Understanding the Effect of Donor Layer Thickness and a MoO3

Subscriber access provided by A.A. Lemieux Library | Seattle University

Article 3

Understanding the Effect of Donor Layer Thickness and a MoO Hole Transport Layer on the Open-Circuit Voltage in Squaraine/C Bilayer Solar Cells. 60

James William Ryan, Thomas Kirchartz, Aurelien Viterisi, Jenny Nelson, and Emilio J Palomares J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/jp406472t • Publication Date (Web): 05 Sep 2013 Downloaded from http://pubs.acs.org on September 14, 2013

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Understanding the Effect of Donor Layer Thickness and a MoO3 Hole Transport Layer on the OpenCircuit Voltage in Squaraine/C60 Bilayer Solar Cells James W. Ryan,1†* Thomas Kirchartz,2 Aurélien Viterisi,1 Jenny Nelson,2 Emilio Palomares1,3* 1

Institute of Chemical Research of Catalonia (ICIQ), Avda. Paisos Catalans 16, Tarragona 43007, Spain

2

Department of Physics and Centre for Plastic Electronics, Imperial College London, South Kensington Campus, London SW7 2AZ, United Kingdom

3

Catalan Institution for Research and Advanced Studies (ICREA), Avda. Lluís Companys, Barcelona, E-08010, Spain

Corresponding Author *Prof. Emilio Palomares. ICIQ, Avda. Paisos Catalans 16, Tarragona 43007, Spain. E-mail address: [email protected]. Tel: +34 977920241. * Dr. James Ryan. The University of Tokyo, Dept. of Chemistry, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. E-mail address: [email protected]. Tel: +81 358411476.

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 34

Present Addresses † Department of Chemistry, School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyoku, Tokyo 113-0033, Japan


2

Page 3 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ABSTRACT

Small molecule organic solar cells are becoming increasingly efficient through improved molecular design. However there is still much to be understood regarding device operation. Here we study bilayer solar cells employing a 2,4-bis[4-(N,N-diisobutylamino)-2,6-dihydroxyphenyl] squaraine (SQ) donor and fullerene acceptor to probe the effect of donor layer thickness and a MoO3 electron transport layer on device performance. The thickness of SQ is seen to drastically affect the open-circuit voltage (VOC) and fill factor (FF), while the short circuit current is not altered significantly. The fact that the VOC of the bilayers with thin (6 nm) donor layers shows a strong dependence on the material and workfunction of the anode cannot be explained with a model for a perfect bilayer. Recombination of electrons from C60 at the anode contact has to be possible to understand the strong effect of the anode workfunction. Using numerical simulations and a simple two-diode model we show that the most likely interpretation of the observed effects is that for thin SQ layers, the roughness of the interface is high enough to allow electrons in the C60 to tunnel through the SQ to recombine directly at the anode. Thicker SQ layers will block most of these recombination pathways, which explains the drastic dependence of VOC on thickness. Bulkheterojunction devices were also fabricated to illustrate the effect of anode material on the VOC.

KEYWORDS Selective contacts, planar heterojunction, nongeminate recombination, organic/inorganic interfaces, small molecule, transient photophysics.


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 34

Introduction Organic solar cells (OSCs) are edging closer to realizing their potential as low-cost energy harvesters with certified efficiencies of 9.2% for a single junction,1 In recent years small molecule based OSCs (SMOSCs) have made particular progress and are steadily catching up on their polymer counterparts, with record efficiencies now approaching 8% for a single junction and as high as 12% for tandem devices.2,

3

Now that small molecular weight organic

semiconductors have demonstrated their ability to produce efficient devices, further progress is expected especially as they offer several favorable characteristics compared to polymers, such as more facile synthesis and purification methods, high tunability of optoelectronic properties,4 and the ability to be fabricated via solution processing and/or highly reproducible vacuum-based methods. The rapid progress of OSCs in recent years has been mainly due to improved synthetic approaches. To further understand and maximize the performance of new and efficient sensitizers a strong effort in studying the device operating physics is required. The open-circuit voltage (VOC) is one of the principle figures of merit that defines solar cell efficiency and its origin is understood to a certain degree, but its controlling factors can change on a case-by-case basis.5-12 The VOC is the voltage at which no net current flows in the device and is controlled by the balance between charge generation and charge recombination in the device. The recombination rate scales with the equilibrium charge-carrier concentrations and therefore exponentially with the energy levels, with changes in energetics having a large influence on VOC.13-15 Understanding the factors controlling the recombination and thus VOC are therefore very important but are non-trivial, as the recombination kinetics may change from one material to the next,16 or as a function of the microstructure of one material.17, 18 While recombination is often


4

Page 5 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


considered to happen at the internal donor/acceptor interface, charges may also recombine at the cathode or anode, especially when contacts are not optimized. In these cases, the VOC might depend quite strongly on the workfunction of the anode or cathode.19 To date most of the studies on OSCs that take into account the energetics of the active layer together with the charge recombination kinetics are based on bulk-heterojunction (BHJ) devices. However, planar heterojunction (PHJ) devices have not received the same attention, with many reports describing the origin of VOC in these devices using steady state measurements or else theoretical calculations. It has been shown by ourselves and others,20, 21 that there is a significant difference between PHJ and BHJ devices in terms of the distribution of charges within the device as well the interfacial non-geminate recombination dynamics. From previous studies of BHJ devices, the relationship between charge carrier density and recombination dynamics changes from material to material,16 not to mention the selective contacts employed.22 Squaraines are excellent NIR-absorbing dyes possessing high extinction coefficients, and high stability.23-25 Several squaraine derivatives have been employed in organic solar cells recently,26-32 with efficiencies of ∼6% having been reported.33,

34

However, although squaraines posses

excellent absorption properties, they are limited by inherently low exciton diffusion lengths, thus restricting their light-harvesting ability, especially in a PHJ device.32, 35 Here we study a bilayer structure consisting of a 2,4-bis[4-(N,N-diisobutylamino)-2,6-dihydroxyphenyl] squaraine (SQ) donor (Figure 1a) and a C60 acceptor, where we change the thickness of the SQ layer and notice a strong shift in VOC with increasing thickness. Furthermore, for each thickness of SQ employed, the effect of employing a MoO3 (HTL) transport layer was investigated (see Figure 1b and 1c for a schematic of the device architecture and relevant energy levels). We find that there is a strong influence of the thickness of the SQ layer employed on performance. With increasing SQ


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 34

thickness, also VOC increases, however, there is a concomitant decrease in fill factor (FF). Furthermore the MoO3 layer increases the VOC of the device for SQ layers less than 9 nm but it has no significant effect on the VOC for higher SQ thicknesses. Transient optoelectronic techniques were used to probe the effects of both SQ thickness and the MoO3 interfacial layer on the charge density in the active layer under working conditions as well as to quantify the rate of non-geminate charge carrier recombination. Finally, numerical simulations are used to analyze and interpret the trends in open-circuit voltage and fill factor.

Figure 1 Chemical structure of 2,4-bis[4-(N,N-diisobutylamino)-2,6-dihydroxyphenyl] squaraine (SQ) (a) and schematic diagrams of device architecture (b) and energy levels for the materials employed (c). The thickness of MoO3 and SQ were varied, with the thickness of MoO3, y, equal to 0 or 8 nm and the thickness of SQ, x, equal to 6, 9 or 12 nm. The thickness of C60, bathocuproine (BCP) and Al were fixed at 40, 8 and 100 nm respectively.


6

Page 7 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Results The current/voltage characteristics of devices without a MoO3 interfacial layer are shown first in Figure 2a, which demonstrate a clear dependence of VOC on the SQ thickness. An opposite trend is observed for the FF. Regarding the photocurrent, except for a small decrease in the shortcircuit current density (JSC) between devices employing MoO3, there is no change in JSC with increasing SQ thickness, as the thickness of the layers exceeds the exciton diffusion length.32, 35 The power conversion efficiency (η) does not differ very significantly due to the concomitant decrease of FF with increase in VOC. Clearly, increasing the SQ thickness therefore has more influence on optimizing the donor/acceptor interface rather than on increasing photocurrent. Employing a MoO3 interfacial layer, an increase in VOC was observed, but only for devices with the thinner 6 nm SQ layer. The thickness of MoO3 did affect the trend in VOC significantly, even with films as thin as 1 nm or as thick as 50 nm (see Supporting Information Figure S1). This suggests that the MoO3/SQ interaction is influencing the VOC rather than MoO3 simply acting as a buffer layer between ITO and SQ. MoO3 improved the FF for the 6 and 12 nm devices but does not significantly affect the 9 nm devices. The improved FF can be in part explained by the increase in shunt (parallel) resistance (RP) although the series resistance does increase with the addition of MoO3 (see Table 1), as measured from the current/voltage curve recorded in dark conditions at short circuit and high forward potential respectively. Comparing the dark current/voltage measurements (Figure 2b) there is a reduction in dark current with increasing the thickness of SQ, dSQ, (and in the case of the 6 nm devices, the addition of MoO3), consistent with the trend in VOC under 1 sun illumination, which can be assigned to a decrease in leakage current.7


7


2

Current Density J (mA/cm )

(a) 6

ITO/SQ/C /BCP/Al 60

4

ITO/MoO /SQ/C /BCP/Al 3 60 = 6 nm

d

SQ

2

d = 9 nm SQ d = 12 nm SQ

0 -2 -4

-6 -0.5

0

0.5

Voltage V (V)

1

(b) 2


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 34

2

10

ITO/SQ/C /BCP/Al

1

60

10

0

10

ITO/MoO /SQ/C /BCP/Al 3

d

= 6 nm

-1

d

= 9 nm

-2

d

= 12 nm

10

SQ SQ

10

SQ

60

-3

10

-4

10

-5

10

-6

10

-0.5

0

0.5

1

Voltage V (V)

Figure 2 Current/voltage curves for devices measured under standard 1 sun conditions (AM 1.5 G, 100 mW/cm2), (a), and dark conditions (b). Each device has the architecture ITO/SQ/C60/BCP/Al (closed circles) or ITO/MoO3/SQ/C60/BCP/Al (open circles) with the SQ thickness varying from 6 to 12 nm. In all cases, the device JSC values showed a linear dependence on light intensity, with the power law relationship, !!" ~!" ! , where ! = 1, meaning none of the devices are limited by space


8

Page 9 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


charge. Plotting VOC vs. light intensity showed that it increased linearly with the log of light intensity (Figure 3); the slopes of these plots were then used to calculate the ideality factor based on the relationship

!!" =

! !"!" !" !"#(!")

(1)

where q is the elementary charge and kT is the thermal energy. The ideality factor is an important parameter to understand how close to ideal the diode behaves and it also allows us to understand the influence of trap states on recombination.36 The ideality factor is between 1 and 2, with an ideal trap-free device having !!" = 1. Increasing values of !!" correspond to the presence of trap states affecting recombination, with traps becoming deeper and more dominant as !!" gets closer to 2.37 Strangely in some devices, the ideality factor supersedes 2 (see Table 1), which for inorganic devices has been suggested to represent either coupled defect recombination or tunneling enhanced defect recombination;38,

39

i.e. in both cases the high ideality factor is

caused by a mechanism based on defect-induced recombination. As we use a simple approximation for calculating !!" by fitting a straight line to VOC versus light intensity, this provides us with an average over several possible recombination mechanisms, which makes it difficult to interpret the exact mechanisms controlling the high ideality factors observed.40 Another explanation for the high !!" could be imagined if we consider that the high leakage currents observed in the dark current/voltage curves may lead to a significant decrease in VOC at low LI, which would steepen the slope of VOC versus LI and thus increase !!" .


9


1.1 1

ITO/SQ/C /BCP/Al 60


60

OC

(V)

0.9

V

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 34

d

= 12 nm

d

= 9 nm

d

= 6 nm

SQ

0.8 0.7

SQ

0.6 0.5 0.4

SQ

10

2

Light Intensity (mW/cm )

100

Figure 3 Light intensity (LI) dependence on the open-circuit voltage for each device. The data is fit using the relationship: VOC = y + m(ln(LI)).

Table 1 Summary of the figures of merit for each device recorded under 1 Sun (100 mW/cm2) irradiation together with the series and shunt resistances obtained from the dark current/voltage curves. The ideality factor was calculated from recording VOC versus light intensity and applying Equation (1). VOC

JSC

V

mA/cm2

dSQ = 6 nm

0.55

5.34

dSQ = 9 nm

0.76

dSQ = 12 nm

0.82

Device

FF

!!"

η

RS

RP

%

Ω/cm2

Ω/cm2

0.67

1.97

0.77

6.4 x 105

1.49

5.27

0.49

1.96

0.72

4.7 x 105

2.56

5.20

0.41

1.77

0.79

4.9 x 105

2.76

ITO/SQ/C60/BCP/Al

ITO/MoO3/SQ/C60/BCP/Al dSQ = 6 nm

0.66

5.09

0.66

2.22

1.36

2.0 x 105

1.67

dSQ = 9 nm

0.76

4.93

0.55

2.05

1.04

3.5 x 106

2.38

dSQ = 12 nm

0.84

4.83

0.47

1.95

0.96

4.6 x 106

1.47


10

Page 11 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Moving from steady state measurements to transient measurements we first look at the influence that the device architecture has on the charge generated in the active layers. For these experiments we employed charge extraction (CE) and transient photovoltage (TPV) techniques using a similar set-up described by Durrant and co-workers,41-43 and in previous studies from our group.20, 22, 44 It must be noted that as we are applying these techniques to bilayer devices special care must be taken in interpreting the data, as recently explained in detail by Foertig and coworkers.21 A key finding in their paper is that the charge density profiles in BHJ and bilayer devices deviate substantially. Nonetheless, they found that a near linear relationship exists between the charge carrier density measured using CE and the calculated value of the charge carrier density at the donor/acceptor interface.21 Therefore the trends we observe in Figures 4 and 6 below are quite accurate but may not be quantitative. Figure 4a shows the total charge per unit area (C/cm2) in the device plotted versus bias obtained using the CE method. We see that the charge increases with increasing SQ thickness. The black line represents the average geometrical capacitive charges, which was calculated by carrying out CE under applied negative bias in the dark,41 and corresponds to a geometrical capacitance of 67 nF/cm2. For the thinnest devices, the total charge recorded is similar in magnitude to the charges at the electrode, meaning that very little charge builds up in the active layer.


11


2

Charge per unit area (C/cm )

a) Total Charges -7

1.4 x 10

ITO/SQ/C /BCP/Al

-7

60

1.2 x 10


-7

60

d = 6 nm SQ d = 9 nm

1 x 10

-8

8 x 10

SQ

d

SQ

-8

6 x 10

= 12 nm

-8

4 x 10

-8

Capacitive Charges

2 x 10

2

(67 nF/cm )

0 0

0.2

0.4

0.6

0.8

1

0.8

1

Voltage V (V)

b) Active Area Charge Density 17

1 x 10

ITO/SQ/C /BCP/Al

-3

Charge Density n (cm )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 34

60

16

ITO/MoO /SQ/C /BCP/Al

8 x 10

d

= 6 nm

d

= 9 nm

d

= 12 nm

SQ

16

6 x 10

3

SQ SQ

60

16

4 x 10

16

2 x 10

0

0

0.2

0.4

0.6

Voltage V (V)

Figure 4 Charge extraction data showing the total charge stored in the device as a function of bias (a) and the charge carrier density present in the active layer (b). Increasing the SQ thickness leads to a larger build up of active layer charge. Importantly, we see a change in the shape of the exponential between ITO and MoO3 devices, which we can relate to a shift in averaged donor/acceptor density of states (DOS).22 As mentioned above, a 1 nm layer is sufficient to induce a change in the VOC for the thin SQ films (dSQ = 6 nm), and this coupled with the shift in the charge density versus bias provides further evidence that there is a chemical interaction between SQ and MoO3 rather than MoO3 simply acting as a physical barrier, consistent with previous studies on organic/MoO3 interfaces.45-47 Correcting for the geometrical


12

Page 13 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


charges,41 the charge in the active layer is plotted in Figure 4b as a function of volume (i.e. charge carrier density). The data fit well to an exponential for values above 0.3 V with the following function

! = !! ! !"!"

(2)

with n being dependent on the applied bias (or splitting of quasi-Fermi levels).42 The values of γ are very low suggesting no defined band edges in contrast to what one would expect for an ideal semiconductor (see Table S1 in the supporting information). The trend observed is that γ decreases with increasing SQ thickness, suggesting a more delocalized distribution of energy states, we also see that there are slightly higher values of γ for the MoO3 devices. However, due to the very low thickness of the active layer, the gradients of charge carrier concentration as a function of position are very steep which means that the effect of γ is difficult to interpret.36 The high charge density observed in these devices is interesting considering that many of the small molecule devices (BHJ and PHJ) reported in the literature and/or investigated in our lab tend to have very low chemical potentials, resulting in most of the charge density observed to arise from the geometrical capacitive charges.20, 21, 41, 48 Small perturbation charge carrier lifetimes (!!! ) are plotted versus bias in Figure 5, obtained from TPV measurements and show an exponential decay with the function !!! = !!!! ! !!!!"

(3)

where !!!! is the lifetime of the charge carriers at short circuit, and β the decay constant. First looking at the dSQ = 6 nm devices, we see that !!! at ∼1 sun is shorter for the ITO device,


13


consistent with the lower VOC obtained for that device. The dSQ = 9 nm devices have almost identical lifetimes and are an order of magnitude longer lived than the dSQ = 6 nm devices.

-3

ITO/SQ/C /BCP/Al 60


60

-4

10

d

"n

(s)

10

!

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 34

SQ

= 12 nm

-5

10

d

SQ

-6

10

d

SQ

= 9 nm

= 6 nm

0.3 0.4 0.5 0.6 0.7 0.8 0.9

1

Voltage V (V)

Figure 5 Small perturbation carrier lifetimes versus bias for each device recorded using the transient photovoltage method. To fully understand this relationship we must compare the lifetimes to the charge density, because as we saw in Figure 4 the charge present in the device changes significantly. Correlating the lifetime and charge carrier density, as shown in Figure 6, has two important advantages in that it decouples the charge recombination dynamics from the voltage highlighting the effect of the charge carrier distribution and it allows for devices of different architecture and materials to be compared. The ITO and MoO3 devices with dSQ = 6 nm, illustrate these points well, with a big difference evident when comparing their charge carrier density plotted versus voltage or charge density. While the device with the MoO3 anode has a lower number of charges present in its active layer for any given voltage, the dependence of the carrier lifetime is far lower than that of the ITO device, which explains the higher VOC observed in the current/voltage curve (Figure 2a). Interpreting the origin of the exponential parameters obtained from the CE and TPV studies is


14

Page 15 of 34

quite complex for bilayer devices as mentioned above, but for reference we have included a table summarizing the data in the supporting information (Table S1).

ITO/SQ/C /BCP/Al 60

-3

10


60

SQ

= 12 nm

10

"n

(s)

d -4

!

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


-5

10

d

SQ

d

SQ

-6

10

15

10

= 9 nm

= 6 nm 16

10

17

10

-3

Charge Density n (cm )

Figure 6 Small perturbation lifetimes plotted versus charge density by using the exponential fit obtained from the CE measurements for each device. The thin film morphology of SQ was also studied using atomic force microscopy (AFM) to see whether there was any significant change in the surface topography with increasing SQ thickness. No major change was observed between the films. The root mean square roughness of ITO is ∼4 nm, and changes very little for the SQ films deposited on it, nor is there any difference of SQ topography on MoO3. The rms roughness is on the same order as the thickness of the SQ film, and this topography illustrates that the SQ/C60 interface is not simply the ideal flat bilayer we imagine for a planar heterojunction device, but a rather rough, bulk-like heterojunction. Another point to consider is the low thickness of the SQ films deposited. It is easy to imagine that there are regions where the C60 may be in extremely close contact with the electrode; close enough for surface recombination to occur (see Scheme 1). Increasing the thickness of SQ


15


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 34

however, reduces the possibility of electrons recombining at the wrong electrode and would help explain the higher VOC observed.

Scheme 1 Illustration of a perfect bilayer (a) in comparison to a more realistic picture of the anode and SQ morphology (b) and (c). For thin SQ layers (b) it is probable that there are regions in which C60 is in contact with the anode leading to surface recombination. Increasing the thickness suppresses surface recombination (c), and explains the increased charge carrier lifetimes and VOC observed. To confirm what happens in this system when C60 is in contact with the anode we fabricated BHJ devices. Here, devices were prepared in exactly the same manner as bilayer devices, except that SQ and C60 were co-evaporated. The thickness of the active layer was 50 nm and the ratio of SQ:C60 was either 1:1 or 1:5, with the latter ratio resembling the ratios used in the bilayer devices. Both devices with a 1:1 and 1:5 SQ:C60 active layer show a clear dependence of VOC on the anode material (see Figure 7) and support our hypothesis that the bilayer devices with thin


16

Page 17 of 34

SQ layers are limited by surface recombination. Figures of merit for the current/voltage curves are presented in Table S2.

2


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


10 1:1 ITO/BHJ/BCP/Al 1:1 ITO/MoO /BHJ/BCP/Al 3

5

1:5 ITO/BHJ/BCP/Al 1:5 ITO/MoO /BHJ/BCP/Al 3

0

1:1 -5

-10 -0.5

1:5 0

0.5

1

Voltage V (V)

Figure 7 Current/voltage curves for co-evaporated SQ:C60 BHJ devices with the SQ:C60 ratio of 1:1 (orange circles) and 1:5 (purple triangles) with (open symbols) and without MoO3 (closed symbols). Active layer thickness is 50 nm for all devices.

Discussion Changing the thickness of the SQ layer changed the VOC significantly and through the transient optoelectronic measurements presented above we can explain this by considering two factors, one being the charge carrier density n and the other being the charge carrier lifetime τ. Increasing the thickness of the SQ increases the charge density generated in the active layer of the device and if we consider a constant value of n, we see that the quasi-Fermi level splitting is largest for the thinner SQ films. Going from dSQ = 6 nm to dSQ = 9 nm, in addition to the increased potential difference between the SQ and C60 quasi-Fermi levels there is also a significant increase in the charge carrier lifetime and decrease in the decay order. Therefore, we have a higher number of charges, which also live longer and thus explain the higher VOC observed. Increasing the SQ


17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 34

thickness further, we generate slightly more free charge carriers, which also have slightly longer lifetimes but a higher decay order, leading to an increase in VOC but not as significant as in the first increment. The changes in the order of decays observed from the TPV measurements hint at different recombination processes dominating the device. The effect of a MoO3 hole transport layer (HTL) has been investigated in the past for several different donor materials, in particular for planar heterojunction devices and has been shown to increase the VOC when certain donors have been employed but not for others.7, 49, 50 Here, for the SQ/C60 devices, we see that MoO3 only plays a significant role in improving the VOC for the thin SQ layers. Comparing this finding with previous reports, most of the devices that showed improved VOC by employing MoO3 utilized donor molecules with low exciton diffusion lengths, which of course imposed a limit to how thick the films could be made. In our study, when we increase the SQ layer thickness to values much higher than the exciton diffusion length we are in effect creating a two tier film, where the region closest to the ITO is simply acting to isolate the SQ/C60 interface from the electrode, preventing unwanted surface recombination and acting as an electron blocking layer (EBL). Therefore the MoO3 is made almost redundant, although it does offer improved shunt resistance, which leads to better FFs, which can be explained by providing a more homogeneous film as ITO films tend to be electrically quite inhomogeneous.6 The fact that the VOC depends on the anode material (at least for the low SQ thicknesses) cannot be explained if we assume a perfect bilayer structure. Indeed, BHJ devices show the same dependency of VOC on the anode material as observed for the devices employing thin SQ layers. To illustrate the effect of changing the anode material for bilayer devices, we performed driftdiffusion simulations of bilayers and varied the energetic distance !!" between the valence band edge and the Fermi level at the anode to simulate the effect of changing the anode workfunction.


18

Page 19 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


From the simulations shown in Figure 8a, we can see that the open circuit voltage is not affected by the workfunction and therefore the value of !!" at all. However, for sufficiently high values of !!" , the VOC will exceed the built-in voltage at some point and S-shaped current voltage curves are expected because the majority carriers have to diffuse against the electric field to be collected.51 Yet, in the experimental data, we neither observe S-shapes nor an insensitivity of VOC to the anode material. Thus, we also studied the case where the donor material does not present a barrier for electron transport from the C60 to the anode. For this purpose we just align the conduction band in the nominally pure SQ layer to the conduction band of the C60. Thereby allowing the electrons to travel freely to the anode and recombine there will lead to a strong dependence of VOC on the workfunction of the anode as seen in Figure 8b.19 However, it will still not explain the strong dependence of VOC on the thickness of the SQ layer. In fact the thickness should hardly affect the open-circuit voltage at all. In addition, it is still unclear why the FFs of the devices with thicker SQ layers would be substantially worse than the FF for the devices with thinner SQ layers.


19

-2 -2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

current density J (mAcm ) current density J (mAcm )


Page 20 of 34

4 (a) bilayer no surface recombination possible

2 0

ϕan = 0.4 eV

-2 -4 -6

dSQ = 9 nm

ϕan = 0.2 eV

-8 4 (b) mixed donor 2

ϕan = 0 eV

electrons can recombine at anode

0 -2

ϕan = 0.4 eV

-4 -6

dSQ = 9 nm

-8 0.0

0.2

ϕan = 0.2 eV

0.4 0.6 voltage V (V)

ϕan = 0 eV

0.8

1.0

Figure 8 Drift-diffusion simulations to illustrate the effect of changing the anode workfunction on the current/voltage curve of (a) a bilayer solar cell and (b) as solar cell with a donor layer that allows electrons to travel efficiently to the anode. In the case of a perfect bilayer, recombination of electrons at the anode is impossible and therefore the VOC is unaltered. However, S-shapes occur if !!" (the distance from the valence band edge to the Fermi-level at the anode) becomes too large, because charges have to diffuse against the electric field to be collected close to open circuit conditions. In case of a mixed donor layer that allows electron transport, surface recombination is possible and the VOC will decrease when !!" becomes too large. A tentative explanation for the observed trend with thickness is a transition between a device that behaves like a bilayer at high SQ thicknesses and a device where recombination of electrons at the anode is easily possible at low SQ thicknesses (as illustrated in Scheme 1). The observed current/voltage curves could be understood as the parallel connection of a solar cell limited by


20

Page 21 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


surface recombination (low VOC but good FF) and a solar cell that is not limited by surface recombination allowing a much higher VOC. Normally, the parallel connection of two solar cells with different VOC would lead to a combined current/voltage curve with a VOC dominated by the lower of the two. However, the higher VOC might dominate the total current/voltage curve if the regions with high recombination are either electrically isolated with high series resistances,52 or if they contribute only a small part to the total area of the device. In these cases, the areas with high surface recombination would act like a non-linear shunt for the rest of the device. Figure 9 illustrates that using a parallel connection of two solar cells in different ratios and with different series resistances, it is possible to reproduce the experiment (in this example, the SQ thickness series with ITO anode). For Figure 9, we used two illuminated diode equations of the type

! = !! exp

! ! − !!s !id !"

− 1 − !sc +

! − !!s !p

(4)

where J0 is the saturation current density. The two diode equations are drawn in green dashed lines and are labeled SC1 and SC2. Here, SC1 represents the diode limited by surface recombination of electrons and the anode and SC2 represents the bilayer where no surface recombination can occur. The experimental current/voltage curve for dSQ = 6 nm can then be fitted by assuming that SC1 dominates the system. The experimental current/voltage for larger SQ thicknesses are fitted by assuming that we connect SC1 and SC2 in parallel and that we vary the series resistance of SC1 as well as the ratio of the area that is made up by SC2 and the area where electrons can reach the anode (SC1).


21

-2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

current density J (mAcm )


4 ITO thickness series 2 symbols: experiments 0

Page 22 of 34

dSQ = 9 nm

lines: simulations SC1 dSQ = 6 nm

-2 -4

dSQ = 12 nm

-6 -8 0.0

SC2

0.2

0.4 0.6 voltage V (V)

0.8

1.0

Figure 9 The changes in current/voltage curves observed with increased SQ thickness are explained here with a parallel connection of two diodes (see Eq. (4)) with different VOC (SC1 and SC2) in different ratios and with different series resistances associated with each diode. In this interpretation of the data, the change in SQ thickness means that the influence of the shunt diode caused by surface recombination (SC1) that controls the device with dSQ = 6 nm is more and more suppressed for higher dSQ. Because the model is quite a crude and simplistic approximation to reality and as the influence of series resistance and the ratio of areas are interchangeable to a certain degree, the parameters obtained from the fit are not particularly meaningful. However, the simple model suggests that the observed current/voltage curves can neither be understood using the picture of a perfect bilayer with suppressed surface recombination nor the picture of a bulk heterojunction, where surface recombination is allowed. Instead, the trends in thickness seem to originate from a transition between two limiting situations. While at low SQ thicknesses, the anode is able to limit VOC due to high surface roughness and or some inter-diffusion of C60 into the SQ layer, this loss pathway gradually disappears with increasing SQ thickness. While the influence of the shunt diode does not affect the VOC much at high SQ thickness it still affects the FF and therefore has an important influence on the device performance. Alternative explanations for the reduced FF at


22

Page 23 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


high SQ thicknesses would be field dependent geminate effects as discussed for instance in references

53-55

. These effects cannot be ruled out but they cannot alone be responsible for the

observed trends. While geminate effects might have a slight effect on VOC in theory, the changes in VOC observed here are accompanied by changes in the dark current that could not be explained by geminate recombination.

Conclusions Through tuning the SQ donor layer thickness we observed a large change in the open-circuit voltage. For the thinnest films studied, dSQ = 6 nm, the VOC was significantly improved by employing a MoO3 hole transport layer. However, increasing the thickness of SQ, the MoO3 did not improve the VOC, evidence that the thin SQ/C60 devices suffer from surface recombination, i.e. electrons from C60 recombining with the ITO electrode. Fabricating BHJ devices and observing the increase in VOC when MoO3 was employed further confirmed this. Once a sufficiently thick SQ layer is employed the C60 is isolated from ITO and thus the MoO3 layer as an EBL is made redundant. Through transient optoelectronic techniques we could study the increased chemical potential in the device with increasing SQ thickness, and the consequent effect on charge carrier lifetime. Although VOC could be improved significantly with increasingly thicker SQ layers the pay off was a lower FF. Through modeling the devices using drift-diffusion simulations we showed that the change in VOC observed for dSQ = 6 nm when adding a MoO3 interfacial layer cannot be explained by assuming a planar heterojunction, but the behavior is more bulk-heterojunction like. Co-evaporated SQ:C60 BHJ devices provided experimental evidence to corroborate these simulations. A basic two-diode model was also used, which illustrated the transition from a more BHJ-like active layer limited by surface recombination at


23


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 34

low SQ thickness, towards a more PHJ-like device at higher SQ thicknesses that features suppressed surface recombination and a higher VOC. These results demonstrate that, despite their apparently simple device architecture, PHJ OSCs can have rather complicated operating mechanisms. Many of the limiting factors observed here for SQ arise from its low exciton diffusion length, which limits the thickness that can be employed. Small molecule donors with higher exciton diffusion lengths should allow thicker donor layers to be utilized that sufficiently isolate the donor/acceptor interface from the anode without introducing unwanted shunt diode behavior.

Experimental 2,4-bis[4-(N,N-diisobutylamino)-2,6-dihydroxyphenyl]

squaraine

(SQ)

was

synthesized

following the procedure of Tian et al.25 High purity greenish shiny crystals of SQ were obtained with purity superior to 99% by HPLC. Devices were fabricated with the architecture ITO/MoO3(0 or 8 nm)/SQ(6, 9 or 12 nm)/C60(40 nm)/BCP(8 nm)/Al(100 nm). MoO3 (99.98% trace metal basis), Bathocuproine (BCP) (>96%) and Al (99.999% trace metal basis) were supplied by Sigma Aldrich and C60 (>99%) was purchased from MER Corp., Texas. BCP and SQ were sublimated twice prior to use. All other materials were used without further purification. ITO substrates (Psiotec, 5 Ohm/square) were sonicated in acetone and twice in isopropanol followed by 20 min of UV/O3 treatment. All layers were deposited by vacuum sublimation at a pressure not exceeding 2 × 10-6 mbar. Devices were placed in an air-tight holder, sealed in a N2 filled glovebox, and measured using an Abet Sun 2000 solar simulator (100 mW/cm2, AM 1.5 G), calibrated with a silicon photodiode (NREL), attached to a Keithley 2400 digital source meter. Transient photovoltage (TPV) measurements were carried out using an array of LEDs and


24

Page 25 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a nanosecond nitrogen laser, exciting the devices at 660 nm. Charge extraction experiments were done using a homebuilt system, where the steady state light source was generated by an array of LEDs and the transient decay of the charge stored in the device when switched from open to short-circuit was recorded using a Yokogawa oscilloscope.

ASSOCIATED CONTENT AUTHOR INFORMATION Corresponding Author James Ryan, The University of Tokyo, Dept. of Chemistry, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. E-mail address: [email protected]. Tel: +81 358411476. Emilio Palomares, ICIQ, Avda. Paisos Catalans 16, Tarragona 43007. E-mail address: [email protected]. Tel: +34 977920241. Notes The authors declare no competing financial interests. ACKNOWLEDGMENT E.P, J.W.R and A.V would like to acknowledge the European Research Council starting Grant ERCstg- POLYDOT and the national project MICINN CTQ-2007-60746-BQU. EP also thanks the ICIQ, ICREA and Catalonian regional government for funding (project 2009 SGR 207). TK acknowledges support via an Imperial College Junior Research Fellowship.


25


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 34

Supporting Information. Additional current/voltage curves, CE and TPV data as well as band diagrams and parameters used for the simulations are included. This information is available free of charge via the Internet at http://pubs.acs.org.

References 1.

He, Z.; Zhong, C.; Su, S.; Xu, M.; Wu, H.; Cao, Y., Enhanced Power-Conversion

Efficiency in Polymer Solar Cells using an Inverted Device Structure. Nature Photon. 2012, 6, 591-595. 2.

Heliatek,

Press

Release.

http://www.heliatek.com/wp-

content/uploads/2013/01/130116_PR_Heliatek_achieves_record_cell_effiency_for_OPV.pdf 2013, Last accessed 27/06/2013. 3.

Kyaw, A. K. K.; Wang, D. H.; Gupta, V.; Zhang, J.; Chand, S.; Bazan, G. C.; Heeger, A.

J., Efficient Solution-Processed Small-Molecule Solar Cells with Inverted Structure. Adv. Mater. 2013, 25, 2397-2402. 4.

Walker, B.; Liu, J.; Kim, C.; Welch, G. C.; Park, J. K.; Lin, J.; Zalar, P.; Proctor, C. M.;

Seo, J. H.; Bazan, G. C.; et al., Optimization of Energy Levels by Molecular Design: Evaluation of Bis-Diketopyrrolopyrrole Molecular Donor Materials for Bulk Heterojunction Solar Cells. Energy Environ. Sci. 2013, 6, 952-962. 5.

Chauhan, V.; Hatton, R.; Sullivan, P.; Jones, T.; Cho, S. W.; Piper, L.; deMasi, A.;

Smith, K., Elucidating the Factors that Determine the Open Circuit Voltage in Discrete Heterojunction Organic Photovoltaic Cells. J. Mater. Chem. 2010, 20, 1173-1178.


26

Page 27 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


6.

Hancox, I.; Chauhan, K. V.; Sullivan, P.; Hatton, R. A.; Moshar, A.; Mulcahy, C. P. A.;

Jones, T. S., Increased Efficiency of Small Molecule Photovoltaic Cells by Insertion of a MoO3 Hole-Extracting Layer. Energy Environ. Sci. 2010, 3, 107-110. 7.

Li, N.; Lassiter, B. E.; Lunt, R. R.; Wei, G.; Forrest, S. R., Open Circuit Voltage

Enhancement Due to Reduced Dark Current in Small Molecule Photovoltaic Cells. Appl. Phys. Lett. 2009, 94, 023307. 8.

Maurano, A.; Hamilton, R.; Shuttle, C. G.; Ballantyne, A. M.; Nelson, J.; O’Regan, B.;

Zhang, W.; McCulloch, I.; Azimi, H.; Morana, M.; et al., Recombination Dynamics as a Key Determinant of Open Circuit Voltage in Organic Bulk Heterojunction Solar Cells: A Comparison of Four Different Donor Polymers. Adv. Mater. 2010, 22, 4987-4992. 9.

Ratcliff, E. L.; Garcia, A.; Paniagua, S. A.; Cowan, S. R.; Giordano, A. J.; Ginley, D. S.;

Marder, S. R.; Berry, J. J.; Olson, D. C., Investigating the Influence of Interfacial Contact Properties on Open Circuit Voltages in Organic Photovoltaic Performance: Work Function Versus Selectivity. Adv. Energy Mater. 2013, 3, 647-656. 10.

Uhrich, C.; Wynands, D.; Olthof, S.; Riede, M. K.; Leo, K.; Sonntag, S.; Maennig, B.;

Pfeiffer, M., Origin of Open Circuit Voltage in Planar and Bulk Heterojunction Organic ThinFilm Photovoltaics Depending on Doped Transport Layers. J. Appl. Phys. 2008, 104, 043107. 11.

Shuttle, C. G.; Hamilton, R.; O’Regan, B. C.; Nelson, J.; Durrant, J. R., Charge-Density-

Based Analysis of the Current–Voltage Response of Polythiophene/Fullerene Photovoltaic Devices. Proc. Natl. Acad. Sci. 2010, 107, 16448-16452. 12.

Bisquert, J.; Garcia-Belmonte, G., On Voltage, Photovoltage, and Photocurrent in Bulk

Heterojunction Organic Solar Cells. J. Phys. Chem. Lett. 2011, 2, 1950-1964.


27


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

13.

Page 28 of 34

Cheyns, D.; Poortmans, J.; Heremans, P.; Deibel, C.; Verlaak, S.; Rand, B. P.; Genoe, J.,

Analytical Model for the Open-Circuit Voltage and its Associated Resistance in Organic Planar Heterojunction Solar Cells. Phys. Rev. B 2008, 77, 165332. 14.

Vandewal, K.; Gadisa, A.; Oosterbaan, W. D.; Bertho, S.; Banishoeib, F.; Van Severen,

I.; Lutsen, L.; Cleij, T. J.; Vanderzande, D.; Manca, J. V., The Relation Between Open-Circuit Voltage and the Onset of Photocurrent Generation by Charge-Transfer Absorption in Polymer : Fullerene Bulk Heterojunction Solar Cells. Adv. Funct. Mater. 2008, 18, 2064-2070. 15.

Blakesley, J. C.; Neher, D., Relationship Between Energetic Disorder and Open-Circuit

Voltage in Bulk Heterojunction Organic Solar Cells. Phys. Rev. B 2011, 84, 075210. 16.

Credgington, D.; Durrant, J. R., Insights from Transient Optoelectronic Analyses on the

Open-Circuit Voltage of Organic Solar Cells. J. Phys. Chem. Lett. 2012, 3, 1465-1478. 17.

Credgington, D.; Hamilton, R.; Atienzar, P.; Nelson, J.; Durrant, J. R., Non-Geminate

Recombination as the Primary Determinant of Open-Circuit Voltage in Polythiophene:Fullerene Blend Solar Cells: an Analysis of the Influence of Device Processing Conditions. Adv. Funct. Mater. 2011, 21, 2744-2753. 18.

Hawks, S. A.; Deledalle, F.; Yao, J.; Rebois, D. G.; Li, G.; Nelson, J.; Yang, Y.;

Kirchartz, T.; Durrant, J. R., Relating Recombination, Density of States, and Device Performance in an Efficient Polymer:Fullerene Organic Solar Cell Blend. Adv. Energy Mater. 2013, DOI: 10.1002/aenm.201300194. 19.

Xia, R.; Leem, D.-S.; Kirchartz, T.; Spencer, S.; Murphy, C.; He, Z.; Wu, H.; Su, S.; Cao,

Y.; Kim, et al., Investigation of a Conjugated Polyelectrolyte Interlayer for Inverted Polymer:Fullerene Solar Cells. Adv. Energy Mater. 2013, 3, 718-723.


28

Page 29 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


20.

Sánchez-Díaz, A.; Burtone, L.; Riede, M.; Palomares, E., Measurements of Efficiency

Losses in Blend and Bilayer-Type Zinc Phthalocyanine/C60 High-Vacuum-Processed Organic Solar Cells. J. Phys. Chem. C 2012, 116, 16384-16390. 21.

Foertig, A.; Wagenpfahl, A.; Gerbich, T.; Cheyns, D.; Dyakonov, V.; Deibel, C.,

Nongeminate Recombination in Planar and Bulk Heterojunction Organic Solar Cells. Adv. Energy Mater. 2012, 2, 1483-1489. 22.

Bolognesi, M.; Sanchez-Diaz, A.; Ajuria, J.; Pacios, R.; Palomares, E., The Effect of

Selective Contact Electrodes on the Interfacial Charge Recombination Kinetics and Device Efficiency of Organic Polymer Solar Cells. Phys. Chem. Chem. Phys. 2011, 13, 6105-6109. 23.

Hu, L.; Yan, Z.; Xu, H., Advances in synthesis and application of near-infrared absorbing

squaraine dyes. RSC Advances 2013, 3, 7667-7676. 24.

Ajayaghosh, A., Chemistry of Squaraine-Derived Materials: Near-IR Dyes, Low Band

Gap Systems, and Cation Sensors. Acc. Chem. Res. 2005, 38, 449-459. 25.

Tian, M.; Furuki, M.; Iwasa, I.; Sato, Y.; Pu, L. S.; Tatsuura, S., Search for Squaraine

Derivatives That Can Be Sublimed without Thermal Decomposition. J. Phys. Chem. B 2002, 106, 4370-4376. 26.

Bagnis, D.; Beverina, L.; Huang, H.; Silvestri, F.; Yao, Y.; Yan, H.; Pagani, G. A.;

Marks, T. J.; Facchetti, A., Marked Alkyl- vs Alkenyl-Substitutent Effects on Squaraine Dye Solid-State Structure, Carrier Mobility, and Bulk-Heterojunction Solar Cell Efficiency. J. Am. Chem. Soc. 2010, 132, 4074-4075. 27.

Beverina, L.; Drees, M.; Facchetti, A.; Salamone, M.; Ruffo, R.; Pagani, G. A., Bulk

Heterojunction Solar Cells - Tuning of the HOMO and LUMO Energy Levels of Pyrrolic Squaraine Dyes. Eur. J. Org. Chem. 2011, 2011, 5555-5563.


29


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

28.

Page 30 of 34

Fan, B.; Maniglio, Y.; Simeunovic, M.; Kuster, S.; Geiger, T.; Hany, R.; Nuesch, F.,

Squaraine Planar-Heterojunction Solar Cells. Int. J. Photoenergy 2009, 2009, Article ID 581068. 29.

Mayerhoffer, U.; Deing, K.; Gruss, K.; Braunschweig, H.; Meerholz, K.; Wurthner, F.,

Outstanding Short-Circuit Currents in BHJ Solar Cells Based on NIR-Absorbing AcceptorSubstituted Squaraines. Angew. Chem. Int. Ed. 2009, 48, 8776-8779. 30.

So, S.; Choi, H.; Ko, H. M.; Kim, C.; Paek, S.; Cho, N.; Song, K.; Lee, J. K.; Ko, J.,

Novel Unsymmetrical Push-Pull Squaraine Chromophores for Solution Processed Small Molecule Bulk Heterojunction Solar Cells. Sol. Energ. Mater.Sol. Cells 2012, 98, 224-232. 31.

Wang, S. Y.; Mayo, E. I.; Perez, M. D.; Griffe, L.; Wei, G. D.; Djurovich, P. I.; Forrest,

S. R.; Thompson, M. E., High Efficiency Organic Photovoltaic Cells Based on a Vapor Deposited Squaraine Donor. Appl. Phys. Lett. 2009, 94, 23304. 32.

Wei, G. D.; Xiao, X.; Wang, S. Y.; Sun, K.; Bergemann, K. J.; Thompson, M. E.; Forrest,

S. R., Functionalized Squaraine Donors for Nanocrystalline Organic Photovoltaics. Acs Nano 2012, 6, 972-978. 33.

Wei, G. D.; Xiao, X.; Wang, S. Y.; Zimmerman, J. D.; Sun, K.; Diev, V. V.; Thompson,

M. E.; Forrest, S. R., Arylamine-Based Squaraine Donors for Use in Organic Solar Cells. Nano Lett. 2011, 11, 4261-4264. 34.

Chen, G.; Sasabe, H.; Wang, Z.; Wang, X.-F.; Hong, Z.; Yang, Y.; Kido, J., Co-

Evaporated Bulk Heterojunction Solar Cells with >6.0% Efficiency. Adv. Mater. 2012, 24, 27682773. 35.

Chen, G.; Yokoyama, D.; Sasabe, H.; Hong, Z.; Yang, Y.; Kido, J., Optical and Electrical

Properties of a Squaraine Dye in Photovoltaic Cells. Appl. Phys. Lett. 2012, 101, 083904.


30

Page 31 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


36.

Kirchartz, T.; Nelson, J., Meaning of Reaction Orders in Polymer:Fullerene Solar Cells.

Phys. Rev. B 2012, 86, 165201. 37.

van Berkel, C.; Powell, M. J.; Franklin, A. R.; French, I. D., Quality factor in a-Si:H nip

and pin Diodes. J. Appl. Phys. 1993, 73, 5264-5268. 38.

Steingrube, S.; Breitenstein, O.; Ramspeck, K.; Glunz, S.; Schenk, A.; Altermatt, P. P.,

Explanation of Commonly Observed Shunt Currents in c-Si Solar Cells by Means of Recombination Statistics Beyond the Shockley-Read-Hall Approximation. J. Appl. Phys. 2011, 110, 014515. 39.

Rau, U., Tunneling-Enhanced Recombination in Cu(In, Ga)Se2 Heterojunction Solar

Cells. Appl. Phys. Lett. 1999, 74, 111-113. 40.

Kirchartz, T.; Deledalle, F.; Tuladhar, P. S.; Durrant, J. R.; Nelson, J., On the Differences

between Dark and Light Ideality Factor in Polymer:Fullerene Solar Cells. J. Phys. Chem. Lett. 2013, 4, 2371-2376. 41.

Credgington, D.; Kim, Y.; Labram, J.; Anthopoulos, T. D.; Durrant, J. R., Analysis of

Recombination Losses in a Pentacene/C60 Organic Bilayer Solar Cell. J. Phys. Chem. Lett. 2011, 2, 2759-2763. 42.

Maurano, A.; Shuttle, C. G.; Hamilton, R.; Ballantyne, A. M.; Nelson, J.; Zhang, W.;

Heeney, M.; Durrant, J. R., Transient Optoelectronic Analysis of Charge Carrier Losses in a Selenophene/Fullerene Blend Solar Cell. J. Phys. Chem. C 2011, 115, 5947-5957. 43.

Shuttle, C. G.; Maurano, A.; Hamilton, R.; O'Regan, B.; Mello, J. C. d.; Durrant, J. R.,

Charge Extraction Analysis of Charge Carrier Densities in a Polythiophene/Fullerene Solar Cell: Analysis of the Origin of the Device Dark Current. Appl. Phys. Lett. 2008, 93, 183501.


31


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

44.

Page 32 of 34

Albero, J.; Zhou, Y.; Eck, M.; Rauscher, F.; Niyamakom, P.; Dumsch, I.; Allard, S.;

Scherf, U.; Krueger, M.; Palomares, E., Photo-Induced Charge Recombination Kinetics in Low Bandgap PCPDTBT Polymer:CdSe Quantum Dot Bulk Heterojunction Solar Cells. Chem. Sci. 2011, 2, 2396-2401. 45.

Matsushima, T.; Jin, G. H.; Kanai, Y.; Yokota, T.; Kitada, S.; Kishi, T.; Murata, H.,

Interfacial Charge Transfer and Charge Generation in Organic Electronic Devices. Org. Electron. 2011, 12, 520-528. 46.

Matsushima, T.; Kinoshita, Y.; Murata, H., Formation of Ohmic Hole Injection by

Inserting an Ultrathin Layer of Molybdenum Trioxide between Indium Tin Oxide and Organic Hole-Transporting Layers. Appl. Phys. Lett. 2007, 91, 253504. 47.

Gwinner, M. C.; Pietro, R. D.; Vaynzof, Y.; Greenberg, K. J.; Ho, P. K. H.; Friend, R. H.;

Sirringhaus, H., Doping of Organic Semiconductors Using Molybdenum Trioxide: a Quantitative Time-Dependent Electrical and Spectroscopic Study. Adv. Funct. Mater. 2011, 21, 1432-1441. 48.

Sánchez-Díaz, A.; Pacios, R.; Muñecas, U.; Torres, T.; Palomares, E., Charge Transfer

Reactions in Near IR Absorbing Small Molecule Solution Processed Organic BulkHeterojunction Solar. Org. Electron. 2011, 12, 329-335. 49.

Hancox, I.; Sullivan, P.; Chauhan, K. V.; Beaumont, N.; Rochford, L. A.; Hatton, R. A.;

Jones, T. S., The Effect of a MoOx Hole-Extracting Layer on the Performance of Organic Photovoltaic Cells Based on Small Molecule Planar Heterojunctions. Org. Electron. 2010, 11, 2019-2025. 50.

Kinoshita, Y.; Takenaka, R.; Murata, H., Independent Control of Open-Circuit Voltage of

Organic Solar Cells by Changing Film Thickness of MoO3 Buffer Layer. Appl. Phys. Lett. 2008, 92, 243309.


32

Page 33 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


51.

Kirchartz, T.; Gong, W.; Hawks, S. A.; Agostinelli, T.; MacKenzie, R. C. I.; Yang, Y.;

Nelson, J., Sensitivity of the Mott–Schottky Analysis in Organic Solar Cells. J. Phys. Chem. C 2012, 116, 7672-7680. 52.

Rau, U.; Grabitz, P. O.; Werner, J. H., Resistive Limitations to Spatially Inhomogeneous

Electronic Losses in Solar Cells. Appl. Phys. Lett. 2004, 85, 6010-6012. 53.

Barker, J. A.; Ramsdale, C. M.; Greenham, N. C., Modeling the Current-Voltage

Characteristics of Bilayer Polymer Photovoltaic Devices. Phys. Rev. B 2003, 67, 075205. 54.

Petersen, A.; Ojala, A.; Kirchartz, T.; Wagner, T. A.; Würthner, F.; Rau, U., Field-

Dependent Exciton Dissociation in Organic Heterojunction Solar Cells. Phys. Rev. B 2012, 85, 245208. 55.

Schwarz, C.; Tscheuschner, S.; Frisch, J.; Winkler, S.; Koch, N.; Bässler, H.; Köhler, A.,

Role of the Effective Mass and Interfacial Dipoles on Exciton Dissociation in Organic DonorAcceptor Solar Cells. Phys. Rev. B 2013, 87, 155205.


33


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 34

TOC:


34

Understanding the Effect of Donor Layer Thickness and a MoO3

Recommend Documents