Thin Films of Chlorinated Vanadyl Phthalocyanines as Active Layers of Chemiresistive Sensors for the Detection of Ammonia

Halogenated metal phthalocyanines are promising materials for the manufacture of active layers of chemiresistive sensors for the detection of various gases. Despite the high interest in such sensors, there are few systematic studies of the position of halogen substituents in phthalocyanine macroring on the chemiresistive response of their films to gases. In this work, we prepared and studied films of novel tetrachlorosubstituted vanadyl phthalocyanine derivatives with Cl substituents in the peripheral (VOPcCl4-p) and nonperipheral (VOPcCl4-np) positions of the phthalocyanine ring as active layers of chemiresistive sensors to reveal the effect of the position of substituents on their structure and sensor response to low concentrations of NH3. It was shown that the films of VOPcCl4-p exhibited a noticeably higher sensor response to NH3 than the VOPcCl4-np ones. The limit of detection of NH3 was 0.7 ppm. The sensing layers demonstrated a reversible sensor response at room temperature with fairly low response/recovery times. It was also demonstrated that NH3 can be detected in the presence of various interfering gases (CO2 and H2) and some volatile organic vapors, as well as in a mixture of gases with a composition close to exhaled air.


Introduction
Ammonia is a colorless gas with a specific smell.The main consumers of ammonia are agricultural enterprises, although it is also used in various household and industrial cleaning products, as well as in industrial metalworking processes, the paper industry, and industrial refrigeration units [1].Ammonia has a strong corrosive effect and causes severe irritation of the skin, eyes, mucous membranes of the mouth, and respiratory and digestive systems [2].Its concentration above 100 ppm can cause burning in the eyes, lacrimation, swelling of the eyelids, corneal abrasion, blurred vision, and even blindness.The detection of ammonia is an important task not only in industry but also in medicine [3,4].Ammonia is used as a biomarker for the diagnostics of disease by analyzing its content in breath air.For example, an increased level of NH 3 in exhaled air of more than 2 ppm is a sign of renal pathologies [5,6].
Currently, electrochemical and luminescent sensors, as well as gas chromatographymass spectrometry are most often used to analyze these gases in exhaled air [7][8][9].The development of new semiconductor sensors for the analysis of gas biomarkers is of particular interest due to their cheapness, portability, and speed.Oxides [10,11], transition metal nitrides [12,13], nanocarbon-containing materials [14], and metal phthalocyanines [4,15,16] are widely used as active layers of chemiresistive sensors.The latter attract attention due to the possibility of widely varying their resistive and electrocatalytic properties by changing their molecular structure (e.g., central cations and substituents in the macroring or axial properties by changing their molecular structure (e.g., central cations and substituents in the macroring or axial positions), as well as the morphology and orientation of their films.Another advantage of MPc layers is the reversibility of their response at room temperature without additional heating, as well as the possibility of creating sensing layers on flexible substrates [17][18][19].
Among the variety of metal phthalocyanines, phthalocyanines with various substituents in axial positions can be distinguished, for example, phthalocyanines of gallium and aluminum MXPc (M = Al, Ga; X = Cl, OH, F), as well as phthalocyanines of titanyl and vanadyl (TiOPc and VOPc).Such phthalocyanines have a packaging type different from phthalocyanines M (II) Pc (M = Mg, Cu, Co, Ni), as they have a flat structure and additional places of interaction with gas analytes [20,21].For instance, unlike other 3D-metal phthalocyanines, VOPc and its substituted derivative, VOPcRn (n = 4,8,16), are nonplanar molecules with vanadyl groups located perpendicular to the plane of aromatic rings [22,23].This leads to two-dimensional π-π stacking in the crystal lattice and, as a consequence, to shorter intermolecular distances and greater overlap of the π-π stacking compared to planar phthalocyanines, which ultimately contributes to the higher mobility of charge carriers [24].These properties open up the prospect of using them in highly productive organic thin-film field-effect transistors (OFETs) and sensors.
Ultrathin vanadyl phthalocyanine films were used as sensing layers of chemiresistors for the detection of nitrogen dioxide at room temperature [25].Surface-type capacitive humidity sensors based on evaporated VOPc films were studied by two different groups of authors [26,27].Ji, et al. [28] fabricated an ultra-thin highly ordered heterojunction-structured device based on multilayers of VOPc, N,N′-diphenyl perylene tetracarboxylic diimide (PTCDI-Ph), and para-hexa-phenyl (p-6P), which was very sensitive to NO2 and had a nearly five times larger response than the corresponding single-heterojunction device.
Many authors show in their works that the introduction of halogen substitutes into the phthalocyanine macrocycles results in improvements in their sensitivity to electron-donating gases [29,30].In tetrasubstituted phthalocyanine molecules, halogen substituents can be introduced into both peripheral (MPcX4-p) and nonperipheral (MPcX4-np) positions of the phthalocyanine macroring (Figure 1).In our previous work, thin films of unsubstituted and tetrafluorosubstituted vanadyl phthalocyanine derivatives (VOPc and VOPcF4-p) were used for the detection of gaseous ammonia and hydrogen.It was shown that VOPcF4-p films demonstrated a 2-3 times higher sensor response to NH3 compared to VOPc films [31].In turn, a comparative analysis of the sensor response to ammonia showed that the films of VOPcF4-p exhibited a greater sensor response than those of VOPcF4-np [32].This result correlated well with DFT calculations, which revealed stronger binding of NH3 and VOPcF4-p molecules.It was shown using the example of zinc phthalocyanines that films of chlorosubstituted metal phthalocyanines demonstrated an even higher chemiresistive sensor response to ammonia than fluorinated ones [30], but the structure and sensor properties of chlorosubstituted oxometallo phthalocyanine films have not yet been studied.
In this work, we prepared and studied films of novel tetrachlorosubstituted vanadyl phthalocyanines with Cl substituents in the peripheral (VOPcCl 4 -p) and nonperipheral (VOPcCl 4 -np) positions of the phthalocyanine ring (Figure 1) as active layers for detecting low concentrations of NH 3 and compared their sensor performance with their fluorinated analogs.The novelty of this work consists of expanding the set of metal phthalocyanines suitable for use as sensor layers and revealing the influence of the position of substituents in the macroring on the structure of films and the sensor response to the gaseous analytes.The sensor characteristics of the sensors (viz reversibility, limit of detection, response, and recovery times) were determined.The sensing layers were tested in the presence of various interfering gases (CO 2 and H 2 ) and some volatile organic vapors, as well as in a mixture of gases simulating exhaled air.
Thicknesses of the VOPcCl 4 -p and VOPcCl 4 -np films were 82 and 87 nm, respectively.The choice of film thickness was based on preliminary studies.It was previously shown that films of smaller thickness often do not form a continuous film on the surface of the electrodes, which prevents their good conductivity.Films with a thickness of more than 100 nm have low sensor sensitivity and a long response time due to the difficulty of gas diffusion into the film.The thickness of the films was determined through spectroscopic ellipsometry, using a spectroscopic ellipsometer ELLIPS 1771 SA (ISP, Novosibirsk, Russia) according to the previously described technique [33].
An XRD study of thin films and bulk powders of VOPcCl 4 -p and VOPcCl 4 -np was performed using a Bruker D8 Advance powder diffractometer (Bruker AXS Inc., Billerica, MA, USA) in the Bragg-Brentano scheme with a vertical θ-θ goniometer (CuKα sealed tube, 0.01 • scan step).
AFM images of the films were obtained using the Ntegra Prima II nanolaboratory (NT-MDT, Moscow, Russia).The HA_NC probe parameters were described in our previous work [30].A scanning electron microscope (JEOL-JSM 6700 F, Tokyo, Japan) was used for the characterization of the films' surface morphology.
Optical absorption spectra were recorded with (OKB SPECTR LLC, Saint-Petersburg, Russia), while Raman spectra were obtained using a LabRAM Horiba single spectrometer (Montpellier, France) (488 nm line of an Ar+ laser).

Measurements of the Sensor Response
The chemiresistive response of VOPcCl 4 -p and VOPcCl 4 -np films was investigated by measuring the change in resistance upon the introduction of ammonia in the concentration range of 1-50 ppm.The resistance was measured at the constant voltage of 10 V, using an electrometer Keithley 236 (Tektronix Inc., Beaverton, OR, USA).To measure the sensor response, ammonia of the required concentration and air used for purging were alternately injected into the flow cell.Most of the experiments were conducted in dynamic mode.In this mode, ammonia was injected at a constant flow rate of 300 mL/min.The exposure time was 15 s.Static mode was used only to determine the response and recovery times of the sensors.In static mode, the airflow was passed through the chamber until the film resistance reached a constant value, and then the chamber valves were closed and NH 3 of the required concentration was injected into the cell.The resistance was recorded after the saturation current was reached.After that, the chamber was opened and blown by a stream of air.All measurements were carried out at room temperature (22 The sensor response values were calculated: (R − R o )/R o .(R is the steady resistance of the film at a certain analyte concentration, and R o is the initial resistance of the film in air).The measurements of the response of three deposited films were used to calculate the standard deviation.The source of NH 3 was a tank with pure commercial ("Chistye Gasy+", Novosibirsk, Russia) gas (1%) diluted in dry argon.The required gas concentrations were set using the mass flow controllers.Dry air was used as a diluting and purging gas.

Theoretical Calculations
The quantum-chemical approaches used here for the calculations of the geometric and electronic structures and total energies of VOPcCl 4 -p/NH 3 -x and VOPcCl 4 -np/NH 3y aggregates, where x = 1-4 and y = 2-4, have already been described in our previous work when performing a similar study of the interaction of ammonia molecule with fluorosubstituted phthalocyanines VOPcF 4 -p and VOPcF 4 -np [32].The values of x and y specify the position of the NH 3 molecule relative to the macrocycle (Figures 2 and 3).From this point of view, the VOPcCl 4 -np/NH 3 -1 aggregate was not considered because it was shown [32] that such an arrangement of the molecule over the phthalocyanine plane cannot be realized due to the small distance between two adjacent macrocycles in the stack of molecules.

Crystal Structure of VOPcCl4-p and VOPcCl4-np
Establishing the structure of single crystals is a necessary step before studying the structural features of thin films.Single crystals of vanadyl phthalocyanines were grown by sublimation in vacuum.Two types of single crystals were isolated from the VOPcCl4-p powder.Crystals in the form of very thin translucent greenish-blue ribbons were tetrag-

Crystal Structure of VOPcCl 4 -p and VOPcCl 4 -np
Establishing the structure of single crystals is a necessary step before studying the structural features of thin films.Single crystals of vanadyl phthalocyanines were grown by sublimation in vacuum.Two types of single crystals were isolated from the VOPcCl 4 -p powder.Crystals in the form of very thin translucent greenish-blue ribbons were tetragonal, while needle-like crystals with a dark purple metallic sheen were in a triclinic phase.The molecular structure and packing diagrams for both VOPcCl 4 -p crystal phases are shown in Figure 4. Unit cell parameters and refinement details for VOPcCl 4 crystal structures are summarized in Table 1.Due to the axial oxygen atom, the VOPcCl 4 -p molecule is not flat and has the shape of a "shuttlecock", while the vanadium atom is displaced from the plane of the macrocycle by 0.399 Å, and the V = O bond is oriented perpendicular to the plane of the macrocycle (Figure 4a).Since VOPcCl 4 -p contains a mixture of regioisomers with peripheral chlorine atoms in different positions, each chlorine atom in the tetragonal crystal structure of VOPcCl 4 -p is disordered in two positions with a ratio of 0.514/0.486.VOPcCl 4p molecules in the tetragonal polymorph pack in vertical stacks along the c axis (Figure 4b,c) with the distance between adjacent molecules in a stack of 3.729 Å.The entire VOPcCl 4 -p molecule appears to be disordered by a mirror plane perpendicular to the tetragonality axis, so it seems that the VOPcCl 4 -p molecules are oriented "up" and "down" at the same time.This means that individual molecular stacks are randomly oriented either "up" or "down", and there is no long-range symmetry in their ordering.This style of packing and disordering is not unique and was previously observed in crystals of halogenated lead phthalocyanines PbPcF 4 -p [38] and PbPcCl 4 -p [39].
plane perpendicular to the tetragonality axis, so it seems that the VOPcCl4-p molecules are oriented "up" and "down" at the same time.This means that individual molecular stacks are randomly oriented either "up" or "down", and there is no long-range symmetry in their ordering.This style of packing and disordering is not unique and was previously observed in crystals of halogenated lead phthalocyanines PbPcF4-p [38] and PbPcCl4-p [39].     Figure 4d shows a VOPcCl 4 -p molecule in triclinic polymorph.While the phthalocyanine core retains the same geometry as in the tetragonal polymorph, there is a clear difference in the distribution of peripheral chlorine atoms in alternative positions.While in the tetragonal polymorph, the distribution is close to 0.5/0.5, in the triclinic polymorph, this ratio is 0.691/0.309,which affects the molecular packaging.In the triclinic polymorph, the VOPcCl 4 -p molecules are assembled into chains, with the molecules alternating between an "up" and "down" orientation (Figure 4f).It is worth noting that the sites of chlorine atoms with low occupancy values (<0.5) are those between neighboring molecules in the chain, which means that close Cl. ..Cl contacts (<2.4 Å) between molecules in the same chain are most likely never present in a real crystal.Thus, the actual closest Cl. ..Cl contact between molecules in adjacent molecular chains in the triclinic VOPcCl 4 -p polymorph is 3.019 Å, which is still significantly shorter than in the tetragonal polymorph (3.364 Å).The distances between adjacent molecules in the chains are 3.217 Å and 3.299 Å.The individual molecular chains are arranged in the form of 2D layers that are packed parallel to the crystallographic plane (−222) (Figure 4e).The distance between molecules in adjacent layers is 3.095 Å, and the angle between the molecular layers and the plane (−222) is ~0.82 • .
VOPcCl 4 -np crystallizes in the space group P-1 with Z = 2 and Z' = 1 and one symmetrically independent molecule in the unit cell (Table 1).The VOPcCl 4 -np molecular structure is shown in Figure 5a.The vanadium atom is displaced from the macrocycle plane by 0.553 Å, and the angle between the V=O bond and the normal to the macrocycle plane is 0.51 • .Four chlorine atoms are disordered in two positions each.Their occupancy ratio is similar to the triclinic VOPcCl 4 -p polymorph and significantly deviates from the ratio 0.5/0.5.VOPcCl4-np molecules are packed into two-dimensional layers parallel to the plane (2−22) (Figure 5b).The distance between molecules in one layer is 3.298 Å, and between adjacent layers, 3.239 Å.The angle between the molecular layers and the plane (2−22) is 0.15°.However, unlike VOPcCl4-p, VOPcCl4-np molecules do not form separate chains within one molecular layer but are arranged in such a way that each molecule faces three other molecules (Figure 5c).

VOPcCl4-p and VOPcCl4-np Thin Films and Powder XRD Study
Experimental XRD patterns of VOPcCl4-p and VOPcCl4-np powders and thin films (before and after annealing) are shown in Figure 6 in comparison with the powder patterns calculated based on the abovementioned single-crystal data.The first three peaks on the XRD pattern of the VOPcCl4-p powder (6.38°, 7.19°, and 7.76° 2θ) correspond to the plane (110) of the tetragonal phase and the planes (001) and (010) of the triclinic phase, indicating that both polymorph phases are present in a significant amount in the powder.The strong peak at 27.03° dominates the entire diffraction pattern and belongs to the plane (−222) of the triclinic polymorph, which suggests that the VOPcCl4-p powder consists mainly of the triclinic phase.terns calculated based on the abovementioned single-crystal data.The first three peaks on the XRD pattern of the VOPcCl4-p powder (6.38°, 7.19°, and 7.76° 2θ) correspond to the plane (110) of the tetragonal phase and the planes (001) and (010) of the triclinic phase, indicating that both polymorph phases are present in a significant amount in the powder.The strong peak at 27.03° dominates the entire diffraction pattern and belongs to the plane (−222) of the triclinic polymorph, which suggests that the VOPcCl4-p powder consists mainly of the triclinic phase.XRD patterns of both VOPcCl 4 -p and VOPcCl 4 -np films contain peaks that do not match the single-crystal data.The diffractogram of the VOPcCl 4 -p film is characterized by one strong diffraction peak at 6.06 • 2θ and a peak at 12.15 • 2θ with a weaker intensity, and the diffractogram of the VOPcCl 4 -np film contains a diffraction peak at 6.02 • 2θ, which belongs to unknown crystal phases.The absence of other strong diffraction peaks indicates that the films have a preferred orientation.The difference in the phase composition of films from single crystals is often observed for phthalocyanines due to the tendency to form different polymorphic modifications depending on temperature [40].To check the effect of the film annealing on their structure, the films were annealed at 200 • C for 2 h in air.Annealing resulted in a small shift in peak positions toward larger angles but did not change the overall appearance of the diffraction pattern, which can be caused by a release of mechanical stress and a decrease in the number of crystal lattice defects in thin films during annealing.
The optical absorption spectra of VOPcCl 4 -p and VOPcCl 4 -np films were also recorded.The spectra of VOPcCl 4 -p and VOPcCl 4 -np films are compared with the spectra of their solutions in dimethylformamide in Supplementary Figure S1.The spectra of VOPcCl 4p and VOPcCl 4 -np solutions are typical for the phthalocyanines, which are present in monomeric form in solutions [41], and have the Q bands with the maxima at 680 and 689 nm, respectively.The spectra of the films become broader because of intermolecular interactions in solids.Similar to the spectra of unsubstituted VOPc [42] and VOPcF 4 [43] films, the maxima of Q bands shift to the higher wavelengths, i.e., to 778 and 795 nm of VOPcCl 4 -p and VOPcCl 4 -np films, respectively.This view of spectra is a characteristic of J-aggregate formation [44].Films' annealing at 200 • C for 2 h does not lead to noticeable changes in the intensity and position of the peaks.
VOPcCl 4 -p and VOPcCl 4 -np films were also characterized by Raman spectroscopy to demonstrate that their chemical composition coincides with that of the powders.Most of the peak positions in the spectra of the films coincide well with those in the spectra of powders (Supplementary Figure S2).The relative intensities of several peaks are different because of the preferred orientation in thin films, which is especially noticeable in the spectra of VOPcCl 4 -p samples.
The morphology of VOPcCl 4 -p and VOPcCl 4 -np films was studied by AFM.The microscopy images of the films before and after annealing (200 • C, 2 h) are shown in Figure 7.The images show that both films are formed by spherical crystallites ranging in size from 50 to 200 nm.The root mean square (RMS) roughness value is 10.4 and 9.9 nm for the VOPcCl 4 -p and VOPcCl 4 -np films, respectively.After annealing in air, the crystallites increase slightly in size and become more clearly defined and more uniform in size.At the same time, the value of the RMS roughness of the VOPcCl 4 -p film decreases to 8.9 nm after annealing, but for VOPcCl 4 -np, it increases to 13.0 nm.The films were also characterized by the SEM method (Supplementary Figure S3).The surface of the films in the SEM images looks rough, with indistinctly pronounced rounded grains that become slightly larger after the heat treatment of the films.The concentration of NH 3 varied in the range from 1 to 50 ppm.The curves of changes in the sensor response, when different concentrations of NH 3 were introduced, are shown in Figure 8a,b.The measurements of the sensor response were conducted in dynamic mode.In this mode, NH 3 was injected at a constant airflow rate, and its exposure time was 15 s.The resistance of the films of both phthalocyanines increased after the introduction of ammonia into the cell, which was usually observed for metal phthalocyanine films exhibiting p-type semiconductor properties, for example, MPc and MPcF 4 films [31].The sensor response to NH 3 was completely reversible at room temperature in the investigated concentration range.The dependence of the sensor response of VOPcCl 4 -p (a) and VOPcCl 4 -p (b) films on NH 3 concentration is shown in Figure 8c.The change in the position of the chlorine substituents in the benzene rings strongly affects the value of the sensor response.For example, the value of the sensor response of a VOPcCl 4 -p film at 10 ppm is 1.6 times higher than that of a VOPcCl 4 -np film.A similar trend was previously found for chlorosubstituted zinc phthalocyanines [30].The dependence of the response on the concentration is linear in the range from 1 to 10 ppm in the case of the VOPcCl 4 -p film and from 1 to 5 ppm in the case of the VOPcCl 4 -np film, thus making it possible to construct calibration curves in these concentration ranges and to calculate the limit of detection (LOD) according to the formula 3σ/m (where σ is the standard deviation of the sensor response to 1 ppm of ammonia, while m is the slope of the calibration plot).The calibration curves with the parameters used for the LOD calculation are presented in Supplementary Figure S4.The LODs were estimated to be 0.7 and 0.9 ppm for VOPcCl 4 -p and VOPcCl 4 -np films, respectively.
Micromachines 2023, 14, x FOR PEER REVIEW 1 standard deviation of the sensor response to 1 ppm of ammonia, while m is the sl the calibration plot).The calibration curves with the parameters used for the LOD lation are presented in Supplementary Figure S4.The LODs were estimated to be 0 0.9 ppm for VOPcCl4-p and VOPcCl4-np films, respectively.To determine the response and recovery times of the sensing layer, the static was used.The average response/recovery times measured at 5 ppm NH3 were 60/30 40/270 s for VOPcCl4-p and VOPcCl4-np films, respectively.The films of both ph cyanines demonstrate good repeatability of the sensor response.For example, the of the sensor response to 1 ppm of NH3 is 0.055 ± 0.003 when gas is repeatedly in into the flow cell (Supplementary Figure S5).

Study of the Nature of Interaction between NH3 and VOPcCl4 Molecules and Comparison of the Sensor Response of VOPcCl4 and VOPcF4 Films
Quantum-chemical calculations of the energy Eb and the nature of NH3 bind the corresponding VOPcCl4-p/NH3-x and VOPcCl4-np/NH3-y aggregates were perfo to explain the difference in the sensor response of thin films of chlorosubstituted va phthalocyanines toward ammonia.In this case, the interaction of a gas molecule macrocycles through their side atoms was considered (Figures 2 and 3), since it wa viously shown that this type of interaction best described the reversible experim sensor response of the films of phthalocyanines with other substituents and m [21,23].It was shown as a result of the calculations that, depending on the position o relative to VOPcCl4-p, the Eb values vary in the range from 0.171 eV to 0.180 eV (Ta and exceed the binding energies in the case of ammonia interaction with VOPc (0.149-0.153 eV).The stronger binding between NH3 molecules and VOPcCl4-p b To determine the response and recovery times of the sensing layer, the static mode was used.The average response/recovery times measured at 5 ppm NH 3 were 60/300 and 40/270 s for VOPcCl 4 -p and VOPcCl 4 -np films, respectively.The films of both phthalocyanines demonstrate good repeatability of the sensor response.For example, the value of the sensor response to 1 ppm of NH 3 is 0.055 ± 0.003 when gas is repeatedly injected into the flow cell (Supplementary Figure S5).

Study of the Nature of Interaction between NH 3 and VOPcCl 4 Molecules and Comparison of the Sensor Response of VOPcCl 4 and VOPcF 4 Films
Quantum-chemical calculations of the energy E b and the nature of NH 3 binding in the corresponding VOPcCl 4 -p/NH 3 -x and VOPcCl 4 -np/NH 3 -y aggregates were performed to explain the difference in the sensor response of thin films of chlorosubstituted vanadyl phthalocyanines toward ammonia.In this case, the interaction of a gas molecule with macrocycles through their side atoms was considered (Figures 2 and 3), since it was previously shown that this type of interaction best described the reversible experimental sensor response of the films of phthalocyanines with other substituents and metals [21,23].
It was shown as a result of the calculations that, depending on the position of NH 3 relative to VOPcCl 4 -p, the E b values vary in the range from 0.171 eV to 0.180 eV (Table 2) and exceed the binding energies in the case of ammonia interaction with VOPcCl 4 -np (0.149-0.153 eV).The stronger binding between NH 3 molecules and VOPcCl 4 -p bearing Cl substituents in peripheral positions correlates well with the higher sensor response of a VOPcCl 4 -p film to ammonia compared to VOPcCl 4 -np.A similar effect was observed earlier for tetrafluorosubstituted phthalocyanines [23].The interaction of ammonia molecules with the side atoms of phthalocyanines occurs due to the formation of hydrogen bonds.This is evidenced by the values of the electron density ρ(r) and its Laplacian ∇ 2 ρ(r) at the corresponding bond critical points (BCPs) (Figures 2 and 3; Table 2).In particular, in the case of VOPcCl 4 -p/NH 3 , hydrogen bonds are formed between the hydrogen atoms of phthalocyanine and the ammonia nitrogen atom (BCP1 and BCP3) and between the bridging nitrogen atom of the macrocycle and one of the hydrogen atoms of the NH 3 molecule (BCP2).The corresponding values of ρ(r) and ∇ 2 ρ(r) are in the ranges of 0.091-0.147e/Å 3 and 0.974-1.559e/Å 5 .Similar results were also obtained in the case of aggregates with chlorine atoms in nonperipheral positions of phthalocyanine macroring.
The difference is that in the VOPcCl 4 -np/NH 3 -2 and VOPcCl 4 -np/NH 3 -3 compounds, one of the hydrogen bonds and the corresponding BCP3 occur between one of the H atoms of NH 3 and the Cl atom.In this case, the values of ρ(r) and ∇ 2 ρ(r) at all considered BCPs are lower and equal to 0.060-0.105e/Å 3 and 0.793-1.566e/Å 5 , respectively.Thus, the stronger NH 3 molecule binding in VOPcCl 4 -p/NH 3 aggregates compared to VOPcCl 4np/NH 3 , in particular, is due to higher electron density values in the corresponding BCPs.The conclusion that ammonia interacts with the side atoms of VOPcCl 4 molecules by forming hydrogen bonds is because the values of ρ(r) and ∇ 2 ρ(r) are in the ranges of 0.013-0.236e/Å 3 and 0.578-3.350e/Å 5 , respectively (Table 2) [45].
The increase in resistance during the adsorption of NH 3 molecules, which are electron donors, is typical for p-type semiconductors [46], which include thin films of tetrahalo-gensubstituted phthalocyanines.Therefore, the mechanism of sensor response can be represented using the scheme shown in Figure 9. Ammonia molecules donate electrons to VOPcCl 4 macrocycles due to interaction with their side atoms, thus leading to a decrease in the concentration of holes, i.e., charge carriers, and, consequently, to an increase in the resistance of phthalocyanine films.The stronger the interaction or the higher the gas concentration, the greater the charge transfer and the higher the sensor response.to VOPcCl4 macrocycles due to interaction with their side atoms, thus leading to a decrease in the concentration of holes, i.e., charge carriers, and, consequently, to an increase in the resistance of phthalocyanine films.The stronger the interaction or the higher the gas concentration, the greater the charge transfer and the higher the sensor response.It should also be noted that the Eb value of the NH3 molecule with VOPcCl4-p is slightly larger than with VOPcF4-p (0.170-0.174 eV), which was estimated earlier using the same computational approaches.This is consistent with a stronger sensor response of VOPcCl4 films compared to VOPcF4 ones.For example, the sensor response ((R − Ro)/Ro) of the VOPcCl4-p film was 0.15 vs. 0.14 (at 5 ppm NH3) in the case of a VOPcF4-p film [32].This difference was even more noticeable at the higher ammonia concentrations (Supplementary Figure S6).

Characteristics of the Sensor Based on VOPcCl4-p Films
More detailed sensor characteristics were obtained for VOPcCl4-p films, which possess higher sensitivity to ammonia than VOPcCl4-np.First, the effect of annealing the films on the value of their sensor response to ammonia was investigated, since it was shown above that the annealing led to a change in the films' morphology (Figure 7).The dynamical curve of the sensor response of the annealed VOPcCl4-p film is shown in Figure 10a.Figure 10b compares the dependencies of the sensor response of VOPcCl4-p films before and after annealing on the NH3 concentration ranging from 1 to 5 ppm.The annealing of VOPcCl4-p leads to an increase in the sensor response and the slope of the linear dependence of the sensor response on the concentration of NH3.As a result, the LOD was reduced to 0.3 ppm.The effect may be associated with an increase in the crystallinity of the films.It is known that the conductivity process in polycrystalline films of metal phthalocyanines is greatly influenced by the grain structure and size, as well as intergranular barriers.In several studies, it has been demonstrated that the higher the crystallinity of the film, the grain size, and the packing density of the crystallites, the greater the mobility of the charge carriers [47][48][49].The electrical properties of polycrystalline films are determined by the capture of charge carriers in areas localized at the grain boundaries, resulting in a decrease in the free charge carriers involved in conduction and a significant decrease in their mobility due to scattering by potential barriers arising at the boundaries.The larger the average grain sizes in a polycrystalline film, the smaller this barrier.An increase in the charge carrier mobility in the VOPcCl4-p film may contribute to a stronger change in the resistance of the film when interacting with NH3.It should also be noted that the E b value of the NH 3 molecule with VOPcCl 4 -p is slightly larger than with VOPcF 4 -p (0.170-0.174 eV), which was estimated earlier using the same computational approaches.This is consistent with a stronger sensor response of VOPcCl 4 films compared to VOPcF 4 ones.For example, the sensor response ((R − R o )/R o ) of the VOPcCl 4 -p film was 0.15 vs. 0.14 (at 5 ppm NH 3 ) in the case of a VOPcF 4 -p film [32].This difference was even more noticeable at the higher ammonia concentrations (Supplementary Figure S6).

Characteristics of the Sensor Based on VOPcCl 4 -p Films
More detailed sensor characteristics were obtained for VOPcCl 4 -p films, which possess higher sensitivity to ammonia than VOPcCl 4 -np.First, the effect of annealing the films on the value of their sensor response to ammonia was investigated, since it was shown above that the annealing led to a change in the films' morphology (Figure 7).The dynamical curve of the sensor response of the annealed VOPcCl 4 -p film is shown in Figure 10a.Figure 10b compares the dependencies of the sensor response of VOPcCl 4 -p films before and after annealing on the NH 3 concentration ranging from 1 to 5 ppm.The annealing of VOPcCl 4 -p leads to an increase in the sensor response and the slope of the linear dependence of the sensor response on the concentration of NH 3 .As a result, the LOD was reduced to 0.3 ppm.The effect may be associated with an increase in the crystallinity of the films.It is known that the conductivity process in polycrystalline films of metal phthalocyanines is greatly influenced by the grain structure and size, as well as intergranular barriers.In several studies, it has been demonstrated that the higher the crystallinity of the film, the grain size, and the packing density of the crystallites, the greater the mobility of the charge carriers [47][48][49].The electrical properties of polycrystalline films are determined by the capture of charge carriers in areas localized at the grain boundaries, resulting in a decrease in the free charge carriers involved in conduction and a significant decrease in their mobility due to scattering by potential barriers arising at the boundaries.The larger the average grain sizes in a polycrystalline film, the smaller this barrier.An increase in the charge carrier mobility in the VOPcCl 4 -p film may contribute to a stronger change in the resistance of the film when interacting with NH 3 .At the same time, the change in the sensor response of the VOPcCl 4 -np film after annealing is less pronounced than in the case of the VOPcCl 4 -p film.The recovery time for both vanadyl phthalocyanine films decreased after annealing.The recovery times for annealed VOPcCl 4 -p and VOPcCl 4 -np films (5 ppm) were 200 and 150 s, respectively.
Micromachines 2023, 14, x FOR PEER REVIEW 14 At the same time, the change in the sensor response of the VOPcCl4-np film afte nealing is less pronounced than in the case of the VOPcCl4-p film.The recovery tim both vanadyl phthalocyanine films decreased after annealing.The recovery time annealed VOPcCl4-p and VOPcCl4-np films (5 ppm) were 200 and 150 s, respectively The sensor response of as-deposited and annealed VOPcCl4-p films to ammonia ppm) was tested at different levels of relative humidity (RH) (Figure 11a,c).With a crease in RH from 5 to 40%, the sensor response of both VOPcCl4-p films practically not change; however, a further increase in RH to 60% leads to the growth of the se response by about 1.6 times.Similar to ammonia, the water molecule is also an ele donor; therefore, the simultaneous exposure of the film by the mixture of NH3 and leads to an increase in its sensor response.The sensor response of as-deposited and annealed VOPcCl 4 -p films to ammonia (1-5 ppm) was tested at different levels of relative humidity (RH) (Figure 11a,c).With an increase in RH from 5 to 40%, the sensor response of both VOPcCl 4 -p films practically does not change; however, a further increase in RH to 60% leads to the growth of the sensor response by about 1.6 times.Similar to ammonia, the water molecule is also an electron donor; therefore, the simultaneous exposure of the film by the mixture of NH 3 and H 2 O leads to an increase in its sensor response.
As already noted in the introduction, an increased level of ammonia in the exhaled air may be a sign of renal pathologies.The study of exhaled air for elevated levels of ammonia can be a convenient tool for the primary diagnosis of kidney diseases.To test the possibility of using phthalocyanine films in exhaled air, the measurements of their sensor response to some other gases (NO, CO 2 , and hydrogen) and volatile organic compounds (VOCs) that may be present in exhaled air, as well as to NH 3 in a mixture of gases similar in composition to exhaled air, were carried out.The diagrams exhibiting the sensor response of as-deposited and annealed VOPcCl 4 -p films to these analytes in comparison with ammonia are shown in Figure 11b,d.Similarly to fluorosubstituted metal phthalocyanines [31], the VOPcCl 4 -p films exhibit a significantly higher response to NH 3 than to the investigated VOCs and other gases, except for nitric oxide (Figure 11b,d).Note that the concentration of these gases is much higher than that of ammonia.In this case, NO is an interfering gas, and the detection of ammonia may be difficult, even if it is present in the gas mixture at the ppb level.This is especially important to take into account when studying biomarker gases in exhaled air.For example, in a healthy person, the NO concentration can be found in the concentration range of 10-50 ppb, while the ammonia concentration can rise from 0.5 ppm in a healthy person to around 2 ppm in the case of liver or kidney failure [50].Therefore, the effect of the presence of nitric oxide on the sensor response to ammonia is the subject of further investigation.As already noted in the introduction, an increased level of ammonia in the exhaled air may be a sign of renal pathologies.The study of exhaled air for elevated levels of ammonia can be a convenient tool for the primary diagnosis of kidney diseases.To test the possibility of using phthalocyanine films in exhaled air, the measurements of their sensor response to some other gases (NO, CO2, and hydrogen) and volatile organic compounds (VOCs) that may be present in exhaled air, as well as to NH3 in a mixture of gases similar in composition to exhaled air, were carried out.The diagrams exhibiting the sensor response of as-deposited and annealed VOPcCl4-p films to these analytes in comparison with ammonia are shown in Figure 11b,d.Similarly to fluorosubstituted metal phthalocyanines [31], the VOPcCl4-p films exhibit a significantly higher response to NH3 than to the investigated VOCs and other gases, except for nitric oxide (Figure 11b,d).Note that the concentration of these gases is much higher than that of ammonia.In this case, NO is an interfering gas, and the detection of ammonia may be difficult, even if it is present in the gas mixture at the ppb level.This is especially important to take into account when studying biomarker gases in exhaled air.For example, in a healthy person, the NO concentration can be found in the concentration range of 10-50 ppb, while the ammonia concentration can rise from 0.5 ppm in a healthy person to around 2 ppm in the case of liver or kidney failure [50].Therefore, the effect of the presence of nitric oxide on the sensor response to ammonia is the subject of further investigation.
Figure 12 shows the sensor response of an annealed VOPcCl4-p film to ammonia (1-4 ppm) in a mixture of gases simulating exhaled air in comparison with that measured in moist air.This mixture contained N2 (76%), O2 (16%), H2O (5%), and CO2 (3%).It should be noted that the relative humidity of the moist air and the gas mixture was the same and was maintained at 80-85%.It can be seen from Figure 11 that the sensor response to ammonia in this mixture is almost the same as in moist air.Thus, films based on Figure 12 shows the sensor response of an annealed VOPcCl 4 -p film to ammonia (1-4 ppm) in a mixture of gases simulating exhaled air in comparison with that measured in moist air.This mixture contained N 2 (76%), O 2 (16%), H 2 O (5%), and CO 2 (3%).It should be noted that the relative humidity of the moist air and the gas mixture was the same and was maintained at 80-85%.It can be seen from Figure 11 that the sensor response to ammonia in this mixture is almost the same as in moist air.Thus, films based on chlorinated vanadyl phthalocyanines can potentially be used to detect the low ammonia concentration in human-exhaled air.Tests of real exhaled samples are beyond the scope of this work, as they require the involvement of other methods to study their composition.chlorinated vanadyl phthalocyanines can potentially be used to detect the low ammonia concentration in human-exhaled air.Tests of real exhaled samples are beyond the scope of this work, as they require the involvement of other methods to study their composition.

Figure 12.
The sensor response of a VOPcCl4-p film to ammonia (1-4 ppm) in moist air (a) and in a mixture of gases similar in composition to exhaled air (b).RH of the moist air and the gas mixture was the same and was maintained at 80-85%.The measurements were carried out at room temperature (22 ± 2 °C).
To test the long-term stability of the sensor, the response of the annealed VOPcCl4-p films was measured after 1, 2, 7, and 30 days (see Supplementary Figure S7).It was shown that the films demonstrate good stability, and the change in their sensor response To test the long-term stability of the sensor, the response of the annealed VOPcCl 4 -p films was measured after 1, 2, 7, and 30 days (see Supplementary Figure S7).It was shown that the films demonstrate good stability, and the change in their sensor response after a fairly long time was within the experimental error.
The sensor characteristics of as-deposited and annealed VOPcCl 4 -p films were compared with other sensors based on metal phthalocyanines and relative compounds studied in recent works (Table 3).It can be seen that both the as-deposited and annealed VOPcCl 4 -p films can compete with sensors based on films of other metal phthalocyanines, and the sensitivity of the annealed films even exceeds the value of the sensor response of the films presented in Table 3.Only halogenated cobalt and zinc phthalocyanine films studied in our previous works [30,32,51] had a lower detection limit than the VOPcCl 4 -p films, but the recovery time of the VOPcCl 4 -p films is lower (150 s) and the value of the sensor response (52%) is higher compared to that of the CoPcF 4 -p films (215 s and 41%, respectively) and ZnPcF 4 -p (210 s and 16%, respectively).Thus, this study makes it possible to expand the range of metal phthalocyanines, which are promising materials for use as active layers of chemiresistive sensors to ammonia.Our analysis of review articles devoted to the study of ammonia sensors showed that resistive sensors based on semiconductor oxides and their hybrid materials have a lower response (4-30 s) and relaxation times (10-90 s) compared to sensors based on fluorosubstituted phthalocyanines; however, their detection limits are usually worse and vary (i.e., 0.1-20 ppm) [57,58].Sensors of ammonia based on OFET structures [57] with conductive polymers as sensing layers have low LODs (usually from tens to hundreds parts per billion) that are comparable with phthalocyanine films, but usually their response and recovery times are quite high and range from several minutes to tens of minutes.
As a rule, the active layers of sensors with better characteristics than those of phthalocyanine films are not single-component films but mixtures of complex organic semiconductors [59,60] or complex heterostructures, for example, oxides or conductive polymers modified with nanoparticles or carbon nanomaterials [61,62].Compared with such heterostructures, the advantage of phthalocyanine films lies in the simplicity of the production method.

Conclusions
In this work, the films of tetrachlorosubstituted vanadyl phthalocyanine derivatives were studied as active layers for detecting low concentrations of NH 3 .Tetrachlorosubstituted vanadyl phthalocyanines with Cl substituents in the peripheral (VOPcCl 4 -p) and nonperipheral (VOPcCl 4 -np) positions of the phthalocyanine ring were investigated to reveal the effect of the position of substituents on the structure of their films prepared by a physical vapor deposition technique and their sensor response to ammonia.It was shown that the films of VOPcCl 4 -p exhibited a noticeably higher sensor response to both NH 3 films when compared to VOPcCl 4 -np films.This was explained with the help of quantum chemical calculations by the stronger interaction of NH 3 molecules with the side atoms of phthalocyanine having peripheral Cl substituents as compared to the phthalocyanines bearing the substituents in the nonperipheral position of the macrocycle.The limit of detection (LOD) of NH 3 was 0.7 ppm.The postdeposition annealing of the films allowed us to increase the sensor response and reduce the detection limit to 0.3 ppm.The sensing layers demonstrated a reversible sensor response at room temperature, with fairly low response/recovery times of 50/300 s at 5 ppm NH 3 .The fact that NH 3 can be detected in the presence of various interfering gases and vapors, as well as in a mixture of gases similar in composition to exhaled air, indicates that VOPcCl 4 films may be good candidates for their use as active layers for detecting ammonia as a biomarker in exhaled air.However, a lot of effort is still needed to improve their selectivity, as well as the response time, which is higher than that of most sensors based on semiconductor oxides.Therefore, further efforts should be directed to testing phthalocyanine-based sensing layers and their arrays in multicomponent gas mixtures and to study hybrid materials of metal phthalocyanines with metal nanoparticles and/or carbon nanomaterials, which are known to contribute to improving sensor performance.https://www.ccdc.cam.ac.uk/structures/, accessed on 13 September 2023 (numbers 2267526, 2267527, and 2267528).

Figure 2 .
Figure 2. Structure of VOPcCl4-p/NH3-x (where x = 1-4) aggregates and bond critical points (BCPs, small green balls) in them.BCPs characterizing the interaction between the ammonia molecule and phthalocyanine atoms are shown in red circles.

Figure 2 .
Figure 2. Structure of VOPcCl 4 -p/NH 3 -x (where x = 1-4) aggregates and bond critical points (BCPs, small green balls) in them.BCPs characterizing the interaction between the ammonia molecule and phthalocyanine atoms are shown in red circles.

Figure 2 .
Figure 2. Structure of VOPcCl4-p/NH3-x (where x = 1-4) aggregates and bond critical points (BCPs, small green balls) in them.BCPs characterizing the interaction between the ammonia molecule and phthalocyanine atoms are shown in red circles.

Figure 3 .
Figure 3. Structure of VOPcCl4-np/NH3-y (where y = 2-4) aggregates and bond critical points (BCPs, small green balls) in them.BCPs characterizing the interaction between the ammonia molecule and phthalocyanine atoms are shown in red circles.

Figure 3 .
Figure 3. Structure of VOPcCl 4 -np/NH 3 -y (where y = 2-4) aggregates and bond critical points (BCPs, small green balls) in them.BCPs characterizing the interaction between the ammonia molecule and phthalocyanine atoms are shown in red circles.

Figure 4 .
Figure 4. Molecular structure (a) and packing diagrams (b,c) for tetragonal polymorph and molecular structure (d) and packing diagrams (e,f) for triclinic polymorph of VOPcCl4-p.

Figure 4 .
Figure 4. Molecular structure (a) and packing diagrams (b,c) for tetragonal polymorph and molecular structure (d) and packing diagrams (e,f) for triclinic polymorph of VOPcCl 4 -p.

Figure 5 .
Figure 5. Molecular structure (a) and packing diagrams (b,c) for VOPcCl 4 -np.VOPcCl 4 -np molecules are packed into two-dimensional layers parallel to the plane (2−22) (Figure5b).The distance between molecules in one layer is 3.298 Å, and between adjacent layers, 3.239 Å.The angle between the molecular layers and the plane (2−22) is 0.15 • .However, unlike VOPcCl 4 -p, VOPcCl 4 -np molecules do not form separate chains within one molecular layer but are arranged in such a way that each molecule faces three other molecules (Figure5c).

3. 2 .
VOPcCl 4 -p and VOPcCl 4 -np Thin Films and Powder XRD Study Experimental XRD patterns of VOPcCl 4 -p and VOPcCl 4 -np powders and thin films (before and after annealing) are shown in Figure 6 in comparison with the powder patterns calculated based on the abovementioned single-crystal data.The first three peaks on the XRD pattern of the VOPcCl 4 -p powder (6.38 • , 7.19 • , and 7.76 • 2θ) correspond to the plane (110) of the tetragonal phase and the planes (001) and (010) of the triclinic phase, indicating that both polymorph phases are present in a significant amount in the powder.The strong peak at 27.03 • dominates the entire diffraction pattern and belongs to the plane (−222) of the triclinic polymorph, which suggests that the VOPcCl 4 -p powder consists mainly of the triclinic phase.

Figure 6 .
Figure 6.XRD patterns of bulk powders and thin films of VOPcCl4-p (a) and VOPcCl4-np (b).Figure 6. XRD patterns of bulk powders and thin films of VOPcCl 4 -p (a) and VOPcCl 4 -np (b).The first two peaks at 6.96 • and 7.21 • 2θ on the XRD pattern of VOPcCl 4 -np powder correspond to the planes (001) and (010), respectively, while the strong peak at 27.00 • 2θ corresponds to the plane(2−22).The difference between the calculated and observed position for the peak (2−22) is caused by the difference in sample temperature during the experiment (150 K for SC-XRD and room temperature for PXRD).In general, the experimental powder patterns coincide well with the calculated ones for both VOPcCl 4 -p and VOPcCl 4 -np.XRD patterns of both VOPcCl 4 -p and VOPcCl 4 -np films contain peaks that do not match the single-crystal data.The diffractogram of the VOPcCl 4 -p film is characterized by one strong diffraction peak at 6.06 • 2θ and a peak at 12.15 • 2θ with a weaker intensity, and the diffractogram of the VOPcCl 4 -np film contains a diffraction peak at 6.02 • 2θ, which belongs to unknown crystal phases.The absence of other strong diffraction peaks indicates that the films have a preferred orientation.The difference in the phase composition of films from single crystals is often observed for phthalocyanines due to the tendency to form different polymorphic modifications depending on temperature[40].To check the effect of the film annealing on their structure, the films were annealed at 200 • C for 2 h in air.Annealing resulted in a small shift in peak positions toward larger angles but did not change the overall appearance of the diffraction pattern, which can be caused by a release of mechanical stress and a decrease in the number of crystal lattice defects in thin films during annealing.The optical absorption spectra of VOPcCl 4 -p and VOPcCl 4 -np films were also recorded.The spectra of VOPcCl 4 -p and VOPcCl 4 -np films are compared with the spectra of their solutions in dimethylformamide in Supplementary FigureS1.The spectra of VOPcCl 4p and VOPcCl 4 -np solutions are typical for the phthalocyanines, which are present in monomeric form in solutions[41], and have the Q bands with the maxima at 680 and 689 nm, respectively.The spectra of the films become broader because of intermolecular interactions in solids.Similar to the spectra of unsubstituted VOPc[42] and VOPcF 4[43] films, the maxima of Q bands shift to the higher wavelengths, i.e., to 778 and 795 nm of VOPcCl 4 -p and VOPcCl 4 -np films, respectively.This view of spectra is a characteristic of

Figure 8 .
Figure 8. Curves of changes in the sensor response of VOPcCl4-p (a) and VOPcCl4-np (b when different concentrations of NH3 (1-50 ppm) are introduced in the flow cell (obtained namic mode).Dependence of the sensor response of VOPcCl4-p (a) and VOPcCl4-np (b) fi NH3 concentration (c).Sensor response of VOPcCl4-p and VOPcCl4-np films to 5 ppm of N measured in the static mode.The measurements were carried out at room temperature (22 ±

Figure 8 .
Figure 8. Curves of changes in the sensor response of VOPcCl 4 -p (a) and VOPcCl 4 -np (b) films when different concentrations of NH 3 (1-50 ppm) are introduced in the flow cell (obtained in dynamic mode).Dependence of the sensor response of VOPcCl 4 -p (a) and VOPcCl 4 -np (b) films on NH 3 concentration (c).Sensor response of VOPcCl 4 -p and VOPcCl 4 -np films to 5 ppm of NH 3 (d), measured in the static mode.The measurements were carried out at room temperature (22 ± 2 • C).

Figure 9 .
Figure 9. Scheme demonstrating the proposed mechanism of the sensor response.NH3 molecules donate electrons to macrocycles (blue arrows) due to interaction with their side atoms, leading to a decrease in the concentration of holes (red "plus" signs), i.e., charge carriers, and, consequently, to an increase in the resistance of phthalocyanine films.

Figure 9 .
Figure 9. Scheme demonstrating the proposed mechanism of the sensor response.NH 3 molecules donate electrons to macrocycles (blue arrows) due to interaction with their side atoms, leading to a decrease in the concentration of holes (red "plus" signs), i.e., charge carriers, and, consequently, to an increase in the resistance of phthalocyanine films.

Figure 11 .
Figure 11.Sensor response of as-deposited and annealed VOPcCl4-p films to ammonia (1-5 ppm) at different levels of relative humidity (a,c) and to various gaseous analytes and volatile organic vapors (b,d).The measurements were carried out at room temperature (22 ± 2 °C).

Figure 11 .
Figure 11.Sensor response of as-deposited and annealed VOPcCl 4 -p films to ammonia (1-5 ppm) at different levels of relative humidity (a,c) and to various gaseous analytes and volatile organic vapors (b,d).The measurements were carried out at room temperature (22 ± 2 • C).

Figure 12 .
Figure 12.The sensor response of a VOPcCl 4 -p film to ammonia (1-4 ppm) in moist air (a) and in a mixture of gases similar in composition to exhaled air (b).RH of the moist air and the gas mixture was the same and was maintained at 80-85%.The measurements were carried out at room temperature (22 ± 2 • C).

*
The sensor response is defined as (R − R o )R o •100% and provided for those concentrations that are indicated in the corresponding reference; ** TPP is tetraphenylporphyrin; *** R is 5-(trifluoromethyl)-2-mercaptopyridine.

Table 1 .
Unit cell parameters and refinement details for VOPcCl4 crystal structure.

Table 2 .
The interaction parameters of the NH 3 molecule with vanadyl phthalocyanines.
* The first atom belongs to the phthalocyanine molecule, while the second atom belongs to the NH 3 molecule.

Table 3 .
Sensor characteristics of as-deposited and annealed VOPcCl 4 -p films in comparison with other sensors based on metal phthalocyanines and relative compounds.