Research Article Vol. 32, No.

6 / 11 Mar 2024 / Optics Express 8614

Enhancing perovskite-silicon tandem solar cells

through numerical optical and electric
optimizations for light management
J INPENG YANG 1,3 AND Q INYE B AO 2,4
1 Collegeof Physical Science and Technology, Yangzhou University, Jiangsu 225009, China
2 Schoolof Physics and Electronic Science, East China Normal University, Shanghai 200241, China
3 [email protected]
4 [email protected]

Abstract: We integrated optical and electrical numerical simulations to precisely investigate

the effectiveness of using a pyramidal perovskite (Cs0.18 FA0.82 Pb(I,Br)3 ) nanostructured film as
an example in perovskite-silicon tandem solar cells to reduce reflective losses and balance the
current densities. Through our calculations, the PCE of tandem solar cells can be improved from
29.2% (the planar structures without texturing) to 36.1% in the best-performing textured tandem
devices under the consistently calculated absorbed and EQE spectrum, where the predicted
open-circuit voltage could reach over 2 V. These findings offer valuable theoretical insights for
the advancement and optimization of perovskite-silicon tandem solar cells.

© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement

1. Introduction
Despite the current dominance of crystalline silicon (c-Si) devices in the photovoltaic (PV) market,
perovskite-silicon tandem architectures have garnered significant attention due to their potential
performance could surpass the Shockley-Queisser limit of single silicon. [1–5] In a typical
perovskite-silicon tandem solar cell with a monolithic (two-terminal) configuration, a crystalline
silicon (c-Si) bottom cell is used in combination with a perovskite top cell. [6–8] The perovskite
top cell absorbs high-energy photons, while the c-Si bottom cell absorbs low-energy photons
transmitted through the top perovskite layers. Recently, a PCE exceeding 31% has been reported
in perovskite-silicon tandem after optimization of the optical structure and bandgap of perovskite
film. [9] However, the theoretical Shockley-Queisser limit, which is based on detailed balance,
suggests that PCE in perovskite-silicon tandem solar cell could reach up to 45% [10], highlighting
the need for further study. Several factors could affect the PCE in perovskite-silicon tandem
devices, such as parasitic absorption limited by energy bandgap, defects induced nonradiative
recombination, undesirable series/shunt resistance for large modules, and structure geometry
induced optical losses. In particular, reflective losses can significantly limit matched photocurrent
density in fully planar tandem devices. Therefore, significant efforts have been dedicated to the
development of textured nanostructures for effective light trapping. [11–15]
Textured interfaces have the ability to reduce losses caused by reflection through two mech-
anisms: enhancing the coupling of light into the structure, and scattering light in a way that
increases the average path length within the absorber, which results in enhanced absorption,
particularly in weak absorbing region. Previous numerical simulations using programs like
GENPRO4, CROWM, or JCMsuite have made attempts to address the light trapping effect.
[15–20] These simulations, with focusing on the optical geometry, have revealed that absorption
in the perovskite top cell can significantly impact the efficiency of perovskite-Si tandem solar
cells, and emphasize the importance of careful photon management. However, there is still a
lack of comprehensive theoretical study that can effectively integrate two following aspects: (i)
the utilization of optical structures to balance photocurrent density in perovskite-silicon tandem

#513887 https://doi.org/10.1364/OE.513887
Journal © 2024 Received 2 Jan 2024; revised 12 Feb 2024; accepted 13 Feb 2024; published 26 Feb 2024
 Research Article Vol. 32, No. 6 / 11 Mar 2024 / Optics Express 8615

solar cells, while considering the angular dependence of incident light, (ii) accurately predicting
current-voltage characteristics with considering additional electric parameters. Such separate
understanding hampers the ability to further optimize and achieve the maximum PCE in realistic
perovskite-Si tandem solar cells.
In this study, we employ a combination of optical and electrical calculations utilizing the finite
element method to investigate the influence of pyramidal nanostructures on the top perovskite
sub-cell (using Cs0.18 FA0.82 Pb(I,Br)3 as an example) on the PCE of perovskite-Si tandem solar
cells. Our objective is to assess the effectiveness of these nanostructures in reducing reflective
losses and balancing current densities in perovskite-Si tandem devices. We systematically
studied three different architectures of perovskite layers. The results demonstrated that the
best-performing nanostructured perovskite-Si tandem device exhibits a significant enhancement
in PCE, reaching up to 23.8% improvement compared to its planar structure. Additionally,
angular-dependent current density change indicates that the textured surface has the potential to
alleviate current density mismatch issues between perovskite and silicon films in tandem solar
cells.

2. Simulation method and device structures

The optical simulations in this study are performed using the COMSOL Multiphysics software,
specifically utilizing its two-dimensional model and the finite element method. The simulated
domain consists of a stack of layers treated as coherent or incoherent depending on the film
thickness, and embedded in semi-infinite air spaces on top and bottom, which are treated as
incoherent layers. Here, we employed the actual thickness of each layer during the calculation
without using any approximation, particularly in the case of thicker silicon wafers. From each
simulated layer, equivalent photocurrent densities can be calculated with the following equation:
∫ 1200 nm
Ji = e Ai (λ)×AM1.5 (λ)dλ (1)
300 nm

where Ji represents the absorbed current density corresponding to the absorption profile of the
i-th layer (Ai (λ)). ΦAM1.5 represents the photon flux based on the AM 1.5 G spectrum [21], and e
denotes the elementary charge. The photocurrent densities obtained for the perovskite and silicon
layers are considered as the maximum achievable current densities. This assumption is made
that all the light absorbed in the solar cell can be converted into electron-hole pairs, which are
then extracted to generate electrical current. On the other hand, the current densities observed in
the remaining layers represent losses attributed to parasitic absorption. The absorbed spectrum
can simultaneously change with variations in the incident light angle, thereby necessitating the
investigation of angular-dependent Ji . To perform accurate simulations, the complex refractive
index spectra (n, k) are taken from and given in [22] Supplement 1, Fig. S1.
The electric calculations were performed using software SCAPS [23–25] for one dimensional
perovskite-Si tandem solar cells. The current-voltage characteristics of solar cells were simulated
under a standard AM1.5 G spectrum with 100 mW cm−2 intensity across a voltage range of 0-2 V.
The electrical parameters used in this simulation are listed in Table S1. The performance of
the top perovskite solar cells is directly simulated using a standard AM1.5 G spectrum after
filtering out reflections (utilizing results of 1-R, calculated from optical simulation). In contrast,
the bottom silicon solar cells make use of the remaining incident light after passing through
the layers of the top perovskite solar cells. In order to achieve a maximum PCE, we strive to
achieve a nearly identical current density match for both the top perovskite and bottom silicon in
monolithic tandem devices.
Figure 1 shows the three main architectures of the perovskite-Si tandem solar cells studied in
this work with flat and/or textured top c-Si surfaces. Figure 1(a) illustrates a simple architecture
commonly found in literature [9,26,27], featuring a “double-side” flat perovskite top cell for
 Research Article Vol. 32, No. 6 / 11 Mar 2024 / Optics Express 8616

a direct comparison. In Fig. 1(b), only the intermediate layers between the Si absorber and
perovskite absorber are textured, while the interfaces on top of the perovskite layers remain flat. In
experimental setups, the spin-coating technique is employed to grow perovskite layers on textured
silicon substrates, resulting in nearly flat perovskite on top. Figure 1(c) depicts the architecture
with a “double-side” textured perovskite top cell, where the front-side textured perovskite can
be achieved through nanoimprint processes or by utilizing continuous vacuum deposition for
conformal growth of the perovskite films. To enhance light trapping and absorption from silicon
in the near infrared wavelength region, textured architectures obtained from experiments [26]
are implemented on the back surface of c-Si. For the top perovskite sub-cell, it is assumed that
all textured interfaces possess the same texture. A period of 300 nm with a depth of 300 nm is
employed for all textured perovskite and other related layers to ensure sufficient light trapping
[16,27]. Film thicknesses are chosen based on experimental devices and optimization as discussed
in the later section.

Fig.Fig.
1 1.
Schematic diagram
Schematic diagram of the monolithic
of the monolithic perovskite-Si
perovskite-Si tandem devices utilizedtandem
for opticaldevices
and electric simulations. (a)- (c) are defined as “textured back-side-only” silicon, “textured
utilized for optical
double-side” and“textured
silicon with electric simulations.
back-side” perovskite,(a)- (c) aredouble-side”
and “textured defined silicon
as “textured
back-side-only” silicon, “textured
with “textured double-side” double-side”
perovskite, respectively. silicon
The film withof“textured
thicknesses back-side”
electrodes and
antireflective coating (AR-LiF) employed in the simulations are also indicated. It should
perovskite, and “textured double-side” silicon with “textured double-side”
be noted that the thickness of the perovskite films varies depending on the optimization of
perovskite, respectively.
matched current densities. The film thicknesses of electrodes and antireflective
coating (AR-LiF) employed in the simulations are also indicated. It should be
noted that the thickness of the perovskite films varies depending on the
3. Results
optimization of matched current densities.
To begin, we examined the effects of an antireflective coating (AR-LiF) on the absorbed spectrum
in3. Results
a perovskite-Si tandem device configuration where both the back side of perovskite and double
sides of silicon are textured (depicted as (b) in Fig. 1). In Fig. 2(a), typical absorbed spectrum of
To begin, we examined the effects of an antireflective coating (AR-LiF)
perovskite and silicon films are illustrated, comparing the cases with and without using AR-LiF.
It on
canthe absorbed
be seen spectrum
that adding AR-LiFinlayers
a perovskite-Si
could result intandem device
increased configuration
absorptions where
for both the top
both the back side of perovskite and double sides of silicon are textured
(depicted as (b) in Figure 1). In Figure 2(a), typical absorbed spectrum of
perovskite and silicon films are illustrated, comparing the cases with and
without using AR-LiF. It can be seen that adding AR-LiF layers could result in
increased absorptions for both the top perovskite and bottom silicon films,
suggesting the effectiveness of reducing reflection loss at the air/ITO surface
 Research Article Vol. 32, No. 6 / 11 Mar 2024 / Optics Express 8617

perovskite and bottom silicon films, suggesting the effectiveness of reducing reflection loss at
the air/ITO surface and emphasizing the necessity of inserting low refractive index materials
between the air and ITO, even when textured interfaces have been utilized (see an additional case
in Supplement 1, Fig. S2(a) for device structure from Fig. 1(a)). Furthermore, analysis of the
occur
1-R in the
spectra wavelength
reveals range
that reflection below
losses 500 nm.
primarily occurThis
in thephenomenon
wavelength rangecanbelow
be partially
500 nm.
attributed
This to the
phenomenon canabsorption characteristics
be partially attributed of ITO,
to the absorption specifically
characteristics its extinction
of ITO, specifically
coefficient
its (k value).(k value).
extinction coefficient

Fig. Fig. 2. illustrates the simulated absorbed and the 1-total reflection (1-R) spectrum for
2 illustrates the simulated absorbed and the 1-total reflection (1-R)
different perovskite-Si tandem architectures. In panel (a), we observe the absorption and 1-R
spectrumchangesfor different tandem
in perovskite-Si perovskite-Si
devices (devicetandem
structure architectures. In panel
is shown in Fig. 1(b)), both with(a), we
observe the absorption
and without the implementationand of1-R changes coating
an antireflective in perovskite-Si
(AR-LiF). Panel (b) tandem
showcases devices
the variation in absorbed spectrum among three tandem devices from Fig. 1, where shaded
(device areasstructure
give spectrum ischange
shown in Figure
for direct comparison. 1(b)), both line
The dash-dot with andthewithout
denotes impact the
implementation
of C60 absorption, of anandantireflective
the dashed-symbol coating (AR-LiF).
line represents absorbedPanel
spectra(b) showcases
change in the the
near-infrared wavelength region due to the absence of textured back side of silicon (device
variation in absorbed spectrum among three tandem devices from Figure 1,
structure is from Fig. 1(b)). It is important to note that a constant perovskite film with
wherethickness
shadedof areas
560 nm give spectrum
is maintained changeoptimized)
(not thickness for direct comparison.
during the calculation.The dash-dot
line denotes the impact of C60 absorption, and the dashed-symbol line represents
In Fig. 2(b), a direct comparison is presented, illustrating the changes in absorbed spectrum for
absorbed spectra change in the near-infrared wavelength region due to the
different interfacial structures (either with planar or textured configurations). The thickness of the
absence films
perovskite of textured
used in all back side
of the of silicon
tandem (device structure
device architectures was fixed is atfrom Figure
560 nm, without1(b)).
any
It is important to note that a constant perovskite film with thickness of 560 nm
additional thickness optimization. However, it should be noted that a typical balanced absorption
after undergoing thickness
is maintained optimization
(not thickness can be found
optimized) duringin previously published results. [5,28] For
the calculation.
the device architecture featuring “textured back-side only Si” (as shown in Fig. 1(a)), relatively
In Figure
low absorption 2(b), ain direct
is observed silicon, comparison is presented,
indicating the great illustrating
loss of incident light in thethe changes
800-1200 nm
in absorbed
wavelength rangespectrum
due to the forreflection.
differentThe interfacial
dependence structures
of silicon(either
absorption with planar on
spectrum or
textured configurations). The thickness of the perovskite films used in all of the
thickness for the device architecture with “textured back-side only Si” is also illustrated in
Supplement 1, Fig. S2(b). Notably, a significant enhancement is achieved when changing
tandem device
architectures from the architectures
“textured back-sidewas onlyfixed Si”at 560“textured
to the nm, without
double-side anySi +additional
back-side
thickness
PVK” optimization.
(as depicted However,
in Fig. 1(b)), it should
particularly be notedregion
in the absorption that from
a typical balanced
the silicon films,
absorption after undergoing thickness optimization can be found in previously
although a minor amount of reflection loss (on the top of spectra) remains. Furthermore, our
findings confirm that (i) the losses in the low wavelength region (below 500 nm) can primarily be
published results. [28, 29] For the device architecture featuring "textured back-
attributed to the absorption of C60 films, as indicated by the dash-dot lines. It is observed that
sidea only
with thinnerSi"C60(asfilm,shown
there isin Figure
higher 1(a)),forrelatively
absorption the perovskitelowfilm absorption is observed
(Two different thickness
in silicon,
of C 60 are indicating the great loss of incident light in the 800-1200
compared during our calculations and presented in Supplement 1, Fig. S3). Wenm
wavelength range due to the reflection. The dependence of silicon absorption
spectrum on thickness for the device architecture with "textured back-side only

6
 Research Article Vol. 32, No. 6 / 11 Mar 2024 / Optics Express 8618

have further shown exemplary comparisons between our calculations and experimental results
obtained from the literature (refer to Fig. S4), revealing a high degree of agreement [8,9,29];
(ii) the implementation of a textured back-side silicon structure improves the spectral response
in the 1000-1200 nm infrared wavelength range (after comparing red-symbol dash line with a
blue shadow), and (iii) further improvements can be achieved by changing the perovskite films to
demonstrate a “textured double-side Si + PVK” architecture (also seen in Fig. 1(c)), leading to the
highest absorption percentage both in silicon and perovskite (highlighted with a pink shadow).
Once we have gained insights into the influences of various device structures on the absorbed
spectrum of top perovskite and bottom silicon sub-cells, we can proceed to calculate the “ideal”
maximum absorbed current density (Jmax Ph ) generated separately by the perovskite and silicon

absorbers using equation (1). To optimize the performance of perovskite-Si tandem devices,
we only need to adjust the thickness of the perovskite films in order to maximize the matching
point of current densities, while keeping the thicknesses of other films unchanged. The current
densities of the limiting sub-cells can be determined by the following equation:
ph ph
min(Jmax,PVK , Jmax,Si ) → max (2)

Figure 3 shows the evolution of perovskite film thickness (d) on the maximum absorbed current
but rather resembles a parabolic trend. According to our simulations, the
densities of the top perovskite and bottom silicon sub-cells according to three tandem devices
maximum
depicted matched
in Fig. currenttrends
1. The observed densities for the across
are consistent device structures
these shown(i)ina gradual
tandem devices: (a)-(c)
of Figure
decrease in J1
ph
are 16.43
and a simultaneous2
mA/cm , 19.54 mA/cm
increase
ph2
, and
in Jmax,Si can 20.45 mA/cm
be observed 2
when, the
respectively.
thickness of
max,PVK
perovskite film is reduced; (ii) the relationship does not follow a linear
These results strongly confirm the efficiency of employing textured interfaces pattern but rather resembles
a parabolic trend. According to our simulations, the maximum matched current densities for the
to realize
device lightshown
structures trapping andofenhance
in (a)-(c) Fig. 1 are absorption.
16.43 mA/cm2Additionally,
, 19.54 mA/cm2 , we demonstrate
and 20.45 mA/cm2 ,
that both textured
respectively. interfaces,
These results stronglynot onlythe
confirm in efficiency
the bottom silicon structure
of employing but also in
textured interfaces to
the top perovskite films, are highly effective, resulting in enhancements of 18.9%
realize light trapping and enhance absorption. Additionally, we demonstrate that both textured
interfaces, not only in the bottom silicon structure but also in the top perovskite films, are highly
(for “textured back-side” perovskite) and 24.5% (for “textured double-side”
effective, resulting in enhancements of 18.9% (for “textured back-side” perovskite) and 24.5%
perovskite)
(for respectively.
“textured double-side” perovskite) respectively.

Fig. 3. The optimization process for adjusting the perovskite thickness to achieve maximum
Fig. 3. The absorbed
matched optimization process
current densities fromfor adjusting
the top perovskitethe perovskite
and bottom thickness to
silicon sub-cells.
achieve
Themaximum matched
results are given absorbed
based on three differentcurrent densities
perovskite-Si from with
tandem devices the the
topchanged
perovskite
thickness of the perovskite films (d) indicated on the inset of the figure.
and bottom silicon sub-cells. The results are given based on three different
perovskite-Si tandem devices with the changed thickness of the perovskite films
(d) indicated on the inset of the figure.
The angular-dependent performance is also an important factor for solar
cells without using active light-tracking. Figure 4 provides valuable insights
into this aspect by illustrating the variation of normalized maximum matched
current density with respect to the angle of incident light for both the top
 Research Article Vol. 32, No. 6 / 11 Mar 2024 / Optics Express 8619

The angular-dependent performance is also an important factor for solar cells without using
active light-tracking. Figure 4 provides valuable insights into this aspect by illustrating the
variation of normalized maximum matched current density with respect to the angle of incident
light for both the top perovskite and bottom silicon sub-cells under the consideration of three
devices. perovskite-Si
optimized Conversely,tandem the "double-side"
devices. The texturedflat tandem device(referred
tandem devices (referred
to astoback-side
as flat
PVK and double-side PVK) exhibit a slight decrease in normalized current densities (JPVK and
PVK) shows different behavior for Jsc. In
JSi ) when the angle of incident light increases
this case, a disparity between JPVK and
from 0 degree to 30 degrees. Such a consistent
JSi can be
tendency observed
in the as the incident
current densities for both Jangle varies from 30 degrees to 60 degrees,
PVK and JSi indicates a good match for maintaining
indicating insufficient light absorption
high short-circuit current density (J sc ) when they andstacked
are clear mismatch
in series andof current
formed densities.
tandem devices.
Conversely, the “double-side” flat tandem device (referred to as flat PVK) shows different
Consequently, the perovskite-Si tandem solar cells under such an architecture
behavior for Jsc . In this case, a disparity between JPVK and JSi can be observed as the incident
wouldvaries
angle exhibit
from low PCE. toTherefore,
30 degrees 60 degrees, these discrepancies
indicating should
insufficient light be taken
absorption into
and clear
account of
mismatch during
currentstructure design. Notably,
densities. Consequently, the advantage
the perovskite-Si in absorption
tandem solar and
cells under such
an architecture would exhibit low PCE. Therefore, these discrepancies
current densities provided by the textured perovskite films, whether "back-side should be taken into
account during structure design. Notably, the advantage in absorption and current densities
textured"
provided by or
the"double-side textured,"
textured perovskite is maintained
films, whether “back-side over a wide-angle
textured” range.
or “double-side This
textured,”
finding
is supports
maintained over athe application
wide-angle range.ofThis using textured
finding interfaces/surfaces
supports for tandem
the application of using textured
solar cells.
interfaces/surfaces for tandem solar cells.

Fig.4 Fig.
The4. angular
The angulardependence
dependence ofof thethe normalized
normalized maximum maximum matched
matched current current
densities
densities for both the top perovskite (JPVK) and bottom silicon (JSi) sub-cells in
for both the top perovskite (J PVK ) and bottom silicon (J Si ) sub-cells in three perovskite-Si
tandem solar cells with architectures depicted in Fig. 1.
three perovskite-Si tandem solar cells with architectures depicted in Figure 1.
Finally,
Finally, we performed
we performed calculations of calculations of the current
the current density-voltage density-voltage
characteristics to estimate
characteristics
the potential PCE to estimate
of the tandem the solarpotential PCE of the
cells after combining our tandem solar cells
obtained optically after
simulated
results and assuming the all absorbed photons could be efficiently
combining our obtained optically simulated results and assuming the all converted into electric current.
Therefore, our calculated electrical results represent the “ideal” maximum value of device
absorbed photons
performance that can becould be efficiently
obtained by combiningconverted
the absorption intofromelectric
opticalcurrent. Therefore,
calculations, without
our calculated
considering electrical
any defects. resultssemiconductor
The detailed represent the "ideal" for
parameters maximum value ofare
device simulations device
listed
performance that can be obtained by combining the absorption from optical
in Supplement 1, Table S1. In Fig. 5, the left side displays the calculated current density-voltage
curves of three different perovskite-Si tandem solar cells, where the tandem architectures are
calculations, without considering any defects. The detailed semiconductor
depicted in Fig. 1. To provide detailed comparison, we also include the individual current
parameters forcurves
density-voltage device simulations
of the top perovskite areand
listed in Table
bottom silicon S1. In Figure
sub-cells. On the5,right
the side
left
side displays the calculated current density-voltage curves of three different
of Fig. 5, we present the calculated external quantum efficiency (EQE) obtained through our
electric simulations using SCAPS, along with the absorptions and reflections (1-R) derived
perovskite-Si tandem solar cells, where the tandem architectures are depicted in
from optical simulations for better comparison. In the perovskite-silicon tandem solar cell
Figure 1. To provide detailed comparison, we also include the individual
current density-voltage curves of the top perovskite and bottom silicon sub-cells.
On the right side of Figure 5, we present the calculated external quantum

9
 Research Article Vol. 32, No. 6 / 11 Mar 2024 / Optics Express 8620

featuring a planar perovskite film (right side of Fig. 5 (a)), we could achieve a PCE of 29.20%,
which is accompanied by a Jsc of 16.19 mA/cm2 , open-circuit voltage (Voc ) of 2.10 V, and fill
factor (FF) of 85.8%. A significant enhancement in the PCE can be observed, reaching 34.61%
(Fig. 5(b)), when we change top perovskite sub-cells from a planar to a “back-side textured”
structure. The noteworthy improvement in PCE is primarily attributed to a substantial increase
in the Jsc which reaches 19.32 mA/cm2 , while having a negligible impact on the Voc and FF. A
slight additional enhancement could be further observed when top perovskite sub-cell changes
from “back-side textured” to a “double-side textured” structure, leading to a minor increase in the
Jsc from 19.32 mA/cm2 to 20.18 mA/cm2 , as illustrated in Fig. 5(c). Table 1 provides a summary
of the calculated results on the tandem device performance obtained from Fig. 5. On the right
side of Fig. 5, we observe a strong agreement between the External Quantum Efficiency (EQE)
and absorbed spectra, both from top perovskite and bottom silicon sub-cells. This agreement
serves as further confirmation of the accuracy from our electric calculations, which appropriately
incorporated optically simulated results. Furthermore, three exemplary comparisons given in
Supplement 1, Fig. S5 highlight the importance of high crystallinity and low defects in preparing
perovskite films for achieving higher PCE in tandem solar cells. A detailed study on defects and
other electrical parameters that could affect device performance will be continuously studied in
the future.

Fig. 5Fig. 5. Calculated current-voltage characteristics of perovskite-Si tandem solar cells with
Calculated current-voltage characteristics of perovskite-Si tandem solar
“planar or textured” perovskite films, displaying both the individual and total contributions
cells with
from the"planar or textured"
top perovskite and bottom perovskite films,
silicon sub-cells. Thedisplaying both
left side of the the
figure individual
includes
and total contributions from the top perovskite and bottom silicon sub-cells.
(a)-(c), which correspond to the same device architectures illustrated in (a)-(c) of Fig. 1. The The
right side of figure presents the calculated External Quantum Efficiency (EQE), as well as
left side of the figure
the absorptions includes
and reflections (1-R)(a)-(c),
exhibited which correspond
by these three tandem solarto cells.
the Note
samethatdevice
architectures illustrated in (a)-(c) of Figure 1. The right side of figure presents
our calculation does not take into account any additional impacts on defects.
the calculated External Quantum Efficiency (EQE), as well as the absorptions
and reflections (1-R) exhibited by these three tandem solar cells. Note that our
calculation does not take into account any additional impacts on defects.

Table 1. A summary of the calculated results on the tandem device

performance obtained from Figure 5.
2
 Research Article Vol. 32, No. 6 / 11 Mar 2024 / Optics Express 8621

Table 1. A summary of the calculated results on the tandem device

performance obtained from Fig. 5.
Voc (V) Jsc (mA/cm2 ) FF PCE
Planar 2.10 16.19 85.8% 29.17%
Back-side textured 2.10 19.32 85.3% 34.61%
Double-side textured 2.10 20.18 85.3% 36.14%

4. Conclusions and outlook

To fully unlock the potential of perovskite-Si tandem solar cells, optimizing their optical
performance is crucial. In this manuscript, we present the results of current density optimization
for perovskite-Si tandem solar cells using Cs0.18 FA0.82 Pb(I,Br)3 as an example with considering
three different textured perovskite top sub-cells: “double-side flat,” “back-side textured,” and
“double-side textured” perovskite films. Because the textured structures can reduce the reflective
losses and increase light absorption, higher balanced Jsc even with large angular dependence
of tolerance in perovskite-Si tandem devices are obtained. An additional cumulative Jsc with
>3 mA/cm2 can be generated in the two sub-cells with respect to the regular tandem architecture
of a “double-side flat” perovskite film. To achieve maximal matched current densities for both
the top perovskite and back silicon sub-solar cells, perovskite films with double-sided textured
architectures are recommended. Through our research, we have observed that employing this
“double-sided textured” perovskite-silicon tandem solar cells can lead to maximum matched
current densities of up to 20.18 mA/cm2 , representing an improvement of up to 24.6% compared
to the planar structure.
In “double-side flat” perovskite films, reflection loss emerges as a major obstacle to effectively
utilizing incident light. To address this issue, light management techniques must be employed.
Firstly, applying an antireflective coating on the front side is essential. Secondly, implementing
light trapping methods such as texturing not only on the back side of the bottom silicon sub-cell to
enhance c-Si absorption in the infrared wavelength range but also on the front perovskite sub-cell
can increase the average path length within the absorber and balance the photon generated current
densities. Thirdly, by texturing both sides of the perovskite films, the reflective loss can be further
reduced compared to only texturing the back side of the perovskites. The double-side textured
tandem structures facilitate current matching, thereby maximizing the overall photocurrent
density (as well as PEC) in perovskite-silicon tandem. This could lead to an increase in the
maximum PEC from 29.20% in the optimized planar reference to 36.14% in the best-performing
textured tandem device.
Additionally, it is important to change the angles of incident light to study the evolution of
“ideally” generated maximum current densities independently for the top perovskite and bottom
silicon sub-cells. Maintaining consistent changes in the maximum generated current densities
between the top and bottom sub-cells as the incident angle varies is also crucial. Tandem solar
cells with textured architectures can help achieve this consistency.
Funding. National Natural Science Foundation of China (62375234, 62322407, 22279034, 52261145698); National
Key Research and Development Program of China (2022YFB3803300); Science and Technology Innovation Plan Of
Shanghai Science and Technology Commission (22ZR1418900).
Acknowledgments. We would like to acknowledge Professor Yadong Xu from Soochow University for the support
of Electromagnetic simulations.
Disclosures. The authors declare no conflicts of interest.
Data availability. Data underlying the results presented in this paper are not publicly available at this time but may
be obtained from the authors upon reasonable request.
Supplemental document. See Supplement 1 for supporting content.
 Research Article Vol. 32, No. 6 / 11 Mar 2024 / Optics Express 8622

References
1. P. Löper, S.-J. Moon, S. M. de Nicolas, et al., “Organic-inorganic halide perovskite/crystalline silicon four-terminal
tandem solar cells,” Phys. Chem. Chem. Phys. 17(3), 1619 (2014).
2. T. Todorov, T. Gershon, O. Gunawan, et al., “Perovskite-kesterite monolithic tandem solar cells with high open-circuit
voltage,” Appl. Phys. Lett. 105(17), 173902 (2014).
3. M. H. Futscher and B. Ehrler, “Modeling the performance limitations and prospects of perovskite/Si tandem solar
cells under realistic operating conditions,” ACS Energy Lett. 2(9), 2089 (2017).
4. K. Bush, A. Palmstrom, Z. Yu, et al., “23.6%-efficient monolithic perovskite/Silicon tandem solar cells with improved
stability,” Nat. Energy 2(4), 17009 (2017).
5. J. Werner, C.-H. Weng, A. Walter, et al., “Efficient monolithic perovskite/silicon tandem solar cell with cell area >1
cm2 ,” J. Phys. Chem. Lett. 7(1), 161 (2016).
6. Hui Li and Wei Zhang, “Perovskite tandem solar cells: from fundamentals to commercial deployment,” Chem. Rev.
120(18), 9835 (2020).
7. L. Duan, D. Walter, N. Chang, et al., “Stability challenges for the commercialization of perovskite-silicon tandem
solar cells,” Nat. Rev. Mater. 8(4), 261 (2023).
8. S. Mariotti, E. Köhnen, F. Scheler, et al., “Interface engineering for high-performance triple-halide perovskite–silicon
tandem solar cells,” Science 381(6653), 63 (2023).
9. X. Y. Chin, D. Turkay, J. A. Steele, et al., “Interface passivation for 31.25%-efficient perovskite/Silicon tandem solar
cells,” Science 381(6653), 59 (2023).
10. M. H. Futscher and B. Ehrler, “Efficiency limit of perovskite/Si tandem solar cells,” ACS Energy Lett. 1(4), 863
(2016).
11. R. Santbergen, R. Mishima, T. Meguro, et al., “Minimizing optical losses in monolithic perovskite/c-Si tandem solar
cells with a flat top cell,” Opt. Express 24(18), A1289 (2016).
12. D. T. Grant, K. R. Catchpole, K. J. Weber, et al., “Design guidelines for perovskite/silicon 2-terminal tandem solar
cells: an optical study,” Opt. Express 24(22), A1454 (2016).
13. K. Jager, L. Korte, B. Rech, et al., “Numerical optical optimization of monolithic planar perovskite-silicon tandem
solar cells with regular and inverted device architectures,” Opt. Express 25(12), A473 (2017).
14. L. Mazzarella, M. Werth, K. Jager, et al., “Infrared photocurrent management in monolithic perovskite/Silicon
heterojunction tandem solar cells by using a nanocrystalline silicon oxide interlayer,” Opt. Express 26(10), A488
(2018).
15. M. Filipič, P. Löper, B. Niesen, et al., “CH3NH3PbI3 perovskite/Silicon tandem solar cells: characterization based
optical simulations,” Opt. Express 23(7), A263 (2015).
16. D. Chen, P. Manley, P. Tockhorn, et al., “Nanophotonic light management for perovskite–silicon tandem solar cells,”
J. Photonics Energy 8(02), 022601 (2018).
17. S. Albrecht, M. Saliba, J. P. Correa-Baena, et al., “Towards optical optimization of planar monolithic
peerovskite/silicon-heterojunction tandem solar cells,” J. Opt. 18(6), 064012 (2016).
18. B. Lipovsek, J. Krc, and M. Topic, “Optical model for thin-film photovoltaic devices with large surface textures at the
front side,” J. Microelectron. Electron. Components Mater. 41, 264 (2011).
19. R. Santbergen and R. J. C. van Zolingen, “The absorption factor of crystalline silicon PV cells: A numerical and
experimental study,” Sol. Energy Mater. Sol. Cells 92(4), 432 (2008).
20. D. Zhang, W. Verhees, M. Dorenkamper, et al., “Combination of advanced optical modeling with electrical simulations
for performance evaluation of practical 4-terminal perovskite/c-Si tandem modules,” Energy Procedia 92, 669 (2016).
21. IEC: 60904–3: Photovoltaic device−Part 3: Measurements principles for terrestrial photovoltaic (PV) solar devices
with reference spectral irradiance data (2008).
22. M. Polyanskiy, Refractive Index INFO (2008). https://refractiveindex.info/
23. M. Burgelman, P. Nollet, and S. Degrave, “Modelling polycrystalline semiconductor solar cells,” Thin Solid Films
361, 527 (2000).
24. K. Decock, P. Zabierowski, and M. Burgelman, “Modeling metastabilities in chalcopyrite-based thin film solar cells,”
J. Appl. Phys. 111(4), 043703 (2012).
25. M. Burgelman, K. Decock, S. Khelifi, et al., “Advanced electrical simulation of thin film solar cells,” Thin Solid
Films 535, 296 (2013).
26. P. Tockhorn, J. Sutter, A. Cruz, et al., “Nano-optical designs for high-efficiency monolithic perovskite–silicon tandem
solar cells,” Nat. Nanotechnol. 17(11), 1214 (2022).
27. A. Callies, M. Hanser, J. C. Goldschmidt, et al., “Structuring of perovskite-silicon tandem solar cells for reduced
reflectance and thermalization losses,” Opt. Express 31(12), 19428 (2023).
28. C. Messmer, B. S. Goraya, S. Nold, et al., “The race for the best silicon bottom cell: Efficiency and cost evaluation of
perovskite–silicon tandem solar cells,” Prog. Photovolt. Res. Appl. 29, 744 (2021).
29. M. De Bastiani, R. Jalmood, J. Liu, et al., “Monolithic perovskite/silicon tandems with >28% efficiency: role of
silicon-surface texture on perovskite properties,” Adv. Funct. Mater. 33, 2205557 (2023).