Synthesis and investigation of the structure of nanocomposites based on nickel nanoparticles dispersed in a phthalocyanine matrix

A method based on doping of pure nickel phthalocyanine (NiPc) with a polycrystalline potassium powder at relatively low temperatures (300°C) has been proposed for the synthesis of a magnetic nanocomposite containing nickel nanoparticles stabilized in the NiPc matrix. The structural analysis of the synthesized nanoparticles and changes in the NiPc initial matrix has been performed using X-ray diffraction, X-ray absorption spectroscopy, and transmission electron microscopy. It has been found that, at the doping level used in this study, the synthesized samples of the KxNiPc nanocomposites contain from 9 to 18% Ni in the form of metallic magnetic nanoparticles with an average size of more than 40 nm. It has been shown that the formation of nanoparticles is accompanied by a relative misorientation of persistent NiPc molecules with the unchanged structure of each of these molecules. The stabilization of nickel nanoparticles by the phthalocyanine matrix leads to the fact that the synthesized nanocomposites acquire time-conserving magnetic properties.


INTRODUCTION
Nanoclusters and nanoparticles of transition and noble metals, both embedded in dielectric matrices and located on their surfaces, have attracted much attention of researchers due to the possibility of using these objects in many fields, including catalysis, surface Raman spectroscopy, development of nonlinear optical devices, biological sensors, nanowhiskers for waveguides, solar cells, light-emitting diodes, and integrated optics [1][2][3][4]. Magnetic nanoparticles stabilized in the phthalocyanine matrix are very promising for use in catalysis [5], optical systems [6], and microwave absorbers [7]. Moreover, as was shown in [7,8], phthalocyanine coatings of metal nanoparticles are resistant to oxidation. At present, an intensive search has been carried out for low-cost effective methods of synthesizing such nanocomposites. In particular, the authors of [9][10][11][12][13] proposed a number of simple effective solid-phase chemical synthesis techniques with the features similar to those of the method used in the present study.
In the method used in our work, magnetic nickel nanoparticles in the matrix of nickel phthalocyanine (NiPc≡NiN 8 C 32 H 16 were synthesized by doping solid NiPc with alkali metal (potassium) vapor (see Section 2). It is noteworthy that the temperature of the reaction did not exceed 300°C, which is significantly less than the characteristic temperature of the nucleation and growth of nanoparticles. A more detailed information about the synthesis conditions can be found in [14,15].
The structure of the synthesized composite materials was investigated using well-tested methods for diagnostics of the atomic structure: (1) transmission electron microscopy (TEM); (2) the powder X-ray diffraction (XRD) analysis, which makes it possible to judge the presence and character of the long-range order in the material; and (3) the X-ray absorption spectroscopy method sensitive to details of the local atomic structure in the vicinity of the absorbing atom, which was used to analyze the extended X-ray absorption fine structure (EXAFS) spectra near the Ni K edge. The analysis of the experimental X-ray diffraction and EXAFS data, as well as the results of the TEM measurements, are presented in Section 3. The brief conclusions are formulated in Section 4.

SAMPLE PREPARATION AND EXPERIMENTAL TECHNIQUE
Pure (undoped) nickel phthalocyanine used in this study was synthesized according to the Linstead method [16] from phthalonitrile and the corresponding nickel salt. The reaction product was chemically purified by washing in acidic and alkaline media and in a medium sublimed twice under vacuum at a temperature of 500-550°C.
Nickel nanoparticles were prepared by doping a polycrystalline NiPc powder with potassium vapor at a relatively low temperature (300°C). In this case, potassium atoms reduce Ni 2+ ions in NiPc with the formation of nickel atoms, which are grouped into nickel clusters and nanoparticles, according to the following reaction: The experimental investigations demonstrated that the potassium powder should be heated to at least 300°C so as to ensure the reduction of nickel by potassium in NiPc. The gas-phase doping of phthalocyanines was described in detail in [14,15]. In this paper, we considered three doped K x NiPc samples with different contents x of potassium atoms (x = K : NiPc): the initially synthesized K 2.5 NiPc sample with the intermediate (between the considered samples) level of doping with potassium and the samples synthesized one year later with the lowest (K 1.8 NiPc) and highest (K 3 NiPc) potassium contents. Immediately after the preparation, the samples were annealed at a temperature of 280°C for 2 h.
The EXAFS spectra above the Ni K edge in pure and potassium-doped NiPc samples were measured on the Structural Materials Science end-station at the Kurchatov Synchrotron Radiation Source of the National Research Centre "Kurchatov Institute" (Moscow, Russia) [17]. The electron accelerator at the Siberia-2 storage ring operated with an electron energy of 2.5 GeV and an average direct current of 100 mA. The measurements were performed in the "transmission" mode using the Si(111) monochromator and two ionization chambers filled with a gas (N 2 -Ar) mixture so as to ensure 20 and 80% of the absorption for I 0 and I t , respectively. The scanning steps were chosen to be δE ~ 0.3 eV and δk ~ 0.05 Å -1 in the X-ray absorption near edge structure (XANES) and EXAFS regions, respectively. The total time for measuring one spectrum was about 30 min.
The X-ray diffraction profiles were measured on the same experimental station. The diffraction measurements were carried out using a FujiFilm Image-Plate2D detector in the "transmission" mode at a radiation wavelength of 1.072 Å.
The TEM micrographs of the samples were obtained on a Tecnai G2 Spirit transmission electron 2 2 NiPc 2K 2H O Ni H Pc 2KOH. + + = + + microscope operating at an accelerating voltage of 120 kV. All three synthesized K x NiPc samples had stable magnetic properties (magnetization, etc.). The discussion of these properties will be presented in a subsequent paper.

TEM Measurements
The TEM micrograph of the K 3 NiPc sample is shown in Fig. 1. It can be seen from this figure that, in the sample, there are homogeneous electron-scattering regions, which can be attributed to the scattering in nickel phthalocyanine matrices in the crystalline or amorphous state. Darker spots with an average linear size of ~40 nm correspond to nickel nanoparticles, which correlates with the results of the X-ray diffraction studies presented in Fig. 2. The inset in Fig. 1 illustrates the size distribution of nickel nanoparticles in the composite, which was obtained by counting 110 particles in several TEM micrographs.

X-Ray Diffraction Profiles
The X-ray diffraction profiles of the K x NiPc samples are shown in Fig. 2. As can be seen from this figure, the X-ray diffraction profiles exhibit two peaks in the range of large angles 2θ > 25°, which are absent in the undoped NiPc. A comparison with the peaks attributed to the nickel foil suggests that these two peaks can be assigned to the face-centered cubic (fcc) nickel reflections (111) and (200). In the range of smaller angles 2θ < 25°, the peaks are observed both in the case of pure NiPc and in the case of the doped samples. An increase in the doping level leads to a broadening of these peaks. The latter fact most likely means that, as the level of doping with potassium increases, the phthalocyanine matrix becomes less ordered. This leads to an increase in the width and to the disappearance of the peaks at small diffraction angles. Since pure NiPc does not exhibit peaks attributed to metallic nickel, it can be concluded that nickel nanoparticles are actually formed upon doping of phthalocyanine with potassium. The size of nickel nanoparticles was estimated from the width of diffraction peaks according to the Scherrer formula [19]. One of the improvements of the method used in this work is that the width of the diffraction peak is determined by fitting with the Voigt function [20], so that the instrumental and physical contributions to the line broadening are taken into account by the Gaussian and Lorentzian parts and can be separated. This fitting was carried out using the Voigt function implemented in the Gnuplot program [21]. The widths of the Ni(111) peaks of the doped samples were found to be equal to 0.12°, 0.13°, and 0.16°, which correspond to linear crystallite sizes of 55 ± 5, 51 ± 5, and 39 ± 5 nm for the samples K 1.8 NiPc, K 2.5 NiPc, and K 3 NiPc, respectively. The obtained values allow us to conclude that, as follows from the X-ray diffraction data, the average linear size of nickel nanoparticles formed in the phthalocyanine matrix is more than 40 nm.

Analysis of the EXAFS Spectra
The EXAFS spectroscopy method was used to confirm the qualitative conclusions about the structure of the K x NiPc composites, which were drawn from the TEM data and X-ray diffraction measurements, as well as to obtain the quantitative characteristics of the structure of these samples. The normal-ized experimental XANES spectra of the three synthesized samples and their oscillating parts χ(k) obtained using the AUTOBK algorithm implemented in the IFEFFIT package [22] are compared in Fig. 3 with the corresponding functions for the metallic nickel foil and undoped NiPc. The similarity between the nearthreshold regions of the XANES spectra of pure NiPc and the doped samples (Fig. 3) indicates the presence of NiPc molecules in all the samples under investigation. Meanwhile, the fine structure of the XANES spectra does not exhibit clear signs of the presence of a metallic nickel phase (nanoparticles or nanoclusters) in the samples. The presence of such objects in the samples is indicated by the comparative analysis of the Fourier transforms (F(R)) of the experimental Ni K EXAFS spectra of the doped samples, the nickel foil, and pure NiPc (Fig. 4), which revealed the manifestation of Ni-Ni interactions in F(R) of the samples at the interatomic distances R Ni-Ni corresponding to the metallic nickel phase.
In the construction of the model of the immediate environment of nickel atoms, which is used for fitting the Fourier transforms F(R), we took into account that, in the NiPc crystal structure shown in Fig. 5c, the Ni-Ni shortest distance is equal to 4.79 Å [24], which is almost two times longer than the Ni-Ni bond length in the nickel foil (2.49 Å [25]). Owing to this difference, it appears to be possible to separate the contributions from the scattering of photoelectrons by nickel atoms in nanoparticles and in the molecular In order to obtain quantitative information about the structure of the immediate environment of nickel atoms in the samples under investigation, the Ni K EXAFS spectra were analyzed based on the Fourier transformation (FT) of the experimental spectra and the subsequent fitting of the obtained Fourier transforms F(R) by the theoretical functions corresponding to the used model of the local atomic structure of the samples in the vicinity of the nickel atoms. The scattering amplitudes and phase shifts for all photoelectron paths were calculated in the Hedin-Lundqvist exchange-correlation approximation taking into account the completely screened 1s hole on the absorbing nickel atom [26].
The Fourier transformation of the functions χ(k) obtained from the experimental Ni K EXAFS spectra and the fitting of their Fourier transforms F(R) were performed using the FEFFIT code [22]. In order to decrease the influence of multiple-scattering processes on the structural parameters to be determined, the lower limit of the k-interval for the Fourier transform spectra, according to [27], was chosen to be k min = 4.5 Å -1 .
The construction of the fitting model based on the assumption that the doped K x NiPc samples contain both nickel nanoparticles and molecules of the initial phthalocyanine requires preliminary consideration of the spectra of the nickel foil and pure NiPc as the basic components of the model.

Ni K EXAFS in the nickel foil.
The analysis of EXAFS spectra of fcc metals is currently the subject of testing and verification of different approximations. It is known that the theoretical description of EXAFS spectra in an extended region of the scale of distances R for fcc metals requires the inclusion of several coordination spheres around the absorbing atom in the fitting as well as the consideration of some multiple-scattering processes, especially significant in the case of focusing of photoelectron waves [27]. The agreement between the theoretical and experimental functions |F(R)| in an extended R region (up to ~5.5 Å) in Fig. 4 (spectrum a) was achieved by taking into account the scattering in the four nearest coordination spheres of the absorbing nickel atom (Fig. 5b), which was supplemented by several paths of few-fold (to the third order) scattering. However, it should be emphasized that, for the lower limit k min used in this work for the Fourier transform spectra, the high accuracy in the determination of the parameters of the first coordination sphere of nickel atoms can be achieved without taking into  account the aforementioned multiple-scattering processes.
The fitting of the Fourier transforms F(R) was performed in the range 1 Å < R < 5 Å with fixed values of the coordination numbers of the fcc structure. We used the following variable parameters: the radii R i of the first four coordination spheres; the Debye-Waller factors , which take into account the thermal and structural disorders in the ith coordination sphere; and the nonstructural parameters, such as the factor , which accounts for the inelastic losses at the ionized atom upon photoionization, and the energy parameter e 0 [28]. The obtained values of the parameters listed in the first row of the table are physically reasonable: the Debye-Waller factor for the first coordination sphere ≈ 0.006 Å 2 is characteristic of room temperature [29] and, as should be expected, turned out to be approximately 1.5 times greater for more distant coordination spheres. The factor also lies within the reasonable range 0.5 < < 1.0 [30]. In

Ni K EXAFS in NiPc.
The molecular crystal structure of nickel phthalocyanine represents a spatial packing of planar molecules with intermolecular distances of ~3.5 Å [31] (Figs. 5a, 5c). Such a large interplanar spacing indicates that there is a weak bonding between individual molecules, which leads to a low stiffness of the bonds, high values of the Debye-Waller factors, and, as a consequence, to a smearing of the contributions from the photoelectron scattering by the neighboring molecules [32]. Therefore, in the construction of the contribution from nickel phthalocyanine, we restricted ourselves to the consideration of only one isolated NiPc molecule. The fitting of the Fourier transform F(R) of the experimental Ni K EXAFS spectrum for NiPc was carried out in the range 1.0 Å < R < 3.5 Å (including three coordination spheres and three few-fold scattering paths illustrated in Fig. 5a). A comparison of the experimental and theoretical Fourier transform moduli |F(R)| for pure NiPc (spectrum e in Fig. 4) and the results of the corresponding fitting presented in the last row of the table demonstrate that the used model of the structure of the immediate environment of nickel atoms in NiPc makes it possible to describe the behavior of F(R) in an extended R range (up to ~3.5 Å). In this case, there is a good agreement between the parameters of the first coordination sphere of the nickel atom and the crystallographic data available in the literature [23,31].

Ni K EXAFS in K x NiPc.
In order to obtain information about the structure of the immediate environment of nickel atoms in the K x NiPc samples, we used a model in which it was taken into account that, according to the X-ray diffraction measurements, a part of the nickel atoms remains in the NiPc molecule, while the other part of the nickel atoms forms nickel nanoparticles. Let the contribution from nickel atoms in the NiPc molecule be designated as χ Ni-NiPc (k) and the contribution from nanoparticles (NP) be denoted as χ Ni-NP (k). Then, the resulting oscillating part χ(k) for the EXAFS spectrum of the doped sample can be written as (1) where C is the relative concentration of the absorbing nickel atoms involved in the composition of the nickel nanoparticle. In order to decrease the number of parameters variable during the fitting based on representation (1) for the function χ(k), we used the values of the nonstructural parameters established for the first and second terms in the analysis of the corresponding test compounds, i.e., the nickel foil and pure NiPc. It should be noted that, for the observed largesized nickel nanoparticles, the average number of atoms in the first coordination sphere of the absorbing nickel atom is slightly smaller than that in the fcc structure of the nickel foil [25]. This is an additional justification for the possibility of using the nonstructural parameters of the nickel foil in the determination of the structural parameters corresponding to the contribution from the first term in representation (1), including the concentration C [27]. For the second term in representation (1), the variable parameters are the radius of the first coordination sphere R Ni-N and the Debye-Waller factor . The values obtained according to the above-described procedure for the parameters of the structure of the immediate environment of nickel atoms in the nickel foil, NiPc, and the synthesized K x NiPc samples are presented in the table.
The presented values of the parameters indicate that the Ni-N bond lengths in the three doped K x NiPc samples remain (within the error) the same as in pure phthalocyanine. The Debye-Waller factors also change insignificantly and correspond to the values observed at room temperature [33,34] without any degree of structural disorder. The obtained results suggest that the atomic structure of the NiPc molecules does not substantially change upon doping with potassium. This conclusion, together with the results of the X-ray diffraction measurements (smearing of the Ni Ni − σ reflections from NiPc with an increase in the number of embedded potassium atoms), allows us to conclude that an increase in the level of doping with an alkali metal leads to an enhancement in the mutual misorientation of the NiPc molecules with the unchanged structure of each of these molecules. It should also be noted that, for all the K x NiPc samples, the concentration C of nickel atoms forming nickel nanoparticles lies in the range of 9-18%. In this case, one gram of the sample of the composite material contains 10 12 nanoparticles. The quality of the fitting corresponding to the structural parameters listed in the table for the K x NiPc samples is illustrated in Fig. 4  (spectra b-d).

CONCLUSIONS
The gas-phase doping of NiPc with potassium atoms at a temperature of 300°C was used to synthesize nickel nanoparticles dispersed in the phthalocyanine matrix and having an average size of 40 nm.
The results of the analysis of the X-ray diffraction profiles and Ni K EXAFS spectra of the K x NiPc samples at different levels of doping with potassium (x = 1.8, 2.5, 3.0) demonstrated that an increase in the number of embedded potassium atoms leads to a misorientation in the relative arrangement of the persistent phthalocyanine molecules with the unchanged structure of each of these molecules.
In the nickel nanoparticles formed in the K x NiPc composites under investigation, the character of the immediate environment of nickel atoms corresponds to the fcc structure of the bulk sample. The concentration of nickel atoms forming the nickel nanoparticles lies within the range of 9-18%. The density of these nanoparticles proved to be sufficient for the manifestation of the time-conserving macroscopic magnetic properties in the synthesized K x NiPc composite materials.

ACKNOWLEDGMENTS
This study was supported by the Russian Foundation for Basic Research (joint project no. 15-52-05051 Arm_a) and the State Committee of Science of the Ministry of Education and Science of Armenia (project no. SCS 15RF-085 51).
The measurements of the X-ray diffraction profiles and Ni K EXAFS spectra of the synthesized materials were performed on a synchrotron radiation source at the National Research Centre "Kurchatov Institute" (Moscow, Russia) and supported in part by the Ministry of Education and Science of the Russian Federation (federal contract no. 16.552.11.7003).