The authors wish to retract the referenced publication.1 The reported gate oxide used in calculating mobilities was approximately 100 times too small leading to mobilities that were nearly 100 times too large. In addition, experimental results such as a full set of transfer curves, hysteresis measurements, and gate leakage currents that would allow others to duplicate the experiments and compare their results to those reported in the paper were not presented. The authors sincerely apologize for these errors and omissions and regret any inconvenience they may have caused.

1.
X.
Song
,
Y.
Zhang
,
R.
Wang
,
M.
Cao
,
Y.
Che
,
J.
Wang
,
H.
Wang
,
L.
Jin
,
H.
Dai
,
X.
Ding
,
G.
Zhang
, and
J.
Yao
,
Appl. Phys. Lett.
106
,
253501
(
2015
).