3 Indoline sensitizers on ZnO
3.1 Absorbance of ZnO films sensitized with indoline dyes
The sensitization of electrodeposited ZnO with the indoline dyes D149, DN91, DN216 and DN285 and a coadsorbate led to intense-colored films, especially for sensitization times of 15 min and 1 h.
Similar observations were made for D149 on the highly porous electrodeposited ZnO films before 16-19. The sensitization for 1 min was too short to achieve a complete dyeing of the film for some of the samples (films [45] and [46]), which was visible by a more whitish appearance of the film from the substrate side. Looking from the film side of the samples, the intensity of the coloring observed by eye increased with increasing sensitization time.
To characterize the adsorbed dye in more detail, the absorbance of the sensitized films was measured before the semiconductor films were assembled to solar cells. As described in detail in 2.4.3, the films were measured with an integrating sphere in transmission setup to gather also scattered light from the strongly-scattering ZnO films, and the absorbance of the non-sensitized substrate is subtracted from the absorbance of the sensitized films. As the scattering of the films also influences the absorbance values, the different scattering of films is briefly discussed prior to the absorbance of the sensitized films vi.
v Similar but slightly different coadsorbates were chosen to ensure a better comparability to the D149-sensitized films of Melanie Rudolph 16,17,127 or the cells sensitized with double anchor dyes in the work of Felix Fiehler 131.
vi Note that “absorbance” in this work includes also backward scattered and reflected light. Usually the absorbance spectra contain only the absorbance of the dye, as the absorbance of the substrate and of the ZnO
3.1.1 Scattering of the semiconductor substrate
Using a simple transmission setup, the absorbance (including scattering and reflection) of different porous films was determined, see Figure 19. For a nanoparticulate TiO2 film, the absorbance or scattering of the film can be simulated assuming Rayleigh scattering, which speaks for a uniform particle size. For an electrodeposited ZnO film, the addition of a linear contribution to the Rayleigh scattering is needed to achieve a good match of simulated and measured data especially at longer wavelengths vii. For a nanoparticulate ZnO film, also an additional linear contribution is observed for the scattering, even though this contribution is smaller than for the electrodeposited film. This deviation from a pure Rayleigh scattering indicates that also other contributions to scattering are present especially for the electrodeposited ZnO films. As seen from SEM cross-sectional images (see for example 3), the oxygen-based electrodeposition of ZnO with EosinY leads to cauliflower-like structures with more dense “stems” and very fine “branches”. Larger structures within the film can lead to additional Mie scattering, which occurs at larger particles, which would then explain the observed deviation from Rayleigh scattering. As the nanoparticulate ZnO film also shows additional Mie scattering, it will also contain larger structures, probably aggregated particles viii. The very good match of the fit with pure Rayleigh scattering of the absorbance of the nanoparticulate TiO2 film indicates a good film quality without larger aggregates of particles, as expected from the preparation procedure. The different structure of the films is not only important for the absorbance or scattering, but it also influences the electrochemical behavior discussed in the following sections, especially due to a different diffusion in differently sized pores.
Even though the strong light scattering of the electrodeposited ZnO films complicates the measurement and evaluation of the absorbance (see also 9.1.1, p. 187), it has advantages for the application of such films in a solar cell. Light will travel a longer way through a highly scattering film than for less scattering films, and thus the probability of absorption by a dye molecule is increased.
That means that for the already strongly scattering electrodeposited ZnO films no additional layer with larger particles is needed, in contrast to more transparent TiO2 films, where often a second layer of larger particles is added to the main layer to increase light scattering 119.
vii The undulation of the graph between 450 and 700 nm can probably be ascribed to the uniform film thickness and thus to interference of light, and not to remnants of EosinY, as the desorption procedure leads to completely discolored films. For lower wavelengths, the deviation of the simulation is most likely caused by the use of a constant value of the refractive index n.
viii The less stable material ZnO gives more probability to differently sized structures than TiO2.
400 600 800 1000 0
1 2
nanoparticulate TiO2 [TiO2-01]
electrodeposited ZnO [35]
nanoparticulate ZnO
Rayleigh scattering + linear contribution
Rayleigh scattering
Absorbance
Wavelength / nm
Figure 19 – Absorbance of different non-sensitized films (see legend for exact designation) measured in a simple transmission setup (without integrating sphere) to obtain scattering. Measurement data are shown as thick dashed lines, while simulations of the scattering are shown as thin solid lines. The absorbance of the substrate was subtracted from the measured absorbance, leaving only the absorbance of the porous films. [TiO2-1] and [35] are the sample IDs of the non-sensitized films. The values for the refractive index for TiO2 and ZnO were used from 248 and 249, respectively.
3.1.2 Influence of the adsorption time and the sensitizer on the absorbance
The absorbance of differently sensitized electrodeposited ZnO films is shown in Figure 20. Due to the problems in measurement and evaluation which arose from the strong scattering of the electrodeposited ZnO films (see detailed discussion in 9.1.1, p. 187), only films sensitized for 1 min and absorbances in the wavelength region between ca. 600 and 650 nm are considered for the following discussion. The absorbance at longer wavelengths can be taken as a rough measure of the amount and aggregation of dye molecules within the films, and values of the absorbance at 635 nm are given in Table 4.
The direct comparison of the absorbance of films sensitized for 1 min with the different indoline dyes, Figure 20(a), shows a very similar curve shape and a comparable height of the absorbance maximum, especially for the dyes containing two anchor groups. For all four sensitizers, the maximum of the spectra is located between 525 nm and 540 nm. The spectrum of the film D1491min
CA [45], which is sensitized with the single-anchored dye D149, differs slightly in shape and position, as it is shifted to lower wavelengths, and the absorbance around 500 nm is slightly increased compared to other dyes.
Such a shape where the left side of the absorbance maximum is slightly higher than the left side, is commonly found for D149 adsorbed on ZnO 16,121,127. For the double-anchor dyes, the absorption maxima are more symmetric, or the right side is slightly higher compared to D149. The spectra resemble more the solution spectra of D149 and DN216, which are also included in Figure 20(a).
400 500 600 700 0.0
0.5 1.0 1.5 2.0
D1491minCA [45]
DN911minLCA [46]
DN2161minLCA [55]
DN2161minLCA [59]
DN2851minLCA [56]
DN2851minLCA [57]
solution of D149 solution of DN216
Absorbance (integrating sphere)
Wavelength / nm
400 500 600 700
0.0 0.5 1.0 1.5 2.0
DN9115minLCA [43]
DN9115minLCA [60]
DN21615minLCA [52]
DN28515minLCA [34]
DN28515minLCA [63]
D14915minCA [61]
DN9115minLCA [35]
Absorbance (integrating sphere)
Wavelength / nm
400 500 600 700
0.0 0.5 1.0 1.5 2.0
D1491hCA [65]
DN911hLCA [69]
DN2161hCA [68]
DN2851hLCA [71]
Absorbance (integrating sphere)
Wavelength / nm
Figure 20 – Absorbance of dyes adsorbed to porous ZnO films, grouped after the sensitization time of (a) 1 min, (b) 15 min and (c) 1 h. The absorbance was measured using an integrating sphere, and the absorbance of the non-sensitized ZnO films was subtracted from the absorbance of the sensitized films. Different colors indicate the different sensitizers, while dashed lines of the same color indicate different films with the same adsorption procedure, according to the legends. Dotted lines in (a) give the absorbance (measured in transmitting setup) of D149 and DN216 in dimethyl formamide.
This means that even though the orbitals of D149 and the double-anchor dyes are very similar, as they show an almost identical absorbance in solution and similar extinction coefficients, the electronic system of the adsorbed dyes on ZnO is different depending on the absence or presence of a second carboxylic anchor group. The closer resemblance of solution spectra by double-anchor dyes speaks of more separated dye molecules, and thus less aggregated molecules for the double-anchor dyes. This difference can be caused either by a different arrangement of the dyes on the ZnO surface because of the second anchor group, or by a different dye-coadsorbate interaction due to the stronger binding 15. The amount of dye adsorbed after 1 min is higher for D149 and lowest for the dye with the longest alkyl spacer at the second anchor group, which for such a short sensitization time will be caused by a hindered diffusion inside the porous film for the sterically more demanding double-anchor dye ix.
ix Even though the different coadsorbate for D149 and the double-anchor dyes might imply that the different shape of the respective spectra is caused by this difference, this is not the case. In 108, deoxycholic acid was used as coadsorbate for both D149 and double-anchor dyes, and the shape of the spectra was comparable to those seen
(c)
1 h
(a) (b)
1 min 15 min
When an adsorption time of 15 min is applied, the amount of adsorbed dyes increases for all four sensitizers, as can be seen from the higher absorbance values at 635 nm, see Table 4. For 15 min adsorption time, D149 shows lower absorbance values than the other double-anchor dyes, which is the opposite of what is observed for 1 min adsorption. Even though it is probable that the spectrum of D149 is shifted to lower wavelengths, as it was observed for 1 min adsorption, still the amount of adsorbed D149 dye molecules is probably lower than for the double-anchor dyes. For the longest adsorption time of 1 h, again the absorbance and thus the amount of adsorbed dye molecules increases for all sensitizers, yet remains lower for D149 than for the double-anchor dyes. For longer adsorption times, the amount of adsorbed dye will probably no longer be defined by diffusion, but by adsorption and desorption equilibria (also in competition to the coadsorbate or other possible adsorbates from the electrolyte). These equilibria will certainly depend on the presence of a second anchor group, so that D149 will be driven from adsorption sites more frequently than for example DN216, which might cause the lower amount of adsorbed D149. It was shown before that for double-anchor indoline dyes the adsorption is slower than for indoline dyes with a single anchor group 128, but on the other hand the double-anchor dyes bind stronger to the ZnO surface 15. This supports well the above explanation of the observed differences in the amount of adsorbed dyes.
The D149 molecules will be aggregated for longer adsorption times, as D149 shows a strong tendency of aggregation on the ZnO surface 126, see also section 1.2.1. The double-anchor dyes are expected to show a lower aggregation tendency, as the second anchor group and the spacer will prevent such close packing of dye molecules as it is possible for D149. In itself the graphs in Figure 20 do not allow an estimation of the aggregation, as the full maximum of the absorbance is beyond the detection limit, and a broadening of the spectrum can be caused by more adsorbed dye and/or increased aggregation.
From the slope of the absorbance at higher wavelength it can be surmised that aggregation also is important for DN91, DN216 and DN285, mostly for the longest adsorption time of 1 h, not so much for shorter sensitization. These trends are even more clearly observed when the absorbance spectra are normalized to the values at 635 nm, where the absorbance is least affected by the depression of the maximum and yet high enough to be significant (Figure 21).
Table 4 – Absorbance values of differently sensitized electrodeposited ZnO films at 635 nm as a measure of dye amount and aggregation, see Figure 20 for the absorbance curves. Numbers in brackets give the cell IDs, while grey values indicate values for cells that are not mainly discussed in this chapter.
Time 1 min 15 min 1 h
D149CA 0.079 [45] 0.091 [61] 0.173 [65]
DN91LCA 0.034 [46] 0.206 [60]; 0.206 [35]; 0.219 [43] 0.451 [69]
DN216LCA/CA 0.071 [59]; 0.052 [55] 0.160 [52] 0.178 [70]
DN285LCA 0.055 [57]; 0.046 [56] 0.160 [63]; 0.217 [34] 0.330 [71]
650 675 700 0.0
0.5
D1491minCA [45]
D14915minCA [61]
D1491hCA [65]
DN911minLCA [46]
DN9115minLCA [35]
DN9115minLCA [43]
DN9115minLCA [60]
DN911hLCA [69]
Corrected absorbance (integrating sphere)
Wavelength / nm
650 675 700
0.0 0.5
DN2161minLCA [55]
DN2161minLCA [59]
DN21615minLCA [52]
DN2161hCA [68]
DN2851minLCA [56]
DN2851minLCA [57]
DN28515minLCA [34]
DN28515minLCA [63]
DN2851hLCA [71]
Corrected absorbance (integrating sphere)
Wavelength / nm
Figure 21 – Normalized absorbance of indoline-sensitized ZnO films (absorbance of the non-sensitized film subtracted). The graphs from Figure 20 were normalized to the absorbance at 635 nm and only the more relevant absorbance at lower wavelengths is displayed to show the aggregation at the right downward slope more clearly. Films sensitized with (a) D149, DN91, (b) DN216 and DN285 are shown. Increasing color depth indicates increasing sensitization time; dashed lines of the same color indicate different films of the same sensitization procedure, according to the legends.
An increase of adsorbed amount of dye (and aggregation) with increasing time of adsorption was also found before for D149 16,126, and from the results shown here, this observation can now be extended also to the double-anchor dyes DN91, DN216 and DN285. The addition of a second anchor group thus changes the absorbance behavior especially because of additional steric hindrance and stronger binding of the molecules, but the observed trends with different adsorption conditions for the double-anchor dyes are similar to the well-established single-double-anchor dye D149.