1. Introduction

Phthalocyanines have many different interesting properties which make them suitable compounds to be used in many useful technological and scientific applications. The recent increasing interest by scientists and engineers for studying different types of organic compounds shed light on the importance and prospective of semiconducting phthalocyanine systems^[1-4]. These compounds are macro-cyclic compounds having an alternating nitrogen-and carbon-atom ring structure. Phthalocyanines have many different interesting properties which make them suitable compounds to be used as gas sensors, solar cells, optoelectronic devices and as prototypes of organic semiconductors^[1-6]. Metal-phthalocyanines have a distinctive crystalline structure in three forms ($\alpha$-, $\beta$-, and $\gamma$-modification) in which a transformation may occur on varying the temperature or pressure of the compound. It is worth noting that the crystalline structure makes a distinction between organic and inorganic materials. Many researchers devoted their attention on the remarkable semiconducting properties of these compounds^[3-6]. Metal–phthalocyanine compounds are chemically stable having dense colors; therefore, they have been employed as colorant (dyes and pigments) in textile and paint industries^[6-8]. Many of these compounds were technologically employed as gas sensors, solar cells, and opto-electronic devices^{[1, 2, 7, 8]}. Compounds such as NiPc, CoPc and CuPc have been intensively studied and then manufactured as gas sensing devices^[6-9], while platinum phthalocyanine solution, PtPc, was utilized to generate various types of optical switching devices^[10]. In addition, copper phthalocyanine, CuPc, was utilized as a buffer-zone layer in manufacturing a white organic light emitting diode^[11], whereas ZnPc-loaded PLGA biodegradable nanoparticles were employed in photodynamic therapy in tumor-bearing mice^[12].

Phthalocyanine crystalline structure has a high stability, which enables them to exist in multiple crystalline polymorphs, including $\alpha$-, $\beta$-, and $\gamma$-structure^[13]. According to an intensive study on metal-free and metal-phthalocyanine electronic structures of $\alpha$-, $\beta$-, and $\gamma$-phases, Orti and Barades^[14] expected to have more than 70 modifications of metal-phthalocyanine (M-Pc) complexes. Those complexes could be manufactured via removing the two-central hydrogen atoms from metal free-phthalocyanine (H$_{2}$Pc) and replacing them by the suitable metal; such as Cu, Zn, Fe, Co, and Ni^[14-16].

In the past few years, a new promising and important era was opened for employing organic phthalocyanine compounds in biomedical and biophysical technology. Zinc phthalocyanine, ZnPc, was employed in medical applications due to its selective binding to tumor-antibodies. Also, it was used in the synthesis of a novel compound applicable in photodynamic therapy^[17]. Moreover, Gao and coworkers^[18] synthesized tetra-trifluoroethoxyl zinc phthalocyanine (an organic compound which dissolves in most of organic solvents) that has the potential to be used in photodynamic therapy of cancer.

In the present study, the dielectric and electronic properties of different phthalocyanine compounds (ZnPc, CuPc, FePc, H$_{2}$Pc) are analyzed and a comparative study is performed. The relaxation time and optical band gap values of these compounds are extracted.

2. Experimental

Phthalocyanine compounds (ZnPc, CuPc, FePc, H$_{2}$Pc) powder were obtained from Kodak, UK, with a purity of about $\sim$97%. Some raw phthalocyanine compounds were purified using the entertainer sublimation method. Metal-phthalocyanine thin films with different thicknesses were prepared by the thermal evaporation technique using Edward's coating unit. The coating unit combines many features which simplify the evaporation process. Firstly, a rotary masking system is used to allow successive evaporation of different materials (such as Au, ZnPc and Au to form Au–ZnPc–Au sandwich structure) without breaking vacuum^[19]. Secondly, a sample holder is deeply attached to the system so that breaking masking system cannot happen. Thirdly, a mechanical shutter is used to control the time period of deposition precisely. Corning glass substrates were thoroughly cleaned before deposition in order to avoid contamination in the deposition champers. A metal–phthalocyanine sandwich structures with different electrode materials, such as Au or Al, were prepared by the thermal-evaporation process. A deposition rate of less than 1 nm/s was maintained throughout the deposition process. The deposition rates and film thicknesses were monitored by a quartz crystal monitor^{[3, 6]}.

After deposition, the accurate thickness measurements were confirmed by using a planer surfometer (SF200) stylus instrument. After film growth, the films were mounted on another subsidiary vacuum chamber for conductance, capacitance and loss tangent measurements. The data were collected in the frequency range 0.1 to 20 kHz using a Hewlett-Packard LCZ meter equipped with a four terminal test fixture to minimize the experimental errors. All measurements were carried on a sandwiched structure of either Au–MPc–Au or Al–MPc–Al, where M $=$ Zn, Fe, Cu or H$_{2}$. More experimental details could be found elsewhere^{[3, 6, 15, 16]}.

3. Results and discussion

The AC capacitance of various thin films of phthalocyanine compounds was measured over a wide range of temperature (90 to 450 K) in a frequency range between 0.1 to 20 kHz. Films of effective area 1.2 $\times$ 10$^{-5}$ m$^{2}$ and thickness varied from 0.1 to 1.7 $\mu$m were employed in the present report. The capacitance data was employed to derive the corresponding real part of the dielectric function (constant) as a function of frequency and temperature $\varepsilon _{1}$($T$, $\omega)$. The variation of the dielectric function with frequency is demonstrated in Fig. 1(a), while the variation with temperature at 1 kHz is displayed in Fig. 1(b). Qualitatively, the dielectric function of the four phthalocyanine compounds has roughly the same frequency behavior. Below 5 kHz, $\varepsilon_{1}$ is strongly dependent on frequency for the four Pc compounds, however, the dependence is less pronounced and approaches a constant value above this frequency. The electronic structure and atomic polarizability of the central atom in the phthalocyanine molecule is responsible for the variation of the values of $\varepsilon_{1}$ for the different compounds. According to Liao and Scheiner^[20] metals like Fe, Cu and Zn are strongly bonded to the phthalocyanine ring, however, the valence electronic structure of FePc or CuPc differ significantly more than the ZnPc molecule. The film thickness, impurities, carrier concentration and electrode materials may also have great influence on the characteristics of the central atom behavior^[21]. By analyzing the curves of Fig. 1(a), the dependence of the dielectric constant, $\varepsilon_{1}$, on frequency at a constant temperature was found to obey a relation of the form:

$\varepsilon _1 =A\omega ^n,$

(1)

where $A$ is a constant, $\omega$ the angular frequency and the power-index $n$ assumes negative values ($n$ < 0). Such behavior has been observed in CoPc thin films^[22], CoPc pellets^[23], MgPc^[24], ZnPc^[25] and also in some inorganic compounds^[26].

Figure 1(b) displays the variation of $\varepsilon_{1}$ as a function of temperature at 1 kHz. At moderate frequencies ( < 8 kHz), the variation is qualitatively the same, while at higher frequencies ($>$ 12 kHz), $\varepsilon_{1}$ saturates to a constant value. At a fixed low-frequency, $\varepsilon_{1}$ of ZnPc, and FePc is strongly temperature dependent above 300 K, while for H$_{2}$Pc, the dependence is weak over the entire temperature range of interest. At low temperatures and high frequencies the real part of the dielectric constant saturates to a constant value for all four compounds. The same behavior of the dielectric function was observed in ZnPc thin films^{[27, 28]}, MgPc^[24] and CoPc pellets^[23]. There is a good qualitative agreement between the present data and other data of phthalocyanine compounds presented by other research groups. It is worth noting that the increase of $\varepsilon_{1}$ for H$_{2}$Pc above 300 K and then followed by a decrease above 350 K (as evident from Fig. 1(b)) could be explained in terms of nomadic polarization. The nomadic polarization arises from the steep increase in the number of free carriers with increasing temperatures as suggested by Nalwa and Vazudevan^[23]. Such nomadic polarization behavior was also observed in CoPc^[23].

The measured values of the complex dielectric function ($\varepsilon$ $=$ $\varepsilon_{1}$ – i$\varepsilon_{2})$ will give valuable information about the dielectric response and energy absorption of the material when affected by an alternating electric field. The imaginary part of the dielectric constant, representing the energy loss, ($\varepsilon_{2}$ $=$ $\varepsilon_{1} \tan\delta)$ as a function of frequency, $f$, for fixed temperature (350 K) is displayed in Fig. 2 on a semi-log graph. At all temperatures, the behavior of $\varepsilon_{2}$ against $f$ is qualitatively almost the same. The variation of the imaginary part, $\varepsilon_{2}$, with frequency and temperature depends on the phthalocyanine compound under consideration. At low frequencies and all temperatures, a strong dependence is observed, while at higher frequencies $\varepsilon_{2}$ approaches to a constant value. The electronic structure of the central metal, impurities, electrode material type, sample thickness and charge accumulation will play an important role in energy dissipation within the compound^{[20, 21]}. Very thin phthalocyanine films tend to have a nearly crystalline structure by having preferential orientations, however, as the thickness increases, the amorphous structure is manifested^[15]. Such amorphous structure tends to have more impurities, vacancies and even trapped free charge carries and hence will in turn affect the polarizability of the sample. Additionally, it is well established that too many phthalocyanine compounds are easily transformed from a monoclinic to a tetragonal crystalline structure^{[3, 11, 22]}.

Careful analysis of the frequency dependence of the dielectric loss (Inset Fig. 2) shows that $\varepsilon_{2}$ follows a power law with the angular frequency, $\omega$, given by the following relation:

$\varepsilon _{2}= A_{1} \omega^{m},$

(2)

where $A_{1}$ is a temperature-dependent constant and the index $m$ is related to the energy needed to remove the electron from its site to infinity, $W_{\rm M}$, {$m=\frac{-4k_{\rm B} T}{W_{\rm M}}$} as given by Giuntini et al.^[27]: The index $m$ was calculated from the slopes of the fitted straight lines of log $\varepsilon_{2}$ versus log $\omega$. The values of m are negative for all temperatures and it decreases with increasing temperature at all different frequency ranges. Similar behavior had been observed in ZnPc^[28], chalcogenide glasses^{[26, 27]} and amorphous Ga$_{2}$S$_{3}$–Ga$_{2}$Se$_{3}$ films^[29].

The temperature dependence of the imaginary part of the dielectric constant, $\varepsilon_{2}$, for the different phthalocyanine compounds at 1 kHz is displayed in Fig. 3. In general, the same behavior of $\varepsilon_{2}$ versus $T$ at other fixed frequencies was obtained. At low temperature, a very weak dependence is observed, while as $T$ approaches room temperature the dependence becomes much stronger. The variation of $\varepsilon_{2}$ behavior for the different phthalocyanine compounds may be attributed to the type of central atom and its electronic configuration^{[20, 21]}, film thickness, type of electrode material and charge accumulation within traps^{[3, 6, 15, 23, 28]}. In some phthalocyanine compounds, a peak (around 310 K) is observed in the imaginary part of the dielectric constant $\varepsilon_{2}$ versus $T$ curves which is probably due to nomadic polarization^[23], and to exhaustion of oxygen molecules at high temperatures as was confirmed in H$_{2}$Pc thin film sample^[30]. On the other hand, Al–ZnPc–Al, Au–Zn–Au, Au–CuPc–Au, and Al–FePc–Al thin film samples did not show any peak in the $\varepsilon_{2}$–$T$ curves as partly shown in Fig. 3. Therefore, it is apparent that the preparation conditions, the central atom, and the presence of impurities such as O$_{2}$, which acts as an acceptor, will greatly affect the electrical properties of the phthalocyanine compounds^{[3, 6, 15, 30]}. It is well established that annealing the sample or heating it to high temperatures (above 350 K) will stabilize the electrical properties due to oxygen desorption and structural modifications^{[3, 14, 16, 31]}. The increase in both $\varepsilon_{1}$ and $\varepsilon_{2}$ values with temperature is mainly attributed to the enhanced conductivity through the thermal excitation of the charge carriers. It is worth recalling that an aluminum electrode may provide a blocking contact (Schottky barrier) to many phthalocyanine compounds since a small Al$_{2}$O$_{3}$ thin layer may be formed during deposition, while Au provides ohmic contact for all samples, hence affecting the dielectric properties^{[3, 16, 22, 30]}.

The relaxation time, $\tau$, was calculated from the values of $\varepsilon_{1}$, $\varepsilon_{2}$ at different temperatures for all phthalocyanine compounds using the following relation^{[28, 33]}:

$\tau =\frac{\varepsilon _2 }{\omega(\varepsilon _1 -\varepsilon _\infty )},$

(3)

where $\varepsilon_{\infty }$ is the optical dielectric constant at very high (infinite) temperature. For all compounds, the relaxation time is strongly temperature and frequency dependent, however, the variation strongly depends on the frequency–temperature range of interest. Figure 4 displays the variation of $\tau$ with frequency for ZnPc, FePc, CuPc, and H$_{2}$Pc at different temperatures around 300 K. At low frequency, $\tau$ is strongly frequency dependent, while at a higher frequency, a weak dependence is observed and $\tau$ appears to saturate to a constant value. At low temperatures and high frequency (around 17 kHz), a slow increase in the relaxation time is observed in most of the previous compounds, so an experimental extension above 20 kHz is necessary to check this behavior of $\tau$. In general, the relaxation time increased with decreasing frequency, but the values of relaxation time for ZnPc are closer to each other more than other compounds. This discrepancy may be related to electronic distribution of the central atom, growth conditions, sample packing fraction, and difference in sample thicknesses^{[20, 21, 25]}. In addition, impurities such as O$_{2}$, humidity, and sample aging are very important in changing the electronic properties of phthalocyanine compounds^[34]. Qualitatively, the behavior of relaxation time and its values are in good agreement with ZnPc films^{[25, 27]}, bulk MgPc^[24], glassy a-Se–Te–Ga system^[26] and amorphous Ga$_{2}$S$_{3}$–Ga$_{2}$Se$_{3}$ films^[29].

For a large number of compounds, the relaxation time is related to a thermally activated process according to the following (Arrhenius-like) relation^{[24, 25, 28, 33]}.

$\tau =\tau _\infty \exp \left(-\frac{E_{\rm r}}{k_{\rm B}T}\right),$

(4)

where $\tau_{\infty}$ is the relaxation time at infinite (very high) temperatures which are approximately equal to 10 ^-13[35], and $E_{\rm r}$ is the relaxation activation energy. The dependency of ln$\tau$ on a reciprocal of temperature for different fixed frequencies is depicted in Fig. 5(a) for H$_{2}$Pc and FePc, while Figure 5(b) displays such dependency for ZnPc. The data of FePc and H$_{2}$Pc samples are fitted reasonably well to a straight line as shown in Fig. 5(a). This is an indication of the variation of $\tau$ exponentially with temperature as expected for a thermally activated (Arrhenius-like) process as expected from Eq. (4). Activation energies of 0.028 eV and 0.007 eV were calculated from the slopes of the fitted lines for FePc and H$_{2}$Pc respectively. For other phthalocyanine compounds, the activation energy values were also comparable to those values of FePc and H$_{2}$Pc. In general, the behavior of $\tau$ as a function of $T$ is qualitatively the same for all Pc-compounds at different frequencies. The relaxation time for ZnPc samples was also investigated and the data showed that two regions are evident in $\tau$ versus 1/$T$ plots as displayed in Fig. 5(b). The experimental data was extended down to 90 K in the case of Au–ZnPc–Au samples. An activation energy value of 0.01 eV was estimated for an Au–ZnPc–Au sample from the slope of the fitted line at low temperatures indicating a hopping conduction. However, a higher activation energy value was derived for the Al–ZnPc–Al sample at a frequency of 1 kHz. It is apparent that the material type of electrodes plays a critical role in the variation of the electronic properties of metal-phthalocyanine compounds. The appearance of two regions in $\tau$ versus 1/$T$ curves is an indication of two types of conduction processes: at low temperatures a hopping conduction process ($E_{\rm r}$ $\sim$ 0.01 eV) is dominant, while at higher temperatures a thermally activated process is evident with activation energy $E_{\rm r}$ $\sim$ 0.12 eV^{[3, 28, 32]}. Activation energy 0.034 eV for Au–CuPc–Au was also derived from the slope of the fitted data of $\tau$ versus 1/$T$. The low values of the activation energies for the various phthalocyanine compounds are consistent with the hopping of charge carriers between localized states^{[3, 6, 32, 35-37]}. The same conduction behavior was also observed in some inorganic materials like AlN$x$^[38], and ZnO^[39], and also in phthalocyanine compounds like ZnPc^{[16, 28, 32]}, MgPc^[40] and CuPc^{[3, 40]}.

The measured data of the frequency dependence of AC conductivity ($\sigma _{\rm ac} =A'\omega^s)$ is employed to derive the frequency exponent index, $s$^{[3, 6, 32, 35, 41]}. For phthalocyanine compounds, the value of index $s$ is less than unity and it is a temperature dependent quantity as expected for hopping conduction^{[3, 22, 32, 35, 36, 42]}. The experimental data of the AC conductivity were analyzed with various theoretical models and the correlated barrier hopping (CBH) model was found to be the appropriate mechanism for the electron transport in phthalocyanine films^[35]. Application of the CBH model reveals that the electronic conduction takes place via a single or bipolaron hopping processes in the whole temperature range of study. According to the CBH model, the index $s$ is related to the maximum barrier height $W_{\rm m}$ to the first order approximation as^[35]:

$s=1-\frac{6k_{\rm B} T}{W_{\rm m }},$

(5)

where $k_{\rm B}$ is the Boltzmann's constant and $T$ is the temperature. The maximum barrier height at infinite separation, $W_{\rm m}$, is also known as the "polaron binding energy", i.e. it represents the binding energy of the carrier in its localized sites^[35]. For a bipolaron $W_{\rm m}$ is approximately equal to the band gap width, while for a single polaron its value is equal to a quarter of the optical band gap^[26].

Figure 6 displays the derived values of the maximum height barrier $W_{\rm m}$ as a function of temperature for different phthalocyanine thin films. The value of $W_{\rm m}$ is dependent on the central element of the compound. Due to variation of electronic structure of various elements, it is expected that for example, CuPc valence electronic configuration, and hence, both molecular polarizability and band structure will be different than those for other metal-phthalocyanine compounds^[20]. Besides, a nomadic polarization may occur at phthalocyanine samples at temperatures above 300 K^[23]. At very low temperatures, $W_{\rm m}$ has a maximum value and starts to decrease as the temperature increases then it starts to increase slowly above room temperature. This increase in $W_{\rm m}$ may be attributed to accumulation of charges due to thermal excitation as the sample temperature increases. In addition, the electrode material type (and hence formation of dipolar layer of various thicknesses and/or its interaction with ambient gases, such as O$_{2})$ will affect the value of the barrier height. In the present investigation, the values of $W_{\rm m}$ for ZnPc with Al electrodes are larger than those with Au-electrodes. It is well established that an Au-electrode will provide ohmic contacts while Al electrodes produced blocking contacts in most phthalocyanine compounds^{[3, 5, 32, 40]}. Usually a very thin aluminum-oxide layer (Al$_{2}$O$_{3})$ may be formed at the interface between the Al-electrode and the phthalocyanine film during deposition^{[15, 16, 32, 36]}. Qualitatively, the derived values and the behavior of $W_{\rm m}$ have a reasonable agreement with the values obtained for ZnPc^{[25, 28]} and for amorphous Ga$_{2}$S$_{3}$–Ga$_{2}$Se$_{3}$ thin films^[29].

4. Summary and conclusions

In this work, the AC-conductivity, capacitance and loss tangent data of different phthalocyanine compounds (ZnPc, CuPc, FePc, H$_{2}$Pc) were employed to extract and study their AC-electronic and dielectric parameters. The dielectric function (constant) is influenced greatly by temperature and frequency. The dependence of both parts of the dielectric constant (real and imaginary), on temperature and frequency varies upon the frequency-and/or temperature-range of interest. At low temperatures, the real part of the dielectric function saturates and approaches a constant value regardless of electrode type (Al and Au). The real part was found to be proportional to $\omega^{n}$, while the imaginary part proportional to $\omega^{m}$, where $\omega$ is the angular frequency, $n$ and $m$ are temperature dependent constants and their values depend on the specific frequency range. Also, both $n$ and $m$ have values less than 0 ($n$, $m$ < 0). Qualitatively, the behavior of the real part of the dielectric constant of phthalocyanine compounds (ZnPc, CuPc, FePc, and H$_{2}$Pc) as a function of temperature and frequency were almost similar, but with some differences in their values. This variation in the $\varepsilon_{1}$ values may be accounted for the electronic structure of the central atom, accumulation of charges in the sample, differences in film thickness, and material type of the electrodes.

The calculated relaxation time, $\tau$, was found to depend on both temperature and frequency for all Pc-compounds. The value of $\tau$ depends on frequency, temperature and the type of the central atom. The relaxation activation energy is then determined from the slopes of the fitted lines of ln $\tau$ and the reciprocal of the temperature. The low values of the activation energy ($\sim$ 0.01 to 0.05 eV) were accounted and consistent with the hopping of charge carriers between localized states.

The maximum barrier height, $W_{\rm m}$, was found to be temperature and frequency dependent for all phthalocyanine compounds. The value $W_{\rm m}$ depends greatly on the nature of the central atom and electrode material type. The correlated barrier hopping (CBH) model was found to be the appropriate mechanism to describe the charge carrier's transport in phthalocyanine films.

In conclusion, the present results demand an extension of the study to include different electrode materials, more metal-phthalocyanine compounds, extending the frequency and temperature ranges. Such a study extension will allow us to refine the theoretical models applicable to such organic compounds and their interesting electrical and dielectric properties. Additionally, comprehensive X-ray diffraction (XRD) and electronic structure studies for such compounds are crucial and demanding to understand their behavior due to presence of a certain metallic element in the core of the phthalocyanine compounds.