TY - JOUR
T1 - Model for thickness dependence of mobility and concentration in highly conductive zinc oxide
AU - Look, David C.
AU - Kevin, D. Leedy
AU - Arnold Kiefer, Kiefer
AU - Bruce Claflin, Claflin
AU - Itagaki, Naho
AU - Matsushima, Koichi
AU - Surhariadi, Iping
N1 - Funding Information:
We wish to thank T. A. Cooper for the Hall-effect measurements and W. Rice for the reflectance measurements. Support for DCL was provided by AFOSR Grant FA9550-10-1-0079 (J. Hwang), NSF Grant DMR0803276 (C. Ying), DOE Grant DE-FG02-11ER46820 (R. Kortan), and AFRL Contract HC1047-05-D-4005 (D. Tomich).
PY - 2013
Y1 - 2013
N2 - The dependences of the 294 and 10 K mobility μ and volume carrier concentration n on thickness (d = 25 to 147 nm) are examined in aluminum-doped zinc oxide (AZO). Two AZO layers are grown at each thickness, one with and one without a 20-nm-thick ZnON buffer layer. Plots of the 10 K sheet concentration ns versus d for buffered (B) and unbuffered (UB) samples give straight lines of similar slope, n = 8.36 × 1020 and 8.32 × 1020 cm-3, but different x-axis intercepts, δd = -4 and +13 nm, respectively. Plots of ns versus d at 294 K produce substantially the same results. Plots of μ versus d can be well fitted with the equation μ(d) = μ(∞)/[1 + d*]/(d - δd)], where d* is the thickness for which μ(∞) is reduced by a factor 2. For the B and UB samples, d* = 7 and 23 nm, respectively, showing the efficacy of the ZnON buffer. Finally, from n and μ(∞) we can use degenerate electron scattering theory to calculate bulk donor and acceptor concentrations of 1.23 × 1021 cm-3 and 1.95 × 1020 cm -3, respectively, and Drude theory to predict a plasmonic resonance at 1.34 μm. The latter is confirmed by reflectance measurements.
AB - The dependences of the 294 and 10 K mobility μ and volume carrier concentration n on thickness (d = 25 to 147 nm) are examined in aluminum-doped zinc oxide (AZO). Two AZO layers are grown at each thickness, one with and one without a 20-nm-thick ZnON buffer layer. Plots of the 10 K sheet concentration ns versus d for buffered (B) and unbuffered (UB) samples give straight lines of similar slope, n = 8.36 × 1020 and 8.32 × 1020 cm-3, but different x-axis intercepts, δd = -4 and +13 nm, respectively. Plots of ns versus d at 294 K produce substantially the same results. Plots of μ versus d can be well fitted with the equation μ(d) = μ(∞)/[1 + d*]/(d - δd)], where d* is the thickness for which μ(∞) is reduced by a factor 2. For the B and UB samples, d* = 7 and 23 nm, respectively, showing the efficacy of the ZnON buffer. Finally, from n and μ(∞) we can use degenerate electron scattering theory to calculate bulk donor and acceptor concentrations of 1.23 × 1021 cm-3 and 1.95 × 1020 cm -3, respectively, and Drude theory to predict a plasmonic resonance at 1.34 μm. The latter is confirmed by reflectance measurements.
UR - http://www.scopus.com/inward/record.url?scp=84879568867&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84879568867&partnerID=8YFLogxK
U2 - 10.1117/1.OE.52.3.033801
DO - 10.1117/1.OE.52.3.033801
M3 - Article
AN - SCOPUS:84879568867
SN - 0091-3286
VL - 52
JO - SPIE J
JF - SPIE J
IS - 3
M1 - 033801
ER -