Stacking single-junction organic solar cells is effective in increasing the power conversion efficiency (PCE) by reducing the thermalization loss and increasing the open circuit voltage. Recent developments of non-fullerene acceptors (NFAs) offer a range of materials whose energy gaps are suited for absorbing relatively narrow slices of the solar spectrum, thus easing requirements for current balance between sub-elements in multijunction stacks. Here, we demonstrate a solution-processed tandem organic solar cell comprising a binary, visible-absorbing sub-cell and a ternary near-infrared (NIR) absorbing sub-cell. The ternary NIR sub-cell utilizes a narrow energy gap NFA that enables a broadened and increased absorption compared to a binary NIR sub-cell. An isopropanol surface treatment is developed to connect the hydrophilic–hydrophobic surfaces in the charge recombination zone (CRZ) located between the sub-cells. The nearly optically and electrically lossless CRZ combined with an anti-reflection coating results in tandem organic photovoltaics with PCE = 15.9% ± 0.2% under AM 1.5G simulated illumination.

Organic photovoltaics (OPVs) are considered a promising means for solar energy harvesting due to their potential for low cost, light weight, transparency, and flexibility.1–4 By stacking wide and narrow energy gap cells into multijunction devices, the power conversion efficiency (PCE) can exceed the thermodynamic limit of single-junction devices due to their broader absorption spectral range with less thermalization losses and increased open circuit voltage.5–10 Solution processed non-fullerene acceptors (NFAs) based on conjugated thiophene backbones have also provided a variety of molecules with spectral coverage useful for achieving current balance in multijunction solar cells while being compatible with a diversity of donor molecules.11–14 In this work, we demonstrate a high efficiency tandem OPV structure that comprises one binary and one ternary solution-processed sub-cell. The development of near-infrared (NIR) absorbing NFA molecules with long wavelength (>1000 nm) cutoffs is critical to achieving a high PCE.8,15–19 Here, we utilize a narrow energy gap NFA BEIT-4F that absorbs at wavelengths up to 1050 nm when combined with the donor, PCE-10, which enables the tandem solar cell to achieve broad spectral coverage from the blue to the NIR. We introduce a surface treatment with isopropanol to improve wetting without the need to add surfactant into poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) in the charge recombination zone (CRZ). The tandem cell device achieves PCE=15.2% ± 0.2% under AM 1.5G simulated illumination. An anti-reflection coating (ARC) deposited on the glass substrate increases the efficiency to 15.9% ± 0.2%.

For the visible-absorbing sub-cell, we employ a wide energy gap NFA SFT8-4F combined with the polymer donor, PM6 (see the supplementary material for detailed information of the molecules employed in the active regions of the sub-cells). Figure 1(a) presents the normalized thin film absorption spectra of these two molecules. As shown, the absorption of SFT8-4F peaks at 750 nm, after which it decreases rapidly.

FIG. 1.

(a) Thin film absorption of PM6 and SFT8-4F used in the visible-absorbing sub-cell. (b) Thin film absorption of PCE-10, BT-CIC, and BEIT-4F used in the ternary near-infrared absorbing sub-cell.

FIG. 1.

(a) Thin film absorption of PM6 and SFT8-4F used in the visible-absorbing sub-cell. (b) Thin film absorption of PCE-10, BT-CIC, and BEIT-4F used in the ternary near-infrared absorbing sub-cell.

Close modal

The ternary NIR absorbing sub-cell comprises a single donor PCE-10 and two acceptors BT-CIC and BEIT-4F. N-annulated perylene and dithienopicenocarbazole-based derivatives have been applied in efficient dye-sensitized and organic solar cells due to their planar frameworks and electron-donating capabilities.16,20–22 The strong electron-donating motif dithienopicenocarbazole was incorporated into the BEIT-4F molecule to shift its absorption into the NIR, which is complementary to the absorption of SFT8-4F. As shown in the thin film absorption spectra in Fig. 1(b), the narrow energy gap NFA BEIT-4F absorption spectrum is 50 nm redshifted compared to BT-CIC.

To systematically study the photogeneration mechanism in the ternary blends, we investigated their electronic states. Figure 2(a) presents the photoluminescence (PL) spectra of the ternary PCE-10:BT-CIC:BEIT-4F active region of the NIR sub-cell at different blend ratios. The PCE-10:BT-CIC binary exhibits the 0–0 transition peak of PCE-10 at 1.73 ± 0.01 eV, and for BT-CIC, at 1.39 ± 0.01 eV. The PCE-10:BEIT-4F binary and all ternary mixtures at different blend ratios have a BEIT-4F 0–0 transition at 1.32 ± 0.01 eV. In the ternary mixtures, excitons generated on BT-CIC rapidly transfer to BEIT-4F, which accounts for the absence of the BT-CIC exciton peak in the PL spectra. Thus, the excited states in the ternary blends are not influenced by the mixture composition. Figure 2(b) shows the electroluminescence (EL) spectrum of the PCE-10:BT-CIC binary blend. The spectrum comprises a BT-CIC exciton peak at 1.37 ± 0.02 eV and a PCE-10:BT-CIC charge transfer (CT) peak at 1.10 ± 0.02 eV. In Fig. 2(c), the EL spectrum of the PCE-10:BEIT-4F binary blend is fit with a BEIT-4F exciton peak at 1.30 ± 0.02 eV and a PCE-10:BEIT-4F CT peak at 1.15 ± 0.02 eV. The small energy offset between the highest occupied molecular orbitals (HOMOs) of PCE-10 and BEIT-4F gives rise to hole transfer from the PCE-10 to the BEIT-4F HOMO. Hence, there is a prominent BEIT-4F exciton peak in the EL spectrum. Figure 2(d) presents the EL spectrum of a PCE-10:BT-CIC:BEIT-4F (1:1.25:0.5, w/w/w) ternary blend, which comprises BEIT-4F, PCE-10:BT-CIC CT, and PCE-10:BEIT-4F CT exciton peaks, which are the same as those in the binaries. A similar fitting procedure is applied to this ternary system with different acceptor blend ratios. The spectral features fit well with the electronic states of the binaries. With these results, we infer that the ternary OPV operates as the combination of two, parallel-connected binary heterojunctions.23 

FIG. 2.

(a) Photoluminescence of PCE-10:BT-CIC:BEIT-4F blends at different blend ratios. (b) Electroluminescence (EL) spectrum of the 1:1.5 PCE-10:BT-CIC binary cell, with Gaussian peak fits (dashed lines) and their sum (blue line). (c) EL spectrum of 1:1.5 PCE-10:BEIT-4F with fits as in (b). (d) EL spectrum of 1:1.25:0.5 PCE-10:BT-CIC:BEIT-4F with fits as in (b).

FIG. 2.

(a) Photoluminescence of PCE-10:BT-CIC:BEIT-4F blends at different blend ratios. (b) Electroluminescence (EL) spectrum of the 1:1.5 PCE-10:BT-CIC binary cell, with Gaussian peak fits (dashed lines) and their sum (blue line). (c) EL spectrum of 1:1.5 PCE-10:BEIT-4F with fits as in (b). (d) EL spectrum of 1:1.25:0.5 PCE-10:BT-CIC:BEIT-4F with fits as in (b).

Close modal

The current density–voltage (JV) characteristics and external quantum efficiency (EQE) spectra of ternary devices with different blend ratios are shown in Figs. 3(a) and 3(b), with the detailed device performance data listed in Table I. Compared to PCE-10:BT-CIC, the PCE-10:BEIT-4F binary OPV shows a broader EQE spectrum that absorbs up to 1050 nm, a higher open-circuit voltage (VOC) of 0.754 V ± 0.005 V vs 0.695 V ± 0.004 V, and a lower fill factor (FF) of 0.56 ± 0.01 vs 0.71 ± 0.01. Consistent with the two binary junction model, adjusting the ratio of the two binary heterojunctions gives rise to EQE, VOC, and FF, which fall between those of the binaries and change monotonically with the blend ratio, as shown in Figs. 3(b) and 3(c).

FIG. 3.

(a) Current density–voltage (JV) characteristics under 1 sun, AM 1.5G simulated illumination and (b) external quantum efficiency (EQE) spectra of ternary cells with different PCE-10:BT-CIC:BEIT-4F blend ratios. (c) Open-circuit voltage (VOC) and fill factor (FF) with different BEIT-4F blend ratios.

FIG. 3.

(a) Current density–voltage (JV) characteristics under 1 sun, AM 1.5G simulated illumination and (b) external quantum efficiency (EQE) spectra of ternary cells with different PCE-10:BT-CIC:BEIT-4F blend ratios. (c) Open-circuit voltage (VOC) and fill factor (FF) with different BEIT-4F blend ratios.

Close modal
TABLE I.

Performance of PCE-10:BT-CIC:BEIT-4F single-junction OPVs at different blend ratios under simulated AM 1.5G illumination.

PCE-10:BT-CIC:BEIT-4FaJSCb (mA/cm2)VOC (V)FFPCE (%)
1:1.5:0 22.5 ± 0.2 0.695 ± 0.004 0.71 ± 0.01 11.0 ± 0.1 
1:1.25:0.25 22.2 ± 0.3 0.703 ± 0.003 0.67 ± 0.01 10.5 ± 0.2 
1:1.25:0.5 22.6 ± 0.2 0.715 ± 0.003 0.65 ± 0.01 10.9 ± 0.1 
1:1.25:0.75 22.5 ± 0.2 0.726 ± 0.004 0.62 ± 0.01 10.4 ± 0.2 
1:0:1.5 22.0 ± 0.3 0.754 ± 0.005 0.56 ± 0.01 9.3 ± 0.1 
PCE-10:BT-CIC:BEIT-4FaJSCb (mA/cm2)VOC (V)FFPCE (%)
1:1.5:0 22.5 ± 0.2 0.695 ± 0.004 0.71 ± 0.01 11.0 ± 0.1 
1:1.25:0.25 22.2 ± 0.3 0.703 ± 0.003 0.67 ± 0.01 10.5 ± 0.2 
1:1.25:0.5 22.6 ± 0.2 0.715 ± 0.003 0.65 ± 0.01 10.9 ± 0.1 
1:1.25:0.75 22.5 ± 0.2 0.726 ± 0.004 0.62 ± 0.01 10.4 ± 0.2 
1:0:1.5 22.0 ± 0.3 0.754 ± 0.005 0.56 ± 0.01 9.3 ± 0.1 
a

4 mm2 device area.

b

The JSC values are from the integrated EQE spectra.

To construct an optimized tandem OPV, the optical field was simulated to determine layer thicknesses and positions within the tandem cell required to achieve current balance. The lack of metal anode reflection of the PM6:SFT8-4F front sub-cell (i.e., the sub-cell adjacent to the transparent ITO cathode) requires a thicker active layer to compensate for the reduced absorption, with the optimized structure shown in Fig. 4(a). The front sub-cell absorbs primarily between wavelengths of λ = 350 nm and 750 nm, whereas the back sub-cell, which is adjacent to the reflective anode, absorbs primarily at λ > 750 nm. A nearly optically and electrically lossless CRZ was introduced between the sub-cells consisting of a polymer layer PEDOT:PSS sandwiched between a MoO3 layer and a ZnO nanoparticle layer. To improve wetting between hydrophobic and hydrophilic surfaces, a thin layer of MoO3 was thermally evaporated and then precoated with isopropanol to initiate PEDOT:PSS deposition. This treatment significantly reduces the contact angle of PEDOT:PSS and thus improves the fabrication yield (see supplementary material Fig. S3).

FIG. 4.

(a) Tandem device structure with optimized layer thicknesses and the corresponding simulated optical field intensity distribution. (b) JV characteristics of single-junction and tandem devices. (c) EQE of single-junction and tandem cells. The symbols are for discrete single-junction devices, whereas the dashed and solid lines are for sub-cells in the stack and their sum; colors represent the same sub-cells as in (b).

FIG. 4.

(a) Tandem device structure with optimized layer thicknesses and the corresponding simulated optical field intensity distribution. (b) JV characteristics of single-junction and tandem devices. (c) EQE of single-junction and tandem cells. The symbols are for discrete single-junction devices, whereas the dashed and solid lines are for sub-cells in the stack and their sum; colors represent the same sub-cells as in (b).

Close modal

Figures 4(b) and 4(c) present the JV characteristics and EQE spectra of the single-junction and tandem devices, with the details listed in Table II. The PM6:SFT8-4F (1:1.5, w/w, 80 nm) single junction device reaches PCE=12.1% ± 0.3% with JSC = 18.4 ± 0.3 mA/cm2, VOC = 0.98 ± 0.01 V, and FF=0.67 ± 0.01. The trade-off between VOC and FF in the PCE-10:BT-CIC:BEIT-4F ternary NIR cell results in the highest PCE at a blend ratio of 1:1.25:0.5 (w/w/w). Furthermore, the active layer was thermally annealed at 120 °C for 8 min to optimize the active layer morphology. The NIR absorbing device with a 75 nm active layer thickness has PCE=11.7% ± 0.3% with absorption up to 1050 nm. This is 7% higher than the PCE of the PCE-10:BT-CIC binary device.

TABLE II.

Discrete sub-cell and tandem device performances under simulated AM 1.5G illumination.

DeviceaJSCb (mA/cm2)VOC (V)FFPCE (%)
PM6:SFT8-4F = 1:1.5 18.4 ± 0.3 0.98 ± 0.01 0.67 ± 0.01 12.1 ± 0.3 
PCE-10:BT-CIC:BEIT-4F = 1:1.25:0.5, TAc 24.3 ± 0.4 0.70 ± 0.01 0.68 ± 0.01 11.7 ± 0.3 
Tandem w/o ARC 13.5 ± 0.2 1.66 ± 0.01 0.68 ± 0.01 15.2 ± 0.2 
Tandem w/o ARC 14.1 ± 0.2 1.66 ± 0.01 0.68 ± 0.01 15.9 ± 0.2 
DeviceaJSCb (mA/cm2)VOC (V)FFPCE (%)
PM6:SFT8-4F = 1:1.5 18.4 ± 0.3 0.98 ± 0.01 0.67 ± 0.01 12.1 ± 0.3 
PCE-10:BT-CIC:BEIT-4F = 1:1.25:0.5, TAc 24.3 ± 0.4 0.70 ± 0.01 0.68 ± 0.01 11.7 ± 0.3 
Tandem w/o ARC 13.5 ± 0.2 1.66 ± 0.01 0.68 ± 0.01 15.2 ± 0.2 
Tandem w/o ARC 14.1 ± 0.2 1.66 ± 0.01 0.68 ± 0.01 15.9 ± 0.2 
a

4 mm2 device area.

b

The JSC values are from the integrated EQE spectra.

c

Thermal annealing at 120 °C.

The optimized tandem OPV with a 120 nm PM6:SFT8-4F front active layer and a 90 nm PCE-10:BT-CIC:BEIT-4F back active layer exhibits JSC = 13.5 ± 0.2 mA/cm2, VOC = 1.66 ± 0.01 V, FF=0.68 ± 0.01, and PCE=15.2% ± 0.2% under AM 1.5G illumination. To reduce the optical loss, an ARC consisting of a 120 nm MgF2 layer (index of refraction nMgF2=1.38±0.01) and a 130 nm low-refractive-index SiO2 layer24 (nSiO2=1.12±0.03) was deposited on the glass substrate, which reduced the reflection of the glass substrate between 400 and 1000 nm for approximately 4%.25 The tandem device with the ARC has an increased JSC and a correspondingly increased PCE. As shown in Fig. 4(c), the tandem solar cell absorbs from 350 nm to 1050 nm. The reduced absorption of the front sub-cell is due to the lack of electrode reflection and back sub-cell absorption within the same wavelength region. This provides for current balance between sub-cells, along with a concomitant reduction in JSC compared to that of the single junction PM6:SFT8-4F device. The integration of the EQE spectra gives a balanced front sub-cell JSC = 14.2 mA/cm2 and a back sub-cell current of JSC = 14.1 mA/cm2. The tandem device with an ARC reaches PCE=15.9% ± 0.2%. The detailed performance parameters are listed in Table II.

With rapid development in material synthesis, single-junction OPV has reached over 16% efficiency.13,14 Multijunction solar cells can exceed the single-junction thermodynamic limit, which suggests that there is still room to improve the tandem OPV efficiency. Previously, a double ternary‐junction tandem cell with a PCE of 17.3% was demonstrated15 although the higher voltage of the tandem cell compared to the sum of the individual sub-cells was left unexplained. Nevertheless, the results on high efficiency tandems point to their benefits of lower currents (and hence reduced series resistance losses) and their potential for higher PCE compared to single junction cells.

To conclude, we demonstrate a high efficiency tandem OPV structure with a NIR-absorbing PCE-10:BT-CIC:BEIT-4F ternary back sub-cell combined with a visible-absorbing binary PM6:SFT8-4F front sub-cell. The NIR-absorbing NFA BEIT-4F extends the absorption of the back sub-cell to 1050 nm and results in a current balance with the shorter wavelength absorbing binary front sub-cell. A nearly optically and electrically lossless CRZ comprising a hydrophilic PEDOT:PSS layer sandwiched between MoO3 and ZnO nanoparticle layers is constructed using an isopropanol surface treatment. The tandem device on an ARC-coated glass substrate reaches a maximum PCE of 15.9% ± 0.2% under AM 1.5G simulated illumination.

See the supplementary material for the experimental details and contact angle measurement.

X.H. and B.S. contributed equally to this work.

This work was supported in part by the U.S. Department of Energy's Office of Energy Efficiency and Renewable Energy (EERE) under Solar Energy Technologies Office (SETO) Agreement No. DE-EE0008561 (X.H. and Y.L.) and by the Department of Navy, Office of Naval Research (ONR) under Award No. N00014-17-1-2211 (S.R.F. and D.F.). This work was also supported by the National Natural Science Foundation of China (No. 21871199), the Collaborative Innovation Centre of Suzhou Nano Science and Technology (Nano-CIC), the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), and the “111” Project of The State Administration of Foreign Experts Affairs of China (B.S., C.J., and J.F.).

The data that support the findings of this study are available within this article and its supplementary material.

1.
Y.
Li
,
C.
Ji
,
Y.
Qu
,
X.
Huang
,
S.
Hou
,
C.
Li
,
L.
Liao
,
L. J.
Guo
, and
S. R.
Forrest
,
Adv. Mater.
31
,
1903173
(
2019
).
2.
Y.
Li
,
G.
Xu
,
C.
Cui
, and
Y.
Li
,
Adv. Energy Mater.
8
,
1800716
(
2018
).
3.
Y.
Li
,
J. D.
Lin
,
X.
Che
,
Y.
Qu
,
F.
Liu
,
L. S.
Liao
, and
S. R.
Forrest
,
J. Am. Chem. Soc.
139
,
17114
(
2017
).
4.
B.
Qu
and
S. R.
Forrest
,
Appl. Phys. Lett.
113
,
053302
(
2018
).
5.
X.
Che
,
Y.
Li
,
Y.
Qu
, and
S. R.
Forrest
,
Nat. Energy
3
,
422
(
2018
).
6.
X.
Che
,
X.
Xiao
,
J. D.
Zimmerman
,
D.
Fan
, and
S. R.
Forrest
,
Adv. Energy Mater.
4
,
1400568
(
2014
).
7.
J.
Gilot
,
M. M.
Wienk
, and
R. A. J.
Janssen
,
Appl. Phys. Lett.
90
,
143512
(
2007
).
8.
Y.
Li
,
J.-D.
Lin
,
X.
Liu
,
Y.
Qu
,
F.-P.
Wu
,
F.
Liu
,
Z.-Q.
Jiang
, and
S. R.
Forrest
,
Adv. Mater.
30
,
1804416
(
2018
).
9.
J.
Gilot
,
M. M.
Wienk
, and
R. A. J.
Janssen
,
Adv. Mater.
22
,
E67
(
2010
).
10.
D.
Di Carlo Rasi
and
R. A. J.
Janssen
,
Adv. Mater.
31
,
1806499
(
2019
).
11.
12.
Q.
Fan
,
T.
Liu
,
W.
Gao
,
Y.
Xiao
,
J.
Wu
,
W.
Su
,
X.
Guo
,
X.
Lu
,
C.
Yang
,
H.
Yan
,
M.
Zhang
, and
Y.
Li
,
J. Mater. Chem. A
7
,
15404
(
2019
).
13.
J.
Yuan
,
Y.
Zhang
,
L.
Zhou
,
G.
Zhang
,
H. L.
Yip
,
T. K.
Lau
,
X.
Lu
,
C.
Zhu
,
H.
Peng
,
P. A.
Johnson
,
M.
Leclerc
,
Y.
Cao
,
J.
Ulanski
,
Y.
Li
, and
Y.
Zou
,
Joule
3
,
1140
(
2019
).
14.
M. A.
Pan
,
T. K.
Lau
,
Y.
Tang
,
Y. C.
Wu
,
T.
Liu
,
K.
Li
,
M. C.
Chen
,
X.
Lu
,
W.
Ma
, and
C.
Zhan
,
J. Mater. Chem. A
7
,
20713
(
2019
).
15.
L.
Meng
,
Y.
Zhang
,
X.
Wan
,
C.
Li
,
X.
Zhang
,
Y.
Wang
,
X.
Ke
,
Z.
Xiao
,
L.
Ding
,
R.
Xia
,
H.-L.
Yip
,
Y.
Cao
, and
Y.
Chen
,
Science
361
,
1094
(
2018
).
16.
Z.
Yao
,
H.
Wu
,
Y.
Li
,
J.
Wang
,
J.
Zhang
,
M.
Zhang
,
Y.
Guo
, and
P.
Wang
,
Energy Environ. Sci.
8
,
3192
(
2015
).
17.
Z.
Xiao
,
X.
Jia
,
D.
Li
,
S.
Wang
,
X.
Geng
,
F.
Liu
,
J.
Chen
,
S.
Yang
,
T. P.
Russell
, and
L.
Ding
,
Sci. Bull.
62
,
1494
(
2017
).
18.
Z.
Xiao
,
X.
Jia
, and
L.
Ding
,
Sci. Bull.
62
,
1562
(
2017
).
19.
F.
Liu
,
Z.
Zhou
,
C.
Zhang
,
J.
Zhang
,
Q.
Hu
,
T.
Vergote
,
F.
Liu
,
T. P.
Russell
, and
X.
Zhu
,
Adv. Mater.
29
,
1606574
(
2017
).
20.
X.
Li
,
F.
Yu
,
S.
Stappert
,
C.
Li
,
Y.
Zhou
,
Y.
Yu
,
X.
Li
,
H.
Ågren
,
J.
Hua
, and
H.
Tian
,
ACS Appl. Mater. Interfaces
8
,
19393
(
2016
).
21.
W.
Jiang
,
Y.
Li
, and
Z.
Wang
,
Chem. Soc. Rev.
42
,
6113
(
2013
).
22.
Z.
Yao
,
X.
Liao
,
K.
Gao
,
F.
Lin
,
X.
Xu
,
X.
Shi
,
L.
Zuo
,
F.
Liu
,
Y.
Chen
, and
A. K. Y.
Jen
,
J. Am. Chem. Soc.
140
,
2054
(
2018
).
23.
X.
Huang
,
X.
Liu
,
K.
Ding
, and
S. R.
Forrest
,
Mater. Horiz.
7
,
244
(
2020
).
24.
J.-Q.
Xi
,
J. K.
Kim
,
E. F.
Schubert
,
D.
Ye
,
T.-M.
Lu
,
S.-Y.
Lin
, and
J. S.
Juneja
,
Opt. Lett.
31
,
601
(
2006
).
25.
M.
Slootsky
and
S. R.
Forrest
,
Opt. Lett.
35
,
1052
(
2010
).

Supplementary Material