**3. Results and discussion**

The capillaries with FePc at the ends and tetramethoxysilane in the centre were allowed to stand at 22°C for 2 weeks. Subsequently, the generated particles were extracted from the capillary within the gel and were observed by SEM. **Figure 2a** and **b** show, at different magnifications, the FePc particles embedded in tetramethoxysilane. Despite being very small, they showed several structures-amorphous particles, regular particles and needles. In all cases, there was a heterogeneous distribution of particles inside the gel. The particles were removed from the tetramethoxysilane, washed and dried in a vacuum. The use of this technique demonstrated its applicability to the *in situ* formation of nanometric-size particles inside the gel. A preliminary TEM study of the nanometric FePc particles was also performed. **Figure 2c** shows a high-resolution bright field image of the FePc sample, where particles ranging in size between 2.8 and 20 nm can be seen. The shape of the particles is irregular, although some quasi-spherical forms can be discerned. A heterogeneous dispersion of the nanoparticles can also be seen. Among the advantages of using this technique for reinforcing particles in the manufacture of composite materials are that a very small sample can be used and the continuous manipulation of particles can be avoided; furthermore, it permits a continuous control of the growth process. It is difficult to determine the crystalline arrangement from TEM imaging in real space, so a wider characterization by IR spectroscopy was required. IR spectroscopy was specifically used to identify the structural nature of FePc, given that the IR spectrum depends on the crystal structure [4]. MPcs are known to have different polymorphs which are strongly identified by the IR absorption technique [4, 5]. It has been reported that the α-form of MPc can be characterized by a band around 720 cm−1, while the β-form can be characterized by a band at a greater wave number at approximately 778 cm−1 [4–7]. In **Table 2**, it can be observed that FePc particles were present in the α and β crystalline structures.

**Figure 2.** Gel with FePc particles (a) 1000×, (b) 7000× and (c) HRTEM micrographs.

Preparation and Structural Characterization of Metallophthalocyanine Particles Embedded in a Polymer Matrix http://dx.doi.org/10.5772/67576 35


**Table 2.** Characteristic FT-IR bands for particles and thin films (cm−1).

there was a heterogeneous distribution of particles inside the gel. The particles were removed from the tetramethoxysilane, washed and dried in a vacuum. The use of this technique demonstrated its applicability to the *in situ* formation of nanometric-size particles inside the gel. A preliminary TEM study of the nanometric FePc particles was also performed. **Figure 2c** shows a high-resolution bright field image of the FePc sample, where particles ranging in size between 2.8 and 20 nm can be seen. The shape of the particles is irregular, although some quasi-spherical forms can be discerned. A heterogeneous dispersion of the nanoparticles can also be seen. Among the advantages of using this technique for reinforcing particles in the manufacture of composite materials are that a very small sample can be used and the continuous manipulation of particles can be avoided; furthermore, it permits a continuous control of the growth process. It is difficult to determine the crystalline arrangement from TEM imaging in real space, so a wider characterization by IR spectroscopy was required. IR spectroscopy was specifically used to identify the structural nature of FePc, given that the IR spectrum depends on the crystal structure [4]. MPcs are known to have different polymorphs which are strongly identified by the IR absorption technique [4, 5]. It has been reported that the α-form of MPc can be characterized by a band around 720 cm−1, while the β-form can be characterized by a band at a greater wave number at approximately 778 cm−1 [4–7]. In **Table 2**, it can be

34 Phthalocyanines and Some Current Applications

observed that FePc particles were present in the α and β crystalline structures.

**Figure 2.** Gel with FePc particles (a) 1000×, (b) 7000× and (c) HRTEM micrographs.

IR spectroscopy was also used in this study to ascertain the presence of the more representative bonds in the FePc compound and to determine whether significant chemical changes took place in this compound during gel nucleation and growth. **Table 2** shows the characteristic bands of the FePc particles deposited in the gel. The band appearing at 1605 ± 4 cm−1 was assigned to the C═C stretching vibration for pyrrole. The peak responsible for carbonnitrogen stretching and bending occurs at 1332 ± 4 cm−1. The peaks located at 1164 ± 2, 1117 ± 2 and 753 ± 2 cm−1 are due to the interaction of carbon atoms with the peripheral-ring hydrogen atoms [8–10]. As mentioned above, spin coating and annealing were carried out to produce the thin films. IR spectroscopy was performed in these films in order to verify that no chemical changes occurred in the FePc when interacting with the polymeric matrix. The results reported in **Table 2** indicate that the MPc did not experience any chemical changes during the deposition; on the other hand, in the thinnest film, the crystalline phase α is not observed. It is worth mentioning that the signals in the MPc film show slight changes in location. This occurs because, in thin films deposited by any method, internal stress affects intramolecular angles and bonding energies. Nevertheless, no significant changes occurred in these films, so we may conclude that the production of thin films from the FePc-polystyrene composite by the spin coating and annealing method is appropriate.

The films obtained by spin-coating were analysed by SEM. **Figure 3** shows the presence of the two phases-polymeric matrix and reinforcement. During the annealing, polymerization of polystyrene generated the needles shown in the images, while the FePc appears as irregular conglomerates. It is possible to observe that the MPc particles have been embedded in the matrix homogeneously, i.e. the particles are not agglomerated or separated, which in turn indicates that polystyrene is an appropriate matrix for this kind of films.

Optical absorption measurements are widely used to characterize the electronic properties of the thin films through the determination of parameters describing the electronic transitions, such as the band gap [11]. Additionally, the absorption spectra of different polymorphs of some Pc compounds show significant differences among each other [7, 12]. MPcs have two typical absorption bands, namely the *Q*-band in the visible region and the *B* or Soret-band in the near-ultraviolet

**Figure 3.** SEM images for spin-coated films (a) 83×, (b) 500× y (c) 1000×.

region [13–17]. The Q-band absorption is responsible for the characteristically intense blue/green colour of the FePc and this band has been interpreted in terms of π-π\* excitation between bonding and antibonding molecular orbitals [7, 18]. The electronic spectrum of the FePc particles obtained in tetramethoxysilane (**Figure 4a**) shows the characteristic Q-band absorption in the 578–750 nm region. The Soret-band of FePc arising from the deeper π levels → LUMO transitions is observed in the UV region at about 400–463 nm. On the other hand, the optical transmittance spectra of the thin-films deposited on quartz were recorded from 200 to 1100 nm and are shown in **Figure 4b**. Differences in the transmittance of the films under examination can be attributed to differences in thickness (see **Table 1**) according to Beer's law [19]. When the thickness of the film increases, its transmittance diminishes. The UV-Vis spectra of FePc-polystyrene thin films exhibited a characteristic B-band in the region between 285 and 305 nm. The observation of a single peak in the Soret band resembles that observed for CoPc, NiPc and other Pc thin films [20, 21]. This may imply that the splitting structure of this peak could be affected by the orbital overlap of the Pc ring with the central metal [21], although this effect could also be attributed to the presence of the polymeric matrix which, while protecting the FePc from oxygen and environmental humidity, also alters its optical properties in the visible region of the spectrum.

Considering the above results, we further apply the Cody and the Tauc models for the determination of the band gaps of the thin films [7, 22, 23]. The Cody model provides an effective option for the determination of the optical band of thin films in terms of its thickness. It uses the dependence between the photon energy (*hν*) and the absorption coefficient (α). The optical gap associated with the thin films is determined by extrapolating the linear trend observed in the spectral dependence of **(α/***h***ν)<sup>n</sup>** on *h*ν. Here, *n* is a number characterizing the transition process, depending upon the nature of the electronic transitions responsible for the absorption; for direct transitions, *n* = ½, and, for indirect transitions, *n* = 2. The intersection with the *x*-axis of this linear extrapolation corresponds to the Cody optical gap for a given thickness of the film [22, 23]. The Cody optical gaps *Egi* and *Egd* for both transitions were obtained from the curves corresponding to those shown in **Figure 5** for the film with the largest thickness (*Thin Film 7*), which was of 348 nm.

For this film, the optical gap value is similar for both transitions, direct and indirect (see **Table 1**); apparently, the high concentration of FePc related to the highest thickness could be the cause of the similar values, but this could also be related to the fact that 4.3 eV is the lower (indirect) gap of the films under examination and may be quantitatively close to the direct gap for that particular film. On the other hand, the Tauc model argues that the optical gap associated with the thin film is determined through an extrapolation of the linear trend observed in the spectral

Preparation and Structural Characterization of Metallophthalocyanine Particles Embedded in a Polymer Matrix http://dx.doi.org/10.5772/67576 37

**Figure 4.** UV-Vis spectroscopy for: (a) FePc and (b) thin films.

region [13–17]. The Q-band absorption is responsible for the characteristically intense blue/green colour of the FePc and this band has been interpreted in terms of π-π\* excitation between bonding and antibonding molecular orbitals [7, 18]. The electronic spectrum of the FePc particles obtained in tetramethoxysilane (**Figure 4a**) shows the characteristic Q-band absorption in the 578–750 nm region. The Soret-band of FePc arising from the deeper π levels → LUMO transitions is observed in the UV region at about 400–463 nm. On the other hand, the optical transmittance spectra of the thin-films deposited on quartz were recorded from 200 to 1100 nm and are shown in **Figure 4b**. Differences in the transmittance of the films under examination can be attributed to differences in thickness (see **Table 1**) according to Beer's law [19]. When the thickness of the film increases, its transmittance diminishes. The UV-Vis spectra of FePc-polystyrene thin films exhibited a characteristic B-band in the region between 285 and 305 nm. The observation of a single peak in the Soret band resembles that observed for CoPc, NiPc and other Pc thin films [20, 21]. This may imply that the splitting structure of this peak could be affected by the orbital overlap of the Pc ring with the central metal [21], although this effect could also be attributed to the presence of the polymeric matrix which, while protecting the FePc from oxygen and environmental

**Figure 3.** SEM images for spin-coated films (a) 83×, (b) 500× y (c) 1000×.

36 Phthalocyanines and Some Current Applications

humidity, also alters its optical properties in the visible region of the spectrum.

**Figure 5** for the film with the largest thickness (*Thin Film 7*), which was of 348 nm.

*Egi*

and *Egd*

Considering the above results, we further apply the Cody and the Tauc models for the determination of the band gaps of the thin films [7, 22, 23]. The Cody model provides an effective option for the determination of the optical band of thin films in terms of its thickness. It uses the dependence between the photon energy (*hν*) and the absorption coefficient (α). The optical gap associated with the thin films is determined by extrapolating the linear trend observed in the spectral dependence of **(α/***h***ν)<sup>n</sup>** on *h*ν. Here, *n* is a number characterizing the transition process, depending upon the nature of the electronic transitions responsible for the absorption; for direct transitions, *n* = ½, and, for indirect transitions, *n* = 2. The intersection with the *x*-axis of this linear extrapolation corresponds to the Cody optical gap for a given thickness of the film [22, 23]. The Cody optical gaps

for both transitions were obtained from the curves corresponding to those shown in

For this film, the optical gap value is similar for both transitions, direct and indirect (see **Table 1**); apparently, the high concentration of FePc related to the highest thickness could be the cause of the similar values, but this could also be related to the fact that 4.3 eV is the lower (indirect) gap of the films under examination and may be quantitatively close to the direct gap for that particular film. On the other hand, the Tauc model argues that the optical gap associated with the thin film is determined through an extrapolation of the linear trend observed in the spectral

**Figure 5.** Plot of (a) **(α/***h***ν)1/2** and (b) **(α/***h***ν)<sup>2</sup>** versus photon energy *hν* of *Thin Film 7*.

dependence of **(***αhν***) <sup>n</sup>** over a limited range of *hν* [1, 2]. The Tauc optical gaps for *Egi* and *Egd* were obtained from the curves corresponding (see **Table 1**) and they are shown in **Figure 6** for the film with the largest thickness (*Thin Film 7*). According to **Table 1** for the thicker film the smaller gap is obtained. At this thickness, the concentration of FePc is sufficient to decrease the gap and increase the overlap between Pc molecules. As the stacking between molecules increases, the electron flux increases significantly with respect to films with small thickness. On the other hand, for each of the remaining films, the indirect transition is the predominant one, with significantly lower values than the direct transition; this may be expected because of the mainly amorphous characteristics of the films and their effect on orbital overlap, despite FePc showing some α or β crystalline forms. It is important to mention that the variations in optical gaps obtained for the different films are of low significance. This may be attributed to the similar morphology of these systems, which differ only in the quantity and size of the FePc particles and the arrangement of their molecules in the polymeric matrix. Additionally, the gap depends on the number of electrons of the metal in the Pc ring [7, 19], which is the same for all these films.

**Figure 6.** Plot of **(α***h***ν)1/2** and **(α***h***ν)<sup>2</sup>** versus photon energy *hν* of *Thin Film 7*.

Finally, in order to evaluate the electrical properties of the thin films, the four-point technique was employed, using the glass substrate with an ITO conducting contact. This study was performed on the sample labelled *Thin Film 7*, which was the one having the lowest optical gap. The film had a surface area of 2.16 cm2 . **Figure 7** shows the *I-V* characteristics of *Thin Film 7* under different illumination types (yellow light, white, blue, orange, green, infrared, UV and dark [no light]). Regardless of the wavelength of the incident radiation, the thin film follows the same behaviour. At lower voltages (around 10 V), ohmic conduction is evident, while space-charge limited conductivity (SCLC) governed by an exponential trap distribution is found at higher voltages. On the other hand, the *I-V* characteristics display symmetric

**Figure 7.** *I*-*V* characteristics of *Thin Film 7:* (a) ITO is positively biased and (b) ITO is negatively biased.

behaviour, both when (a) the current due to hole injection from positively biased ITO was measured and also when (b) the current due to hole injection from silver was measured by reversing the polarity of the bias voltage. This can be explained by a negligible energy barrier at the *ITO/FePc-polystyrene* and *FePc-polystyrene/Ag* interfaces leading to a SCL bulk current when either the ITO or silver electrode is positively biased [24–26].
