A. Optical properties

In situ monitoring of optical transmission (absorbance) in UV–VIS was applied to examine variation of the optical properties with film thickness, fluence, and repetition rate. Absorbance was also used to determine fluence threshold (F _th) for deposition of ZnPc thin film. The absorbance measurement was usually taken after each 15,000 shots. The evaluation of absorbance as a function of both fluence (which varied from 10 to 100 mJ/cm²) and number of shots (which varied from 15,000 to 45,000) at repetition rate f _L of 50 and 200 Hz is depicted on Figs. 1–3.

FIG. 1.

In situ absorbance evaluation for ZnPc film growth at repetition rate of 50 Hz and fluence: (a) 10 mJ/cm², (b) 20 mJ/cm², (c) 40 mJ/cm², (d) 70 mJ/cm², (e) 100 mJ/cm² and at repetition rate of 200 Hz and fluence: (f) 20 mJ/cm². Absorbance was measured after each 15,000, 30,000, and 45,000 laser shots.

FIG. 2.

Evolution of absorbance of (a) B, (b) Q₁ and (c) Q₂ bands during the ZnPc film growth for laser fluence (10–100 mJ/cm²) and repetition rate of 50 and 200 Hz.

FIG. 3.

Dependence of growth rate on fluence at repetition rate of 50 Hz derived from in situ absorbance (B, Q₁ and Q₂ absorption bands).

We can observe representative phthalocyanine absorption B and Q bands in the spectra in Fig. 1. Both these bands arise due to the π–π* transitions and the lowest allowed n–π* transition might also be found in the B band. ^{Reference Senthilarasu, Sathyamoorthy and Kulkarni45–Reference Edwards and Gouterma47} ZnPc does not present ligand to metal and metal to ligand transfer bands due to fully filled Zn²⁺ d ¹⁰-configuration, which simplifies the spectra in relation to other metal phthalocyanines. B band is equivalent to γ or the Soret band in porphyrins ^{Reference El-Nahass, Zeyada, Aziz and El-Ghamaz48} and is due to a _2u(π) → e _g(π*) together with b _2u(π) → e _g(π*) transitions. B band wave length of maximum intensity λ_max was varied in range 337–342 nm in the spectra. Q band, equivalent to the α band in porphyrins, is due to a _1u(π) → e _g(π*) transitions, with the dipole moment in the plane of the molecule. ^{Reference Henriksson and Sundbom46} Hence this band is very sensitive to the crystal structure and can also be exploited to calculate the average molecular tilt angle of the ZnPc molecules. ^{Reference Senthilarasu, Sathyamoorthy and Kulkarni45,Reference Louis, Lehmann, Friedrich and Zahn49,Reference Barrett, Borkowska, Humphreys and Parsons50} Q band is split due to so-called Davydov splitting (exciton coupling between two nonequivalent molecules in a unit cell) ^{Reference Kadish, Smith and Guilard10} into two bands Q₁ and Q₂, λ_max around 630–638 and 681–695 nm, respectively. The Q₁ and Q₂ bands are identified as the Q band of the monomer and the dimer or higher aggregates, respectively. ^{Reference Maggioni, Manera, Spadavecchia, Tonezzer, Carturan, Quaranta, Dejulianfernandez, Rella, Siciliano and Dellamea11} Davydov model suggests shifts of the absorption bands—a blue shift of the Q-band is predicted in the case of a cofacial alignment of two or more molecules and a red shift of the Q-band, on the contrary, is expected in the case of coplanar dimers or stripes, as long as a strong coupling of electronic states still occurs.

We could notice a very weak increase of absorbance at F around 10 mJ/cm², shown in Figs. 1(a) and 2, while a significant increase was obtained for F above 20 mJ/cm², shown in Figs. 1(b)–1(f) and 2. The dependence of absorbance on the number of shots is rather linear and its slope is increasing with increasing F that might be attributed to increasing growth rate. ^{Reference Guy, Bensalah-Ledoux, Lambert, Guillin, Guy and Mulatier28} We used linear regression fit to visualize absorbance trend in Fig. 2. Absorbance of a sample deposited at F = 40 mJ/cm² was measured after each 7500 laser shots to verify more accurately evaluation of the absorbance with number of shots. The plots of the absorbance of B, Q₁, and Q₂ bands (proportional to the deposition rates) versus F are presented in Fig. 3 for f _L = 50 Hz. We assume that the deposited material thickness is proportional to the optical absorption following the Beer–Lambert law. The proportionality factor is deduced by comparing the thickness of “thick” samples measured by profilometry with the intensity of particular absorption band (B, Q₁, and Q₂).

We could establish a fluence threshold F _th which defines two regimes: below F _th, the deposition rate is very low and above F _th, it increases rapidly and depends quasi-linearly on the fluence (Fig. 3). From the curves, we define the ablation threshold F _th as the fluence value where the line describing the regime at high fluence crosses the abscissa axis. Based on absorbance evaluation we could estimate the fluence threshold F _th ∼ 10 mJ/cm².

We could reach the highest growth rate GR = 10 nm/min at F = 100 mJ/cm² at f _L = 50 Hz. The growth rate can be also substantially enhanced when elevated f _L = 200 Hz is applied. We can follow this fact at the samples grown at same F = 20 mJ/cm² and different f _L values (ZnPc_B and ZnPc_B1) comparing Figs. 1(b), 1(f), and 2 and the data in Table I. We reached higher absorbance at f _L = 200 Hz than at f _L = 50 Hz for the same number of shots. The enhancement of the GR (the ratio of GR for these samples is ∼5.3) is even higher than expected ratio of particular f _L (200/50 = 4), as one can see in Fig. 2 and Table I. We can attribute this effect to deposition of higher amount of material due to elevated heating and melting of the target surface at f _L = 200 Hz.

The relative intensities of Q₁ and Q₂ bands give insight to the phases of ZnPc molecules. ^{Reference Kadish, Smith and Guilard10,Reference Debe51–Reference Chowdhury, Biswas, Majumder, Sanyal and Mallik55} The spectra depicted on Fig. 1, where Q₂ band dominates, would correspond to β-form structure of ZnPc films. On the other hand λ_max of Q₂ band would rather correspond to α-form ΖnPc comparing with experimental data in literature because λ_max of Q₂ band for β-ZnPc was reported at wavelengths longer than 700 nm. ^{Reference Wojdyła, Derkowska, Łukasiak and Bała54,Reference Senthilarasu, Hahn and Lee56,Reference Cai, Crossley, Reimers, Kobayashi and Amos57}

The absorbance values of B and Q bands, when we take into account film growth rate depicted in Fig. 3, are higher than those for films evaporated on substrate at room temperature ^{Reference El-Nahass, Zeyada, Aziz and El-Ghamaz48} but they are rather comparable with those measured on films evaporated at elevated substrate temperature ^{Reference Senthilarasu, Hahn and Lee56} or annealed films. ^{Reference El-Nahass, Zeyada, Aziz and El-Ghamaz48} This finding confirms the growth of different ZnPc phase by PLD compared to evaporation. We can see in the spectra in Fig. 1 that λ_max of Q₂ band is increasing with increasing fluence while λ_max of Q₁ band is not varying with fluence only is slightly decreasing with increasing film thickness. Red shift of λ_max of Q₂ was observed with increasing substrate temperature ^{Reference Senthilarasu, Sathyamoorthy and Kulkarni45} and after thermal annealing ^{Reference El-Nahass, Zeyada, Aziz and El-Ghamaz48} of vacuum evaporated ZnPc films. It could be also correlated with variation of crystalline quality together with surface roughness.

We performed DFT calculations to shed more light on relationship between the optical and structural properties. The DFT functionals B3LYP, PBE0, CAM-B3LYP fail to correctly describe measured UV–VIS spectra. B3LYP and PBE0 functionals do not correctly describe this system, because these functionals include incorrect asymptotic forms of the exchange-correlation potential. ^{Reference Cai, Crossley, Reimers, Kobayashi and Amos57} These functionals can fail to describe charge transfer, especially when electron is moved over large distances. CAM-B3LYP functional is based on long-range correction of the exchange potential developed by Tawada et al. ^{Reference Tawada, Tsuneda, Yanagisawa, Yanai and Hirao25} and applied to modify B3LYP functional. However, UV–VIS spectrum calculated by CAM-B3LYP overestimated experimental UV–VIS spectra and shifts the first and second allowed electronic states to higher transition electronic energy. UV–VIS spectrum simulated by LC-BLYP DFT functional shows very good agreement with experimental UV–VIS spectra (extinction coefficient k of sample ZnPc_E deposited at F = 100 mJ/cm²) as shown in Fig. 4. For this purpose, we have modeled a UV–VIS spectrum using a sequence of calculations with different number of monomers units, atomic basis sets, pseudopotentials and different values of cutoff of plane-wave kinetic energy, see Table SI. Our results show that the calculations for monomer units have only negligible value of Davydov splitting (up to 0.016 eV) and that oscillator strength for first and second allowed electronic transitions are almost the same. The calculations of UV–VIS spectra for tetramer and hexamer systems depicted in Fig. 5 show significant improvement. Davydov splitting is shifted up to 0.140 eV (tetramer systems) and 0.168 eV (hexamer system). Moreover, calculated oscillator strengths for first and second allowed electronic transitions are in the correct ratio (oscillator strength of first allowed electronic transition is higher about ∼24% than oscillator strength of second allowed electronic transition for tetramer system and oscillator strength of first allowed electronic transition is higher about ∼40% than oscillator strength of second allowed electronic transition for hexamer system). The comparison between experimental and theoretical position of λ_max of Q₁ and Q₂ bands and their intensities splitting ratio show that theoretical values calculated for hexamer systems lies very close to experimental values (630 up to 642 nm for Q₁ bands and 688 up to 702 nm for Q₂ band in comparison to experimental values of 630–638 nm for Q₁ band and 681–695 nm for Q₂ band) when considering β-form ZnPc. Hence, DFT/TD-DFT calculations revealed that the spectra in Fig. 1 could be attributed to β-phase ZnPc.

FIG. 4.

Comparison of Q band—extinction coefficient k for sample deposited at laser fluence of 100 mJ/cm² (ZnPc_E) and DFT simulation.

FIG. 5.

ZnPc hexamer used for VIS absorption spectra modelled by TD-DFT method. The figure shows the stacking axe b related to monoclinic crystal system.

B. FTIR spectroscopy

IR spectroscopy measured in fingerprint region (600–1800 cm⁻¹) was involved to characterize chemical structural properties and evaluate deterioration of the deposited ZnPc. FTIR spectra of the ZnPc films deposited at laser fluence 20, 70, and 100 mJ/cm² (f _L = 50 and 200 Hz) and the spectra of the target are depicted in Fig. 6. The experimental band positions and DFT calculated values are listed in Table II. Considering the spectrum of the target, the molecules are randomly distributed and the spectrum shows peaks at 725 and 775 cm⁻¹ corresponding to the out-of-plane vibrational modes of ZnPc (wagging of C_c–H bonds), while the band at 750 cm⁻¹ corresponds to bending motion of C_a–N_b–C_a and C_a–N_i–C_a ^{Reference Saini, Singh, Kaur, Kumar, Sathe and Tripathi58} and all other bands above 800 cm⁻¹ correspond to various in-plane vibrational modes of the molecules. ^{Reference Saini, Singh, Kaur, Kumar, Sathe and Tripathi58} Dominating bands at 1331 and 1485 cm⁻¹ correspond to metal–nitrogen stretching mode and displacement of C–N–C bridge bond for ZnPc molecules, respectively. ^{Reference Bala, Wojdyla, Rebarz, Szybowic, Drozdowski, Grodzicki and Piszczek59} The absorption at 1608 cm⁻¹ can be assigned to a general skeletal vibration of porphyrin ring system. ^{Reference Farag60} The shoulder appearing around 800 cm⁻¹ that was clearly pronounced only in the deposited films can be attributed to Zn–N _i vibration and isoindole breathing vibration. ^{Reference Saini, Singh, Kaur, Kumar, Sathe and Tripathi58} We can observe in the IR spectra of the films another absorption band of unknown origin at 1261 cm⁻¹ that is not presented in the spectra of the target. A comparison of the IR spectra of the target and the films shown in Fig. 6 and Table II generally suggest low deterioration of ZnPc material in the films.

FIG. 6.

FTIR spectra of ZnPc films and target—in dependence on laser fluence: 20, 70, and 100 mJ/cm² and laser repetition rate 50 and 200 Hz.

TABLE II.

Position of observed IR bands of ZnPc in PLD target and films (deposited at different fluence and repetition rate) and calculated by DFT.

Considering the out-of-plane peak positions reported for β-ZnPc films [728, 771 cm⁻¹ (Ref. Reference Louis, Lehmann, Friedrich and Zahn49); 729, 775 cm⁻¹ (Ref. Reference Bala, Wojdyla, Rebarz, Szybowic, Drozdowski, Grodzicki and Piszczek59); 727, 779 cm⁻¹ (Ref. Reference Chowdhury, Biswas, Majumder, Sanyal and Mallik55)] and α-ZnPc [719, 769 cm⁻¹(Ref. Reference Louis, Lehmann, Friedrich and Zahn49); 722, 771 cm⁻¹ (Ref. Reference Saini, Singh, Kaur, Kumar, Sathe and Tripathi58); 721, 769 cm⁻¹ (Ref. Reference Chowdhury, Biswas, Majumder, Sanyal and Mallik55)] makes difficult to unambiguously attribute the phase of the ZnPc target and the films according to the IR spectra. We might suggest the presence of both phases. The peak position could be significantly influenced due to the average crystallite size. In case of nanocrystalline phthalocyanine films of the average crystalline size smaller than 3 nm the contribution of the surface molecules may become considerable with respect to that of the core molecules and a perturbation of the IR features takes place. ^{Reference Maggioni, Quaranta, Carturan, Patelli, Tonezzer, Ceccato and Della Mea61}

C. Crystalline structure

The results from PXRD of prepared samples are shown in Fig. 7. There is an obvious presence of Bragg peaks at the 2θ positions of 6.88 and 11.4°, which clearly show that the prepared thin layers of ZnPc are crystalline. The phase analysis from PXRD patterns leads to the monoclinic or triclinic ZnPc phase with the space group C2/n and P ¹ respectively, ^{Reference Schunemann, Elschner, Levin, Levichkova, Leo and Riede62} they can be assigned using PDF-4+ 2013 database source. These Bragg maxima can be indexed as the h00 diffractions of the different orders which depends on phase. For triclinic phase, the order is of h = 1, 2 and for the monoclinic phase h = 2, 4 for the 6.88 and 11.4° of 2θ, respectively. The small intensity broader feature with at 9.39° of 2θ can be indexed as a $\left( {hkl} \right) = \left( {\bar 202} \right)$ reflection of the monoclinic phase or as a $\left( {hkl} \right) = \left( {\bar 101} \right)$ reflection of triclinic phase. This reflection was not found in thermally evaporated films at room temperature but appeared only in annealed samples. ^{Reference El-Nahass, Zeyada, Aziz and El-Ghamaz48} We could notice appearance of a broad peak centered at 2θ = 22.3° in XRD patterns that might imply that some of the ZnPc molecular macrocycles lay parallel to the substrate surface plane and their stacking axes are inclined to it. ^{Reference Senthilarasu, Hahn and Lee63} The broad peak can be also related to the contribution of amorphous phase. ^{Reference Saini, Singh, Kaur, Kumar, Sathe and Tripathi58} It should be seen that the full-widths of half maxima of the reflections in XRD patterns are broader, which indicates the presence of rather small crystallites of ZnPc.

FIG. 7.

XRD pattern of ZnPc film [hkl notation for triclinic $\left( {100} \right)$ , $\left( {\bar 101} \right)$ , $\left( {200} \right)$ and monoclinic $\left( {200} \right)$ , $\left( {\bar 202} \right)$ , $\left( {400} \right)$ crystalline phase, respectively].

D. Electrical properties

We noticed a variation of the resistivity of orders of magnitude therefore the measurements were performed at constant voltage (typically −30 and 30 V), which avoided large changes of the potential difference on voltage contacts. The resistivity was then calculated using current values monitored by the source-meter. We observed slow charge relaxation, usually involving several hours. In the relaxed state, we obtained resistivity ρ ≈ 1.5 × 10⁹ Ω cm measured in air ambient for the sample deposited at F = 70 mJ/cm² (ZnPc_D). When the sample was placed in vacuum (pressure of 10 Pa) the resistivity increased, at least, by one order of magnitude (ρ > 10¹⁰ Ω cm), while the observed resistance became comparable with the measurement limit of our experimental setup. Sample deposited at higher fluence F = 100 mJ/cm² (ZnPc_E) revealed lower resistivity ρ ∼ 5 × 10⁸ Ω cm in air. It might suggest formation of trapping sites at higher laser fluence, which reduces the forbidden energy band ^{Reference Novotny, Bulir, Bensalah-Ledoux, Guy, Fitl, Vrnata, Lancok and Moine16} and resistivity. Measured resistivity values are in agreement with ρ ≈ 4.7 × 10⁷ Ω cm for α-ZnPc and ρ ≈ 4.3 × 10¹¹ Ω cm for β-ZnPc mentioned in literature. ^{Reference Kadish, Smith and Guilard10}