Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes
The influence of non-radiative exciton recombination (NRER) on the photoconversion efficiency in silicon solar cells with short Shockley–Read–Hall lifetimes τSRH has been studied. It has been shown that the efficiency reduction due to this effect is stronger the shorter τSRH. The influence of NRER i...
Saved in:
| Published in: | Semiconductor Physics Quantum Electronics & Optoelectronics |
|---|---|
| Date: | 2017 |
| Main Authors: | , , , , , , |
| Format: | Article |
| Language: | English |
| Published: |
Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
2017
|
| Online Access: | https://nasplib.isofts.kiev.ua/handle/123456789/214915 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Journal Title: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Cite this: | Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes / A.V. Sachenko, V.P. Kostylyov, V.M. Vlasiuk, R.M. Korkishko, I.O. Sokolovskyi, V.V. Chernenko, and M.A. Evstigneev // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2017. — Т. 20, № 1. — С. 34-40. — Бібліогр.: 8 назв. — англ. |
Institution
Digital Library of Periodicals of National Academy of Sciences of Ukraine| _version_ | 1860301490427527168 |
|---|---|
| author | Sachenko, A.V. Kostylyov, V.P. Vlasiuk, V.M. Korkishko, R.M. Sokolovskyi, I.O. Chernenko, V.V. Evstigneev, M.A. |
| author_facet | Sachenko, A.V. Kostylyov, V.P. Vlasiuk, V.M. Korkishko, R.M. Sokolovskyi, I.O. Chernenko, V.V. Evstigneev, M.A. |
| citation_txt | Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes / A.V. Sachenko, V.P. Kostylyov, V.M. Vlasiuk, R.M. Korkishko, I.O. Sokolovskyi, V.V. Chernenko, and M.A. Evstigneev // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2017. — Т. 20, № 1. — С. 34-40. — Бібліогр.: 8 назв. — англ. |
| collection | DSpace DC |
| container_title | Semiconductor Physics Quantum Electronics & Optoelectronics |
| description | The influence of non-radiative exciton recombination (NRER) on the photoconversion efficiency in silicon solar cells with short Shockley–Read–Hall lifetimes τSRH has been studied. It has been shown that the efficiency reduction due to this effect is stronger the shorter τSRH. The influence of NRER is most evident when the NRER time becomes shorter than τSRH. At sufficiently short τSRH, NRER substantially limits the optimal base doping levels of silicon solar cells, at which the photoconversion efficiency is maximal.
|
| first_indexed | 2026-03-21T19:35:52Z |
| format | Article |
| fulltext |
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2017. V. 20, N 1. P. 34-40.
doi: https://doi.org/10.15407/spqeo20.01.034
© 2017, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
34
PACS 72.20.J, 78.60.J
Influence of non-radiative exciton recombination
in silicon on photoconversion efficiency.
2. Short Shockley–Read–Hall lifetimes
A.V. Sachenko1,*, V.P. Kostylyov1, V.M. Vlasiuk1, R.M. Korkishko1, I.O. Sokolovskyi1,2,
V.V. Chernenko1, and M.A. Evstigneev2
1V. Lashkaryov Institute of Semiconductor Physics, NAS of Ukraine,
41, prospect Nauky, 03028 Kyiv, Ukraine
2Department of Physics and Physical Oceanography,
Memorial University of Newfoundland, St. John’s, NL, A1B 3X7 Canada
*E-mail: sach@isp.kiev.ua
Abstract. The influence of non-radiative exciton recombination (NRER) on the
photoconversion efficiency in silicon solar cells with short Shockley–Read–Hall
lifetimes τSRH has been studied. It has been shown that the efficiency reduction due to this
effect is the stronger the shorter τSRH. The influence of NRER is most evident when the
NRER time becomes shorter than τSRH. At sufficiently short τSRH, NRER substantially
limits the optimal base doping levels of silicon solar cells, at which the photoconversion
efficiency is maximal.
Keywords: solar cells, photoconversion, non-radiative exciton recombination, Shockley–
Read–Hall lifetime.
Manuscript received 15.11.16; revised version received 02.02.17; accepted for
publication 01.03.17; published online 05.04.17.
1. Introduction
In the first part of this work [1], the influence of non-
radiative exciton recombination (NRER) on the
photoconversion efficiency in silicon was analyzed for
the case of long Shockley–Read–Hall lifetimes
τSRH > 100 μs. It was shown that upon reduction of τSRH
the role of this recombination channel increases, which
leads to stronger lowering the photoconversion
efficiency. Therefore, it is interesting to investigate this
effect in the opposite case τSRH < 100 μs. The respective
analysis has been performed in this work.
In contrast to [1], the following criteria are
fulfilled: (1) excess electron-hole pair density Δn is
much lower than the equilibrium density of charge
carriers n0 in the base region; (2) diffusion length Ld of
electron-hole pairs is much smaller than the base
thickness d.
Experimental part of this work includes the
illuminated current-voltage curves and photocurrent
quantum yield results measured using two p-n junction-
based silicon solar cells (SC) with short values of τSRH.
Apart from determination of the main characteristics of
the SCs studied under AM1.5 conditions, temperature
dependence of the short-circuit current ISC, open-circuit
voltage VOC , current-voltage curve fill factor FF, and
photoconversion efficiency η were measured. As our
analysis has shown, at very low base doping levels,
NRER has practically no effect on these parameters and
on their sensitivity to temperature. At the same time, at
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2017. V. 20, N 1. P. 34-40.
doi: https://doi.org/10.15407/spqeo20.01.034
© 2017, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
35
moderate base doping levels this recombination
mechanism leads not only to the decrease of these
characteristics, but also to a small change in the slope of
the curves VOC (T), FF (T), and η (T). In the case of
small Shockley–Read–Hall lifetimes and heavy base
doping, the reduction of photoconversion efficiency
caused by the NRER becomes quite essential. For the SC
parameters used in this work, it can exceed 40% of the
efficiency value obtained without taking NRER
mechanism into account.
2. Experimental procedure
SC samples with the structure p+-n-n+ and n+-p-p+ were
based on n- and p-type silicon wafers fabricated using
Czochralski method. Because the effect of non-radiative
exciton Auger recombination via the deep levels on the
photoelectric SC parameters is the stronger the shorter
Shockley–Read–Hall lifetime [1], the original Si wavers
were chosen with τSRH of about 10–6…10–5 s. The
dividing frontal p+-n and isotypic rear p-p+ junctions
were formed by diffusing boron into the phosphorus-
doped wafers, and the n+-p and n-n+ junctions were
produced by diffusion of phosphorus into boron-doped
wafers. As an antireflecting coating, thermal silicon
dioxide layers with the thickness ca. 110 nm were used.
The initial parameters of silicon and the main
properties of the samples studied under AM1.5
conditions are indicated in Table. Here, n0 is the doping
density, d – sample thickness, S0 – surface
recombination velocity, and RS – series resistance.
The obtained in this manner SC samples were
characterized by very short Shockley–Read–Hall
lifetimes, and quite large p-n junction depth, which
rendered the photoconversion efficiency below 10%.
They were produced using the technology, which allows
one to reduce the original lifetime in the base region in
order to be able to determine the diffusion length of
minority carriers from the open-circuit voltage.
To imitate solar radiation, tungsten lamps with the
radiation temperature of 2800 K were used. The
illuminated I-V curves for two SC samples, one with the
n-type and the other with the p-type base, are shown in
Fig. 1. The spectral dependence of the internal quantum
yield is presented in Fig. 2. The temperature dependence
of the short-circuit current, JSC, open-circuit voltage,
VOC, the fill factor, FF, and photoconversion efficiency,
η, measured in the range from 25 to 60 °С, are shown in
Figs. 3-6.
0.0 0.2 0.4 0.6
0
20
40
C
ur
re
nt
I,
m
A
Voltage V, V
n-type
p-type
Fig. 1. Theoretical (lines) and experimental (symbols)
illuminated current-voltage curves obtained on the SCs with
n- and p-type base region.
3. Theoretical analysis
In our theoretical analysis, the following assumptions
were employed to develop a relatively simple analytical
description of the experiment.
First, apart from the criterion Ld < d we assume that
the heavily doped region of the p-n junction has a little
effect on the photocurrent internal quantum yield
q (λ, T). This allows us to use the following expression
for this parameter:
( )
( ) ( )p
d
d dT
TLT
TLTTq ),(exp
),(1
),(),( λα−
λα+
λα
=λ , (1)
where α (λ, T) is the absorption coefficient for light of
the wavelength λ at temperature T, and dp – “dead layer”
thickness, i.e. the thickness of the heavily doped region.
In general, the short-circuit current density can be
written as
( ) ∫
λ
λ
λλλ−=
)(
0
)(),(1)(
T
LSC
m
dITqrqTJ , (2)
where q is the elementary charge, rL – mean coefficient
of photocurrent losses related with the incomplete
absorption of light caused by reflection and availability
Table. Silicon parameters and the main characteristics of the samples investigated.
Base
type
n0,
cm–3
τSRH,
s ASC, cm2 d,
μm RS, Ohm S0,
cm/s JSC, mA/cm2 VOC, V FF, % η, %
n 3.1·1015 2.5·10–5 2 380 0.6 2·103 22.4 0.568 76.6 8.31
p 3·1015 7·10–7 2 350 1.5 2.25·104 23.1 0.503 67.5 7.86
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2017. V. 20, N 1. P. 34-40.
doi: https://doi.org/10.15407/spqeo20.01.034
© 2017, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
36
of the contact grid on the frontal surface of SC,
)(/eVm24.1)(/)( TETEhcT ggm ⋅μ==λ – absorption
threshold wavelength, Eg(T) – bandgap value, λ0 –
shortest wavelength that can be absorbed by SC, I (λ) –
spectral photogeneration intensity of the electron-hole
pairs, which depends on the irradiation conditions.
For maximal power conditions of the SCs
considered here, the recombination velocity in the space-
charge region (SCR) SSCR does not exceed 30 cm/s. It is
much smaller than the bulk recombination velocity, Sd ,
and the front surface recombination velocity, S0.
Therefore, in the generation-recombination balance
equation under the open-circuit conditions, and in the
expression for the current-voltage curve, the term related
to SSCR can be neglected. Then, the generation-
recombination balance equation can be written as
( ) nSSqJ dSC Δ+= 0 , (3)
where Sd = D/Ld, D is the minority carriers’ diffusion
coefficient,
2/11
AugerSRH
1111
⎥
⎥
⎥
⎦
⎤
⎢
⎢
⎢
⎣
⎡
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
τ
+
τ
+
τ
+
τ
=
−
nrr
d DL (4)
is the diffusion length, τr = (An0)–1 – radiative
recombination time, ( )0SRH nnxnr τ=τ – NRER time
with the parameter nx = 8.2·1015 сm–3 found in [1], A =
6.3·10–15 сm3/s – radiative recombination parameter [1],
Augerτ – band-to-band Auger recombination time in Si
determined by Eqs. (18) and (19) of Ref. [2].
400 600 800 1000 1200
0.0
0.2
0.4
0.6
0.8
1.0
In
te
rn
al
q
ua
nt
um
y
ie
ld
q
Wavelength λ, nm
n-type
p-type
Fig. 2. Theoretical (lines) and experimental (symbols) spectral
dependences of the internal quantum yield in SCs.
The illuminated current-voltage curve with the
assumptions formulated above, and with the series
resistance RS taken into account, has the form:
( )
( )
,1
)(
exp
)(
)(
0
2
0
⎥
⎦
⎤
⎢
⎣
⎡
−⎟⎟
⎠
⎞
⎜⎜
⎝
⎛ −
×
×+−=
kT
RVIVq
n
Tn
SSqAIVI
si
dSCSC
(5)
where ni (T) is the intrinsic charge carrier density and ASC
is SC area.
30 40 50
44
46
48
50
S
ho
rt
ci
rc
ui
t I
S
C
, m
A
Temperature T, oC
n-type
p-type
Fig. 3. Theoretical (lines) and experimental (symbols)
temperature dependences of the short-circuit current.
30 40 50
0.40
0.45
0.50
0.55
0.60
O
pe
n
ci
rc
ui
t v
ol
ta
ge
V
O
C
, V
Temperature T, oC
n-type
p-type
Fig. 4. Theoretical (lines) and experimental (symbols)
temperature dependences of the open-circuit voltage.
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2017. V. 20, N 1. P. 34-40.
doi: https://doi.org/10.15407/spqeo20.01.034
© 2017, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
37
30 40 50
60
65
70
75
80
Fi
ll
Fa
ct
or
F
F,
%
Temperature T, oC
n-type NRER
p-type NRER
p-type
n-type
Fig. 5. Theoretical (lines) and experimental (symbols)
temperature dependences of the fill factor.
From the maximum-power condition
0/))(( =dVVVJd , one can find Vm, substitution of
which into (5) allows one to determine Jm. As a result,
we obtain the usual expression for the photoconversion
efficiency
SSC
mm
PA
VI
=η , (6)
where PS is the incident irradiation power density.
The fill factor is
OCSC
mm
VI
VI
FF = , (7)
where the open-circuit voltage is given by the following
expression
( ) ⎟
⎟
⎟
⎟
⎟
⎠
⎞
⎜
⎜
⎜
⎜
⎜
⎝
⎛
+
+
= 1
)(
ln
0
2
0 n
Tn
SSq
J
q
kTV
i
d
SC
OC . (8)
4. Experimental results and discussion
As seen from Fig. 1, the theoretical I-V curves (for the
illuminated SC) obtained using Eq. (5) and the parameters
from Table agree well with the experimental data.
Fig. 2 shows the experimental and theoretical
spectral dependences of the internal quantum yield for
two investigated samples. As seen from this figure, the
theoretical curves plotted using Eq. (1) agree well with
the experiment. For the sample with the n-type base, the
agreement was obtained for the “dead layer” thickness
dp = 0.7 μm, and for the sample with the p-type base, we
took dp = 2.5 μm.
Our analysis has shown that the sensitivity of the
short circuit current to the temperature of the
environment depends on the irradiation temperature. The
lower Teff , the steeper JSC (T) curve. The experimental
JSC (T) curves can be described theoretically, if one
approximates the imitator spectrum with the Planck
spectrum of temperature Teff. It should be noted that the
black-body spectrum is always realized in the long-
wavelength region near the temperature-dependent
absorption edge wavelength )()( TEhcT gm =λ .
If the irradiation source can be modeled as a black
(or grey) body, then
( ) ∫
λ
λ
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
−⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
λ
λ
λλ
−=
)(
4
0
0 1exp
),(1)(
T
eff
LSC
m
kT
hc
dTqBrqTJ , (9)
where B0 is a constant. In particular, in the case of solar
spectrum, ( )20 2 ss DRcB π= [3], where RS is the
radius of the Sun, DS – distance from the Sun to the
Earth, h – Planck’s constant, c – speed of light, and k –
Boltzmann’s constant.
The luminosity of our solar imitator was chosen so
as to obtain the same photocurrent as under AM1.5
conditions:
,
1exp
)298,(
)(
)298,(2
)298(
4
)298(
5.1
2
0
0
∫
∫
λ
λ
λ
λ
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
−⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
λ
λ
λλ
=
=λλ⎟⎟
⎠
⎞
⎜⎜
⎝
⎛
π
K
eff
L
K
AM
s
s
m
m
kT
hc
dKq
Rf
dKqI
D
r
c
(10)
where IAM1.5 is the solar spectrum under AM1.5
conditions and f (RL) is the function describing
illumination of the sample at the distance RL to the
irradiation source.
To model the theoretical JSC (T) curves, we need to
take into account the temperature dependences of
α (λ, T), Eg(T), ni (T), D(T), τSRH (T), τAuger (T) and S0(T).
The absorption coefficient temperature dependence
inherent to Si was thoroughly investigated in the work
[4]. Near the absorption edge, it can be approximated as
rTT )298/)(298,(),( λα=λα . (11)
The temperature dependence of the bandgap in
silicon is well known (see, e.g., [5]). In view of
Einstein’s relations and due to the fact that at and above
room temperature the mobility is determined by phonon
scattering and decays with temperature according to the
power law mT − , the temperature dependence of the
diffusion coefficient can be well approximated by an
expression ( ) 1)K298/()K298()( +−⋅= mTDTD . The
expression for the Shockley–Read–Hall lifetime in an
n-type semiconductor has the form
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2017. V. 20, N 1. P. 34-40.
doi: https://doi.org/10.15407/spqeo20.01.034
© 2017, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
38
where p0 is the equilibrium hole density,
)()( ,,, TVTC pnpnpn σ= are the electron and hole
capture coefficients by a recombination center, Ei and Nt
– energy and concentration of the deep levels, and Vn,p –
thermal velocities of electrons and holes. As shown in
[6], the typical temperature dependence of the capture
cross section for holes by recombination centers varies
as k
p TT −σ ~)( with k = 2, if the centers are neutral
and k = 1…3, if they are negatively charged. Therefore,
for low excitation levels, taking the temperature
dependence of the thermal velocity into account, for
2)( −∝σ TTp we obtain 2/3
SRH T∝τ [7]. If hole
capture is due to positively charged centers,
( )( )3/1
00 /exp)( TTTTp −∝σ [6] and τSRH (T)
decreases with temperature.
Because for our SCs, the inequality AugerSRH τ≤τ
holds, the dependence )(TLd in the case of neutral or
attractive recombination centers is given by the relation
,
4
12
2
1,K)298/()K298(
)()()( SRH
−
+
−
=⋅=
=τ=
kmxTL
TTDTL
x
d
d
(13)
where the parameter k varies between 1 and 3 depending
on the exponent of the temperature-dependent capture
cross section for the minority carriers.
Note that, generally speaking, the analysis of
τSRH (T) should take into account the shift of the Fermi
energy towards the middle of the bandgap as the
temperature increases. This effect leads to a non-
linearity of τSRH (T), and thus to the non-linearity of
ISC (T). In a sufficiently narrow temperature range from
25 to 60 °С used in our measurements, the nonlinearity
of the short-circuit current vs. temperature was not
observed. As our estimates show, this is possible, if the
recombination level is sufficiently close to the middle of
the bandgap.
Fig. 3 shows the experimental short-circuit current
vs. temperature curves. The theoretical ISC (T) curves
were obtained from the relations (9), (11), and (13) by
adjusting the parameter r+x, where r and x are the
exponents in Eqs. (11) and (13). In the theoretical
curves, the sum of these two exponents 1≥+ xr ; it
mainly defines the slope of the ISC (T) curves. As seen
from this figure, the agreement between theory and
experiment is rather good.
Shown in Figs. 4, 5, and 6 are the experimental
temperature curves of VOC , FF, and η of our SCs.
The theoretical counterparts were obtained using the
above-introduced formulas with the parameter nx =
8.2·1015 cm–3. The expressions for JSC , VOC , FF, and η
in the absence of NRER can be obtained as a limiting
case of infinite nx.
When fitting the experimental results for VOC , FF,
and η under AM1.5 conditions, the fit parameters were S0
and RS. Their values that allow one to obtain good
agreement of the theory with experiment are given in
Table. Then, determination of the temperature dependence
of VOC , FF, and η was performed taking into account the
temperature dependence of JSC (T) (see Fig. 3 and the
discussion above), as well as the temperature dependence
of ni (T), which plays a crucial role.
As seen from Figs. 4 to 6, the theoretical curves
obtained by taking into account NRER agree well with
the experimental curves within the temperature range
studied. Also, the theoretical expressions for the
photoconversion efficiency obtained with and without
taking this recombination process into account differ by
as much as 3.8% for n-type and 5.1% for p-type
samples. Likewise, the slopes of the efficiency vs.
temperature curves obtained with and without NRER
somewhat differ.
Fig. 7 shows the experimental and theoretical
temperature dependence of the photoconversion
efficiency temperature coefficient defined as
( ) %100
)(
)()(
)(
00
0
TTT
TT
T
−η
η−η
=β , (14)
where T0 is the initial temperature. As seen in this figure,
the magnitude of ( ) C%5.0 °≥β T is typical for silicon
SCs with graded p-n junctions at relatively low
photoconversion efficiency [7].
30 40 50
6.0
6.5
7.0
7.5
8.0
8.5
9.0
E
ffi
ci
en
cy
η
, %
Temperature T, oC
n-type
n-type NRER
p-type
p-type NRER
Fig. 6. Theoretical (lines) and experimental (symbols)
temperature dependences of photoconversion efficiency.
( )( ) ( )( )
0
00
SRH )()(
/exp)()(/exp)()(
nNTCTC
kTETnnTCkTETnpTC
tnp
iiniip −+++
≡τ , (12)
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2017. V. 20, N 1. P. 34-40.
doi: https://doi.org/10.15407/spqeo20.01.034
© 2017, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
39
30 40 50
0.0
0.2
0.4
0.6
0.8
1.0
E
ffi
ci
en
cy
d
ec
re
as
e
β,
%
/K
Temperature T, oC
p-type
n-type
Fig. 7. Theoretical (lines) and experimental (symbols)
temperature dependences of photoconversion efficiency
temperature coefficient.
1014 1015 1016 1017 1018 1019
0
4
8
12
without
accounting for NRER
NRER
τSRH=10-4s
τSRH=10-5s
E
ffi
ci
en
cy
η
, %
Dopant concentration p0, cm-3
(a) n-type
τSRH=10-6s
1014 1015 1016 1017 1018 1019
0
4
8
12
16
20
τSRH=10-5 s
τSRH=10-6 s
without
accounting for NRER
NRER
E
ffi
ci
en
cy
η
, %
Dopant concentration p0, cm-3
(b) p-type τSRH=10-4 s
Fig. 8. Theoretical photoconversion efficiency as a function of
doping level for (a) n-type base and (b) p-type base.
Let us now return to the question about the
influence of NRER on the SC efficiency with
sufficiently low τSRH. As follows from Eq. (4), this
influence can become apparent, if the condition xnn ≥0
is fulfilled. Because for our samples the opposite is true,
i.e. n0 < nx, it can be expected that in this case, the effect
of NRER on the efficiency should not be very
significant. However, as seen from Fig. 6, the theoretical
efficiencies obtained with and without incorporating τnr
differ quite noticeably, and this difference increases with
increasing the doping level.
To validate this observation, we have plotted the
theoretical η(n0) curves using all the parameters
characterizing our samples except for the Shockley–
Read–Hall lifetime, which were set to 100, 10, and 1 μs,
respectively. These curves are shown in Fig. 8 for (a)
n-type and (b) p-type base for AM1.5 irradiation
spectrum. As seen in Fig. 8a, the η(n0) curves obtained
with and without accounting for NRER are completely
different. In particular, as τSRH gets smaller, the
maximum of these curves shifts towards lower doping
levels, if NRER is taken into account, and it shifts
towards larger doping levels, if it is discarded. At the
same time, the difference in the efficiency peak values
obtained with and without accounting for NRER effect
becomes larger. In particular, for τSRH = 1 μs, this
difference can be as high as about 26% of the efficiency
peak value obtained without NRER.
These curves are similar in the case of a p-type
base, see Fig. 8b. Here, the displacement of the
efficiency peaks plotted with NRER taken into account
is smaller than in the n-type base case. This is caused by
the higher surface recombination velocity. The
difference between the efficiency values for τSRH = 1 μs
is about 45% of the efficiency peak value obtained
without NRER.
The curves in Fig. 8 indicate that it is impractical to
use high base doping levels in silicon SC with short τSRH.
Although this fact has been known before from
experience, it remained unexplained theoretically.
Last but not least, let us discuss the criteria
allowing one to establish the degree of influence of
NRER on the photoconversion efficiency of silicon SCs.
In order for this recombination mechanism to have no
effect on the photoefficiency is that τr < τnr, i.e. the
radiative recombination should dominate. Using the
expressions for τr and τnr (see above Eq. (4)), we find
that this condition holds for τSRH > 20 ms.
In order for NRER to have a weak effect on the
photoconversion efficiency, the inequalities τr > τnr and
τnr > τAuger must hold. With the relevant parameter
values, we find that the maximum on the η(n0) curves at
not too high surface recombination velocities (below
103 cm/s) is at 317
0 cm10 −≥n . The effect of NRER on
photoconversion efficiency is weak for τSRH between 0.1
and 20 ms. We note that in this case the maxima of the
η(n0) curves plotted with and without taking NRER into
account coincide.
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2017. V. 20, N 1. P. 34-40.
doi: https://doi.org/10.15407/spqeo20.01.034
© 2017, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
40
Finally, in the case τnr < τAuger, i.e. τSRH < 0.1 ms,
the effect of NRER on photoconversion efficiency is
strong. In this case, the maxima of the η(n0) curves
plotted with and without taking this effect into account
do not coincide.
5. Conclusions
NRER has a strong effect on the photoconversion
efficiency in silicon SC under the condition n0 > nx, in
which case the characteristic time of this process
becomes shorter than the Shockley–Read–Hall lifetime.
The smaller τSRH, the stronger the effect of NRER
on photoconversion efficiency. Therefore, at τSRH of the
order of 1 μs, this recombination mechanism is also
pronounced when n0 < nx.
At sufficiently short τSRH, NRER is responsible for
the shift of the optimal base doping level, at which the
photoconversion efficiency has a maximum.
References
1. Sachenko A.V., Kostylyov V.P., Vlasiuk V.M.,
Sokolovskyi I.O., and Evstigneev M.A. The
influence of the exciton non-radiative recombi-
nation in silicon on the photoconversion efficiency.
1. The case of a long Shockley–Read–Hall lifetime.
SemiconductorPhysics, Quantum Electronics and
Optoelectronics. 2016. 19, No. 4. P. 334–342.
2. Richter A., Glunz S.W., Werner F., Schmidt J.,
Cuevas A. Improved quantitative description of
Auger recombination in crystalline silicon. Phys.
Rev. B. 2012. 86. P. 165202.
3. Pikus G.E. Fundamentals of The Theory of
Semiconductor Devices. Moscow: Nauka, 1965.
4. Hguyen H.T., Rougieux F.E., Mitchell B.,
Macdonald D. Temperature dependence of the
band-band absorption coefficient in crystalline
silicon from photoluminescence. J. Appl. Phys.
2014. 115. P. 043710.
5. Green M.A. Intrinsic concentration, effective
densities of states, and effective mass in silicon. J.
Appl. Phys. 1990. 67, No. 6. P. 2944–2954.
6. Abakumov V.N., Perel V.I., Yassievich I.N.
Radiationless Recombination in Semiconductors.
St. Petersburg: B.P. Konstantinov St. Petersburg
Institute of Nuclear Physics of RAN, 1997 (in
Russian).
7. Fahrenbruch A.L., Bube R.H., Fundamentals of
Solar Cells: Photovoltaic Solar Energy
Conversion. New York: Academic, 1983.
8. Skoplaki E., Palyvos J.A. On the temperature
dependence of photovoltaic module electrical per-
formance: A review of efficiency power corre-
lation. Solar Energy. 2009. 83, No. 5. P. 614–624.
|
| id | nasplib_isofts_kiev_ua-123456789-214915 |
| institution | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| issn | 1560-8034 |
| language | English |
| last_indexed | 2026-03-21T19:35:52Z |
| publishDate | 2017 |
| publisher | Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України |
| record_format | dspace |
| spelling | Sachenko, A.V. Kostylyov, V.P. Vlasiuk, V.M. Korkishko, R.M. Sokolovskyi, I.O. Chernenko, V.V. Evstigneev, M.A. 2026-03-03T11:09:52Z 2017 Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes / A.V. Sachenko, V.P. Kostylyov, V.M. Vlasiuk, R.M. Korkishko, I.O. Sokolovskyi, V.V. Chernenko, and M.A. Evstigneev // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2017. — Т. 20, № 1. — С. 34-40. — Бібліогр.: 8 назв. — англ. 1560-8034 PACS: 72.20.J, 78.60.J https://nasplib.isofts.kiev.ua/handle/123456789/214915 https://doi.org/10.15407/spqeo20.01.034 The influence of non-radiative exciton recombination (NRER) on the photoconversion efficiency in silicon solar cells with short Shockley–Read–Hall lifetimes τSRH has been studied. It has been shown that the efficiency reduction due to this effect is stronger the shorter τSRH. The influence of NRER is most evident when the NRER time becomes shorter than τSRH. At sufficiently short τSRH, NRER substantially limits the optimal base doping levels of silicon solar cells, at which the photoconversion efficiency is maximal. en Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України Semiconductor Physics Quantum Electronics & Optoelectronics Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes Article published earlier |
| spellingShingle | Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes Sachenko, A.V. Kostylyov, V.P. Vlasiuk, V.M. Korkishko, R.M. Sokolovskyi, I.O. Chernenko, V.V. Evstigneev, M.A. |
| title | Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes |
| title_full | Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes |
| title_fullStr | Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes |
| title_full_unstemmed | Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes |
| title_short | Influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. Short Shockley–Read–Hall lifetimes |
| title_sort | influence of non-radiative exciton recombination in silicon on photoconversion efficiency. 2. short shockley–read–hall lifetimes |
| url | https://nasplib.isofts.kiev.ua/handle/123456789/214915 |
| work_keys_str_mv | AT sachenkoav influenceofnonradiativeexcitonrecombinationinsilicononphotoconversionefficiency2shortshockleyreadhalllifetimes AT kostylyovvp influenceofnonradiativeexcitonrecombinationinsilicononphotoconversionefficiency2shortshockleyreadhalllifetimes AT vlasiukvm influenceofnonradiativeexcitonrecombinationinsilicononphotoconversionefficiency2shortshockleyreadhalllifetimes AT korkishkorm influenceofnonradiativeexcitonrecombinationinsilicononphotoconversionefficiency2shortshockleyreadhalllifetimes AT sokolovskyiio influenceofnonradiativeexcitonrecombinationinsilicononphotoconversionefficiency2shortshockleyreadhalllifetimes AT chernenkovv influenceofnonradiativeexcitonrecombinationinsilicononphotoconversionefficiency2shortshockleyreadhalllifetimes AT evstigneevma influenceofnonradiativeexcitonrecombinationinsilicononphotoconversionefficiency2shortshockleyreadhalllifetimes |