Alq$_{3}$ layer could be the magnetoresistance (MR) mechanism, i.e. the reported MR may be due to artifacts rather than spin transport via the molecular levels in Alq$_{3}$.\footnote{W. Xu, et al. Appl. Phys. Lett. 90, 072506 (2007).} We present transport measurements on Alq$_{3} $-based spin valves and unipolar devices where the Alq$_{3}$ thickness is beyond the tunneling limit. The I-V characteristic is highly asymmetric and strongly temperature- dependent, different from the behaviors of devices where GMR has been reported. The charge transport in the Co/Alq$_{3}$/Fe spin valves is by holes only and is injection- limited. More importantly, we observe no measurable MR in our non-tunneling Co/Alq $_{3}$/Fe spin valves, or in Co/AlO$_{x}$/Alq$_{3}$/Fe structures where spins can be injected via the AlO$_{x}$ barrier. These results indicate that spin transport in Alq$_ {3}$ is unlikely.