Transcript Electronic structure and phase stability of MgTe, ZnTe

Systematical study of doping properties of MgTe
Ji-Hui Yang, Shiyou Chen, and X.G. Gong
Department of Physics and MOE laboratory for computational physical sciences, Fudan University, 200433
Su-Huai Wei
National Renewable Energy Laboratory, Golden, Colorado 80401, USA
Abstract
We studied the general chemical trends of defect formation in MgTe using first-principles band structure methods. The formation energies
and transition energy levels of intrinsic defects and extrinsic impurities and some defect complexes in zinc-blende MgTe were calculated
systematically using a new hybrid scheme. The limiting factors for n-type and p-type doping in MgTe were investigated. Possible solutions
to overcome the doping limitation of MgTe are proposed. The best n-type dopant is suggested to be N with nonequilibrium growth process
and the best p-type dopant is suggested to be I with its doping complex VMg+4ITe.
I. background
V. Results C: Compensation defects
I. Introduction:
• Optoelectronic device applications require that material can be doped both ntype and p-type.
• MgTe of zinc-blende structure and its ternary alloys might be good candidates
for low-cost thin films or high efficiency multi-juction solar cell materials.
• MgTe-related alloys have potential applications for the fabrication of short wavelength light-emitting devices and high efficiency solid-state light-emission devices.
• No theoretical and systematical research on its doping properties.
II. Doping limits:
(i) the formation energies of the desirable dopants are relatively large
(ii) the desirable dopants have sufficient solubility, but they produce deep levels,
which are not ionized at working temperature
(iii) their compensating defects could form spontaneously.
II. Calculation methods
Defect formation energy is defined as:
H f  , q   E , q    ni i  qEF
where
E , q   E , q   EMgTe    ni Ei   qEVBM
  q / q'  E  , q   E  , q' / q'q 
For acceptors:
VBM
VBM

MgTe   E  , q   E  ,0  q Dk 0/( q)
  0 / q    D 0   VBM
(2)
For donors:

MgTe   D 0 E , q E , q  E ,0  q Dk 0/ q
 g MgTe   0 / q   CBM
Then the formation energies can be given:
H f  , q   H f  ,0  q 0 / q   qEF
C
C
M
G
(1)
Defect transition energy level is defined as:
relax
Intrinsic compensation
(3)
C
C
M
G
Mg i2  is the main intrinsic
defect limiting p2
type doping while VMg plays the role of ntype compensating intrinsic defect. As a
result, for p-type doping, it’s better to be
under Te rich condition, which will cause
high formation energies for N and P doping
and limit the solubility of N or P. However,
this can be solved by epitaxial growth.
dopant compensation
If Na
𝑀𝑔Mgis used as an acceptor, the Fermi
energy level will be lowered as the
concentration
of Na increases and the

defect Na i becomes more stable. Thus,
more and more Na will move to the
interstitial site, compensating the p-type
NaMg
dopant
𝑔. This occurs under both the
Mg-rich and Te-rich conditions.
AX center
DX center
(4)
III. Results A: formation energy
Formation energies of neutral point defects: The effect of chemical potential and
potential limits requirement:
(i)Avoiding the precipitation of Mg, Te,
and elemental dopant A:
 Mg  0, Te  0,  A  0
(ii) MgTe should maintain stable:
 Mg  Te  H f MgTe 
(iii) Secondary compound should be
avoided:
or
n A  m Mg  H f Mg m An 
n A  m Mg  H f Mg m An 
C
C
M
G
IV. Results B: dopant selection
Acceptor levels:
Donor levels:
C
C
M
G
N and P could be important p-type dopants. I could be important n-type dopants.
Na could create both shallow acceptor and donor levels, which cause self-compensation.
AX formation energy is defined as:
E AX   E AX , q  E , q
For N, AX center is not a limiting factor
while for P, the AX center
is more stable.


 / AX 
The negative-U
transition energy
level is too deep and limits P doping.
DX formation energy is defined as:
EDX   EDX , q  E , q 
DX center is found to limit I doping.
VI. Results D: defect complex
For p-type doping, forming defect complex in general can’t low the
transition energy levels; for n-type doping, the defect complex or defect
cluster can low the transition energy levels to some extent. The doping
VMg  3BrTe
VMg  3ITe
complex
and
have a transition energy levels (0/-) of -0.09
VMg  4 I Te
eV and 0.16 eV. Besides, the doping cluster
has a (0/-) level of
VMg
0.07 eV, which may overcome the doping limitation of
compensation.
VII. Summary
The doping properties of MgTe have been systematically studied using
first-principles band structure methods. The formation energies and
transition energy levels of intrinsic defects and extrinsic impurities are
calculated using a new hybrid scheme, as well as the defect complexes.
Overall, our calculations suggest that the best p-type dopant should be
N with the formation energy lowered through nonequilibrium growth
VMg the
4 I Te best n-type dopant should be I with its defect
process and
complex
.
To be submitted. Thank you for your attention!