those achieved with OPVs, DSCs or SS-DSCs, with reported state-of-the-art efficiencies near 4 %.[100, 101] One of the remaining challenges is to understand and optimize the photocurrent contribution of the polymer, an issue that is essential to take full advantage of the HSC concept.[102, 103]
Furthermore, there are other HSCs more closely related to organic BHJ solar cells. In these devices, metal oxide nanoparticles are directly blended with a conjugated poly-mer and a mixed film is solution processed, similar to blend films of polypoly-mer and fullerene.[104–106] This approach leads to good contact between polymer and metal oxide and incomplete pore filling cannot occur. However, the missing interconnection between the metal oxide nanoparticles can lead to unfavorable donor-acceptor mor-phologies, enhancing charge carrier recombination and limit the device efficiencies.
2.3 Extremely Thin Absorber Solar Cells
ETA-SCs are closely related to SS-DSCs and have been extensively studied during the past decade. Instead of a self-assembled monolayer of dye molecules, a thin coating (typically less than 10 nm) of a highly absorbing crystalline inorganic material is used in order to create a p-i-n-like device structure.[107–109] Combination of absorbing layers with nanostructured metal oxide electrodes and infiltrated hole transporters allows the use of absorbers with high extinction coefficients and broad spectrum but poor electronic properties since charge transport through the absorber is necessary only over very short distances.
TiO2 NP with Dye
E
Absorber TiO2
HTM
Ag
a) b)
FTO TiO2 HTM
Ag
Figure 2.4: Extremely thin absorber solar cell.(a) Schematic of the geometry of an ETA solar cell with a TiO2nanoparticle film as n-type electrode. (b) Energy landscape of a typical ETA solar cell.
The device geometry of a typical ETA-SC is shown in Figure 2.4 (a). A TiO2 nanopar-ticle film similar to the electrodes used in DSCs and SS-DSCs is coated with absorbing materials, such as CdSe, CdTe, CuInS2, or Sb2S3.[109–114] The electrodes are then in-filtrated with a HTM, most commonly Spiro-OMeTAD, CuSCN, or P3HT. Depending on the absorber material, ETA-SCs are not necessarily excitonic, since excited states
Chapter 2. Excitonic Solar Cells might be able to thermally separate into free electrons and holes. However, since the absorber coating is thin, similar to DSCs, exciton diffusion does not play a role even if exciton separation occurs only at interfaces. Nevertheless, an important difference to DSCs is that charge traps in the inorganic absorber can promote charge carrier recom-bination, which is believed to be the reasons why ETA-SCs to date operate with VOCs that are significantly lower than the theoretical built-in potentials.[115, 116]
A relatively new class of ETA-SCs features organic-inorganic perovskites as absorbers.
Impressive efficiencies beyond 9 % were reported in 2012 by the groups of Snaith and Grätzel.[117, 118] The Grätzel group used a (CH3NH3)PbI3 perovskite on TiO2 nanoparticle film with Spiro-OMeTAD as HTM, yielding PCEs of9.7 %. A similar ap-proach was used by the Snaith group for the mixed halide perovskite (CH3NH3)PbI2Cl.
Interestingly, the latter perovskite exhibits n-type properties and gave even higher effi-ciencies, up to10.9 %, if used on an insulating network of Al2O3 nanoparticles instead of TiO2. Further optimization of these devices by the Grätzel group lead to certified record efficiencies beyond 14 % in 2013. Technically, these devices are not ETA-SC, since the perovskite can play a dual role as absorber and n- or p-type electrode.
2.3 Extremely Thin Absorber Solar Cells
3 Working Mechanisms of
Nanostructured Hybrid Solar Cells
This chapter gives a more detailed overview over the working principles of solid-state dye-sensitized and hybrid solar cells. Loss mechanisms in these devices are discussed and work on nanostructured metal oxide electrodes, which potentially help to overcome these loss mechanisms, is summarized. The chapter is based on the book chapterSolid state dye-sensitized solar cells (J. Weickert and L. Schmidt-Mende, 2013), the book chapter Metal Oxides: New Science and Novel Applications(J. Weickert and L. Schmidt-Mende, 2013), the review article Nanostructured Organic and Hybrid Solar Cells (J. Weickert et al., Advanced Materials 2011) and the perspective article Hybrid Solar Cells: How to Get the Polymer to Cooperate? (J. Weickert and L. Schmidt-Mende, APL Materials 2013).[1, 96, 119]
3.1 Light Absorption and Charge Separation in Solid-State Dye-Sensitized Solar Cells
Besides sufficiently thick dye-sensitized films the absorption profile of the dye itself is crucial for a high photon harvesting efficiency. Since solar cells are operated under illumination with sunlight, the absorption has to be optimized especially in spectral regions where the power input of the sun on the earth is strong, i.e., in the visible and near IR. Since light absorption takes place only in a monolayer of dye, absorption coefficients have to be as high as possible. Over the past 20 years, tremendous effort has been put into increasing the extinction of DSC dyes and efficiency improvements could be mostly attributed directly to more complete light harvesting. Nevertheless, still mesoporous films of10−20µmthickness are necessary in order to absorb almost all incident photons covered by the absorption spectrum of the dye.
Many dyes for SS-DSCs are well-known from applications in conventional DSCs. Fig-ure 3.1 shows the chemical structFig-ures of commonly used dyes. N3 is a Ru complex dye and was the first material, which allowed to realize PCEs beyond10 %in conventional DSCs since it absorbs strongly in the visible (13900 l mol−1cm−1 at 541 nm) and up to800 nm.[76] Owing to its absorption spectrum it reaches theoretical photocurrents of more than25 mA cm−2 under AM 1.5G solar illumination.[120] Today, there is a huge variety of N3 derivatives as nicely summarized in the DSC review article by Hagfeldt et al.[18] Figure 3.1 also shows chemical structures of the two most important N3-based dyes, N719 and Z907. N719 is today’s standard dye for application in conventional DSCs, whereas Z907 has proven to be better suited for SS-DSCs.
3.1 Light Absorption and Charge Separation in Solid-State Dye-Sensitized Solar Cells
Spiro-OMeTAD
N3
Z907
N719
Figure 3.1: Chemical structures of the most common hole conductor for SS-DSCs Spiro-OMeTAD and important Ru dyes.N3 was the first Ru dye which allowed efficiencies of about 10 % in conventional DSCs. To date, its derivative N719 is the most widely used dye for DSC applications and established as a standard material. For SS-DSCs, the Z907 dye has proven more efficient due to its two alkyl side moieties, which serve as physical spacers and help to reduce charge carrier recombination in these solar cells.
indoline dye
squarine dye
Figure 3.2: Chemical structures of two typical organic dyes, an indoline dye and a squaraine dye.Organic dyes have proven highly efficient and show performances even superior to Ru metal complex dyes in SS-DSCs.
Besides metal complex dyes, fully organic metal-free dyes have emerged during the past years. Organic dyes like indoline or squaraine dyes exhibit even higher extinction coefficients than their conventional counterparts and allow PCEs comparable to val-ues achieved with Ru dyes in liquid electrolyte-based DSC.[120, 121] Typical chemical
Chapter 3. Working Mechanisms of Nanostructured Hybrid Solar Cells structures of two important classes of organic dyes are exemplarily shown in Figure 3.2.
In SS-DSCs, even higher performances were possible when using organic dyes instead of metal complex materials, suggesting that this class of dyes is especially well-suited for applications with solid state hole transporters.[101, 122]
hn
TCO TiO2 dye Spiro-OMeTAD Au/Ag e-
h+
Einj
Enet E
Ereg
Figure 3.3: Schematic representation of the mechanism of charge separation in a SS-DSC.Light is absorbed in the dye and an exciton is formed. Provided there are sufficient energy offsets, this exciton can then be separated either via electron injection into the TiO2or hole tranfer to the HTM. Free electron (e-) and hole (h+) are subsequently transported through the TiO2 and the HTM, respectively, and collected at the external contacts.
Since the dye is located directly at the TiO2-HTM interface, it is not only responsible for light absorption, but also plays a crucial role in the mechanisms of charge separation, which is schematically shown in Figure 3.3.[33, 103] Following to the absorption of a photon in the dye chromophore, an excited state is forming within the dye. This excited state can either directly separate into free charges via ultrafast electron injection into the TiO2 or relax to the lowest excited energetic state via emission of a photon. The binding energy of the resulting excitons is typically much larger than kBT (in the order of 0.3 eV), so charge separation cannot be thermally induced as in the case of conventional Si solar cells. Instead, the energy necessary for charge separation is provided by the energy difference between the lowest unoccupied molecular orbital (LUMO) of the dye and the conduction band of the TiO2. This energy is the minimum driving force for electron injection, denoted as Einj. Electron injection from the LUMO of the dye into the conduction band of TiO2 happens on timescales of ps or fs, which is extremely quick compared to the lifetime of excited states in the dye, which is in the order of 20−60 ns.[123–127] After electron injection, the positively charged dye molecule is regenerated by the HTM, i.e., the hole residing in the highest occupied molecular orbital (HOMO) of the dye is transferred to the HOMO of the HTM. This transfer is driven by the energy difference between the two HOMO levels and is denoted as Ereg.
3.1 Light Absorption and Charge Separation in Solid-State Dye-Sensitized Solar Cells In addition to the described mechanism of charge separation it is also possible that the hole transfer occurs quicker than the electron transfer. In conventional DSCs, dye regeneration via the liquid electrolyte is a two electron process, which is significantly slower than electron injection from the dye’s LUMO into the conduction band of the TiO2. In SS-DSCs, however, where the HTM is a solid organic material, the hole transfer can also be very quick.[89, 128–130] The charge separation is then mainly driven by Ereg. The final state, where the electron has reached the conduction band of the TiO2 and the hole is in the HOMO of the HTM allows a net energy conversion of Enet as shown in Figure 3.3. In a non-kinetic case Enet/q is also the maximum open circuit voltage (VOC) that can be generated by the solar cell (where q is the elementary charge). Since this voltage depends on Ereg, it can be tuned by shifting the HOMO of the HTM with respect to the HOMO of the dye. This is an advantage of SS-DSCs compared to liquid electrolyte DSCs, since the properties of the HTM can be chemically adjusted in order to minimize potential losses. However, the commonly used I-/I3- electrolyte in conventional DSCs exhibits electronic properties that lead to relatively large values of Eregfor most common dyes. Therefore, once optimized, higher values ofVOC should be possible for SS-DSCs compared to their liquid electrolyte counterparts.
Even though the final state is the same for both pathways of charge separation, there is a huge difference depending on which type of charge carrier is transferred first. If electron injection is the quicker mechanism, i.e., the electron is in the conduction band of the TiO2 and the hole remains in the HOMO of the dye, the Coulomb interaction between electron and hole is weak due to the high dielectric constant of TiO2. Accordingly, spatial separation of the charge carriers is relatively easy. On the other hand, if the hole is transferred first, both charge carriers reside in organic materials with rather low dielectric constants resulting in a strong Coulombic attraction between them. In this case, the probability of recombination of the resulting electron-hole pair is enhanced and spatial separation becomes less likely.[131] Therefore, not only HOMO and LUMO levels are important, but also their location inside the dye and the chemical structure of the dye, e.g. the presence of spacing moieties like alkyl chains. In so-called push-pull dyes, in the excited state the LUMO is located close to the TiO2, whereas the HOMO points towards the HTM.[89, 132] Thus, electron injection and subsequent complete charge separation becomes more likely since electron and hole are already partly separated in the excited state of the dye. It should be noted, however, that push-pull dyes are optimized for liquid electrolyte DSCs, i.e., for a situation, where electron injection is much quicker than hole transfer to the electrolyte as HTM. If the latter occurs first, the LUMO location of the dye has to be close to the conduction band of the TiO2 even if the dye is in its reduced state, i.e., negatively charged after the hole is transferred.
It might therefore be necessary to design new types of push-pull dyes especially for applications in SS-DSCs.
Additionally, physical spacers like alkyl side chains can influence the time constants for charge separation and recombination. For SS-DSCs, where the hole transfer from the dye to the HTM is quick, also the back reaction becomes more likely, i.e., hole move-ment from the HTM to the HOMO of the dye inducing charge carrier recombination.
Therefore, Z907 is the better suited N3 derivative than N749, the standard in liquid
Chapter 3. Working Mechanisms of Nanostructured Hybrid Solar Cells
Time
a) b)
c)
TiO2
HTM
Ag/Au
dye
TCO
Figure 3.4: Possible mechanisms of charge separation in HSCs. a)Light is absorbed in the dye and charges are separated either via electron injection into the TiO2or hole transfer to the HTM.
b)Upon light absorption in the HTM, an exciton is forming. The excited state then travels into the dye molecule via an energy transfer from the HTM to the dye and charges are separated as in case a). c)Excitons that are formed in the HTM can also be directly separated between HTM and dye or HTM and TiO2.
electrolyte DSCs, due to its alkyl side chains, which point away from the TiO2 and build a physical spacer between electrons in the TiO2 and holes in the HTM.[133]
In contrast to SS-DSCs with transparent HTM there is another possible mechanism of charge separation in HSCs. Since light is not only absorbed in the dye but mainly in the hole conducting polymer, excitons are also generated in the HTM. Possible routes towards charge separation in HSCs are depicted in Figure 3.4. If light is absorbed in the dye, as in Figure 3.4 a), charge separation is initialized either by electron injection into the TiO2 or hole transfer into the HTM as described for SS-DSCs (with trans-parent HTM). However, light absorption in the HTM itself can result in two different mechanisms of charge separation. Necessary for this, of course, is that the exciton is generated within the exciton diffusion length away from the dye-decorated TiO2 - oth-erwise, the exciton recombines directly without separating into free charge carriers. For matching energy levels and close proximity of the excited region and the dye molecules, the excited state can be transferred from the HTM into the dye molecule as depicted in Figure 3.4 b). The underlying mechanism of Förster Resonance Energy Transfer (FRET) demands for overlapping emission and excitation spectra of the polymer and the dye, respectively, and is conceivable, e.g., for a visible light absorbing HTM and a near-IR dye.[134–136]
As shown in Figure 3.4 c) it is also possible that charges are separated directly between the HTM and the dye, i.e., an electron is transferred from the LUMO of the HTM into