Tải bản đầy đủ - 0 (trang)
5  Dye-Sensitized Solar Cells

5  Dye-Sensitized Solar Cells

Tải bản đầy đủ - 0trang

Solar Photovoltaics



251



for the development of dye-sensitized solar cells (DSSC), an area of research that has

been growing ever since (Meyer 2010).

How are DSSCs different from the solar cells we have examined thus far? Here’s

a quick recap:

• In crystalline silicon solar cells, a bilayer architecture between a p-type

material and an n-type material sets up the photocurrent via a bandgap promotion of an electron from the valence band to the conduction band. Both

materials are essentially the same (a homojunction).

• In inorganic thin-film solar cells, the layer architecture and mechanism are

similar to silicon cells, but heterojunctions between different p- and n-type

materials are the norm.

• In organic solar cells, the p- and n-type inorganic semiconducting materials

are replaced with organic materials: a conducting polymer donor (or electronrich small molecule) and a fullerene acceptor. Band theory no longer applies;

the pertinent energy differential is between HOMO and LUMO energy levels.

The bilayer architecture is replaced with the interspersed bulk heterojunction.

• In DSSCs, the number of working materials increases and both inorganic

and organic materials are part of the cell’s key components. An organic

or organometallic dye sensitizer is used to absorb the light and supply an

excited state electron to a metal-oxide semiconductor. A redox couple (commonly referred to as the electrolyte) is required to regenerate the dye. The

electrons and holes are conducted to the appropriate electrodes as usual.

DSSCs are basically our attempt to mimic photosynthesis: instead of chlorophyll in

a leaf absorbing light energy and injecting an electron into an electron transport chain,

a synthetic dye is used in a solar cell with the result being generation of electricity

instead of synthesis of carbohydrates. In conventional PV processes, the semiconductor is both the light absorber and the charge carrier. In DSSC, the absorber (the dye)

and the charge carrier (the metal oxide) are separate species. In this section, we will

focus on n-type DSSCs, where the semiconductor involved is an n-type semiconductor. DSSCs in which the sensitized semiconductor is a p-type material have also been

developed, but we will leave those to the interested reader to pursue independently.



7.5.2 Architecture

The DSSC is built up in layers upon a glass substrate just as for other solar cells, but

the addition of the dye sensitizer makes the architecture slightly more complex, as

shown in Figure 7.56. The glass substrate is coated with a transparent conducting

oxide (TCO) [indium tin oxide (ITO) or fluorine-doped tin oxide (FTO)]; the glass/

TCO combination makes up the anode. The metal-oxide semiconductor is deposited

on the anode to a thickness of about 4–10 μm (Meyer 2010). It will come as no

surprise to learn that the nanoscale morphology of the metal-oxide layer is important; a mesoporous titanium oxide is the most commonly used semiconductor. A

mesoporous solid is intermediate in porosity between macroporous (>50 nm) and



252



Chemistry of Sustainable Energy

Pt contact



TiO2

Dye



Redox couple

I3–



Incident light



3I







Glass/TCO cathode



Glass/TCO anode



Electron

diffusion

Ion

diffusion



Load

External current



FIGURE 7.56  Schematic of a typical DSSC.



microporous (<2 nm) materials (Rouquerol et al. 1994). Applied on top of the metal

oxide is a thin (ca. 10 μm) layer of the light-absorbing dye, of which there are many

variants. By using a mesoporous semiconductor material, the surface area of the

dye is increased enormously, markedly improving the light-harvesting ability of the

device. However, the thickness of the dye must be optimal for a good balance of

light harvesting and electron injection to the semiconductor. If the layer is too thick,

the electron injection is impeded. An atomic monolayer of dye sensitizer is ideal for

good diffusion of the exciton to the metal oxide. Some kind of redox electrolyte—

typically the I−3 /I−  couple—is applied next, followed by the cathode. In the case

of DSSC, a “stronger” working electrode is required, thus platinum-coated glass is

often used to collect the electrons. More detail on the TCO, dye, and redox mediator

is presented in Section 7.5.4.



7.5.3 Mechanism

How the various materials work together in a DSSC to create a photocurrent is

graphically illustrated in Figure 7.57.

• Absorption. First and foremost, the incident light excites the dye molecules

to an excited state (1), indicated by S* (S because the dye molecules are the

sensitizers in a DSSC).

• Injection. The excited dye molecule can then inject an electron [on the picosecond time scale, (2)] into the conduction band of the metal-oxide semiconductor (in this example, TiO2) (Listorti et al. 2011).



253



Solar Photovoltaics

(2)

Electron

injection



S*/S+



TiO2

CB

(3)

Hole

transfer



Energy



Fermi level

(1)

+ hν



VOC



I3– / I–



S/S+



Dye regeneration



Dye



FIGURE 7.57  Diagram of the conduction process in a DSSC.



• Collection. The circuitous path through the metal-oxide semiconductor to

the anode must be fast. The mesoporous structure means that the diffusion

path is relatively long; failure to diffuse quickly to the electrode leads to

recombination and back-electron transfer mechanisms.

• Regeneration. Upon injection of an electron into the mesoporous TiO2,

the dye becomes oxidized (as indicated by S+) and the hole is transferred to the redox couple (3). Something must be present in the cell to

reduce S+  back to the ground, unoxidized state. In this figure, I– plays

that role, rapidly (≈10 μs) reducing the dye (Boschloo and Hagfeldt

2009). Reduction of the oxidized dye must take place fairly quickly lest

a buildup of S+ results increasing the likelihood of back-electron transfer

to the oxidized dye. More detail regarding this step of the mechanism is

presented in Section 7.5.4.3.

• The I3−  anion is reduced back to I– by electrons collected at the platinum

electrode to complete the electrochemical circuit.

Key to the success of the DSSC is appropriate matching of the energy levels

of the metal-oxide semiconductor conduction band, the HOMO–LUMO of the dye

sensitizer and the redox potential of the mediator. The maximum voltage that can

be obtained is based on the Fermi level of the metal oxide and the redox potential of

the I3− /I−  couple. Given the complexity of the DSSC, the potential for inefficiencies

are legion. For example, the excited state dye can decay and the injected electron can

recombine with the oxidized dye or the I3−. Despite these issues, the efficiencies of the

DSSC are above 10% and increasing.



254



Chemistry of Sustainable Energy



7.5.4 Materials

7.5.4.1  Metal Oxide

Tin, zinc, and titanium oxides are three wide bandgap semiconductor metal oxides

that have found widespread use in DSSC and Figure 7.58 gives their HOMO and

LUMO energies. It is titania (TiO2), however, that is far and away the most commonly used n-type semiconductor. As noted above, the morphology of the TiO2 layer

is a crucial aspect, since the high surface area of the metal-oxide layer—and, consequently, the dye—is required for efficient light harvesting. However, there is much

more to the morphology of TiO2 than meets the eye and the progression from the

empirical study of TiO2 performance in DSSCs to development of a mechanistic

understanding is taking place both theoretically and empirically.

Titanium dioxide exists in nature in one of three forms: rutile is the most common

and the most stable form, where the oxygen atoms are octahedral about the Ti atom.

The other two forms—anatase and brookite—exhibit distorted octahedra about Ti,

with the most stable form of anatase existing as a tetragonal bipyramid (Lazzeri et al.

2001). Evidence suggests that anatase is the most photocatalytically active form of

TiO2 and, therefore, the best choice for use in DSSC and a commercial product that

is primarily anatase is available. A method for the controlled synthesis of anatase

TiO2 in either microflowers or microspheres is shown in Figure 7.59 (Li et al. 2013).

Other syntheses of anatase nanocrystals often make use of the sol–gel method. In

the sol–gel method, a precursor material (often Ti(IV) isopropoxide) is hydrolyzed

and heated in the appropriate solvent(s) to generate a colloid which is then refluxed to



–0.05



TiO2



Energy (volts) vs. NHE



ZnO

0.00

0.05



3.2 eV



SnO2



3.2 eV



3.00

3.8 eV

3.50



FIGURE 7.58  The wide bandgaps of metal-oxide semiconductors used in DSSC. (Adapted

from Renew. Sustain. Energy Rev., 16, Gong, J., J. Liang, and K. Sumathy, Review on dyesensitized solar cells (DSSCs): Fundamental concepts and novel materials. 5848–5860.

Copyright 2012 with permission from Elsevier.)



255



Solar Photovoltaics



Ti4+ + H2O2



O

O



O



O



O



O



Ti



Ti



O



O



O



O



O



O



H2O



O



O



O



Condensation O



Ti



O



O



O

O



O

O i.e.,



O

Ti

O



O

O



Ti

O



O



O

O



H2O Condensation



O

Ti



O

O

O



O

Ti



O



O

O

Ti

O



i.e.,



Growth



O



O



O



(Anatase)



FIGURE 7.59  An overview of controlled anatase synthesis. (Reprinted with permission

from Li, T., B. Tian, J. Zhang et al. 2013. Facile tailoring of anatase TiO2 morphology by

use of H2O2: From microflowers with dominant {101} facets to microspheres with exposed

{001} facets. Ind. Eng. Chem. Res. 52 (20):6704–6712. Copyright 2013 American Chemical

Society.)



generate the TiO2 nanoparticles as a gel (Isley and Penn 2006). Coating the gel onto

the TCO/glass substrate requires subsequent heating to sinter the particles together

slightly for electrical contact between the particles. Small changes in the synthetic

parameters—solvent, temperature, pH, and so on—can result in substantial changes

in the shape, surface, size, and, therefore, properties of the resultant TiO2 nanoparticles and the PV device.

Much more elaborate methods continue to be developed for preparation of more

strictly controlled TiO2 morphologies. Templating, that is, growing the TiO2 film in

the presence of something that will cause it to grow with a defined porosity, can lead

to mesoporous structures of well-defined and -controlled pore sizes. For example, by

using a nonionic block copolymer made up of ethylene oxide and propylene oxide

repeat units (Figure 7.60), miscelles form in the mixture upon which the mesoporous

structure can build. A three-layer film of this templated mesoporous TiO2 was found

to give an enhancement of about 50% in the photon conversion efficiency due to

a significant increase in the short-circuit photocurrent, likely a result of increased

light-harvesting attributable to the improved porosity of the titania (Zukalová et al.

2005).

New methods of synthesis have been developed that allow the controlled growth

of metal-oxide semiconductors in various shapes, sizes, and with varying facets to

maximize photocatalytic activity, and XRD and electron microscopy have been

used extensively to characterize these materials and morphologies. It was known

that rod-like anatase gave better performance in a DSSC than ball-shaped anatase

(6.2% photoelectric conversion efficiency vs. 5.4%) (Wu et al. 2007) and two facets

of the anatase crystal—{001} (flat) and {101} (somewhat corrugated)—have been of



256



Chemistry of Sustainable Energy



O

*



O



O



O



O



Ethylene oxide block



O



*



Propylene oxide block



FIGURE 7.60  A representative ethylene oxide–propylene oxide block copolymer.



particular interest in unraveling the source of the photocatalytic activity (Figure 7.61)

(Lazzeri et al. 2001). The synthesis of nanocrystals of anatase from TiCl4 or TiF4

(Figure 7.62) can be carried out to give a preponderance of either {001} or {101} facets depending upon the surfactant used as seen in the TEM images shown in Figure

7.63. Research is ongoing to determine which shape of nanocrystal gives the best

results for a DSSC (Gordon et al. 2012).

Even more elaborate high-surface-area structures for the metal-oxide semiconductor dye support have been investigated. “Bamboo-like” ridged TiO2 nanotube

arrays were prepared by the relatively mild anodic oxidation of titanium foil in the



{001}

{101}



FIGURE 7.61  Descriptors for facets of the anatase crystal structure.



HC

TiX4 + surfactant + CH3(CH2)7



O



CH

(CH2)7



C



CH3(CH2)15CH=CH2

OH



X = F, Cl



NH2

Surfactant



or

OH



FIGURE 7.62  Synthetic preparation of anatase nanocrystals.



Anatase

nanocrystals



257



Solar Photovoltaics

TiF4



Cosurfactant

ROH

RNH2



(a)



50 nm



(d)



001

101



(b)



50 nm



(e)



100 nm



50 nm



TiF4 + TiCI4



TiCI4



(c)



50 nm



(f )



50 nm



FIGURE 7.63  Impact of surfactant and halide on TiO2 crystal shape; a–f represents different

combinations of cosurfactant and titanium halide precursor. (Reprinted with permission from

Gordon, T.R., M. Cargnello, T. Paik et al. 2012. Nonaqueous synthesis of TiO2 nanocrystals

using TiF4 to engineer morphology, oxygen vacancy concentration, and photocatalytic activity. J. Am. Chem. Soc. 134 (15):6751–6761. Copyright 2012 American Chemical Society.)



presence of an ethylene glycol-based electrolyte. The relevant chemical transformations are shown in Equations 7.11 through 7.13 (Luan et al. 2012).









H 2 O → 2H + + O2 − (decomposition of water )



(7.11)



Ti + 2O2 − → TiO 2 + 4e − (metal oxidation )



(7.12)







TiO 2 + 6 F − + 4H + → TiF62 − + 2H 2 O (oxide dissolution )







(7.13)



As indicated in Equation 7.13, the growing TiO2 nanotube is dissolved in the

presence of fluoride ions; by alternating the voltage, pH, and electrolyte composition, the TiO2 nanotube growth can be controlled to make the sectioned bamboolike structure with pronounced ridges between the sections. The nanotubes were

converted into anatase TiO2 by annealing in air at 450°C as confirmed by the XRD

results shown in Figure 7.64; an SEM image of an array is shown in Figure 7.65.

Apparently, the rough edges of the nanotubes increased their ability to adsorb the

dye sensitizer. Smooth-walled TiO2 nanotubes gave a PCE of 3.90%, while the same

length bamboo-like nanotubes showed a PCE of 5.64% (Luan et al. 2012).

A final example of titania morphology that has shown considerable promise

for DSSC is that known as the inverse opal. The inverse opal morphology can be

described as the fused spherical cavities that are left behind after templating a material on close-packed spheres. Figure 7.66 graphically illustrates the process, and

Figure 7.67 shows a scanning electron micrograph of such a structure prepared by

dipping a substrate templated with close-packed polymethylmethacrylate spheres

into a TiO2 precursor solution. The inverse opal voids were then filled with mesostructured titania. A DSSC prepared with this TiO2 showed a 10-fold improvement

in PCE over the regular mesoporous titania (Mandlmeier et al. 2011).



258



Chemistry of Sustainable Energy



Intensity/au



Anatase TiO2

Ti



20



450°C annealed TiO2 nanotube



As-prepared TiO2 nanotube



30



40



50

60

2 theta/degree



70



80



FIGURE 7.64  XRD results for bamboo-like TiO2 showing transformation to anatase upon

annealing. (Reprinted with permission from Luan, X., D. Guan, and Y. Wang. 2012. Facile

synthesis and morphology control of bamboo-type TiO2 nanotube arrays for high-efficiency

dye-sensitized solar cells. J. Phys. Chem. C 116 (27):14257–14263. Copyright 2012 American

Chemical Society.)



FIGURE 7.65  Scanning electron micrograph showing cluster of bamboo-like TiO2.

(Reprinted with permission from Luan, X., D. Guan, and Y. Wang. 2012. Facile synthesis and

morphology control of bamboo-type TiO2 nanotube arrays for high-efficiency dye-sensitized

solar cells. J. Phys. Chem. C 116 (27):14257–14263. Copyright 2012 American Chemical

Society.)



259



Solar Photovoltaics



Opal template



Add metal oxide



Remove template

Mesoporous inverse

opal material



FIGURE 7.66  Process for synthesis of templated inverse opal TiO2. (Adapted with permission from Wu, G., Y. Jiang, D. Xu et al. 2010. Thermoresponsive inverse opal films fabricated with liquid-crystal elastomers and nematic liquid crystals. Langmuir 27(4): 1505–1509.

Copyright 2011 American Chemical Society.)



FIGURE 7.67  An SEM image of the inverse opal TiO2 scaffold. (Adapted with permission

from Mandlmeier, B., J. M. Szeifert et al. 2011. Formation of interpenetrating hierarchical

titania structures by confined synthesis in inverse opal. J. Am. Chem. Soc. 133(43): 17274–

17282. Copyright 2011 American Chemical Society.)



260



Chemistry of Sustainable Energy



7.5.4.2  Dye Sensitizer

As with any solar cell, maximizing the capture of photon flux from the solar spectrum is an important goal and in the DSSC, this is the role of the dye. The ideal dye

should absorb as much sunlight as possible, ideally capturing all wavelengths below

1000 nm. The LUMO level of the dye must be high enough for thermodynamically

favorable electron transfer to the metal oxide, and for regeneration of the dye the

reduction potential of the oxidized sensitizer must be more positive than that of the

redox mediator. Of course, the dye needs to be stable in sunlight and at elevated

temperatures, but with DSSCs comes a new requirement: the dye should be functionalized with a group that allows it to anchor strongly to the metal-oxide surface in a

tight monolayer to prevent access of the redox mediator to the metal oxide. The most

common covalent linker is the carboxyl group, but phosphonates, sulfonates, and even

borates and silyl linkers (Figure 7.68) have been used (Hagfeldt et al. 2010).

Octahedral ruthenium pyridyl systems are the historical dye sensitizers, with the

“gold standard” for DSSC being the so-called N3 dye, shown in Figure 7.69; the

O



O

R



O



R



P



O

R



O



OH

Carboxylate



Phosphonate



S



R

R



O



OH



Si

R



O



R



Sulfonate



Silane

R = C, H



B



O



Borate



FIGURE 7.68  Functional group classes used as linkers for DSSC.

CO2H



HO2C



CO2X



N

N



N

Ru



N



N

C



N



S



XO2C



C

S

X=H



N3



X=nBu4N



N719



FIGURE 7.69  Ruthenium dyes for DSSCs; N3 = cis-bis(isothiocyanato)bis(2,2′-bipyridyl4,4′-dicarboxylato)ruthenium(II) dye, a.k.a. N3.



Solar Photovoltaics



261



analogous tetra-n-butylammonium salt is the widely used N719 dye. The carboxyl

groups on the bipyridyl ligands react with hydroxy functionality on the metal-oxide

surface to anchor the dye. The isothiocyanate ligands (−N=C=S) on ruthenium

increase the energy level of the HOMO so that the complex has a red-shifted absorption to a longer wavelength (Hagfeldt et al. 2010). The primary reason these dyes

work so well for DSSC is because of the fact that, upon absorption of a photon and

promotion of an electron to the dye LUMO, the metal complex can shift electron

density between the metal center and the ligands to help stabilize the excited complex. This is known as a low-energy metal-to-ligand charge transfer state. Because

of this ability to stabilize the excited state, many other successful Ru dye sensitizers

have been designed, including the broadly successful “black dye” (Figure 7.70). As

Figure 7.71 illustrates, these dyes are much more effective at absorbing light through

the visible region into the near IR than the wide bandgap material TiO2, leading to

the high efficiencies characteristic of the DSSC.

Ruthenium is one of the Earth’s rare metals (Section 1.2.3), so use of the Ru dyes

adds to the cost of the solar cell and its extensive use is not sustainable. Other metal

complexes have been explored as alternatives (osmium, rhenium, iron, platinum,

and copper) and research into alternatives proceeds apace (Hagfeldt et  al. 2010).

Porphyrins and phthalocyanines are good candidates because of their overall stability and excellent absorption in the near-IR region (>800 nm). More inexpensive and

abundant metals like zinc have also yielded promising results as sensitizers. The zinc

phthalocyanine in Figure 7.72, for example, showed an IPCE of 45% in the near-IR

(Nazeeruddin et al. 1999).

Another alternative to rare, toxic, or expensive metals is to avoid them altogether,

and the use of metal-free dyes is increasing. These organic dyes offer immense structural diversity subject to design and synthesis, leading to:

• Tunable HOMO–LUMO levels

• Very high molar absorptivity

• Broad absorption spectrum into the far red and near-IR portion of the electromagnetic spectrum

• Low cost and low toxicity

• Sustainability

There are several classes of metal-free dyes and, not suprisingly, several share

structural features with the conducting polymer donor materials used in OPVs

(triarylamines, phenylvinylenes, fluorenes, thiophenes, and pyrrole derivatives).

A representative example from a few of the more common classes of metal-free

dyes is given in Table 7.2. Most possess a general donor–π system–acceptor framework, injecting the electron into the semiconductor through the acceptor group.

Natural dyes (e.g., porphyrins and anthocyanin-based flavonoids (Figure 7.73)

have also been explored. For example, a solar cell capable of achieving a Jsc of

1–2 mA/cm2 (in bright sunlight) can be constructed from the natural anthocyanin

dyes found in a variety of berries (Smestad and Grätzel 1998; Wang and Kitao

2012). Extraction of natural dyes sidesteps the potential agony of synthesis, but also



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

5  Dye-Sensitized Solar Cells

Tải bản đầy đủ ngay(0 tr)

×