2. Materials and Methods
Experimental procedures and analytical methods were the same as previously described [
1], except for the substrate temperature (
= 100
C); the thickness of films was also 100 nm. For the photoelectrical measurements, the films were deposited on ceramic (Al
O
) substrates with a pre-patterned array of planar interdigital electrodes, IDE [
3,
4,
5,
6]. The actual temperature of both the quartz crucible, from which the InCl-EtioP-I powder was evaporated using resistive heating, and the massive copper stage, on which the substrates were fixed, was controlled by a thermocouple. In each case, the temperature was precisely set by an AC-DC power supply. The current-voltage characteristics were measured with a Keithley SCS-4200 station in the dark and under illumination of 100 mW/cm
in a shielded stainless steel chamber filled with ultrapure argon (99.998% ) after proper purging to avoid the environmental influences. We also tested the photoconductivity of samples after exposure to air and found no great difference, at least during measurements. The values of specific conductivity
at room temperature were derived from the linear section of the I–V curve—
Table 1. Given the IDE geometry, the transition from the ohmic I–V section to the space charge-limited currents occurs at a bias of 12–14 V or greater depending on a sample [
4]. Then, the films were heated in air to 120
C and cooled back to room temperature with a stop every 10
C so as to obtain steady-state I–V dependencies. Using the points on the cooling arm, the Arrhenius-type dependence of logarithm
vs. reciprocal absolute temperature
T was plotted for each sample and the value of thermal activation energy
was calculated using the equation
[
4,
7]. Typically, 2–4 samples of each type were measured until reproducible results were obtained. We decided to use only copper complexes in this study since for nickel etioporphyrin-I, the deposition process on both cold and hot substrates is different [
1], which lowers the relevance of the comparison. The results on the morphology and properties of vacuum-deposited films of Ni-EtioP-I and -II complexes will be published elsewhere.
3. Results and Discussion
As seen from
Figure 1, there are minor differences in the electronic absorption spectra (EAS) for the three Cu-EtioP isomers deposited on the hot glass/ITO substrate. Moreover, temperature does not affect the position of the absorption bands—
Table 1. The only remarkable feature is a decrease in intensity of Q-bands relative to the B-band in the spectra of the hot-grown films (
Figure S1), which might suggest a change in intermolecular interaction [
8]. However, the bathochromic shift of bands due to J-type aggregation [
9] is about 10 nm for all three isomers, i.e., the same as in the cold-grown films [
1]. The solution spectra of I, II, and III-isomers are nearly identical—
Figure 1, inset. Spectra of films deposited on other transparent substrates, such as optical glass and quartz glass, do not qualitatively differ from those deposited on glass/ITO.
As shown previously [
1], the similarity of spectral profiles of different isomers in the solid state does not necessarily imply a similar morphology of films deposited on an unheated (cold) substrate. In
Figure 2, SEM images of the films deposited on cold and hot substrates are compared. The morphology of hot-grown films is coarser and more clearly reflects the unique packing motif inherent in each isomer. This motif must be associated with the structural features of the molecules that are essential for their self-assembling, with symmetry in this case. So, the Cu-EtioP-I forms the rounded objects with a lateral size of up to 1
m—
Figure 2. These objects are composite and they closely resemble the balls of threads, which makes one wonder how such threads progressively grow on a flat substrate (a spiral or spindle-like growth?). Notably, the molecule of the I-isomer has a rotation-type
symmetry. The II-isomer forms large ingrown needles that are oriented mostly parallel to the substrate surface, although in some cases, they are inclined and protrude above the surface of the film—
Figure 2. The surface of films consisting of molecules of the least symmetrical Cu-EtioP-III features variously shaped grains whose size does not exceed 0.3–0.4
m.
Unevenly spaced huge protruded grains and the softness of molecular aggregates make it difficult to scan the surface of hot-grown films by atomic-force microscopy in the conventional semi-contact mode. The height difference on the surface relief ranges from a few hundred nanometers to a micron, depending on the probability of an AFM tip meeting a large-sized object; some results are shown in
Figure S2. The XRD patterns of films deposited on cold and hot substrate are compared in
Figure 3. In all cases, no new diffraction maxima were detected in the hot-grown films. However, the Cu-EtioP-I films deposited on a hot substrate show a tremendous increase in the intensity of the low-angle diffraction peak at 6.9 degrees, while other peaks practically disappear. The peak at 20.8 degrees is most likely a tripled diffraction order of a strong peak at 6.9 degrees.
In the previous article, we tentatively assigned the peak at 6.9 degrees to a metastable phase that appears in the films of I-isomers of metal-etioporphyrins [
1]. It now becomes clear that this phase is quite stable and tends to dominate in hot-grown Cu-EtioP-I films. For a strong diffraction peak at 6.9 degrees arising in a hot-grown film, the rocking curve (omega scan) can be reliably measured. The full width at half maximum is 0.6 degrees, which indicates a very high degree of texture. This fact disagrees with the morphology of Cu-EtioP-I film as viewed with SEM/AFM in
Figure 2 and
Figure S2, and we cannot offer a consistent explanation of such a disagreement. Perhaps, the internal structure of a “ball of threads” is actually rather complex, such that it allows molecules to retain their preferred orientation with respect to the substrate. Since the lattice parameters for this thin-film phase are unknown, it is not possible to determine the orientation. In the XRD patterns of the films of II and III isomers, the intensity of all peaks related to hot-grown films decreases, whereas the number and relative intensity of the peaks remain unchanged—
Figure 3. Given equal nominal thickness of molecular layers in all batches, this can be explained by decreased strength of the crystallites texture in hot-grown films, which, again, is not consistent with the SEM/AFM images in
Figure 2 and
Figure S2. Crystallinity of Cu-EtioP-III is stronger suppressed by deposition on a hot substrate: the intensity of peaks is reduced by a factor of five (
Figure 3).
It seems natural to assume that the morphological singularity of etioporphyrin isomer films would result in variations of their macroscopic properties, the most interesting of which is electrical conductivity. Therefore, we examined the conductive properties of the films that were similarly deposited on the IDE substrates maintained at either cold of hot temperature
—
Table 1. These are commercially available substrates consisting of a high-dielectric support (finely polished polycorundum p-Al
O
plates) equipped with a pair of photolithographically patterned 60 nm thick multi-pin electrodes (nickel) spaced 30
m apart [
1]—
Figure S6. Thus, the growth surface in the electrode channel is not the same as in the Si(111) wafers used for SEM/AFM and XRD measurements; in its dielectric nature and roughness, it is close to the glass used for optical measurements. The morphology of films grown on dissimilar substrates (glass and Si) is compared in Figures S3–S5. It may be concluded that the substrate material definitely affects the morphology of the growing etioporphyrin layer; in particular, those grown on insulating technical substrates are more textured—cf. [
10]. However, the general trends demonstrated by the isomers continue.
The values of
are low, which is typical of weakly bound molecular solids [
6,
8], with no lucid correlation with an isomer type. Similarly to phthalocyanines, the dark conductivity in copper etioporphyrins is most likely of extrinsic nature and associated with the presence of thermally activated charge carriers (holes) [
2,
4,
5].
increases in the isomeric row I⇒II⇒III, remaining basically the same for cold- and hot-grown films—
Table 1. The most obvious distinction between cold- and hot-grown films consists in the two-orders-of-magnitude lower conductivity and in a largely increased photosensitivity; these parameters are likely to be interrelated. A decrease in the dark conductivity in hot-grown films can be explained by the fact that, despite the appearance of large-sized grains with a presumably high conduction ability, there exist well-defined intergrain boundaries, deep voids, etc., which limit the charge transport—
Figure 2. Nonetheless, the results in
Table 1 are somewhat surprising, since the morphology of films composed from different isomers, e.g., the size, shape, composition, and orientation of grains, differ very strongly, whereas the conductivity parameters change either slightly or monotonically.
Note that the value of
corresponds to the lateral transport of charge carriers across the film, it may change when measured in the vertical direction, i.e., in the sandwich electrode geometry [
5], as the films are textured. Therefore, a slight variation of
depending on the isomer type can be assigned to the existence of a very thin underlayer that bridges the electrodes and, hence, determines the charge transport in the channel. This layer could be amorphous or somehow oriented [
10], so that the correlation of
with the bulk film morphology would be minimal; in hot samples, it becomes thinner. To rule this hypothesis out, one needs to do measurements on the sandwich-type cells with a varying bottom electrode material (this work is underway currently).
On the other hand, such a strong increase in the photo-assisted conductivity of hot-grown Cu-EtioP-III films cannot be solely due a decrease in the dark-current signal—
Table 1. The mechanisms underlying the photoconductivity in the polycrystalline porphyrin-type materials are recently reviewed in [
2]. The photosensitivity reaches
for the III-isomer, which is a very high value for for an IDE-based cell with a top-evaporated molecular layer [
6,
8]. Therefore, the photo-activated transfer of charge carriers in moderately biased cells seems to be more efficient across a film consisting of densely packed, highly textured crystallites of Cu-EtioP-III (
Figure 2 and
Figure S3), probably due to the reduced barrier heights. In other words, the bulk morphology of films begins to play its role in the conductivity of illuminated devices.