Modeling of GalnP/GaAs Dual-Junction Solar Cells including Tunnel Junction

Mathieu Baudrit and Car/os A/gora

Instituto de Energia Solar, Universidad Politecnica de Madrid, Spain

[email protected]

ABSTRACT TUNNEL JUNCTION MODEL

This paper presents research efforts conducted at the The local band-to-band tunneling models use the
IES-UPM in the development of an accurate, electric field value at each node to give a generation
physically-based solar cell model using the general­ rate at that point due to the tunneling. In reality, the
purpose ATLAS® device simulator by Silvaco. Unlike tunneling process is non-local and is necessary to take
solar cell models based on a combination of discrete into account the spatial profile of the energy bands. It
electrical components, this novel model extracts the is also necessary to take into account the spatial
electrical characteristics of a solar cell based on virtual separation of the electrons generated in the
fabrication of its physical structure, allowing for direct conduction band from the holes generated in the
manipulation of materials, dimensions, and dopings. valence band [2]. ®
As single junction solar cells simulation was yet A model for this process has been created for ATLAS
achieved, the next step towards advanced simulations [3]. It assumes that the tunneling can be modeled as
of multi-junction cells (MJC) is the simulation of the being one-dimensional in nature so that it can be
tunnel diodes, which interconnect the subcells in a calculated using a special rectangular mesh
monolithic MJC. The first results simulating a Dual­ superimposed over and coupled to the ATLAS® mesh.
Junction (DJ) GalnP/GaAs solar cells are shown in this This mesh needs to include the junction region of
paper including a complete Tunnel Junction (TJ) model interest and the direction of the band-to-band
and the resonant cavity effect occurring in the bottom
tunneling, which is generally perpendicular to
cell. Simulation and experimental results were
compared in order to test the accuracy of the models junction interface, must be specified.
employed.

Index Terms: Multi-Junction Solar Cells, Tunnel Eft

1
r - .- - - - - - - - - - - - -

Junction, Simulation Ev
I I

Enermr Range
lor Band4lrBand
Tunneling

// j
I
INTRODUCTION
LEDs, lasers and Multi-junction Solar Cells can all
employ tunnel junctions to improve performance. E ~
Calculating the effects of this junction is tricky, but ~-- ___ ~ _____ ~~7-----------
there are ways to accurately simulate chip
characteristics and cost-effectively optimize the
structure's design. After the successful simulation of
IIIN single junction solar cells [1], we tested the known Figure 1. Schematic band diagram of a tunnel junction.
tunneling models such as Hurkx, Kane and Klaassen.
Despite their good behavior in reverse bias they do not In order to explain how the tunneling current is
manage to reproduce the JV curve of a tunnel diode in calculated, let us consider an energy band profile
forward bias. along each slice in the tunneling direction when
An intense work was then done to create a reliable and applying a reverse bias across the junction. Figure 2
flexilble model to allow the complete electrical shows a schematic of this, together with the allowed
simulation of tunnel diodes, especially in forward bias. range of valence band electron energy for which
After the development of a non-local tunneling model, tunneling is permitted.
room was opened to Dual-Junction solar cell modeling. At moderate doping levels, a tunneling effect can be
In this paper we present the first results obtained by seen in reverse bias, but if the junction doping levels
IES-UPM regarding MJC simulation using Silvaco are high enough, then this energy range may also exist
ATLAS®. We show that when correctly adjusted, the in forward bias and tunneling effect can also be
software allows a good fit between simulation and appreciated. If we consider only elastic scattering
experimental results for both Tunnel Junction and mechanisms, then electrons from anywhere in the
Dual-Junction solar cell. permitted energy range can tunnel from the valence
band to the conduction band. ATLAS® considers each

978-1-4244-1641-7/08/$25.00 ©2008 IEEE

energy in the allowed range and determines the spatial local one, affecting the current after the Jvall ey and
start and end positions for the tunneling at each increasing its value
energy, E, which we label Xbeg and Xend. respectively,
the contribution to tunneling current for an electron in
the energy range from E-11E12 to E+11E12 (where l1E is
a small energy increment) is

where T(E) is the tunneling probability, T the

temperature, En and E'r are defined on Figure 1 and

Figure 2: Simulated tunnel junction

(2)

Non-local TAT
- - - - - i.. ~ ~.
f~2l\~. .:> .

.~'-~'i:t
In equilibrium, En = E'r and the current is zero as
expected. This contribution to the tunneling current is J
calculated and coupled into the mesh at Xbeg and Xend.
ATLAS® uses a transmission matrix method to <d/ -
calculate the tunneling probability for direct quantum _~..r I:.... "'-~ • .,
tunneling simulations through an insulator. In the case . - --. ( . r
of band-to-band tunneling, however, a carefully applied
Wenzel-Kramers-Brillouin (WKB) method was found to ___________
:
~\._
,?::.~
c/' Local TAT
give equivalent results and is computationally more
efficient.
Figure 3: Trap Assisted Tunneling mechanisms.

Figure 4 shows the very good fit obtained as a result of

TJ: COMPARISON WITH EXPERIMENTAL the tunnel junction simulation made with Silvaco
RESULTS ATLAS® software including the non local tunneling
model described before as well as TAT and non local
In order to calibrate the nonlocal tunnel model, an TAT. Results have been obtained after adjusting the
isolated p++-GaAs / n++-GaAs tunnel diode was grown material parameters affecting the tunneling effect such
by MOVPE has been investigated. The simulated as the effective mass for both holes and electrons, the
structure is shown in Figure 2. trap concentration and the limits of the superimposed
The tunnel diode consists of the tunnel junction itself mesh. In this case, a high specific contact resistance
made by two degenerately doped n ++- an d++GA p - a s was put in evidence, reaching actually the value of
layers as well as two enclosing barrier layers with the O.3e-3 mohm'cm 2 in the case of the tunnel junction
purpose to minimize dopant diffusion [4]. The cap layer presented in figure 4 but reaching sometimes the value
and the substrate have been included in the of 1.5e-2 ohm·cm 2 . This very high value is due to the
simulation. For the purpose of this paper, a tunnel metal/semiconductor contact, actually we saw that the
diode with a sharp uniform doping profile was metallization grid on some measured tunnel diode was
modeled. of very poor quality, inducing then a very high series
Local and non local Trap Assisted Tunneling (TAT) resistance. We also have to note that the real doping
mechanisms have been included in the models profile, which has been measured for other tunnel
involved in tunneling effects because of their influence junctions, is important to allow a faster fit between
on the simulated IV curve. experimental and simulation data.
Figure 3 resumes how the TAT works in the case of
tunnel junction. It is indeed divided in two different
effects: the local one affecting mainly the peak current
and the decreasing slope of the IV curve and the non-

978-1-4244-1641-7/08/$25.00 ©2008 IEEE

 E)(p.t"im .... tal Data
Sim .... latio ...
i I
~ ---- ------ f--
t
-.-- CiaI ... p E)(pet"ime ... tal Data +

i I
CiaI ... p Sim .... latio ...
CiaAs E)(pet"ime ... tal Data
CiaAs,Sim .... latio ...
~
x
~

!
. 1

F
i [
i

~---------!.--~----
\;:\\,
_._---- ------- i-~-~ --._-_.---
I-
<I, i i
i fi \+ i i
f , .
.
I "-
.i I +.:1"
I I
-----L .'
['~
I
..*'"
~/ ....-!.-••••
.... ~
i
I

.,
/,j'/
.
j
I
.., ,
i
•• '-:;-.3_ _ -!-:__ -!-:-_ _ -,L_ _l~....~,---...L_ _..J
\ 1

'.7
L

Figure 2: Experimental (dots) and simulated wavele ... gth <. . m)

(solid line) Tunnel Junction IV curve. Figure 3: Experimental (dots) and simulated
(solid lines) External Quantum Efficiency of the
Dual-Junction solar cell.
DUAL-JUNCTION SOLAR CELL SIMULATION
Once we are able to simulate the tunnel diode, the Table 1 describes the layer structure of the simulated
next step is the modeling of a complete Dual-Junction Dual-Junction solar cell. The results in Figure 5 show a
solar cell. We will focus in this paper on the External good agreement with experimental data. The small
Quantum Efficiency (EQE), the IV curve at 1 sun and mismatch is probably due to the differences between
the dark IV curve. nominal thicknesses (introduced in the simulation) and
real values which were not measured exactly. We will
further make an in depth characterization (Doping
External Quantum Efficiency concentration and thicknesses) to perfect the fit
between experimental and simulated results.
In a GalnP/GaAs dual-junction solar cell, GaAs bottom
cell suffers oscillations of its External Quantum
Efficiency because of the resonant cavity effect JV curve at 1 sun
occurring between the top cell BSF and the bottom cell
~indow layer. In this case, using traditional ray-tracing
Once the EQE simulation agreed well with the
IS useless because reflected rays in each layer should
measurement and we can do do a good J sc estimation
be set to a very large number and simulation time using a simulation software, it was necessary to
increases exponentially with the internal reflection achieve an JV curve simulation to see how the models
number. The only way to achieve a good accuracy is to can predict the Voc, the fill factor and the efficiency of
use the Characteristic Matrix Method or the Transfer the solar cell.
Matrix Method.
Silvaco ATLAS® uses the Characteristic Matrix
approach that relates total tangential components of
the ele~tric and magnetic fields at the multilayer
boundanes. The structure of a multilayer completely
determines the characteristic matrix of this multilayer.
The transfer matrix also contains information about the
media on both sides of the multilayer.

Name Material Thickness N [cm-~l

topfsf AlinP 50 nm n - 3.010
topem GaO.51I nO.49P 170 nm n =1.8 10+ 18
top base GaO.51I nO.49P 800nm p =1.0 10+ 17
topbsf AIGalnP 100 nm p =3.010+ 17
phighTD GaAs 50 nm p =5.010+ 19
nhighTD GaAs 50 nm n =3.010+ 19
botfsf Alo.4Gao.eAs 50 nm n =2.010+ 18
botem GaAs 100 nm n =1.0 10+ 10 V (V)

botbase GaAs 3500 nm p =2.010+ 17 Figure 4: Experimental (dots) and simulated (solid line)
botbsf AI0.3GaO.7As 100 nm p =2.010+ 18 IV curve
subs GaAs 300 IJm p =2.010+ 18
As demonstrated before by the EQE simulation, we
Table 1. Nominal semiconductor structure of
see a very good fit of the J sc using the AM1.5 Low­
the Dual-Junction solar cell.
AOD spectra (see Figure 6), however, slight

978-1-4244-1641-7/08/$25.00 ©2008 IEEE

disagreements can be observed. Table 2 below As seen before with the IV curve under illumination,
resumes the main characteristics of the cell. there is a misfit between experimental and simulation
data, especially for middle range voltage. If we
compare Figure 6 and Figure 7, we also see the
Experimental Simulation mismatch voltage range is different but the dark JV
results also shows a shunt resistance effect.
Jse (mAlem<) 8.4 8.4 As in this case there is no light input to the device and
Voe (V) 2.12 2.12 differences can be observed, light interaction as
Jmax (mAlem 2) 7.9 7.66 nothing to do with this. However, viewing the results
Vmax (V) 1.9 1.82 and the good results obtained before simulating single
FF (%) 84 78 junction solar cells, we think the mismatch is due the
11 (%) 15 13.9 tunnel model once introduced inside a complete dual­
junction solar cell structure. More, we think this is due
Table 2: Parameters extracted from the to how the tunneling model manage traps as we saw
experimental and simulated IV curve the kind of curve seen in Figure 7 in some structures
with a very high traps concentration, with the difference
that in our case the traps concentration cannot explain
The OJ solar cell simulated had no Anti-Reflective such a difference.
Coating layers and is a low/medium quality solar cell The revision of the tunnel model will be our priority in
which can explain the low efficiency at 1 Sun and the next months as, despite of the good results, we
shows that the models are not only able to reproduce need it to be totally reliable.
high quality solar cells but also lower quality devices.
Once again the simulation shows a very good fit with
the experimental data on the Voc prediction. However,
some differences can be seen on the FF value and so CONCLUSIONS
on on the J max, Vmax and Efficiency.
Before the existence of reliable non-local tunneling
As good results were obtained before modeling single model, the simulations of dual-junction solar cells were
junction solar cells at IES-UPM, we suspected the
limited to optically coupled simulations, separating
tunneling model used in the OJ solar cell. To be exact then the cells electrically. With the development of a
we suspected that the model doesn't behave as it
reliable non-local tunneling model, a complete
should under illumination, this why we simulated a electrical simulation is now possible. It was shown that
dark IV curve to localize the origin of the differences
in contrast to local tunneling models the non local
under illumination.
tunneling model reproduces the measured JV curve of
a tunnel diode structure in a large voltage range very
well, especially in the decisive range of operation when
applied to multi-junction solar cells.
We also proved the validity of the model, which
coupled with an adequate optical modeling method,
Dark JV curve allows to reproduce very well the EQE of a OJ solar
cell, making the J sc prediction accurate. Regarding the
JV curves, despite the good results obtained on Voc
In contrast with the other OJ solar cells simulations estimation, we detect a problem affecting the values of
presented in this paper, the dark IV curve only relies the FF and the efficiency. We think this problem is due
on the electrical models as no light is input to the to the tunneling model behavior and a revision of it has
device. to be done.
However, this paper shows that solar cells simulation
under a TCAO environment is possible and can be
Da..k IV curve
predictive for single junction solar cells as well as for
OJ solar cells. Room is open for Multi-Junction solar
8.1 Cells simulation as we now know the critical part of this
structure, the tunnel junction, is well managed.
The future works at the IES-UPM will be the
B.8.
enhancement of the simulated structure to triple
junction solar cells and the inclusion of real operation
0.06 conditions as they can strongly change the
~
optimization of the cell.
~
8.04

Acknowledgements
0.02

The authors would like to thank the Silvaco Grenoble

Research Center (G.RE.CE.) for its collaboration.
Thanks also to Ignacio Rey-Stolle, Beatriz Galiana,
8.' 1.' 2.' Ivan Garcia and Pilar Espinet of the IES-UPM who
Tension (V) made the experimental data possible.
Figure 5: Experimental (dots) and simulated (solid line)
dark IV curve

978-1-4244-1641-7/08/$25.00 ©2008 IEEE

This paper has been supported by the European
Commission under contract SES6-CT-2003-502620
(FULLSPECTRUM project). The Spanish Ministerio de
Educaci6n y Ciencia has also contributed with the
CONSOLIDER- INGENIO 2010 program by means of
the GENESIS FV project (CSD2006-004) and also with
the research projects with references TEC2005-02745
and TEC2004-22300-E as well as the Comunidad de
Madrid under NUMANCIA programme (S-
05050/ENE/031 0).

REFERENCES
[1] C.Algora, et ai, Pending Issues In The Modeling of
Concentrator Solar Cells, Proceedings 19th
European Photovoltaic Solar Energy Conference,
Paris, 2004.
[2] G. A. M. Hurkx, D. B. M. Klaassen, and M. P. G.
Knuvers, A new recombination model for device
simulation including tunneling in IEEE
Transactions on Electron Devices 39, 331-8
(1992).
[3] Silvaco International, Silvaco User's Manual, ed.
Silvaco, 2006.
[4] T. Takamoto, et ai, Mechanism of Zn and Si
diffusion from a highly doped tunnel junction for
InGaP/GaAs tandem solar cells in Journal of
Applied Physics 85, 1481-6 (1999).

978-1-4244-1641-7/08/$25.00 ©2008 IEEE