See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/281344307

Copper Oxide as Inorganic Hole Transport

Material for Lead Halide Perovskite Based Solar
Cells of Enhanced Performance

Article in Solar Energy · October 2015

Impact Factor: 3.47 · DOI: 10.1016/j.solener.2015.07.040

CITATIONS READS

2 217

3 authors, including:

Fahhad Alharbi Nouar Tabet

Qatar Environment and Energy Research Inst… Qatar Foundation
70 PUBLICATIONS 238 CITATIONS 131 PUBLICATIONS 989 CITATIONS

SEE PROFILE SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Nouar Tabet
letting you access and read them immediately. Retrieved on: 12 July 2016
 Available online at www.sciencedirect.com

ScienceDirect
Solar Energy 120 (2015) 370–380
www.elsevier.com/locate/solener

Copper oxide as inorganic hole transport material for lead halide

perovskite based solar cells
Mohammad I. Hossain, Fahhad H. Alharbi, Nouar Tabet ⇑
Qatar Environment & Energy Research Institute, Doha, Qatar

Received 19 May 2015; received in revised form 6 July 2015; accepted 24 July 2015

Communicated by: Associate Editor Takhir M. Razykov

Abstract

The use of inorganic layers as hole transport materials (HTM) has been suggested to enhance the resistance to degradation of methy-
lammonium lead halide perovskite based solar cells. Few inorganic materials have been tested with limited success as they led to a lower
power conversion efficiency (PCE) than the value obtained using spiro-OMETAD. In this work, we used wxAMPS and SCAPS software
to compute the key characteristics of CH3NH3PbI3 based solar cells with various HTM layers including spiro-OMETAD, NiO, CuI,
CuSCN, and Cu2O. The computations were carried out by considering defect free perovskite and HTM layers. The results show that
solar cells containing Cu2O as HTM outperform all other devices with organic or inorganic HTM hitherto tested. A power conversion
efficiency exceeding 24% was obtained. These results indicate that there is a possibility to further increase the performance of perovskite
based cells and reduce their cost by replacing the expensive and moisture sensitive spiro-OMETAD by copper oxide.
! 2015 Elsevier Ltd. All rights reserved.

Keywords: Perovskite solar cell; Hole transport material; Cu2O; Solar cells simulation; SCAPS

1. Introduction made of perovskites show gains that exceed those of the best
devices made of organic materials (Laquai, 2014).
Methylammonium lead halide perovskites (CH3NH3 However, there are few challenges that need to be
PbX3, X1=4 Cl, Br, I) have recently emerged as promising addressed before perovskites dislodge silicon from its dom-
materials for optoelectronics. These materials are solution inant position in photovoltaic industry: resistance to degra-
processable and made of earth abundant elements which dation, replacement of the expensive hole transport
made them materials of choice for the design of cost effec- material (HTM) made of 2,20 ,7,70 -tetrakis(N,N-di-p-meth
tive devices. Since the pioneering work of Miyasaka on oxyphenylamine)-90 ,9-spirobifluorene (spiro-OMETAD)
Dye Sensitized cells in 2009 (Kojima et al., 2009), the use and the substitution of lead by a non-toxic element.
of perovskites as absorbing materials has enabled research- Intensive efforts are being deployed worldwide to tackle
ers to design solar cells with power conversion efficiency the above issues. Moisture and UV light were found respon-
exceeding 15% (Burschka et al., 2013). Furthermore, lasers sible of the degradation of the lead halide perovskite based
cells. Spiro-OMETAD HTM has been routinely used in
organic photovoltaics (OPV) and organic optoelectronics.
For that reason, it was suggested and used for perovskites
⇑ Corresponding author. based solar cells. However, it is known that
E-mail address: ntabet@qf.org.qa (N. Tabet). spiro-OMETAD causes degradation in the devices

http://dx.doi.org/10.1016/j.solener.2015.07.040
0038-092X/! 2015 Elsevier Ltd. All rights reserved.
 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380 371

performance and the field is still research active (Nogueira 2. Cu2O properties
et al., 2004; Pudzich et al., 2006). There is some published
work dedicated to the HTM layer and its effect on the sta- Cuprous oxide (Cu2O) is a p-type semiconductor with a
bility of perovskite solar cells (PSC) (Malinauska et al., direct bandgap of (Eg ! 2.17 eV) (Gou and Murphy, 2003;
2015; Badia et al., 2014; Nguyen et al., 2014). There is also Wang et al., 2002; Liao et al., 2009; Scanlon et al., 2009;
a reasonable literature on the degradation of the perfor- Hossain and Alharbi, 2013). The first attempt to use this
mance of Dye Sensitized cell (DSC) due to the use of material for PV application was done by national science
spiro-OMETAD (Bailie et al., 2014; O’Brien et al., 1998). foundation and joint centre for graduate studies in 1978
It is reasonable to expect that the same conclusions apply (Abdu and Musa, 2009). Since then various techniques
to PSC as the reported issues are highly associated with were used to fabricate schottky junction, hetero-junction,
the spiro-OMETAD itself. These issues are mainly related and homo-junction devices using Cu2O as a light harvester
to the thermal stress effects (Malinauska et al., 2015; (Minami et al., 2014; McShane et al., 2010; Nishi et al.,
Bailie et al., 2014; O’Brien et al., 1998) and oxidization 2012). However, low device performance was observed
effects (Badia et al., 2014; Nguyen et al., 2014). due to low absorption in the long wavelength region
Additionally, Snaith et al. reported recently TiO2-free because of high energy gap. Due to low electron affinity
device that showed stability of the photocurrent over (3.2 eV) (Jeong et al., 2008) and very high hole mobility
1000 h of exposure to simulated sunlight (Leijtens et al., (256 cm2 V"1 s"1) (Li et al., 2009), this p-type semiconduc-
2013). The loss of photocurrent was assigned by the authors tor has been suggested as a potential hole transport mate-
to carrier recombination process at oxygen vacancy related rial in heterojunction based solar cells (Lee et al., 2013a).
sites on the surface of TiO2. Cu2O thin films can be prepared using various techniques
Niu et al. reported that the insertion of thin layer of alu- including sputtering (Lee et al., 2013a,b; Li et al., 2011,
mina improves the stability of the CH3NH3PbI3 films when 2012; Nakano et al., 2009; Siah et al., 2012; Malerba
exposed to moisture and reduces the photo carrier recom- et al., 2012), copper oxidation (Biccari, 2009), and atomic
bination (Niu et al., 2014). The use of inorganic p-type Layer deposition (ALD) (Waechtler et al., 2009; Lee
materials as hole transport media offers the double advan- et al., 2014). Unintentionally doped films are naturally
tage to reduce the cost of the cell and enhance its resistance p-type because of the native defects identified as negatively
to degradation. Few inorganic materials have been tested charged copper vacancies (V0 Cu) rather than interstitial
such as copper iodide (CuI) (Christians et al., 2014), copper oxygen (O0 i) (Scanlon et al., 2009; Biccari, 2009; Raebiger
thiocyanate (CuSCN) (Qin et al., 2014), and nickel oxide et al., 2007). Experimental work reported hole trapping
(Subbiah et al., 2014). level between 0.40 eV and 0.55 eV above the valence band
The quest for good Inorganic hole transport material (Ev). A native defect state has been investigated as well at
must be guided by the need for high carrier mobility and 0.45 eV above the valence band assigned to copper vacan-
defect free interface with the absorbing layer to minimize cies (Scanlon et al., 2009). Nitrogen doping has been used
carrier recombination. Computer simulation is a valuable to prepare samples containing density of holes as high as
tool that can provide useful information for the screening 1018 cm"3 (Lee et al., 2013a,b; Li et al., 2011, 2012;
process of materials to be tested experimentally. Nakano et al., 2009; Siah et al., 2012; Malerba et al.,
In this work, we calculate the key parameters of lead 2012; Biccari, 2009). We carried out our computations
halide perovskite solar cell with the following configura- using materials parameters that were extracted from pub-
tion: Glass/FTO/TiO2/CH3NH3PbI3/HTM/Au, HTM lished studies. These values are reported in Table 1.
being NiO, CuI, CuSCN, Spiro-OMETAD or Cu2O. The The energy levels alignment is an important factor that
computations were carried out using wx Analysis of affects the performance of the cell. Photoelectrons are
Microelectronic and Photonic Structures (wxAMPS) and injected from the perovskite to TiO2 and holes from per-
1D-Solar Cell Capacitance Simulator (SCAPS) software. ovskite to HTM. The extraction of photoelectrons at the

Table 1
Material properties of Cu2O.
Material properties p-Cu2O
Dielectric permittivity 7.11 (Zhu et al., 2011)
Electron mobility (cm2/V s) 2.000E+2 (Korzhavyi and Johansson, 2011)
Hole mobility (cm2/V s) 8.000E+1 (Matsumura et al., 1996)
Acceptor concentration (1/cm3) 1.00E+18 (Lee et al., 2013a,b; Li et al., 2011, 2012; Nakano et al., 2009;
Siah et al., 2012; Malerba et al., 2012)
Bandgap (eV) 2.17 (Hossain and Alharbi, 2013)
CB DOS (1/cm3) 2.02E+17 (Zhu et al., 2011)
VB DOS(1/cm3) 1.10E+19 (Zhu et al., 2011)
Affinity (eV) 3.20 (Hossain and Alharbi, 2013)
Defect level (above the edge of EV) 0.45 eV (Paul et al., 2006)
372 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380

Fig. 1. Device structure and energy level diagram of the TiO2/CH3NH3PbI3/Cu2O/Au.

TiO2/perovskite interface requires that the electron affinity (Soon et al., 2009; Tretyakov et al., 1972). The careful thick-
(EA) of the perovskite be higher than that of the TiO2 ness selection of an inorganic hole transport material like
while the extraction of holes at the HTM/perovskite inter- Cu2O can act as a capping layer which prevents the contact
face requires that the ionization energy of HTM be lower between the perovskite and gold. The numerical analysis of
than that of the perovskite. The energy levels mismatches the cell performance was carried out assuming first, defect
at both interfaces affect both the short circuit current and free Cu2O and perovskite. Subsequently, the effect of defects
the open circuit voltage. The energy diagram of the device in both layers was considered.
using Cu2O as HTM is displayed in Fig. 1.
Note that a 0.7 eV energy barrier for electrons exists at the 3. Simulation procedure
perovskite/Cu2O interface and prevents the transfer of pho-
toelectrons to the copper oxide layers. The electronic and The numerical analysis has been carried out using wx
optical properties of Cu2O are strongly affected by point Analysis of Microelectronic and Photonic Structures
and structural defects which are material growth dependent (wxAMPS) software which has been developed at
Table 2
Material properties of ETM, absorber, and four types of HTMs.
TiO2 CH3NH3PbI3 Spiro-OmeTAD p-CuSCN p-NiO p-CuI
Dielectric 10 (Xosrovashvilia 10 (Homes, 3 10 (Dhakal 10.7 (Kurnia, 6.5
permittivity and Gorjib, 2013) 2001) (Kalyanasundaram, et al., 2011) 2010) (Madelung,
2010) 2004)
Electron mobility 100 (Xosrovashvilia 1.00E+1 1.000E"4 (Schmidt- 100 (Dhakal 12 (Chia- 100
(cm2/V s) and Gorjib, 2013) (Wehrenfennig Mende and Gratzel, et al., 2011) ChingW, (Fonash,
et al., 2014) 2006) 2013) 2010)
Hole mobility 25 (Xosrovashvilia 1.0E+1 2.0E"4 (Schmidt- 25 (Dhakal 2.8 (Tyagi 43.9 (Chen
(cm2/V s) and Gorjib, 2013) (Wehrenfennig Mende and Gratzel, et al., 2011) et al., 2012) et al., 2010)
et al., 2014) 2006)
Acceptor 0 1.0E+9 1.0E+18 1.0E+18 1.0E+18 1.0E+18
concentration
(1/cm3)
Donor 1.0E+17 1.0E+9 0 0 0 0
concentration (Xosrovashvilia and
(1/cm3) Gorjib, 2013)
Bandgap (eV) 3.26 (Xosrovashvilia 1.5 (Jeng et al., 2 3.06 (Leijtens et al., 3.6 (Ahirrao 3.8 (Subbiah 3.1 (Wu and
and Gorjib, 2013) 013) 2012) et al., 2011) et al., 2014) Wang, 1997)
3
CB DOS (1/cm ) 2.0E+17 2.75E+18 2.80E+19 (Fonash, 2.2E+19 2.8E+19 2.8 = E+19
(Xosrovashvilia and Gorjib, (Giorgi 2010) (Dhakal (Fonash, (Fonash,
2013) et al., 2013) et al., 2011) 2010) 2010)
VB DOS (1/cm3) 6.0E+17 3.9E+18 (Giorgi 1.0E+19 (Fonash, 1.8E+18 1.0E+19 1.0E+19
(Xosrovashvilia and et al., 2013) 2010) (Dhakal (Fonash, (Fonash,
Gorjib, 2013) et al., 2011) 2010) 2010)
Affinity (eV) 4.2 (Xosrovashvilia 3.9 (Jeng et al., 2.05 (Leijtens et al., 1.7 (Ahirrao 1.46 (Wu 2.1 (Hodes,
and Gorjib, 2013) 2013) 2012) et al., 2011) and Wang, 2013)
1997)
Band to band 2.30E"9 (Manser 2.30E"9 2.30E"9
recombination and Kamat, 2014)
rate
(cm3 s"1)
 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380 373

limited to, transport mechanism, individual carrier current

densities, electric field distributions and recombination
profiles. The properties of the materials are obtained from
the literature. The parameters of materials other than
Cu2O used in our simulations are listed down in Table 2.
The effective density of states in the conduction and valence
bands of perovskite have been calculated using the effective
masses obtained from electronic structure calculation on
the pseudocubic phase, where electron effective mass,
me*/mo = 0.23 and hole effective mass, mh*/mo = 0.29.
Fig. 2 captures the energy band alignment of various
Fig. 2. Energy band alignment of lead iodide perovskite and four different materials used in our simulations.
HTMs.
4. Results and discussion

4.1. Optimization of the layer thicknesses

The HTM layer, absorber, and ETM layer thicknesses

were optimized considering defect free layers. The initial
values of thicknesses of the absorber and ETM layer were
set equal to 300 nm and 100 nm respectively. These values
were reported as optimum values for perovskite based cells
including spiro-OMETAD as HTM (Snaith, 2013; Edri
et al., 2014). The iteration process was launched to opti-
mize the HTM thickness. Then using the obtained opti-
mum value of HTM, iterations were carried out to
compute the new optimum ETM layer thickness. The pro-
cess was repeated many times to determine the optimum set
of thicknesses values (ETM, HTM) for the five considered
cell structures.
Finally the optimized values of HTM and ETM are used
to compute the optimum thickness of the absorbing layer
as shown in Fig. 3. The optimized values for the three lay-
ers obtained through the iteration process are reported in
Table 3. One can note that the optimum values rare about
150 nm for ETM and HTM and 350–450 nm for the per-
Fig. 3. Iteration process flow chart to optimize layer thickness.
ovskite absorbing layer. It should be pointed out that these
optimum values are sensitive to the carrier mobilities. For
University of Illinois (Liu et al., 2011, 2012). The software instance, the electron mobility in TiO2 layer used in per-
solves numerically the three main equations that govern the ovskite based solar cells can be as low as 10"1 cm2/V s.
carrier transport, namely Poisson and continuity equations Simulations performed using such value lead to an opti-
for electrons and holes. Computations were also carried mum thickness of 100 nm for the ETM layer in agreement
out using solar cell capacitance simulator (SCAPS) devel- with the experimental values reported in the literature.
oped at the Department of Electronics and Information
Systems of the University of Gent in Belgium (Burgelman 4.2. Simulation of cell parameters
et al., 2000; Decock et al., 2012, 2011) to consolidate the
results. Like most of solar cell simulators, SCAPS captures The key characteristics of the optimized solar cells were
the analytical physics of the solar cell device, but not computed using both softwares. These include the fill
Table 3
Optimized thicknesses of ETM, absorber layer, and HTM.
HTM types ETM thickness (nm) Absorber thickness (nm) HTM thickness (nm)
Cu2O 140 350 150
Spiro-OMETAD 140 450 150
CuSCN 145 450 200
NiO 135 450 200
CuI 145 400 200
 374 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380

factor (FF), the open circuit voltage (Voc), the short circuit

7.26% (Subbiah et al.,

73 (Burschka et al.,
current (Jsc) and the power conversion efficiency (PCE)

12.4% (Qin et al.,

6.0% (Christians
corresponding to different types of HTMs. The obtained

et al., 2014)
values are compiled in Table 4.
The results show clearly that the device including Cu2O
2013)

2014)

2014)
Exp.

as hole transport material has the highest performance.

–

The PCE values obtained by wxAMPS are slightly higher

than those obtained by SCAPS but the two softwares pro-
wxAMPS

vide consistent ranking of the calculated performances for

24.40
79.73

22.38

19.89

19.60
different HTMs: Cu2O (highest), CuSCN, OMeTAD, CuI,
then NiO (lowest performance).
PCE (%)

22.64571

17.28833

17.54064
20.0576
SCAPS

We have also reported the experimental values that have

80.567

been hitherto reported in the literature. The experimental

values are significantly lower than the simulated ones, espe-
60 (Christians et al.,
62 (Qin et al., 2014)

cially in the case of NiO and CuI. This is expected because

75 (Subbiah et al.,

point defects in the bulk of the absorbing layer and at the

20.0 (Burschka

interfaces TiO2/perovskite and perovskite/HTM can act as

et al., 2013)

recombination centres that lower both the collected current

2014)

2014)

and the open circuit voltage.

Exp.

Lead Halide perovskites are characterized by a high

–

light absorbance and a relatively high carrier diffusion

wxAMPS

length reaching 1 lm (Stranks et al., 2013). A layer of

82.14
24.31

77.72

79.38

78.64

few hundred nanometer thickness is enough to absorb most

of the incident sun light. Therefore, most of the photo car-
SCAPS
FF (%)

83.057

79.797

81.467

80.777

riers can be collected as they are generated at distances less

21.83

than the diffusion length away from the interfaces

perovskite/TiO2 and perovskite/HTM.
17.8 (Christians et al.,
19.7 (Qin et al., 2014)

14.9 (Subbiah et al.,

We have calculated the variation of the key parameters

0.993 (Burschka

that determine the solar cell efficiency as a function of the

absorber layer thickness in the range of 200–600 nm using
et al., 2013)

wxAMPS. The results are displayed on Fig. 4. It can be

noted that, as expected, the short circuit current increases
2014)

2014)
Exp.

as the absorber layer thickness increases for all types of

–

cells up to a certain thickness (!400 nm) and then saturates

wxAMPS

with very slight decay due to the complete absorption of

24.75

23.19

22.00

23.08
1.18

the incident photons giving a maximum generated current

Jsc (mA/cm2)

(Fig. 4a). As the defect assisted carrier recombination

SCAPS

was neglected, the slight decrease beyond 400 nm is related

22.74
1.139

20.62

20.23

21.31

to the very small band to band recombination rate which is

given by the following equation (Green and Cells, 1982)
0.55 (Christians et al.,
0.936 (Subbiah et al.,

U ðcm"3 s"1 Þ ¼ Bðnp " n2i Þ

1.016 (Qin et al.,
(Burschka et al.,
OMETAD

where n and p are the carrier densities under excitation and

ni the intrinsic density of the perovskite, B is a constant
Optimized performances with various HTMs.

characteristic of the material. We have used the value

2013)

2014)

2014)

2014)
Exp.

B = 2 & 10"9 cm3 s"1 reported in Manser and Kamat

–

(2014). A clear decrease of the open circuit voltage Voc

wxAMPS

and the power conversion efficiency is observed as the

15.0%

thickness increases beyond 400 nm as shown in Fig. 4d

1.20

1.24

1.13

1.07

while the fill factor remains constant.

In order to confirm that the decrease of Voc and PCE
SCAPS
Voc (V)

1.199

22.99

1.219

1.049

1.019

beyond 400 nm is related to current loss due to the carrier

recombination, we have run simulations using two different
values of the recombination constant: B = 7.2 & 10"9
cm3 s"1 and B = 7.2 & 1012 cm3 s"1 (Fig. 5).
20.03248

CuSCN
Table 4

Spiro-
HTM

Cu2O

It can be clearly observed that both Jsc and Voc and thus
NiO

CuI

PCE saturates beyond 400 nm thickness of the absorbing

 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380 375

Fig. 4. Variation in solar cells performance in the range of 200–600 nm of absorber layer thickness for recombination value of 2.30E"9 cm3 s"1. (a) Open
circuit voltage (Voc), (b) short circuit current (Jsc), (c) fill factor (FF), and (d) efficiency (g) of perovskite solar cells.

layer for the small value of the recombination constant B is absorbed in the perovskite layer, the observed difference
and decrease for the higher value. It is also worth noting between different devices may be assigned to the band mis-
that the Voc decrease for large thickness of the perovskite match at the interfaces: DEc = Ec (Cu2O) " Ec (perovskite)
layer is more significant than that of the short circuit cur- and DEV = EV (Cu2O) " EV (perovskite). An increase of
rent. This can be understood by recalling that Voc is func- DEC would increase the blocking barrier for electrons at
tion of both Jsc and the saturation current of the device J0 the perovskite/HTM interface. Similarly, an increase of
which depends also on the thickness of the layer (Green DEv is expected to reduce the flow of holes back from the
and Cells, 1982) HTM to the perovskite, hence an increase of the net cur-
! " rent of holes collected at the gold contact.
kT J sc
V oc ¼ Ln þ1 In order to analyse the effect of the band mismatch on
e Jo
the current extraction we simulated the variation of Jsc as
The dependence of the saturation current of the device a function DEV. A rigid shift of both the valence band
versus the thickness of the absorbing layer is displayed and conduction band of Cu2O was used while keeping
on Fig. 6. It appears clear that the variation of the open cir- the band gap constant.
cuit voltage Voc is mainly determined by the dependence of The results show an increase of the current as DEV
the saturation current of the device J0. Furthermore, increases from 0.01 to 0.03 eV which is about the thermal
Fig. 4(c) shows that FF remains quasi constant as the energy at room temperature. Beyond this value, Jsc satu-
thickness of HTM layer varies. The fill factor of Cu2O is rates because the reverse flow of holes becomes negligible.
relatively closer to other HTM devices, which indicates that However, an increase of DEV corresponds to a shift of EV
the highest efficiency obtained for the Cu2O based device is (Cu2O) above the level EV (perovskite). This leads to a
mainly due to high short circuit current and open circuit reduction of the open circuit voltage as expected
voltage. The short circuit current depends mainly on the (Seungchan et al., 2014) and shown in Fig. 7.
light absorption and the carrier transport properties in We have also considered a possible effect of the trans-
the perovskite layer and the bands mismatch at the inter- port properties of holes in the Cu2O layer on the calculated
faces perovskite/TiO2 and perovskite/HTM. The efficiency. We have computed the variation of Jsc, Voc and
TiO2/perovskite is unchanged in all devices which were FF as function of the hole mobility in Cu2O. Fig. 8 shows
considered in this study. Since most of the incident light that an increase of the hole mobility increases Jsc through
376 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380

Fig. 5. Variation in solar cells performance in the range of 200–600 nm of absorber layer thickness for recombination value of 7.20 & 10"9 and
7.20 & 10"12 cm3 s"1 using Cu2O HTM. (a) Open circuit voltage (Voc), (b) short circuit current (Jsc), (c) fill factor (FF), and (d) efficiency (g) of perovskite
solar cells.

out simulations to study the distribution of the electric field

across the device as a function of the carrier mobility. The
results are shown in Fig. 9. A drastic increase of the electric
field can be clearly noted at both interfaces
HTM/perovskite and ETM/perovskite as the mobility
exceeds 50 cm2/V s which indicates a strong correlation
with the observed increase of the short circuit current.

4.3. Effects of defect states in perovskite absorber and Cu2O

HTM layers

The above simulations were carried out assuming defect

free device. In reality, various point defects such vacancies
or extended defects such as dislocations and grain bound-
Fig. 6. Saturation current of solar cells versus the thickness of the aries are always present in the perovskite absorbing layer
perovskite layer for different HTMs. and the HTM layer. These defects induce carrier recombi-
nation and lifetime reduction and carrier mobility decrease
in the HTM layer. In order to compute the effect of defects,
an enhancement of the carrier extraction at the interface we have carried out simulations of device performance in
perovskite/HTM. Similar effect has been reported in presence of selected point defects suggested in the pub-
William et al. (2014). The contributions of the ETM and lished literature.
HTM to the series resistance of the device are quite small It must be pointed out that the carrier recombination in
because of the high majority carrier concentrations consid- perovskites has been the subject of intensive discussions.
ered in our simulations (1017 and 1018 cm"3 respectively) The recombination rate is usually expressed as a polyno-
and the small thicknesses of the layers. We have carried mial dependence of the carrier density including three
 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380 377

Fig. 7. Cell parameters versus the valance band alignment mismatch at Cu2O/perovskite interface.

Fig. 8. Cell parameters versus hole mobility of Cu2O.

378 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380

k2 and k3: k2 = 9.9&10"10 cm3 s"1, and k3 = 1.3 &

10"28 cm6 s"1.
However, Wetzelaer et al. (2015) suggested a dominance
of trap assisted recombination based on a detailed analysis
of the ideality factor of the perovskite cell for a density of
traps of 5 & 1016 cm"3 which is considered as a low value
according to Xing et al. (2014). In order to assess the effect
of SRH recombination on the cell performance, we have
considered various densities of traps in the perovskite
and HTM.
Simulations were performed assuming the presence of a
trap located as (0.45 + Ev) in the HTM layer and a level
located at (0.05 + Ev) in the perovskite layer. These defect
levels were suggested in Wang et al. (2002) and Yin et al.
(2014), respectively.
Fig. 9. Electric field distribution at HTM/perovskite (0.145 lm) and First we calculated the effect of an increasing defect con-
perovskite/ETM (0.495 lm) with increasing hole mobility in Cu2O HTM centration in Cu2O layer assuming a defect free perovskite.
layer. Then similar calculations were carried for an increasing
density of defects in perovskite while Cu2O layer was
terms (Manser and Kamat, 2014; Wehrenfennig et al., assumed defect free. The results are displayed in Fig. 10.
2014; Trinh et al., 2015) It can be noticed that the efficiency is reduced to 18% as
a result of monomolecular recombination (SRH) for a den-
dn sity of traps of 1016 cm"3. This result can be compared to
¼ "k 3 n3 " k 2 n2 " k 1 n the data displayed on Fig. 5 which shows that the efficiency
dt
of the cell is reduced from 28% to 26% only as the bimolec-
ular recombination rate constant B is increased from
where n is the excess carrier density. The cubic term refers 7.2 & 10"12 cm3 s"1 to 2.7 & 10"9 cm3 s"1. These results
to three body Auger recombination, the quadratic term to suggest that SRH recombination can still have a significant
the bimolecular recombination, and the linear term to contribution to carrier loss if the material includes high
monomolecular recombination i.e. trap assisted recombi- density of traps.
nation called also Shockley Read Hall (SRH) process. As expected, all cell parameters are sensitive to high
Based on transient absorption study, Trinh et al. defect density in the absorbing layer. This efficiency varia-
reported that Auger process becomes significant only at tion is mainly due to presence of defect states thus a reduc-
high injection levels with carrier density exceeding tion of short circuit current, open circuit voltage, and fill
1018 cm"3 (Trinh et al., 2015) and suggested the following factor. However, it is clear that defects states in p-Cu2O
values for the charge decay constants k1 = 0, k2 = have a smaller effect on the device performance as com-
1.1 & 10"9 cm3 s"1, and k3 = 3.8 & 10"28 cm6 s"1. pared to the defects in the absorbing layer. This can be
Wehrenfennig et al. (2014) concluded that the bimolecu- understood as most of the light is absorbed in the per-
lar recombination dominates based also on transient ovskite layer thus a more significant effect is expected on
absorption measurements and reported similar values for the carrier lifetime and the recombination rate.

Fig. 10. Device performances in the range 1014–1017 cm"3 defect states. (a) Defect in Cu2O, perovskite defect free. (b) Defect in perovskite, Cu2O defect free.
 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380 379

5. Conclusion methylammonium lead iodide perovskite: a density functional analysis.

J. Phys. Chem. Lett. 4, 4213–4216.
Gou, L., Murphy, C.J., 2003. Solution-phase synthesis of Cu2O
We have shown that Cu2O constitutes a promising hole nanocubes. Nano Lett. 3, 231–234.
transport material that offers the possibility to enhance fur- Green, Martin A., Cells, Solar., 1982. Operating Principles, Technology,
ther the efficiency of perovskite based solar cells. High and System Applications. Prentice-Hall Inc., Englewood Cliffs.
quality thin films of copper oxide with hole mobility as Hodes, G., 2013. Perovskite-based solar cells. Science 342, 317–318.
high as 256 cm2/V s can be prepared by sputtering tech- Homes, C.C., 2001. Optical response of high-dielectric-constant per-
ovskite-related oxide. Science 293, 673–676.
nique. Furthermore the p-type conductivity of the oxide Hossain, M.I., Alharbi, F.H., 2013. Recent advances in alternative
can be enhanced by nitrogen doping in order to enhance material photovoltaics. Mater. Technol.: Adv. Perform. Mater. 28, 88–
the carrier extraction. In addition, the use of copper oxide 97.
is expected to provide a protection to the perovskite against Jeng, J., Chiang, Y., Lee, M., Peng, S., Guo, T., Chen, P., Wen, T., 2013.
moisture leading to an improvement of the device resis- Methylammonium Lead Iodide Perovskite/Fullerene-Based Hybrid
Solar Cells. SPIE Newsroom. http://dx.doi.org/10.1117/
tance to degradation. However, the quality of the copper 2.1201307.005033.
oxide/perovskite interface is expected to play a critical role Jeong, S.S., Mittiga, A., Salza, E., Masci, A., Passerini, S., 2008.
on the actual performance of the device. Electrodeposited ZnO/Cu2O heterojunction solar cells. Electrochim.
Acta 53, 2226–2231.
References Kalyanasundaram, K., 2010. In: Dye-sensitized Solar Cells (Fundamental
Sciences: Chemistry). EPFL Press, ISBN-10: 143980866X.
Kojima, A., Teshima, K., Shirai, Y., Miyasaka, T., 2009. Organometal
Abdu, Y., Musa, A.O., 2009. Copper (I) oxide (Cu2O) based solar cells – a
halide perovskites as visible-light sensitizers for photovoltaic cells. J.
review. J. Pure Appl. Sci. 2, 8–12.
Am. Chem. Soc. 131, 6050–6051.
Ahirrao, P.B., Gosavi, S.R., Sonawane, S.S., Patil, R.S., 2011. Wide band
Korzhavyi, P.A., Johansson, B. Literature review on the properties of
gap nanocrystalline CuSCN thin films deposited by modified chemical
cuprous oxide Cu2O and the process of copper oxidation. Technical
method. Archives Phys. Res. 2 (3), 29–33.
Report 2011, TR-11-08.
Badia, L., Mas-Marza, E., Sánchez, R.S., Barea, E.M., Bisquert, J., Mora-
Kurnia, F., Hadiyawarman, Jung, C.U., Liu, C., 2010. Effect of NiO
Seró, I., 2014. New iridium complex as additive to the spiro-OMeTAD
growth conditions on the bipolar resistance memory switching of Pt/
in perovskite solar cells with enhanced stability. APL Mater. 2, 081507.
NiO/SRO structure. J. Korean Phys. Soc. 57, 1856–1861.
Bailie, C.D., Unger, E.L., Zakeeruddin, S.M., McGehee, M.D., 2014.
Laquai, F., 2014. Materials for lasers: all-round perovskites. Nat. Mater.
Melt-infiltration of spiro-OMeTAD and thermal instability of solid-
13, 429.
state dye-sensitized solar cells. Phys. Chem. Chem. Phys. 16, 4864–
Lee, Y.S., Heo, J., Winkler, M.T., Siah, S.C., Kim, S.B., Gordon, R.G.,
4870.
Buonassisi, T., 2013a. Nitrogen-doped cuprous oxide as a p-type hole
Biccari, F. Defects and doping in Cu2O. A thesis submitted in partial
transporting layer in thin-film solar cells. J. Mater. Chem. A 1, 15416.
fulfillment of the requirements for the degree of doctor of philosophy
Lee, S.H., Yun, S.J., Lim, J.W., 2013b. The Characteristics of Cu2O Thin
in physics, December 2009, Universita di Roma.
Films Deposited Using RF-Magnetron Sputtering Method with
Burgelman, M., Nollet, P., Degrave, S., 2000. Modelling polycrystalline
Nitrogen-Ambient. ETRI J. 35, 1156.
semiconductor solar cells. Thin Solid Films 361–362, 527–532.
Lee, S.W., Lee, Y.S., Heo, J., Siah, S.C., Chua, D., Brandt, R.E., Kim,
Burschka, J., Pellet, N., Moon, S., Humphry-Baker, R., Gao, P.,
S.B., Mailoa, J.P., Buonassisi, T., Gordon, R.G., 2014. Improved
Nazeeruddin, M.K., Grätzel, M., 2013. Sequential deposition as a
Cu2O – based solar cells using atomic layer deposition to control the
route to high-performance perovskite-sensitized solar cells. Nature
Cu oxidation state at the p–n junction. Adv. Energy Mater. 4, 0. http://
499, 316–319.
dx.doi.org/10.1002/aenm.20130191.
Chen, D., Wang, Y., Lin, Z., Huang, J., Chen, X.Z., Pan, D., Huang, F.,
Leijtens, T., Ding, I., Giovenzana, T., Bloking, J.T., McGehee, M.D.,
2010. Growth strategy and physical properties of the high mobility p-
Sellinger, A., 2012. Hole transport materials with low glass transition
type cui crystal. Cryst. Growth Des. 10 (5), 2057–2060.
temperatures and high solubility for application in solid-state dye-
Chia-ChingW, Cheng.-Fu Y., 2013. Investigation of the properties of
sensitized solar cells. ACS Nano 6, 1455–1462.
nanostructured Li-doped NiO films using the modified spray pyrolysis
Leijtens, T., Eperon, G.E., Pathak, S.A., Abate, A., Lee, M.M., Snaith,
method. Nanoscale Res. Lett. 8, 33.
H.J., 2013. Overcoming ultraviolet light instability of sensitized TiO2
Christians, J.A., Fung, R.C., Kamat, P.V., 2014. An inorganic hole
with meso-superstructured organometal tri-halide perovskite solar
conductor for organo-lead halide perovskite, solar cells. Improved hole
cells. Nat. Commun. 4.
conductivity with copper iodide. J. Am. Chem. Soc. 136, 758–764.
Li, B.S., Akimoto, K., Shen, A., 2009. Growth of Cu2O thin films with
Decock, K., Khelifi, S., Burgelman, M., 2011. Modelling multivalent
high hole mobility by introducing a low-temperature buffer layer. J.
defects in thin film solar cells. Thin Solid Films 519, 7481–7484.
Cryst. Growth 311, 1102–1105.
Decock, K., Zabierowski, P., Burgelman, M., 2012. Modeling metasta-
Li, H.J., Pu, C.Y., Ma, C.Y., Li, S., Dong, W.J., Bao, S.Y., Zhang, Q.Y.,
bilities in chalcopyrite-based thin film solar cells. J. Appl. Phys. 111,
2011. Growth behaviour and optical properties of N-doped Cu2O
043703.
films. Thin Solid Films 520, 212–216.
Dhakal, R., Huh, Y., Galipeau, D., Yan, X., 2011. AlSb compound
Li, B.B., Lin, L., Shen, H.L., Boafo, F.E., Chen, Z.F., Liu, B., Zhang, R.,
semiconductor as absorber layer in thin film solar cells. In: In Book:
2012. Effect of N doping on hole density of Cu2O: N films prepared by
Solar Cells – New Aspects and Solutions. InTech.
the reactive magnetron sputtering method. Eur. Phys. J. – Appl. Phys.
Edri, E., Kirmayer, S., Mukhopadhyay, S., Gartsman, K., Hodes, G.,
58.
Cahen, D., 2014. Elucidating the charge carrier separation and
Liao, L., Yan, B., Hao, Y.F., Xing, G.Z., Liu, J.P., Zhao, B.C., Shen,
working mechanism of CH3NH3PbI3"xClx perovskite solar cells.
Z.X., Wu, T., Wang, L., Thong, J.T., Li, C.M., Huang, W., Yu, T.,
Nat. Commun. 5, 3461.
2009. P-type electrical, photoconductive, and anomalous ferromag-
Fonash, 2010. Solar Cell Device Physics, second ed. Academic Press.
netic properties of Cu2O nanowires. Appl. Phys. Lett. 94, 113106.
Giorgi, G., Fujisawa, J., Segawa, H., Yamashita, K., 2013. Small
Liu, Y., Heinzel, D., Rockett, A. A New Solar Cell Simulator:
photocarrier effective masses featuring ambipolar transport in
wxAMPS.37th IEEE PVSC, 002753–002756; 2011.
380 M.I. Hossain et al. / Solar Energy 120 (2015) 370–380

Liu, Y., Sun, Y., Rockett, A., 2012. A new simulation software of solar solar cells with high open-circuit voltage and fill factor. Energy
cells—wxAMPS. Sol. Energy Mater. Sol. Cells 98, 124–128. Environ. Sci. 7, 2614–2618.
Madelung, O., 2004. Semiconductors Data Handbook, Basic Data, third Siah, S.C., Lee, Y.S., Segal, Y., Buonassisi, T., 2012. Low contact
ed. Springer. resistivity of metals on nitrogen-doped cuprous oxide (Cu2O) thin-
Malerba, C., Ricardo, C.L., D’Incau, M., Biccari, F., Scardi, P., Mittiga, films. J. Appl. Phys. 112, 084508.
A., 2012. Nitrogen doped Cu2O: a possible material for intermediate Snaith, H.J., 2013. Perovskites: the emergence of a new era for low-cost,
band solar cells? Sol. Energy Mater. Sol. Cells 105, 192–195. high-efficiency solar cells. J. Phys. Chem. Lett. 4, 3623–3630.
Malinauska, T., Tomkute-Luksiene, D., Sens, R., Daskeviciene, M., Send, Soon, A., Cui, X., Delley, B., Wei, S., Stampfl, C., 2009. Native defect-
R., Wonneberger, H., Jankauskas, V., Bruder, I., Getautis, V., 2015. induced multifarious magnetism in nonstoichiometric cuprous oxide:
Enhancing thermal stability and lifetime of solid-state dyesensitized first-principles study of bulk and surface properties of Cu2"O. Phys.
solar cells via molecular engineering of the hole transporting material Rev. B 79, 035205.
spiro-OMeTAD. ACS Appl. Mater. Interfaces 7, 11107–11116. Stranks, S.D., Eperon, G.E., Grancini, G., Menelaou, C., Alcocer, M.J.,
Manser, J.S., Kamat, P.V., 2014. Band filling with free charge carriers in Leijtens, T., Herz, L.M., Petrozza, A., Snaith, H.J., 2013. Electron-
organometal halide perovskites. Nat. Photonics 8, 737–743. hole diffusion lengths exceeding 1 micrometer in an organometal
Matsumura, H., Fujii, A., Kitatani, T., 1996. Properties of high-mobility trihalide perovskite absorber. Science 342, 341–344.
Cu2O films prepared by thermal oxidation of Cu at low temperatures. Subbiah, A.S., Halder, A., Ghosh, S., Mahuli, N., Hodes, G., Sarkar,
Jpn. J. Appl. Phys. 35, 5631. S.K., 2014. Inorganic hole conducting layers for perovskite-based solar
McShane, C.M., Siripala, W.P., Choi, K.S., 2010. Effect of junction cells. J. Phys. Chem. Lett. 5, 1748–1753.
morphology on the performance of polycrystalline Cu2O homojunc- Tretyakov, Y.D., Komarov, V.F., Prosvirnina, N.A., Kutsenok, I.B.,
tion solar cells. J. Phys. Chem. Lett. 1, 2666–2670. 1972. Nonstoichiometry and defect structures in copper oxides and
Minami, T., Miyata, T., Nishi, Y., 2014. Cu2O-based heterojunction solar ferrites. J. Solid State Chem. 5, 157–167.
cells with an Al-doped ZnO/oxide semiconductor/thermally oxidized Trinh, M.T., Wu, X., Niesner, D., Zhu, X.Y., 2015. Many-body
Cu2O sheet structure. Sol. Energy 105, 206–217. interactions in photo-excited lead iodide perovskite. J. Mater. Chem.
Nakano, Y., Saeki, S., Morikawa, T., 2009. Optical bandgap widening of A 3, 9285.
p-type Cu2O films by nitrogen doping. Appl. Phys. Lett. 94, 022111. Tyagi, M., Tomar, M., Gupta, V., 2012. Influence of hole mobility on the
Nguyen, W.H., Bailie, C.D., Unger, E.L., McGehee, M.D., 2014. response characteristics of p-type nickel oxide thin film based glucose
Enhancing the hole-conductivity of spiro-OMeTAD without oxygen. biosensor. Anal. Chim. Acta. http://dx.doi.org/10.1016/
J. Am. Chem. Soc. 136, 10996–11001. j.aca.2012.03.027.
Nishi, Y., Miyata, T., Nomoto, J., Minami, T., 2012. Influence of Cu2O Waechtler, T., Oswald, S., Roth, N., Jakob, A., Lang, H., Ecke, R.,
surface treatment on the photovoltaic properties of Al doped ZnO/ Schulz, S.E., Gessner, T., Moskvinova, A., Schulze, S., Hietschold, M.,
Cu2O solar cells. Thin Solid Films 520, 3819–3822. 2009. Copper oxide films grown by atomic layer deposition from Bis
Niu, G., Li, W., Meng, F., Wang, L., Dong, H., Qiu, Y., 2014. Study on (tri-n-butylphosphane) copper (I) acetylacetonate on Ta, TaN, Ru,
the stability of CH3NH3PbI3 films and the effect of post-modification and SiO2. J. Electrochem. Soc. 156, 453–459.
by aluminum oxide in all-solid-state hybrid solar cells. J. Mater. Chem. Wang, W.Z., Wang, G.H., Wang, X.S., Zhan, Y.J., Liu, Y.K., Zheng,
A 2, 705–710. C.L., 2002. Synthesis and characterization of Cu2O nanowires by a
Nogueira, A.F., Longo, C., Paoli, M.D., 2004. Polymers in dye sensitized novel reduction route. Adv. Mater. 14, 67–69.
solar cells: overview and perspectives. Coord. Chem. Rev. 248, 1455– Wehrenfennig, C., Eperon, G.E., Johnston, M.B., Snaith, H.J., Herz,
1468. L.M., 2014. High charge carrier mobilities and lifetimes in organolead
O’Brien, D.F., Burrows, P.E., Forrest, S.R., Koene, B.E., Loy, D.E., trihalide perovskites. Adv. Mater. 26, 1584–1589.
Thompson, M.E., 1998. Hole transporting materials with high glass. Wetzelaer, G.J., Scheepers, M., Sempere, A.M., Momblona, C., Ávila, J.,
Adv. Mater. 10, 1108–1112. Bolink, H.J., 2015. Trap-assisted non-radiative recombination in
Paul, G.K., Nawa, Y., Sato, H., Sakurai, T., Akimoto, K., 2006. Defects organic-inorganic perovskite solar cells. Adv. Mater. 18, 1837–1841.
in Cu2O studied by deep level transient spectroscopy. Appl. Phys. Lett. William, H.N., Colin, D.B., Eva, L.U., Michael, D.M., 2014. Enhancing
88, 141901. the hole-conductivity of spiro-OMeTAD without oxygen or lithium
Pudzich, R., Fuhrmann-Lieker, T., Salbeck, J., 2006. Spiro compounds salts by using spiro (TFSI)2 in perovskite and dye- sensitized solar
for organic electroluminescence and related applications. Adv. Polym. cells. J. Am. Chem. Soc. 136, 10996–11001.
Sci. 199, 83–142. Wu, H., Wang, L., 1997. A study of nickel monoxide (NiO), nickel dioxide
Qin, P., Tanaka, S., Ito, S., Tetreault, N., Manabe, K., Nishino, H., (ONiO), and Ni(O2) complex by anion photoelectron spectroscopy. J.
Nazeeruddin, M.K., Gratzel, M., 2014. Inorganic hole conductor- Chem. Phys. 107.
based lead halide perovskite solar cells with 12.4% conversion Xing, G., Mathews, N., Lim, S.S., Yantara, N., Liu, X., Sabba, D.,
efficiency. Nat. Commun. 5. Grätzel, M., Mhaisalkar, S., Sum, T.C., 2014. Low-temperature
Raebiger, H., LanyS, Zunger, A., 2007. Origins of the p-type nature and solution-processed wavelength-tunable perovskites for lasing. Nat.
cation deficiency in Cu2O and related materials. Phys. Rev. B 76, Mater. 13, 476–480.
045209. Xosrovashvilia, G., Gorjib, N.E., 2013. Numerical analysis of TiO2/
Scanlon, D.O., Morgan, B.J., Watson, G.W., 2009. Modelling the Cu2ZnSnS4 nanostructured PV using SCAPS-1D. J. Mod. Opt. 60,
polaronic nature of p-type defects in Cu2O: the failure of GGA and 936–940.
GGA+U. J. Chem. Phys. 131, 124703. Yin, W., Shi, T., Yan, Y., 2014. Unusual defect physics in CH3NH3PbI3
Schmidt-Mende, L., Gratzel, M., 2006. TiO2 pore-filling and its effect on perovskite solar cell absorber. Appl. Phys. Lett. 104.
the efficiency of solid-state dye-sensitized solar cells. Thin Solid Films Zhu, L., Shao, G., Luo, J.K., 2011. Numerical study of metal oxide
500, 296–301. heterojunction solar cells. Semicond. Sci. Technol. 26.
Seungchan, R., Jun, H.N., Nam, J.J., Young, C.K., Woon, S.Y.,
Jangwon, S., Sang, S., 2014. Voltage output of efficient perovskite