Siemens Research Report

A high-performance tandem organic solar cell with novel active layers providing visible
and near-IR spectrum absorption
Introduction
Faced with the combined threat of depleting non-renewable energy sources and their
disastrous impact on the planet’s ecosystem, the international community has realized the
urgency of developing clean, affordable, and convenient renewable energy sources.1
The sun
offers much more energy than humanity could ever use, giving solar power the potential to
outgrow competing clean energy sources. A harbinger of the viability of solar technology for
energy generation was the creation of efficient crystalline silicon solar cells at Bell Labs in the
1950s, a discovery which sparked the modern field of photovoltaic research. 2
Traditional silicon-based solar cells harvest energy through semiconductor physics,
where photonic energy drives charge separation at the junction of positively and negatively
doped silicon to create harvestable electric current. Solar cells made with elements in groups III-
V of the periodic table demonstrate the best efficiencies in crystalline photovoltaics, achieving
high power conversion efficiencies (PCE 25%).3
However, they remain expensive due to limited
sources for materials and complex fabrication processes.1
There has been remarkable
improvement in inorganic solar technology since the 1950s;4
the majority of recent
improvements in efficiency have come from devices employing multi-layer (tandem) device
architecture. Tandem device structure has shown success in inorganic photovoltaic research, and
has produced astronomical efficiencies in triple-junction GaInP/GaInAs/Ge solar cells (PCE ≈
41%).3
With a tentative promise to solve the problems of traditional solar cells based on
inorganic materials, organic photovoltaics (OPVs) have emerged as a viable technology for solar
1

energy production thanks to developments in conducting and semiconducting polymers.5
They
have recently shown promising results due to their relatively easy production and low application
costs, prompting intensive research to optimize OPVs for stability and efficiency. OPVs have
several advantages over conventional p/n junction crystalline silicon solar cells; they can be
made on flexible substrates, employ high-throughput printing techniques, provide wide chemical
engineering possibilities, and are environmentally sound due to carbon-based semiconducting
materials (conjugated organic polymers). Driven by the promise of low cost, large area solar cell
production using roll-to-roll techniques, solution processability has become a goal for OPV
research alongside the improvement of efficiency and durability.8
In this way, organic solar
systems would employ low installation costs and multi-surface compatibility to offset their
relative inefficiency. Thus OPVs reduce the initial investment necessary to implement solar
technology, and make solar power a more universally accessible renewable energy source.9-11
Traditional OPV devices are fabricated on top of a transparent conductive electrode,
usually indium tin oxide (ITO). Active layers consist of electron-rich organic semiconductors
that absorb light and generate electron-hole pairs, (excitons), paired with electron-deficient
organic materials that separate and transport the generated charges. Electron rich polymers (e.g.
PTB7) are known as “donor” polymers, and electron-deficient compounds (fullerene derivatives,
e.g. PC70BM) are referred to as acceptors.
Connected by a specially designed interlayer, tandem device structure utilizes a
combination of two active layers in series within one device, where each layer is responsible for
absorbing certain wavelengths of light — from visible to near infrared wavelengths for carefully
designed devices.17
— and providing additive voltage (Voc) without compromising current (Jsc)
loss compared to single cells.13-16
Since Hadipour et al. 18
first demonstrated a tandem OPV with
2

two different bandgap polymers, achieving an additive voltage of 1.4 V and a PCE of 0.7%,
tandem devices have attracted attention in the past years. In fact, tandem OPVs have recently
pushed the organic device record to PCE>10%, creating optimism in this emerging field. Despite
presenting an intriguing strategy for efficiency enhancement, only few successful tandem works
have been published due to several reasons, namely: 1. lack of efficient polymers for single cells;
2. significant overlap of absorption spectra between donor polymers; 3. lack of efficient
interlayer between the two sub-cells; 4. fabrication procedures compatible with the multilayer
structure. Thus research in the design of materials and device fabrication must be done for
further advancement of tandem OPV technology.
After gaining an operational understanding of OPV device theory and fabrication
technique through the optimization and testing of the novel polymer PBTI3T, which
demonstrates a record high efficiency of 8.7% for single-layer OPV (manuscript currently under
review), this project decided to utilize this material within a tandem structure in combination
with other high performance materials to further enhance OPV PCEs.
In this work, after achieving a record high efficiency of 8.7% for single-layered OPV
(manuscript currently under review), this project decided to implement the record efficiency
material in a tandem structure combined with other high performance active layer materials to
further enhance OPV PCEs. A few published low bandgap polymers are successfully
implemented in tandem device architecture. All of the polymers later introduced in this work are
implemented together for the first time as part of a tandem OPV structure. Engineering tandem
OPVs to harness their full potential introduces a new set of challenges to the designing of
organic devices. As the complexity of solution processed devices increases, controlling and
optimizing layer morphology, texture, exclusivity, charge transfer, and thickness becomes
3

difficult. Thus materials selection for this research was non-trivial and focused on bandgap
characteristics of the p-type donor polymers, macro and microscopic polymer characteristics, as
well as single-layer cell performance.21
Within its scope, this project achieved one of the highest Voc (>1.4V) in OPV literature
and an overall PCE of >6% in this work. These results exemplify the potential of tandem
organic devices employing next generation high efficiency (single junction device PCE >7%)
polymers. Additionally, this work leads to future paths of work for tandem device optimization
to achieve efficiencies near those of commercially available silicon-based solar devices, while
retaining the hallmark characteristics of OPVs.
Materials & Methods
Figure 1: (a) donor and acceptor compounds used in this study (b) donor/acceptor bandgap diagrams.
Materials choice for the construction of tandem devices in this research focused on an
unpublished novel high efficiency semiconducting polymer poly(bithiopheneimide-trithiophene)
(PBTI3T), used as a donor polymer in one active layer for all of the tandem cells produced. In
4

the first section of this tandem device study the polymer Poly[[4,8-bis [ (2- ethylhexyl) oxy]
benzo [1,2-b:4,5-b'] dithiophene-2,6- diyl] [3- fluoro- 2-[(2-ethylhexyl) carbonyl] thieno [3,4-b]
thiophenediyl]] (PTB7) 22
was chosen for the active layer complementary to PBTI3T. For the
second part of this study, poly(4,4-dioctyldithieno(3,2-b:2',3'-d)silole)-2,6-diyl-alt- (2,1,3-
benzothiadiazole)-4,7-diyl) (PSBTBT-Si)23
was chosen as the complementary active layer. Each
polymer was blended with the molecule [6,6]-phenyl C71-butyric acid methyl ester (PC70BM) to
form a photoactive layer. PC70BM is a derivative of the carbon buckminster fullerene family,
and is a well-established, widely used electron acceptor in most OPVs.5
The chemical structures
and energy alignment of these materials are shown in Figure 1.
Figure 2: Device fabrication procedure for tandem OPVs.
Incorporating these materials into the production of a tandem BHJ solar cell is a multi-
step process with layers deposited upon an indium tin oxide (ITO) coated glass substrate as
follows (Figure 2): a) spin coated 20 nm zinc oxide (ZnO); b) spin coated first polymer active
5

layer (AL1); c) 5 Å vapor-deposited layer of silver (Ag); d) spin coated ~40 nm layer of
PEDOT:PSS; e) second 5 Å vapor-deposited layer of silver (Ag); f) second spin coated layer of
ZnO; g) spin coated second polymer active layer (AL2); h) 7.0 nm vapor deposited molybdenum
oxide (MoO3);120 nm vapor deposited Ag electrodes.
For the fabrication of OPV devices, a 10 Ω/☐ pre-patterned ITO is used as substrate. It is
cleaned by sequential sonication in hexane, DI water, methanol, isopropanol and acetone, and
finally UV/ozone treated (Jelight Co.). The ZnO interfacial layer is fabricated by spin-coating at
5000 rpm for 30 seconds onto ITO substrate from a precursor solution prepared with 220 mg
zinc acetate, 62 mg ethanolamine in 2 ml 2-methoxy-ethanol after sufficient stirring. Electrical
contact areas on substrates are cleaned with isopropyl alcohol (IPA) for anode or cathode
deposition. Deposited ZnO is annealed on a hot plate set to 170o
C. Devices are then transferred
to glovebox under nitrogen vacuum. Active layers are prepared under N2 condition in solutions
of 8 mg/mL PBTI3T in chloroform, 10 mg/mL PTB7 in chlorobenzene, and 10 mg/mL
PSBTBT-Si in chloroform. PC70BM is subsequently added with a polymer:fullerene weight ratio
of 1:2, 1:1.5, and 1:1 for optimal performance, respectively. 2%, 3% and 3% 1,8-diiodooctane
(DIO) are added to PBTI3T, PTB7, and PSBTBT-Si solutions, respectively, as processing
additives to improve the resulting active layer film morphology. All solutions are stirred at 70o
C
for >2 hrs. In the fabrication of PBTI3T+PTB7 or PBTI3T+PSBTBT-Si tandem devices, first
layer of PBTI3T is spun cast at 4000-6000 rpm for 30 seconds and second layer of PTB7 or
PSBTBT-Si solution is spun at 1500 rpm for 30 seconds.
The interlayer between the two separate active layers consists of a thin layer o f Ag (5 Å),
m-PEDOT:PSS (~40 nm), a second layer of Ag (5 Å), and ZnO (20 nm). The m-PEDOT:PSS is
fabricated according to the procedure developed by Yang Yang et al.14
PSS (Mw=40,000) is
6

added into 2g of PEDOT:PSS solution to increase the interlayer conductivity. 0.5% by weight
TritonTM
is subsequently added as a surfactant to enhance the surface wettability of the solution
on top of the first active layer. The solution is spun cast at 5000 rpm for 30 seconds and annealed
at 150O
C for 5 min. A thin layer of 5 Å Ag is thermally deposited under ultrahigh vacuum prior
to the deposition of ZnO which is prepared under the same condition as the first interfacial layer.
After the deposition of the second layer, thin layers of 7.0 nm MoO3 and 140 nm of Ag
are then thermally evaporated through a shadow mask at ~10-6
Torr. For device characterization,
J-V characteristics are measured under AM1.5G light (100 mW/cm2) using the Xe arc lamp of a
Spectra-Nova Class A solar simulator. The light intensity is calibrated using an NREL-certified
monocrystalline Si diode coupled to a KG3 filter to bring spectral mismatch to unity. A Keithley
2400 source meter is used for electrical characterization. The area of all devices is 6 mm2, and
an aperture with size of 6 mm2 is used on top of cells during all measurements. EQEs are
characterized using an Oriel model QE-PV-SI instrument equipped with a NIST-certified Si
diode. Monochromatic light is generated from an Oriel 300W lamp source.
7

Figure 3: The structure of a finished tandem device and an actual finished tandem device.
Discussion, Results, and Illustration
Figure 4: (a) band alignment diagram (b) a representative current-voltage (J-V) curve.
8

The discussion of OPV devices begins with four basic processes within each cell: 1) light
absorption and exciton formation; 2) exciton migration to a donor/acceptor interface; 3) exciton
separation due to sufficient interfacial potential energy drop; 4) charge carrier transport to device
electrodes (Figure 4a). These essential processes drive materials and structural optimization for
OPVs, and the refinement of each step of solar energy generation is crucial to the development
of efficient single layer and tandem OPVs. Bandgap engineering is essential to the beginning of
tandem OPV discussion because it correlates directly to the defining characteristic of additive
voltages in multi-junction device structure. Polymer bandgap is defined as the energy difference
between a semiconducting polymer’s highest occupied molecular orbital (HOMO) and lowest
unoccupied molecular orbital (LUMO) (Figure 1b). In donor polymers, photons are absorbed and
provide the energy to promote electrons from the HOMO level to the LUMO level, forming a
bound electron-hole pair (an exciton). The exciton then diffuses to a donor/acceptor interface,
and separates due to the potential energy difference between the LUMO of the n-type acceptor
material and the LUMO of the donor polymer. The resulting free-charge carriers, electrons and
holes, can travel to their respective electrodes (Figure 4a). The difference in polymer HOMO and
PC70BM LUMO levels corresponds to the theoretical Voc of the device. Therefore, the lower the
HOMO levels of polymers, the higher the Voc that can be expected.
The efficiency of tandem PSCs has been hindered by the availability of suitable low
bandgap materials and spectrally-mismatched high bandgap materials. In an effort to optimize
the absorption and exciton formation characteristics of tandem OPVs, this project chose to
explore the compatibility of a previously published and well-tested low bandgap polymer,
PSBTBT-Si, and to continue the study of PTB7, along with this project’s novel high-bandgap
polymer, PBTI3T.
9

Figure 5: Images of polymer solutions without and with PC70BM, and their respective absorption spectra in
solution.
As a primary evaluation of the selected polymers’ compatibility, the materials were
prepared in solution with and without the addition of PC70BM, and tested for absorption.
Spectrum absorption mismatch between donor polymers in tandem OPVs is crucial to ensure the
additive voltage of active layers. This relationship is demonstrated by the measured absorption
curves of the polymers chosen for this study, which portray offset absorption spectra of ~500-
675 nm, ~550-750 nm, and ~575-825 nm for polymers PBTI3T, PTB7, and PSBTBT-Si,
respectively (Figure 5). In preliminary studies of this research, PBTI3T has achieved 8.7%
efficiency in a single-layer cell, and its demonstrated shorter-wavelength absorbance spectrum is
ideal for working in tandem with both PTB7 and PSBTBT-Si.
Spectral exclusivity is important within the scope of tandem cells because it allows
stacked active layer placement with minimal impact on individual layer performance. Although
10

the intensity of light attaining the second active layer may decrease due to top layer interference,
it still retains photons energetic enough to efficiently activate exciton formation in the donor
polymer. However, increased spectral overlap, such as that of PBTI3T and PTB7 in comparison
to PBTI3T and PSBTBT-Si (Figure 5), causes the first active layer to absorb a portion of the
wavelengths necessary for efficient performance of the second active layer. The addition of
PC70BM to the donor polymers in solution simulates the absorption of an active layer comprised
of both materials. It is important to note that in addition to the polymer peaks, PC70BM adds to
the absorption graphs of PBTI3T, PTB7, and PSBTBT-Si in the ~350-550 nm wavelength range,
emphasizing the difference of each donor polymer’s light absorption at higher wavelengths. In
this view, PSBTBT-Si:PC70BM in combination with PBTI-3T:PC70BM are logical choices for
tandem device active layers because of their large ~150 nm spectral difference. Moreover, this
material choice encompasses visible wavelengths and extends into near-IR. Although the
difference in absorption between PTB7:PC70BM and PBTI3T:PC70BM is much smaller (~50
nm), there is enough difference in the absorption of these active layers for PTB7: PC70BM to
provide complementary voltage gains in combination with the first active layer, especially given
each polymer’s strong single layer Voc and FF. In this scenario, however, the first active layer
dominates the second and skews results towards the performance of the dominating layer
(PBTI3T) supplemented by additional active layer (PTB7) characteristics.
11

Figure 6: Single layer cell performance of PBTI3T, PTB7, and PSBTBT-Si.
Polymer characteristics provide insight into device performance thanks to the Planck
relation E = hc/λ, which relates the wavelength of the light absorbed by active layer polymers to
the energy of their bandgap. As previously mentioned, donor polymer and acceptor polymer
bandgaps work within the device active layer to create voltage and, to some extent, current.
Performance of a solar cell is characterized by its J-V curve, a graphical representation of the
device’s current and voltage output in response to a regulated light source (Figure 4b). The
graph’s y-axis translates to current output (J), and the x-axis translates to voltage output (V).
Notable points on the J-V curve include the device’s short-circuit current (Jsc), which is the y-
intercept of the J-V graph, and is a representation of the cell’s maximum current under no
reverse voltage bias. Similarly, the device’s open-circuit voltage (Voc) is the x-intercept of the J-
V curve, and represents the cell’s maximum voltage output under full reverse bias. A final
notable characteristic of the J-V curve is the fill factor (FF), which is a measure of the curve’s
“squareness”, a representation of the power that the cell could produce under optimal load. Using
these three parameters, it is possible to determine the efficiency of a given solar cell by the
equation described in Figure 4b.
12

Single cell characterization was performed for each of the polymer active layers chosen
in this study; the results are shown above as J-V curves of devices representing the average
tested OPV employing each different donor polymer (Figure 6). The best efficiencies of
PSBTBT-Si, PTB7, and PBTI3T were 8.66%, 8.12%, and 4.33%, respectively (Table 2). This
project was unable to reproduce PSBTBT-Si’s best published single-cell PCE of 5.1% with
polymers sourced from 1-Materials Co.8,9
The decision to use PSBTBT-Si was based on its
optimal absorption spectrum rather than on its PCE. Although the Jsc of PSBTBT-Si is ~14.7
mA/cm2
, its inadequate fill factor (~48.8 %) and lower Voc (.604 V) bring down performance.
Ideally, its implementation in a tandem device structure would allow it to contribute a Voc close
to that of its single cell performance due to minimal spectral overlap with other materials.
This project’s medium-bandgap material, PTB7, has a published PCE of 7.4%.20
This
single-cell efficiency is the main reason for choosing PTB7 as a second active layer option,
especially given that this project’s optimization of PTB7 single cells pushed the polymer’s
efficiency to PCE = 8.12% (Table 2). Its lower HOMO energy level (Figure 1b) also leads to a
higher Voc, which helps to move its single-layer J-V curve closer to the optimal fill factor
(figure 6). PTB7’s high single-cell efficiency can be attributed to its especially high Jsc (16
mA/cm2
), good Voc (0.73 V) and decent FF of 69.4%. Previous tandem cells have made use of
PTB7 because of its high single layer performance, but, for lack of a better higher-bandgap
polymer choice, have paired it with a standard high-bandgap P3HT active layer.10
This project
sought to pair PTB7 with the novel high bandgap polymer PBTI3T to increase spectral mismatch
compared to previous tandem efforts with this medium-bandgap polymer, and thus provide more
distinctly additive characteristics. The performance of PBTI3T can be attributed to its
combination of high Voc (0.862 V), high FF (77.7%) and good Jsc (12.9%) (Table 2).
13

Figure 7: Optimization of PBTI3T polymer. Atomic force microscope study of blend films prepared under
differing conditions
(varying amounts of 1,8-diiodooctane). Table 1. Summary of polymer: PC70BM ratio and
DIO concentration.
The studied single layer devices employed bulk-heterojunction (BHJ) active layers,
which are a heterogeneous mixture of p-type polymer semiconductors as electron donors and n-
type semiconductors, mostly fullerene derivatives, as electron acceptors. The randomized
structure of the p/n junctions in BHJ OPVs provides interesting opportunities for device
morphology optimization. An example of morphology optimization on PBTI3T polymer is
illustrated in Figure 7. The resulting device performance is summarized in Table 1.
A challenge posed by the solution processing of multiple organic layers in succession is
the maintenance of layer uniformity and exclusivity throughout the tandem device. The use of
additives such as DIO provides control over the distribution of linear p-type polymers and
spherical n-type buckyball polymers within the active layer. The optimization of DIO
concentration to control domain size homogeneity to within the active layer solution is depicted
14
Polymer:
PC71BM
(w/w)
Solvent (v/v) Voc
(V)
Jsc
(mA/cm2
)
FF
(%)
PCE
(%)
1:1 CF 0.901 9.18 58.7 4.89
1:1.5 CF 0.877 8.14 60.3 4.30
1:2 CF 0.875 7.51 61.4 4.03
1:1 97%CF+3%DIO 0.869 12.4 70.0 7.54
1:1.5 97%CF+3%DIO 0.866 11.3 73.8 7.22
1:2 97%CF+3%DIO 0.863 9.59 75.4 6.24
1:1 95%CB+5%DIO 0.857 10.7 71.2 6.54
1:1.5 95%CB+5%DIO 0.861 10.2 76.5 6.74
1:2 95%CB+5%DIO 0.847 10.7 76.3 6.90

by AFM imaging (Figure 7). DIO provides selective solubility of PC70BM, allowing the
progressive increase in distribution of these molecules evenly amongst the film, thus reducing
the domain size of PC70BM and distance necessary for charges to travel before reaching
interface. Due to the exciton recombination length of ~10 nm, it is advantageous to distribute
both n and p-type polymers evenly for efficient charge separation along well-spaced p/n
interfaces. This corresponds to a generally enhanced device FF. Moreover, the highest current
(lowest charge recombination) in the DIO optimized PBTI3T occurs with an even (1:1)
PBTI3T:PC70BM ratio and 3% DIO concentration (Table 1). 6
To some extent, it is possible to control layer morphology by employing thin (~5 Ǻ)
layers of vacuum deposited silver. This method provides a physical barrier to reduce film surface
roughness, and especially reduce miscibility between successive solution deposited layers. This
project’s tandem devices employed these silver interlayers between both active layers, reflected
by data as a small but statistically insignificant improvement in Jsc and PCE. To determine
whether this method is valuable to OPV structure theory, more testing must be done. This type of
interlayer aims to improve charge carrier transport between active layers.
Ideally, a device interfacial layer (IFL) should have low electrical resistance, high optical
transparency in the visible and infrared range, low potential energy barriers for both electron and
hole extractions, easy-fabrication process, and protection for the prior-deposited active layer in
solution-processed tandem OPVs.13,14
Zinc oxide, an established n-type transparent conducting oxide (TCO) in OPV literature,
was used multiple times within these tandem devices (Figure 2) for selective charge transport at
device cathodes.13
Once charges are separated and have traveled to the device’s cathode, it is
15

possible to analyze cell performance. ZnO exhibits many of the optimal interlayer characteristics
mentioned previously, notably forming a tough near-crystalline film once annealed.
Additionally, poly(ethylene-dioxythiophene) doped with polystyrene-sulfonic acid
(PEDOT:PSS) is a p-type material that, and facilitates hole (h+
) transport between active layers.7
PEDOT:PSS is a non-ideal IFL in the sense that it is not highly transparent, and provides
minimal protection of lower layers as it remains amorphous once annealed.
However, two different types of PEDOT:PSS were used in the device fabrication. One
was only the PEDOT:PSS solution purchased from an outside source with TritonTM
X-100
surfactant added for wettability and discrete layer deposition. The other was the same
PEDOT:PSS/Triton solution, but modified according to Yang et al.14
to improve conductivity.
The production process itself is an important component of OPV production, as it
determines (to some extent) how well the devices will conduct current. There are fewer layers to
process in single cells, avoiding damage to other layers. In tandems, the production process
involves at least twice as many steps, all of which affect the preceding layers. This causes the
interlayers to be less uniform than is preferable, allowing charges to bleed through and diffuse to
unwanted layers or recombine, decreasing the Jsc and PCE of the cell. If the solution processing
and spin coating is done imperfectly, the surface of the cell becomes even less uniform and more
rough, with each layer added aggravating the problem. An additional issue with solution
deposited active layers is the possibility of sub-layer dilution by the solvent of the second active
layer, which would have generally detrimental effects to cell performance. Finally, roughness
decreases the ohmic contact and efficiency of electrodes because neither silver nor MoO3 can
cover edges of surface imperfections that have angles >90 degrees. -
Results
16

Table 2. Summary of single-layered and tandem device performance.
Device Solvent Thickness (nm) Voc (V) Jsc (mA/cm2
) FF (%) PCE (%)
PBTI3T CF 140 0.862 12.9 77.7 8.66
PTB7 CB 90 0.730 16.0 69.4 8.12
PSBTBT-Si CB 100 0.604 14.7 48.8 4.33
PBTI3T CF 60 0.880 7.89 74.8 5.20
PTB7 CB 60 0.739 12.9 70.3 6.71
PSBTBT-Si CB 70 0.600 8.38 50.3 2.53
PTB7+PBTI3T CB+CF 60 + 140 1.20 8.21 65.9 6.49
PBTI3T+PSBTBT-Si CB+CF 60 + 70 1.41 4.85 48.0 3.28
In order to produce the best possible tandem cells, this research investigated multiple
pathways (polymer bandgap and spectral engineering, single layer device optimization,
morphology control, interfacial layer design) for the optimization of Voc, Jsc, and FF in search of
maximum tandem PCE. The best tandem cells recorded had a PCE of 6.49%, a VOC of 1.2 V, and
a JSC of 8.21 mA/cm2 (Table 2, Figure 8). The two materials used in tandem to produce these
results, PBTI3T and PTB7, demonstrated high efficiency and JSC in single cells, and are shown
here to perform very well in tandem (Table 2, Figure 7). These findings show the successful
17
Figure 8. J-V curve of single-layer performances of
PBTI3T and PTB7, with tandem performance of
the combined polymers.
Figure 9. J-V curve of single-layered cell
performances of PBTI3T and PSBTBT-Si,
with tandem performance of the combined.

partial Voc addition of the respective polymers’ single layer performance, which combine with
adequate FF retention to balance the unfortunate decrease in Jsc, perhaps due to inadequate IFL
treatment or poor device structure.
The major achievement of such a tandem pairing is the resulting voltage of the device,
which is the additive combination of each respective active layer’s Voc. Although the
PBTI3T/PSBTBT-Si tandem cells did not render as high PCE devices as PBTI3T/PTB7 cells, the
resulting voltages of PBTI3T/PSBTBT-Si devices were completely additive (VPBTI3T + VPSBTBT-Si =
VPBTI3T+PSBTBT-Si) (Table 2). Unfortunately, the increased voltage was unable to overcome the
initial handicap of low FF and PCE of PSBTB-Si, which is reflected by the large curve of the
tandem J-V graph (Figure 9). The superiority of PTB7 and PBTI3T over PSBTBT-Si is
demonstrated in the tandem cell efficiency results and graphical comparison of the respective
tandem cell graphs (Figure 8,9). This is a reflection of multiple factors -- the PTB7 and the
PBTI3T both have vastly superior single cell J-V curves in comparison to PSBTBT-Si (figure 6).
The shapes of the PBTI3T and the PTB7 single-cell J-V curves show that they work well in
tandem; the PBTI3T has a higher voltage than the PTB7, and the PTB7 has a higher current than
the PBTI3T (a result of its lower bandgap) (Table 2). The combination of these two polymers,
which are either slightly Jsc or Voc deficient, creates a solar cell that has an incremental increase
of both, making the new J-V curve move closer to a better FF (square graph) (Figure 8).
Conclusions & Future Work
In this study, the exploration of high-performance single cell polymers has been extended
to tandem cells, advancing the usefulness and applicability of OPVs in solar energy technology.
These cells not only improve upon existing organic tandem technology through increased Voc,
but also leave room for optimization and perfection of devices using this project’s two most
18

successful polymers in tandem, PBTI3T and PTB7. Although the goal of record breaking device
PCEs was not attained, the unsatisfactory overall results obtained from PBTI3T/PSBTBT-Si
cells shroud the fact that distinctly additive voltages were observed in these devices. This largely
increased Voc is a characteristic unique to tandem devices, and provides a very high theoretical
PCE limit despite the lack of tandem Jsc and FF optimization. Overall, this project achieved
results that came close to matching published single cell results, and, with a novel high bandgap
active layer (PBTI3T) that has previously never been paired in tandem cells, achieved results that
are consistent with the current literature on tandem OPVs.
These results validate the methods and techniques adopted in this project, despite the fact
that the ideal processing conditions for single cells were not attainable for tandem cell production
during the project. In particular, the tandem cell processing conditions did not allow for
preventing surface roughness and layer mixing, thus decreasing the cell’s Jsc and FF. Specific
areas for improvement would further optimization of single-layer cells to improve individual
polymer performance, more precise tandem device fabrication, and the use of increasingly
advanced interfacial materials as they become available. Building on the successful additive
voltages of the PBTI3T/PTB7 and PBTI3T/PSBTBT-Si tandem cells and refining production
processes to match single cells’ Jsc and FF levels, tandem cells have the potential to attain or
exceed 15% PCE. This research provides necessary data on tandem device optimization, as well
as promising results for the implementation of previously untested polymer combinations,
including a new high efficiency polymer PBTI3T. Added to the versatility and ease of
installation that characterizes organic photovoltaics, these results light the way for eventual
breakthroughs in organic solar technology.
19

Siemens Research Report

More Related Content

What's hot

Similar to Siemens Research Report

Siemens Research Report