Defect states and morphological evolution in mechanically processed ZnO + xC nanosystems as studied by EPR and photoluminescence spectroscopy

Abstract

An evolution of electron paramagnetic resonance (EPR) and photoluminescence (PL) spectra of various active states (hydrogen donor D_H EPR centers, Zn vacancy-related EPR centers , EPR D_S centers from shallow donors (g = 1.9640) in ZnO_W, Mn²⁺ ions in ZnO_ZB, C EPR centers in carbon nanoparticles, forming the near-band-edge (NBE) PL emission, PL emission typical for Zn-, O- and N-enriched ZnO_W particles, as well as oxidized carbon nanodots (OCN)) was observed in the mixtures of ZnO + xC nanoparticles during prolonged high-energy mechanical processing (MP) in a hermetically sealed grinding chamber. The results of EPR and PL spectroscopy, X-ray diffraction analysis, as well as morphological analysis by means of atomic force microscopy (AFM) and laser particle sizer (LPS) measurements show a wide variety of interrelated series–parallel processes in the samples with increasing MP processing time (t_MP). These processes include: (a) dramatic reduction in intensity of the D_H EPR signal and PL bands at 3.14 (1.57), 2.53 and 2.3 eV during the first minutes of MP, which correlate with sample disaggregation; (b) grinding of ZnO particles and formation of Zn vacancy-related EPR centers in the area of destruction (AD); (c) an increase in sample temperature; (d) annealing of the Zn vacancy-related EPR centers formed; (e) initiation of carbon nanoparticle interaction with oxygen in the grinding chamber; (f) formation and growth of the EPR signal due to carbon nanoparticles; (g) formation of a reducing environment in the grinding chamber; (h) stabilization of donor D_S-centers in AD of ZnO nanoparticles; (i) an increase in CO concentration in the grinding chamber; (j) inhibition of D_S paramagnetic centers in ZnO; (k) initiation of the EPR signals due to Mn²⁺ ions in ZnO_ZB sphalerite phase in the sample with the lowest carbon content; (l) inhibition of nitrogen PL centers in ZnO; (m) the formation of oxidized carbon nanodots (OCN) showing a PL band at 2.8 eV. A detailed analysis for the localization of EPR and PL centers in the MP samples is presented.

---

1. Introduction

Nanocomposite materials (in which at least one of the material components has one or more dimensions in the nanometer range) are referred to as materials of the 21st century, due to an opportunity to create in them a combination of properties that cannot be found in conventional bulk composites and to broaden design possibilities.^1,2 In particular, it is known that the decoration of ZnO nanostructures by nanoparticles of other materials allows to modify the surface electronic structure, creating new properties of the composite material (as exemplified for the case of ZnO + Ag³).

Nanocrystalline composite materials of ZnO with different forms of carbon are studied in the context of development of new catalytic and photocatalytic materials,^4–7 low-voltage phosphor applications,⁸ engineering of magnetic ZnO with nonmagnetic C dopants,⁹ assembly of functional nanoscale electronic and photonic structures,¹⁰ cathode materials with improved field emission properties,¹¹ luminescent materials,¹² selective solar absorber coatings,¹³ among others. The effect of synthesis conditions on physico-chemical properties of the nanoparticles are indicative of a close relationship between the structure of their newly formed agglomerated forms and electronic states of the latter. Nanocomposites are not mechanical mixtures of nanoparticles of different phases: an important part of their properties is governed by interactions at interfaces between the phases. In other words, when considering the properties of nanocomposites it is necessary to account for the existence of (often crystalline) nanoparticle cores, their highly-defective surface layer, as well as for chemical interactions between the particles and for those with the surrounding atmosphere.

Earlier electron paramagnetic resonance (EPR) and photoluminescence (PL) studies revealed the existence of a number of active sites in ZnO, whose content depends on the sample preparation conditions.^14–21 It is important to note that, while in single-crystal ZnO samples most of the EPR and PL centers can be identified rather unambiguously, in nanoparticulate samples a number of centers with similar parameters (g-factor in EPR; maximum peak position in PL) and properties can be observed. As a result, UV, violet, green and yellow PL bands in nanoparticulate ZnO are well known, but the luminescence mechanisms remains unclear. Varying synthetic conditions and post-processing (annealing under different atmospheres, irradiation, mechanical processing, etc.^22–25) of ZnO-samples in a broad range led, on one hand, to frequently observed similarities in PL spectra, and to a very controversial interpretation of the nature of luminescence centers, on the other hand. In particular, the near-band-edge (NBE) UV-emission at ν ∼ 3.10–3.35 eV has been attributed to donor–acceptor pair transition,²⁶ zinc vacancies,²⁷ donor-bound excitonic emission,^28,29 electron–hole recombination,³⁰ an excitonic level which is lower than the conduction band,^31,32 free exciton recombination,³³ free-to-bound transition,³⁴ electron–hole plasma recombination,³⁵ surface-bound ionized acceptor–exciton complexes,³⁶ and exciton–exciton collision process.³⁷ These data indicate that NBE emission in each case is defined by the defect structure and chemical nature of the impurities, which in turn depend on the history of sample synthesis, subsequent storage, and even on the sample preparation procedure for PL measurements. EPR and luminescence characterization of carbon materials was performed as well.^38–44

The main focus of our work is the modification of material properties by means of different ways of sample processing, including mechanical processing (MP). The first description of the luminescence spectra evolution in MP samples of ZnO + xC was presented in.^45,46 We observed that the intensity of PL bands (at 3.14 eV, 2.42–2.50 eV and 1.57 eV) present in the spectra of unprocessed samples decreases dramatically in the first 3 min of MP. This effect was attributed to the generation of nonradiative recombination centers (increase of the center concentrations) due to MP.⁴⁷ The appearance of PL band at 2.84 eV attributed to graphite oxide was observed at processing time t_MP ≥ 30 min.^45,46

The origin and evolution of several paramagnetic states in the mixtures of ZnO_W + xC nanoparticles during intensive MP was previously described by us as well.^48,49 These states include three types of Zn vacancy-related centers and shallow donor centers (g ≈ 1.96) in ZnO wurtzite structure (ZnO_W), a C EPR signal due to carbon nanoparticles, an EPR feature due to Mn²⁺ ions in the zinc blende (sphalerite) ZnO phase (ZnO_ZB). All these EPR and PL changes reflect the diversity of physical and chemical processes occurring in the samples during MP.

In the present study, the evolution of EPR signals and luminescence spectra of different active centers (states) present in the untreated samples and appearing in the mixtures of ZnO with carbon nanoparticles during prolonged high-energy mechanical activation were explored in relation to morphological, thermal, chemical and other changes taking place in the samples during their MP. An especial emphasis is done on the localization of EPR and luminescence centers in the samples.

2. Materials and methods

An analytical grade commercial ZnO powder Reasol was used. A maximum limit of impurities was as follows: insoluble sulfuric acid, 0.010%; chlorides, 0.001%; carbonates, 0.001%; nitrates, 0.003%; sulfur compounds (as sulfates SO₄), 0.010%; As, 0.0002%; Fe, 0.001%; Pb, 0.005%; Mn, 0.0005%. The average initial size of ZnO particles was d_ZnO ∼ 250 nm (according to scanning electron microscopy^48,49).

Carbon nanoparticles were purchased from EPRUI Nanoparticles & Microspheres Co. Ltd. According to high resolution transmission electron microscopy (TEM) images, the nanoparticles are shaped as quasi-spheres or convex polyhedra with diameters ranging from 10 to 60 nm. The structure of carbon nanolayers is typical for nongraphitized amorphous carbons.⁵⁰ A more detailed characterization of the carbon nanoparticles by means of TEM, X-ray diffraction (XRD), EPR, etc., will be reported in a separate paper.

The starting mixtures ZnO + xC (x = 0.1%, 1.0% and 3.0% by weight) were prepared by mechanical mixing of the components. Both the mixing and subsequent high energy MP was carried out in a Planetary Ball-Mill PM 400/2 from Retsch Inc., using a grinding chamber of 50 ml volume made of tungsten carbide. For the mixing, seventeen balls of 10 mm diameter, also made of tungsten carbide, were used, at a ball-to-powder sample ratio of about 5 : 1 by weight. The mechanical mixing was carried out at a low rotation speed of about 200 rpm and processing times (t_MP) of 20 min. Hereafter, these mixtures will be referred to as initial samples (note that earlier we did not describe in detail the procedure for mixing powders, since it had no effect on the EPR and X-ray characteristics obtained⁴⁸). The high energy MP was carried out in the same hermetically sealed grinding chamber with another set of tungsten carbide balls: three of 20 mm and ten of 10 mm diameter, at a ball-to-powder sample ratio of about 28 : 1 by weight. MP was performed at a maximum rotation speed of 400 rpm and t_MP of 1, 3, 9, 30, 90 and 390 min.

Morphological analysis (particle and agglomerate sizes) of the samples was performed by atomic force microscopy (AFM; by using Nanoscope IV from Digitial Instruments and a microscale cantilever with a radius of tip curvature of 5 nm), XRD (Bruker D8 Advance diffractometer with CuK_α radiation), and by laser particle sizer (LPS; Analysette 22 COMPACT from Fritsch; lower measuring limit of 0.3 μm). Two types of AFM mode were used: non-contact imaging (constant height mode, h) and the tapping mode (phase regime, f). The powder samples were prepared (a) by dispersing the powder onto a flat disc with a double face adhesive tape on it; (b) by dispersing the powder and its triturating by pressing between two flat discs; (c) by dispersing the powder in a glass vial containing isopropyl alcohol in an ultrasonic bath, then a drop was deposited on a flat disc and the solvent evaporated in air. LPS analysis of particle/agglomerate size distribution with and without the use of ultrasonic processing (USP) were performed.

EPR spectra were recorded on a SE/X 2547 Radiopan instrument (9.38 GHz frequency; modulation frequency of 100 KHz; microwave power 2 mW; modulation amplitude of 1 gauss; cylindrical cavity (TE₀₁₁ mode)) at room temperature. Double integration of EPR-signals was performed for measuring its intensity.

PL spectra were acquired at 300 K, with He–Cd laser excitation at a wavelength of 325 nm and a beam power of 46 mW, by employing the PL setup based on a SPEX500 spectrometer as described in ref. 51.

3. Results

3.1. Atomic force microscopy

The AFM micrographs (Fig. 1; exemplified for ZnO + 1% C) were used to determine an average ZnO particle size, D^ZnO_AFM. With increasing the duration of MP, D^ZnO_AFM decreased from 250 ± 40 nm (initial) to 35 ± 6 nm (t_MP = 90 min) and further to 13 ± 3 nm (t_MP = 390 min).

Fig. 1 AFM micrographs (phase regime) of MT samples of ZnO + 1% C for different MT duration: (a) initial sample; (b) t_MT = 90 min; (c) t_MT = 390 min. (a) Sample was prepared by dispersing the powder and its triturating by pressing between two flat discs; (b and c) samples were prepared by dispersing the powder in a glass vial containing isopropyl alcohol upon USP, then drop casting onto a flat disc and drying in air.

3.2. X-ray diffraction

Fig. 2a shows XRD patterns of ZnO + xC samples as a function of milling time. When t_MP increases, the height of XRD peaks decreases and their width increases. An average ZnO crystallite size (D^ZnO_X) was calculated from XRD peak (101) broadening (Fig. 2a) using Scherrer's equation. Fig. 2b shows the changes in D^ZnO_X versus MP duration. Within the accuracy of peak width measurements, these changes are independent of the carbon content in ZnO + xC samples. ZnO crystallite size changed from about 250 nm for the initial sample, which agrees with the individual particle size estimated by AFM, down to 14 nm for t_MP = 390 min (Fig. 2b). Note that in chemically synthesized samples Zn_1−xMn_xO there was a significant difference between the data on the size of ZnO nanoparticles, determined by AFM and XRD methods.⁵²

Fig. 2 (a) X-ray diffraction patterns of ZnO + 0.1% C, ZnO + 1.0% C and ZnO + 3.0% C samples after MP during different t_MP. (b) Changes in ZnO crystallite average size, D^ZnO_X, and sample temperature, T_S, versus MP duration.

3.3. Laser particle analysis

The plots of size distribution for individual particles and agglomerates (which behave as individual particles), D_LPS, for undoped ZnO sample measured without and with USP, are presented in Fig. 3a and b, respectively. According to Fig. 3a, for the starting sample as measured without USP, two maxima are observed in the range of 3 × 10³ to 4 × 10⁴ nm. At the same time, the measurements performed with USP produced a single maximum for particle size distribution (average size of nanoparticle agglomerates, D̄_LPS) at D_LPS of 1 × 10³ to 2 × 10³ nm (Fig. 3b). Changes of D̄_LPS in undoped ZnO, ZnO + 0.1% C, ZnO + 1.0% C and ZnO + 3.0% C samples as a function of t_MP measured with USP are shown in Fig. 3c. As one can see, for undoped ZnO D̄_LPS values are much higher as compared to the size of individual particles in the sample found by AFM (D^ZnO_AFM). In other words, the particle size distribution as measured by LPS reflects the presence of strongly agglomerated ZnO nanoparticles (Fig. 4), similarly to our earlier studies on mechanically processed ZnO + MnO₂ mixtures.⁵³ Adhesion and autohesion phenomena are caused by superficial forces which depend on the surface properties of contacting particles and the surrounding gas environment. Since the surface properties of particles are formed during their preparation and storage, one can suggest that the sample agglomeration is a consequence of its previous history.⁵³ Upon MP of undoped ZnO powder samples, the destruction of agglomerates take place mainly during first few minutes (t_MP ≤ 9 min). At longer t_MP, the size of ZnO agglomerates increases again (Fig. 3c).

Fig. 3 (a) LPS histogram for the size distribution (D_LPS) in initial undoped ZnO sample as measured without USP; (b) LPS curves for different t_MP values measured for initial ZnO sample with USP; (c) changes of D̄_LPS in ZnO, ZnO + 0.1% C, ZnO + 1.0% C and ZnO + 3.0% C samples as a function of t_MP (measured with USP).

Fig. 4 (a) Models of ZnO particle aggregate, (b) the initial individual particle, (c) the individual particle in the aggregate, (d) destroyed individual particle. Allocated defect areas: DSA (black) – defect surface area; AIC (red) – area of interparticle contacts; AD (blue) – area of destruction.

Fig. 3c also shows that the degree of ZnO agglomeration varies greatly in ZnO + xC mixtures as well. The position of D̄_LPS in D_LPS histograms is shifted toward lower values with increasing the content of carbon additives at the step of mechanical mixing (Fig. 3c). This shift is even more pronounced after high energy MP. Even after as short t_MP as 3 min, D̄_LPS values decrease down to the lower measuring limit for LPS of 0.3 μm.

3.4. Electron paramagnetic resonance

EPR spectra of ZnO + xC samples before and after MP are shown in Fig. 5. In Fig. 6, the integral intensities of EPR signals observed are plotted against MP duration for different carbon content. A detailed description of EPR spectra for carbon containing ZnO samples was presented previously.⁴⁸ Here we analyze the relationship between EPR spectra observed and the evolution of defects in the samples.

Fig. 5 EPR spectra for ZnO + 0.1% C (a), ZnO + 1.0% C (b) and ZnO + 3.0% C (c) samples at different MP times.

Fig. 6 Changes in the integral intensity of EPR signals of Zn vacancy-related centers (II and III) and D_S centers in ZnO nanoparticles, as well as C centers in carbon nanoparticles at different MP times.

3.4.1. Hydrogen donor D_H EPR centers in ZnO nanoparticles. Only a small D EPR signal (D_H) with g = 1.9565 ± 0.0005 and a width line ΔB ≈ 5 G was observed for initial ZnO powders (Fig. 5). Other researchers^54,55 showed that EPR signal at g = 1.957 and g = 1.956 in nominally undoped ZnO is related to shallow, effective-mass-like donor interacting with a H nucleus. Observation of this signal in the present paper allows us to associate its occurrence with the processes of interaction between particles and interaction with the surface layers of ZnO with ambient atmosphere during prolonged storage. The signal disappears after about 3 min of MP (Fig. 6), which indicates that the D_H centers are localized in the ZnO defect surface area (DSA), namely in the area of interparticle contacts (AIC) see Fig. 4.

3.4.2. Zn vacancy-related EPR centers in grinded ZnO nanoparticles. MP results in the initiation and evolution of several paramagnetic centers in the samples. The formation (further increase of their number to a maximum, then decrease down to disappearance) of some hole centers I (g = 2.0190, g < g), II (g = 2.0130, g = 2.0140) and III (g₁ = 2.0075, g₂ = 2.0060, g₃ = 2.0015) (Fig. 5 and 6) takes place at t_MP of 1–30 min. Signal I originates from centers;⁵⁶ signal II, from centers;⁵⁷ and signal III, from (V_Zn⁻)₂⁻ centers.⁵⁶ All of them form in the area of destruction (AD see Fig. 4) of ZnO crystalline particles (on the newly formed surfaces) as a result of movement and interaction between dislocations during MP. From the analysis of peak widths for I (ΔB_I ≈ 2.5 G), II (ΔB_II ≈ 3.0 G) and III (ΔB_III ≈ 2.7 G), one can estimate that local concentration of I, II and III centers does not exceed 0.1 atomic%. The corresponding signal amplitudes I_i increase with increasing MP duration, reaching their maxima very fast, and then each of them weakens and finally disappears at different MP times (Fig. 6). No significant differences in qualitative and quantitative changes in EPR signals from I, II and III centers depending on t_MP and x are observed in ZnO + xC samples.

Such changes of the intensities of signals I, II and III in the sample can be associated with the development of different processes during MP (deformation, fracture, interparticle friction, collision, etc.) leading to a consistent increase in sample temperature. In particular, it is known that signal II disappears after thermal treatment at T_II ∼ 453 K; signal I, after thermal treatment at T_I ∼ 493 K; and signal III, after thermal treatment at T_III ∼ 533 K.⁵⁸ Therefore, sequential disappearance of EPR signals from II, I and III centers when increasing MP times indicates that T_S exceeds annealing temperature of the corresponding defect. Fig. 2b gives an idea on the sample temperature during continuous MP (see also ref. 59).

3.4.3. D_S EPR centers from shallow donors (g = 1.9640) in ZnO nanoparticles. The EPR signal with g = 1.9640 ± 0.0005 from donor D_S centers is observed for ZnO nanoparticles at t_MP > 9 min (when sample temperature reaches T_S ≈ 550 K), then number of the centers increases to a maximum at t_MP = 90 min, and finally disappears at t_MP = 390 min (Fig. 5 and 6). The formation of D_S signals was not observed during high-energy MP (at room temperature under air atmosphere) of ZnO samples with oxide dopants such as SnO₂, NiO, TiO₂, Gd₂O₃, etc. At the same time, the signal was observed when milling ZnO powder in a liquid nitrogen medium.¹⁷ Heating ZnO powder under a non-oxidative atmosphere (vacuum, hydrogen, zinc vapor et al.) produced a significant increase in intensity of the EPR signals at g ≈ 1.96.^17,60 On the other hand, the signal at g ≈ 1.96 is absent in ZnO crystals irradiated with electrons, protons, and high-energy ions, while in the above samples the signals from other point defects, including centers, were observed.⁶¹ That is, the signals at g ≈ 1.96 cannot be attributed to single zinc or single oxygen vacancies. Thus, it seems more appropriated to refer to a group of similar centers that exhibits paramagnetic signal at g ∼ 1.955–1.964, rather than to a single center. The basis for such a group of defects should be specific native point defects, whereas the associated impurities influence the value of g-factor and the peak width. The EPR signal at g = 1.9640 was observed previously in undoped ZnO powders, subjected to a high-pressure compression.¹⁸ In this case, paramagnetic states formed are isolated from contact with the air. This allows us to consider centers responsible for EPR signal with g = 1.9640 as the basic defect complex. It is known⁶² that during plastic deformation of a “two-layered structure” of the diamond-type lattice, the formation of double vacancies takes place. At the initial step of MP, the basic paramagnetic complexes, formed in the zones of destruction, interact with air oxygen molecules in the grinding chamber and convert into nonparamagnetic state (EPR signal at g = 1.9640 is absent). Thus, one can expect that in a sealed grinding chamber after 30-90 min MP, a reducing environment forms, which stabilizes D_S centers in the area of destruction of ZnO particles. One should remember that the sample temperature at such t_MP values reaches ∼600 K, and that in the grinding chamber conditions are created for interaction of carbon both with air oxygen and with surface oxygen atoms of ZnO. Carbon oxidation in the air is accompanied by the formation of two oxides: carbon monoxide (CO) and carbon dioxide (CO₂). With increasing carbon content in the sample (and with increasing t_MP) it is natural to expect an increased content of CO in the milling atmosphere. Thus, the phenomenon of CO inhibition of paramagnetic D_S centers can explain the above-mentioned quantitative changes in the number of these centers in the sample with increasing both t_MP and carbon content in the sample (Fig. 6).

3.4.4. C EPR signal in carbon nanoparticles. No EPR signals were detected in the initial nanocarbon powders. This means that the concentration of unpaired spins in the original carbon nanoparticles is less than ∼10¹⁷ spins per cm³. Such spin density is typical for amorphous carbon samples, exhibiting high values of optical gap E₀₄ ∼ 5 eV.⁶³ MP of ZnO + xC samples to t_MP = 9 min did not lead to the generation of carbon paramagnetic centers. That is, mechanical grinding of initial carbon nanoparticles does not result in the formation of unpaired spins. The origin of carbon EPR signal with Lorentzian shape (Fig. 7) and peak-to-peak width of ΔB_C ≈ 3.1 G at g = 2.0021 ± 0.0003 was observed in ZnO + xC samples only after 9 min of MT (T_S ≈ 550 K). This links carbon oxidation, formation of reducing environment in the grinding chamber and the formation of paramagnetic centers in carbon nanoparticles. The same width and Lorentzian shape of the signals in ZnO + xC samples mechanically processed for 9 min points to formation of the same types of spin clusters in nanocarbon particles. The signal intensity increases during 9–30 min of MP, and the increment is approximately proportional to the content of carbon nanoparticles in the samples (Fig. 6). Between 30 and 90 min of MP, the increase in C EPR signal intensity slows down and then stops, regardless of the differences in carbon content. As a result, in the range of 90–390 min of MP, the intensity of C EPR signals in all the samples is roughly the same. These data suggest that the content of C centers is defined by an approximately equal content of air oxygen in the hermetically closed grinding chamber for each experiment.

Fig. 7 Comparison of the shapes of experimental C EPR signal (black) for ZnO + 0.1% C sample (t_MP = 30 min) with a Lorentzian lineshape (red) having the same width and amplitude.

Three main types of paramagnetic defects in carbon materials can be involved to give an interpretation of EPR signals observed. σ- and π-defects are typical for graphitized carbon, whereas dangling sp³-bonds are typical for diamond-like and amorphous carbon. According to,³⁹ the signals with g-factors lower than the free electron value of g_e = 2.00232 are characteristic for the materials where sp² hybridization dominate. During MP of ZnO + xC samples, the following phenomena take place: (a) breakdown of the initial carbon nanoparticles; (b) coating of ZnO particles with carbon layers is promoted; (c) thermal and mechanical processes induce the reaction of carbon atoms with air oxygen from the grinding chamber and the surface O atoms of ZnO particles. Carbon atoms removed from the surface of amorphous carbon nanoparticles, which initially did not have unpaired electrons (dangling bonds), transform nonparamagnetic particles into particles producing EPR signals. The increase in C EPR line width in all the samples from ΔB_C ≈ 3 G at 30 min of MP to ΔB_C ≈ 4.05 G in ZnO + 3.0% C, ΔB_C ≈ 4.4 G in ZnO + 1.0% C and ΔB_C ≈ 6.8 G in ZnO + 0.1% C after 390 min of MP is apparently due to different kinetics of formation of reducing environment in hermetically sealed grinding chamber, which in turn influences the geometry of spin carbon clusters formed, as a function of carbon content in the samples.

The change in D_S and C signal intensities in the t_MT range of 9–90 min is almost synchronous (Fig. 6). Moreover, the integral intensity of these signals for ZnO + 0.1% C sample is nearly the same. For ZnO + 3.0% C and ZnO + 1.0% C, the intensity of C EPR signals was larger than the intensity of D_S signal (Fig. 6). In ZnO + 1.0% C and ZnO + 3.0% C, the difference in intensities of C lines increased with increasing carbon content. However, as noted above, the intensities of D_S lines decreases somewhat with increasing carbon content (Fig. 6).

3.4.5. EPR detection of MP-generated ZnO sphalerite phase. A typical EPR spectrum of Mn²⁺ ions in manganese-doped ZnO_W is described by the rhombic spin-Hamiltonian (S-H)where g is the electron g-tensor, β denotes the Bohr magneton, B the external magnetic field, S (=5/2) is ground electron spin numbers, D and a are axial and cubic zero-field splitting parameters, A represents the hyperfine parameter, I (=5/2) is nuclear spin. Parameters of the S-H, obtained from EPR-measurements of ZnO single crystal samples, are: g = 2.0012, A = −74.1 × 10⁻⁴ cm⁻¹, D = −236.1 × 10⁻⁴ cm⁻¹, a = 6.6 × 10⁻⁴ cm⁻¹.⁶⁴ The EPR spectrum of Mn²⁺ ions in polycrystalline ZnO_W widely known^{17,48,49,52,55,59} however, accuracy of measurement of S-H parameters from EPR spectra of polycrystalline samples is slightly lower: g = 2.0012 ± 0.0002, A = (74.1 ± 0.2) × 10⁻⁴ cm⁻¹, D = (236 ± 30) × 10⁻⁴ cm⁻¹.⁴⁹ In our samples of ZnO_W + 0.1% C after 390 min of MP the formation of a new type of polycrystalline EPR Mn²⁺ hyperfine sextet superimposed on a broad (ΔB ≈ 500 G) low intensity signal at g ≈ 2.00 were observed (see Fig. 5a and ref. 48 and 49). S-H parameters of the new Mn²⁺ observed spectrum were determined by computer modeling of the Mn²⁺ spectra in polycrystalline samples: g = 2.0010 ± 0.0004, |A| = (73.6 ± 0.3) × 10⁻⁴ cm⁻¹, |D| = (102 ± 30) × 10⁻⁴ cm⁻¹ (see also ref. 49). The comparison of S-H parameters of EPR spectra of Mn²⁺ ions in the newly formed phase and in ZnO_W indicates that the A parameter of hyperfine structure in the new phase is slightly smaller, and that the D parameter is more than by half lower than that for Mn²⁺ in ZnO_W. The former indicates that the secondary phase has a higher symmetry than ZnO_W, whereas the latter implies that the secondary phase shows more covalent character than ZnO_W.⁶⁵ The content of the newly formed phase (∼1%) is significantly higher than the content of external impurity states in initial ZnO sample.

The EPR spectrum observed is attributed to Mn²⁺ ions in the zinc blende (sphalerite) ZnO phase (ZnO_ZB), generated on the surface of ZnO_W nanoparticles during MP. Then the question arises, what is the surface composition of the MP dispersed sample? Even in the initial sample, there are two types of particle surfaces: (a) an exposed surface with a high defect concentration (defect surface area, or DSA); (b) a semi-closed AIC (see Fig. 4). These two surface states formed as a result of specific chemical synthesis of nanoparticles of ZnO (DSA) and their subsequent consolidation into a compact sample (AIC). Time and conditions of storage of samples may serve as a consolidating factor. During MP, the degradation of interparticle area and the formation of new surface implies the area of destruction (AD). This AD with newly formed fracture surface differ significantly from the first two, DSA and AIC. The absence in MP-samples of the EPR signature of Mn²⁺:ZnO_W suggests that the impurities of manganese in the trivalent state are present in surface layer DSA of initial ZnO_W particles. It is possible that most of other residual impurities are located in this layer. The Mn³⁺ → Mn²⁺ reduction directly result from the formation of reducing environment in the grinding chamber. The secondary phase is formed due to rearrangement and subsequent crystallization in DSA zones of ZnO_W nanoparticles. The data presented above allow us to estimate the thickness of DSA layer, h, of ZnO initial particles, in which the observed changes take place, to be h ∼ 0.4–0.6 nm.

The formation of biphasic ZnO_W–ZnO_ZB nanomaterials was observed during thermal oxidation of ZnS,⁶⁶ pulse laser deposition,⁶⁷ molecular beam epitaxy,⁶⁸ chemical vapor deposition,⁶⁹ etc. However, it is worth to mention that the cubic structure of zinc blende (ZnO_ZB) is unstable under normal conditions and that it can be stabilized in epitaxial thin films, grown on cubic substrates.⁷⁰ In this case, the stabilizing factor is the difference in lattice parameters of the substrate and ZnO_ZB film phase, resulting in the emergence of unstable ZnO_ZB structural stresses. It is natural to expect that for different types of substrates the degree of stabilizing structural stresses will be different (different D values). In the case of ZnO_ZB formation on ZnO_W surface, ZnO_ZB creates structural stresses leading to the appearance of Mn²⁺ ions rhombic components with D and E in the S-H parameter of EPR spectrum. The average value of the parameter “stabilizing” D in this case is about 70 × 10⁻⁴ cm⁻¹.

The absence of Mn²⁺:ZnO_ZB EPR signature in the samples with 1% and 3% carbon added indicates that a large carbon content prevents surface transformation ZnO_W → ZnO_ZB. It is possible that observation of intense Mn²⁺:ZnO_ZB EPR signal during MP of ZnO_W + xMnO₂ (ref. 59 and 71) (i.e. in the absence of carbon in the sample) is also due to the above presented mechanism of phase formation in DSA of initial ZnO particles (see Fig. 4). Occurrence of Mn²⁺ ions of the MnO₂ particles within the bulk of ZnO (formation a solid solution of Mn²⁺:ZnO_W)^59,71 is observed at slightly higher sample temperature (at higher t_MP values).

3.5. Photoluminescence measurements

3.5.1. PL sensing disaggregation of ZnO + xC powder at short MP. PL spectra of ZnO + xC samples for carbon nanoparticle concentrations (x) of 0.1%, 1.0% and 3.0% by weight measured at 300 K are shown in Fig. 8. The PL spectrum of initial samples (after mixing) exhibits three well pronounced peaks: in the UV (3.14 eV), visible (green-orange luminescence, ∼2.42 eV) and IR (1.57 eV) range. The luminescence band intensities, I_PL, in initial ZnO samples with 0.1%, 1.0% and 3.0% C (i.e., after mechanical mixing) changed have a ratio of about 5 : 2 : 1, respectively (Fig. 8). The narrow UV peak at 3.14 eV is associated with the near-band-edge (NBE) emissions, and the narrow IR PL peak at 1.57 eV (III) correspond to the second order diffraction peak of the 3.14 eV.⁷² The visible-range luminescence (at ∼2.42 eV) in ZnO samples is typically attributed to defect related (II) emissions.¹⁴ After Gaussian fitting on visible-range PL spectra (Fig. 9), we found that the green-orange luminescence represents a superposition of two intensive PL peaks at 2.53 eV and 2.30 eV and low intensive peak at ∼2.10 eV (Fig. 8a and b). The bands at 2.52 eV and 2.30 eV are typical for Zn-enriched and O-enriched samples, respectively, in nonstoichiometric zinc oxide.⁷³ The full width at half maximum (FWHM) of the emission bands at 3.14 eV, 2.53 eV, 2.30 eV and 1.57 eV (0.16 eV, 0.57 eV, 0.49 eV, 0.35 eV and 0.074 eV, respectively) did not depend on MP duration and nanocarbon content in the samples.

Fig. 8 PL spectra of MP ZnO + 0.1% C, ZnO + 1% C and ZnO + 3% C samples for different t_MP: (1) initial, (2) 1 min, (3) 3 min, (4) 9 min, (5) 90 min, (6) 390 min.

Fig. 9 Gaussian fitting of visible PL spectra for ZnO + 0.1% C samples: (a) initial; (b) t_MP = 1 min; (c) t_MP = 3 min.

The changes in PL band intensities observed in ZnO + xC samples depending on t_MT are shown in Fig. 10. I_PL bands at 3.14 eV and 2.53 eV weakened with increasing t_MP almost equally fast, while the peak at 2.3 eV weakened somewhat slower. This explains the above-noted ‘red-shifted’ superposition maximum of visible emission (Fig. 8 and 9).

Fig. 10 Changes in intensities of PL bands, I_PL, observed in ZnO + xC samples as a function of t_MT.

The dependence of intensities of the above-mentioned PL bands (approximately 5 : 2 : 1) on the content of nanocarbon additives in samples after preparing the initial mixtures deserves special attention. Such changes may be attributed, on one hand, to a change in absorption–emission properties of the samples, depending on the content of carbon nanoparticles and, on the other hand, can result from the processes occurring during preparation of the mixtures. However, the optical gap of E₀₄ ∼ 5 eV for initial carbon nanoparticles, which is higher than the energy range comprising excitation and luminescence spectra, does not imply their significant effect on PL band intensities in the samples at short MP times. Noteworthy, the intensity of FTIR absorption at 447 cm⁻¹ (0.055 eV) due to ZnO vibrationals was roughly the same for all starting mixtures ZnO + xC.

Another important observation is an approximately 100-fold decrease in intensity (without changing the width) of PL bands at 3.14 eV, 2.53 eV and 2.30 eV taking place after first 3 min of MP. The weakening of PL bands due to MP was noted previously^22,24,25 without explaining a mechanism of attenuation; it was linked to the surface localization of the centers responsible for PL. Explanation of such dramatic changes due to the formation of nonradiative recombination centers in ZnO involves the formation of high concentration of these centers upon MP. However, ZnO crystallite size decreases about by half after t_MT = 3 min. The defects form in AD of ZnO (see Fig. 4). As noted above, local concentration of Zn vacancy-related EPR centers in the area does not exceed 0.1 at%. Major changes in the morphological structure of the samples at initial stage of MP is related to particle disaggregation. Changes in the average aggregate size, D̄_LPS(t_MT), for ZnO samples are illustrated in Fig. 3c.

Fig. 11 presents a comparison of changes in the intensity of PL band at 3.14 eV with morphological changes in the samples: namely, with the average size of nanoparticle aggregates, D̄_LPS, (reflecting the disaggregation process) and the medium-sized crystallites, D^ZnO_X, (reflecting the fracture of ZnO nanoparticles) occurring in the first 9 minutes of MP. D^ZnO_X changes, as changes in vacancy-related centers, do not depend on the nanocarbon content. Fig. 11 shows a good correlation between changes in the I_PL(t_MP) and D̄_LPS(t_MP) allowing to link together the changes of I_PL(t_MP) with the process of disaggregation. Intensities of the bands at 2.53 eV and 2.30 eV behave similarly. That is, the centers responsible for luminescence bands at 3.14 (1.57) eV, 2.53 eV and 2.30 eV are localized on the surface of ZnO nanoparticles in the area of interparticle contacts (AIC see Fig. 4) in aggregates collapsing first during MP. In our opinion, these PL centers formed as a result of intergrain (interparticle) interactions during ZnO synthesis (and possibly also during powder storage). For example, in ref. 74 it was shown that the annealing treatment (at 500 and 800 °C) in air of undoped ZnO granular films deposited by magnetron sputtering (grains size ∼ 20 nm) leads both to an increase in grains size to ∼100 nm, and to a clear UV emission peak at 380 nm with simultaneously emerging broad green emission.

Fig. 11 Comparison of changes in the intensity of PL band at 3.14 eV (blue) with morphological changes in the samples: the average size of aggregates of nanoparticles, D̄_LPS, (green) and the medium-sized crystallites, D^ZnO_X, (black) occurring in the first 9 minutes of MP.

In other words, the evolution of surface defect structure of chemically synthesized ZnO particles in the zone of interparticle contacts during synthesis (and storage) results in the formation of strong cohesive area interparticle contacts (aggregation) and in the formation (involving local molecular environment) specific luminescence centers. The constant position and width of these PL bands in combination with ultra-fast changes in intensity during the destruction of interparticle contact (disaggregation) point to the existence of structure-specific luminescence centers, depending on the factors mentioned above as well as on the distance between the contacting surfaces, where separation of the contact zones destroys the luminescence centers.

The nanoparticle aggregates formed have variable mechanical strength. Therefore, partial disaggregation of the powder takes place during ZnO + xC mixture preparation. Nanocarbon additives increase disaggregation of initial ZnO powder at short MP times, and slow down the process of secondary aggregation after prolonged MP.

3.5.2. Evolution of PL band ∼2.1 eV of N-doped ZnO defect states. A strict analysis of the changes in peak intensity at ∼2.10 eV (Fig. 8–10) is complicated due to an overlap with intensive adjacent PL bands. The low intensity band at ∼2.10 eV can be detected in PL spectra after t_MP ≥ 3 min (Fig. 9c), that is, after the weakening of bands at 2.53 eV and 2.30 eV. Nevertheless, it should be noted that with increasing duration of MP (t_MP = 9–390 min) one can observe weakening of the band at ∼2.10 eV, and in samples with a high content of particles of nanocarbon this weakening is faster (Fig. 10). PL emission band at ∼2.10 eV was previously observed in N-doped ZnO.⁷⁵ Nitrogen-related emission centers are more stable than emissions at 3.14 (1.57) eV, 2.53 eV and 2.3 eV, and they may be localized in the bulk of ZnO particles. One cannot exclude that other formation mechanisms for these centers can take place upon MP. An accelerated rate of intensity weakening in the samples with high nanocarbon content (t_MP = 390 min, T_s ∼ 700 K) is apparently due to CO-inhibited nitrogen-related PL centers in ZnO.

3.5.3. PL sensing of nanocarbon oxidation in ZnO + xC at prolonged MP. After 9 min MP, the decrease ZnO nanoparticle size is about four-fold, down to ∼60 nm (Fig. 1b), and the sample temperature reaches ∼600 K (Fig. 1b). After 390 min MP, the size of ZnO nanoparticles decreases to ∼15 nm, and the sample temperature reaches ∼700 K, promoting evolution of defect structures. That is, the conditions created in ZnO + xC samples are favorable for different interactions to occur.

New bands with the maxima at 2.84 and 1.42 eV appear in PL spectra of all samples at t_MP ≥ 9 min (Fig. 2b and 5b). The infrared PL band at 1.42 eV is attributed to the second order diffraction peak of the band at 2.84 eV.^45,46 As follows from Fig. 10, the band intensity at 2.84 eV increases for longer MP duration and higher carbon concentration. Note that correlation between the PL intensity increases (Fig. 10), and the C EPR signal intensity increases (Fig. 6) as well.

A number of emission bands in UV and visible spectral ranges were reported for ZnO nanoparticles prepared by sol–gel technique⁷⁶ and for ZnO films prepared by radiofrequency magnetron sputtering.⁷⁷ Different hypotheses were proposed to explain the origin of PL bands around 2.80 eV. In particular, the blue band with maximum at 2.80 eV in ZnO was attributed to Cu-related defects,^78,79 to interstitial Zn species,⁸⁰ or to the donor–acceptor pairs including shallow donor and oxygen vacancies.⁸¹ At the same time, in mechanically processed commercial undoped ZnO powders, PL bands around 2.80 eV were not observed.^17,22,24,25

Instead, such PL bands are typical for oxidized forms of carbon:graphite oxide, graphene oxide,⁴² carbon nanodots prepared by electrochemical etching of carbon fibers,⁴² as well as for ZnO + C₆₀ fullerene mixtures⁸² (Fig. 12). The different PL spectra for carbon nanodots can be attributed to the presence of surface energy traps, which become emissive upon stabilization as a result of surface passivation.⁸³ The factors resulting in tunable performance are different particle size and different oxygen-containing groups.⁸⁴ The PL peak found at 2.80 eV in the present case is very close to the ones observed for graphite oxides and carbon nanodots with 2.7 nm size.⁴⁴ Thus, one can associate PL band at 2.80 eV appearing during high-energy ball milling of ZnO + xC mixtures with the formation of some oxidized carbon nanodots (OCN) with a size of ∼2.7 nm.

Fig. 12 Comparison of PL spectra for graphite and graphene oxides,⁴² ZnO mixtures with fullerene C₆₀,⁸² carbon nanodots with different sizes,⁴⁴ and mechanically processed ZnO + 3% C sample (t_MP = 90 min).

Monitoring in the temperature range of 20–300 K showed that PL maximum at 2.80 eV does not change its position.⁴³ On the contrary, other PL bands (at 3.14 eV, 2.58 eV and 2.23 eV at 300 K) shifted monotonously to higher energies when temperature decreased. The latter behavior is typical for PL spectra of ZnO:Ag nanorods and is related to shrinking the energy band gap at higher temperatures.⁸⁵

4. Discussion

The data of EPR and PL spectroscopy, X-ray diffraction analysis, and morphological analysis by means of AFM and LPS are indicative of a variety of interrelated parallel processes in ZnO + xC mixtures subjected to high-energy MP in a hermetically sealed grinding chamber. The data obtained allow for identification of defect states in the initial mixture components, for detecting newly formed defect states, as well as for monitoring their evolution. In turn, this makes it possible to suggest an approximate scheme of the evolution of physical, chemical and morphological changes which occur during specific conditions of MP treatment of the samples.

4.1. Initial state of reagents used and principal details of sample preparation

Data on the chemical composition of reagents used and main technical details of the sample preparation are presented in section Materials and methods. Here we focus on the principal details necessary for understanding the results obtained.

Initial ZnO powder represents a system of strongly bound particle aggregates (Fig. 4a). An average size of the individual constituent particles in the aggregates is about 250 nm, and an average size of the aggregates, determined by non-destructive ultrasonic treatment of LPS measurements, is ∼1.7 μm. In untreated ZnO + xC samples, the intense luminescence bands in the UV (3.14 eV), visible (green-orange luminescence, ∼2.42 eV) and IR (1.57 eV) ranges (Fig. 8) and weak D_H EPR signal (Fig. 5) were observed. The narrow UV peak at 3.14 eV has to do with NBE emission. The narrow IR PL peak at 1.57 eV (III) corresponds to the second order diffraction peak of 3.14 eV band.⁷² The visible luminescence spectra represent a superposition of two intense PL peaks at 2.53 eV and 2.30 eV and the low intensive peak at ∼2.1 eV (Fig. 8a and b). The I_PL intensities for the initial samples of ZnO with 0.1% C, 1.0% C and 3.0% C added (i.e., after the mechanical mixing) changed according to the ratio of about 5 : 2 : 1, respectively (Fig. 8). The spectral features observed are associated with the sample history, namely, with the sample storage, preparation and synthesis conditions.

The second important circumstance is oxygen content in the grinding chamber. The chamber volume is 50 ml. When the sample is loaded for MP and the chamber is hermetically sealed, it remains to be partially filled with air. The calculated volume of pure oxygen in the chamber is about 5 ml, which is about 2.6 × 10²⁰ oxygen atoms. On the other hand, the carbon content in the samples ZnO + xC, where x = 0.1%, 1.0% and 3.0% by weight, is ∼5 × 10²⁰, ∼5 × 10²¹ and 1.5 × 10²² atoms respectively. That is, the available oxygen content in the chamber is insufficient for complete oxidation even of the minimal (0.1%) amount of carbon in the sample.

The third important circumstance is the high-energy mechanical action on ZnO + xC mixtures. In this context, one should take into account not only the processes of grinding and defects formation, but also development of thermal processes in the samples.

4.2. Physico-chemical processes and rheological changes at t_MP = 1–9 min

During the first minutes of sample MP, a dramatic (more than 100-fold) reduction in intensity of the above-mentioned PL bands and D_H EPR signal is observed, which correlates with the sample disaggregation (Fig. 5). This allows to attribute the localization of PL and D_H EPR centers to AIC area (Fig. 4a and c). Partial destruction of the aggregates occurs already during ZnO + xC mixture preparation, which explains the difference in PL band intensities in the original samples. Noteworthy, AIC represents a special interest for understanding the cohesion mechanisms. The high sensitivity of PL spectra observed can be employed to study the processes of formation/destruction of this area.

The grinding reduction of initial ZnO particles from 250 nm to about 60 nm size (Fig. 2b) is accompanied by the formation of Zn vacancy-related EPR centers (Fig. 5 and 6) in the area of destruction (Fig. 4d). Simultaneously, the sample temperature increases to about 420 K (Fig. 2b).

Milling of the carbon particles did not result in the detection of C EPR signal, apparently due to a limited sensitivity of our EPR measurements.

4.3. Physico-chemical processes at t_MP = 9–90 min

At these processing times, the particle size decreased from ∼60 to ∼38 nm, and the sample temperature increased from about 420 K to ∼600 K (Fig. 2b). The increase in sample temperature leads to the following phenomena:

(a) Annealing of Zn vacancy-related EPR centers in ZnO nanoparticles.

(b) Initiating the interaction of carbon nanoparticles with oxygen in the grinding chamber. Two carbon oxide species form in the chamber: CO and CO₂. Since oxygen content in the chamber is limited, the relative (with respect to CO₂) CO content in the gas medium becomes higher for the samples with high carbon nanoparticle content. The CO content in the gas medium increases also with increasing t_MP.

(c) Simultaneously, the process of carbon oxidation promotes the initiation and growth of C EPR signal of Lorentzian shape (spin clusters) in carbon nanoparticles (Fig. 5–7). The intensity of this signal is roughly proportional to nanocarbon content in ZnO + xC samples.

(d) Graphite (graphene) oxides begin to form on the surface of carbon nanoparticles, which exhibit luminescence band at 2.80 eV (Fig. 8 and 12).

(e) MP, the formation of reducing environment in the grinding chamber, and increase in sample temperature result in the initiation and stabilization of donor D_S centers in AD of ZnO nanoparticles (Fig. 4d).

4.4. Physico-chemical processes at t_MP = 90–390 min

At these processing times, ZnO particle size further reduces from ∼38 nm to ∼14 nm, and the sample temperature increases from ∼600 K to ∼700 K (Fig. 2b). The integral intensities of C EPR signal in the samples ZnO + xC with different carbon content (x = 0.1%, 1.0% and 3.0%) become roughly equal (Fig. 6), thereby emphasizing the role of limited oxygen content in the grinding chamber in generating a certain number of broken valences in carbon nanoparticles. Variable (depending on x) CO/CO₂ ratio influences geometry of the C-spin clusters formed. The intensities of luminescence band of graphite oxide at 2.80 eV for these three samples increase somewhat and become roughly equal also (Fig. 10). Within the frame of “carbon nanodots” model⁴⁴ the results observed are typical for PL band of nanodots with a size of 2.7 nm. We can assume the formation of nanocomposite particles with ∼14 nm ZnO core coated with carbon nanoparticles of ∼2.7 nm size.

The increase of CO content with increasing t_MP completely inhibits D_S paramagnetic centers in ZnO + xC samples. The conditions created in the grinding chamber at prolonged MP promote emergence of the EPR spectral features due to Mn²⁺ ions in ZnO_ZB sphalerite phase. Such a spectrum was observed for the sample with minimal carbon content only. The newly formed ZnO_ZB phase is located in DSA zone of initial ZnO particles (Fig. 4). The layer thickness of this phase is ∼0.4–0.6 nm.

5. Conclusions

In summary, our EPR and PL studies show that ZnO + xC samples exhibit a number of EPR signals , and D_S in ZnO_W, Mn²⁺ ions in ZnO_ZB, C in carbon nanoparticles) and PL emission bands (NBE, Zn-, O- and N-enriched ZnO_W particles, as well as oxidized carbon nanodots) under different duration of their MP.

Dramatic reduction in intensity of D_H EPR and PL bands at 3.14 (1.57), 2.53 and 2.30 eV during the first minutes of MP are consistent with the localization of corresponding centers in the area of ZnO-interparticle contacts within strongly bound particle aggregates of initial ZnO powder. In other words, reduction in the band intensities correlates with the sample disaggregation.

Increasing t_MP results in the following phenomena: (a) grinding of ZnO particles and the formation of Zn vacancy-related EPR centers in in the area of destruction; (b) increase in sample temperature; (c) annealing of the Zn vacancy-related EPR centers formed; (d) initiating the interaction of carbon nanoparticles with oxygen in the grinding chamber; (e) formation and growth of the C signal in carbon nanoparticles; (f) formation of reducing environment in the grinding chamber; (g) stabilization of the donor D_S centers in the area of destruction of ZnO nanoparticles; (h) increase in CO content in the grinding chamber; (i) inhibition of D_S paramagnetic centers in ZnO; (j) emergence of the EPR spectral features due to Mn²⁺ ions in ZnO_ZB sphalerite phase in the sample with minimal carbon content only; (k) inhibition of N PL centers in ZnO; (l) formation of oxidized carbon nanodots exhibiting PL band at 2.80 eV.

Acknowledgements

This work was supported by the National Council of Science and Technology of Mexico (CONACYT, project number 155731) and the CIICAp-Universidad Autonoma del Estado de Morelos.

References

1. P. Mele, T. Endo, S. Arisawa, C. Li and T. Tsuchiya, Oxide Thin Films, Multilayers, and Nanocomposites, Springer International Publishing, Switzerland, 2015.
2. P. H. Cury Camargo, K. G. Satyanarayana and F. Wypych, Mater. Res., 2009, 12, 1–39.
3. S. A. Ansari, M. M. Khan, S. Kalathil, A. Nisar, J. Lee and M. H. Cho, J. Phys. Chem. C, 2013, 117, 27023–27030.
4. R. Gurwitz, R. Cohen and I. Shalish, J. Appl. Phys., 2014, 115, 033701.
5. H. Yu, H. Zhang, H. Huang, Y. Liu, H. Li, H. Ming and Z. Kang, New J. Chem., 2012, 36, 1031–1035.
6. P. R. Potti and V. C. Srivastava, Ind. Eng. Chem. Res., 2012, 51, 7948–7956.
7. S. Adhikari, D. Sarkar and G. Madras, RSC Adv., 2014, 4, 55807–55814.
8. Z. Fan and J. G. Lu, J. Nanosci. Nanotechnol., 2005, 5, 1561–1573.
9. H. S. Hsu, Y. Tung, Y. J. Chen, M. G. Chen, J. S. Lee and S. J. Sun, Phys. Status Solidi RRL, 2011, 5, 447–449.
10. Z. Chen, N. Zhag and Y.-J. Xu, CrystEngComm, 2013, 15, 3022–3030.
11. N. Pan, H. Xue, M. Yu, X. Cui, X. Wang, J. G. Hou and J. Huang, Nanotechnology, 2010, 21, 225707.
12. Y. Hu and H.-J. Chen, J. Appl. Phys., 2007, 101, 124902.
13. G. Katumba, L. Olumekor, A. Forbes, G. Makiwa and B. Mwakikunga, Sol. Energy Mater. Sol. Cells, 2008, 92, 1285–1292.
14. A. B. Djurišić, A. M. C. Ng and X. Y. Chen, Prog. Quantum Electron., 2010, 34, 191–259.
15. L. S. Vlasenko, Appl. Magn. Reson., 2010, 39, 103–111.
16. S. Yildirimcan, K. Ocakoglu, S. Erat, F. M. Emen, S. Repp and E. Erdem, RSC Adv., 2016, 6, 39511–39521.
17. H. Kaftelen, K. Ocakoglu, R. Thomann, S. Tu, S. Weber and E. Erdem, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 86, 014113.
18. M. G. Kakazey, G. N. Kakazei and J. G. Gonzalez-Rodriguez, Cryst. Res. Technol., 2001, 36, 429–439.
19. V. Ischenko, S. Polarz, D. Grote, V. Stavarache, K. Fink and M. Driess, Adv. Funct. Mater., 2005, 15, 1945–1954.
20. S. Das, C. K. Ghosh, R. Dey and M. Pal, RSC Adv., 2016, 6, 236–244.
21. N. Vorobyeva, M. Rumyantseva, D. Filatova, E. Konstantinov, D. Grishina, A. Abakumov, S. Turner and A. Gaskov, Sens. Actuators, B, 2013, 182, 555–564.
22. M. Šćepanović, M. Grujić-Brojčin, K. Vojisavljević and T. Srećković, J. Appl. Phys., 2011, 109, 034313.
23. T. K. Kundu, N. Karak, P. Barik and S. Saha, International Journal of Soft Computing and Engineering, 2011, 1, 19.
24. P. K. Giri, S. Bhattacharyya, D. K. Singh, R. Kesavamoorthy, B. K. Panigrahi and K. G. M. Nair, J. Appl. Phys., 2007, 102, 093515.
25. R. Radoi, P. Fernández, J. Piqueras, M. S. Wiggins and J. Solis, Nanotechnology, 2003, 14, 794–798.
26. J. G. Lu, L. P. Zhu, Z. Z. Ye, Y. J. Zeng, F. Zhuge, B. H. Zhao and D. W. Ma, J. Cryst. Growth, 2005, 274, 425–429.
27. S. B. Zhang, S.-H. Wei and A. Zunger, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 63, 075205.
28. A. B. M. A. Ashrafi, I. Suemune and H. Kumano, Jpn. J. Appl. Phys., 2002, 41, L1281–L1284.
29. B.-P. Zhang, L.-H. Manh, K. Wakatsuki, T. Ohnishi, M. Lippmaa, N. Usami, M. Kawasaki and Y. Segawa, Jpn. J. Appl. Phys., 2003, 42, 2291–2295.
30. B. Cao, W. Cai and H. Zeng, Appl. Phys. Lett., 2006, 88, 161101.
31. H. Li, Y. Zhang, X. Pan, H. Zhang, T. Wang and E. Xie, J. Nanopart. Res., 2009, 11, 917–921.
32. L. Irimpan, V. P. N. Nampoori, P. Radhakrishnan, A. Deepthy and B. Krishnan, J. Appl. Phys., 2007, 102, 063524.
33. Y. Zhang, B. Lin, Z. Fu, C. Liu and W. Han, Opt. Mater., 2006, 28, 1192–1196.
34. J. S. Reparaz, F. Güell, M. R. Wagner, A. Hoffmann, A. Cornet and J. R. Morante, Appl. Phys. Lett., 2010, 96, 053105.
35. H. C. Hsu, C.-Y. Wu and W.-F. Hsieh, J. Appl. Phys., 2005, 97, 064315.
36. W.-K. Hong, G. Jo, M. Choe, T. Lee, J. I. Sohn and M. E. Welland, Appl. Phys. Lett., 2009, 94, 043103.
37. A. Umar, B. Karunagaran, E.-K. Suh and Y. B. Hahn, Nanotechnology, 2006, 17, 4072–4077.
38. R. C. Barklie, Diamond Relat. Mater., 2001, 10, 174–181.
39. R. C. Barklie, M. Collins and S. R. P. Silva, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 61, 3546–3554.
40. F. Beuneu, C. L'Huillier, J.-P. Salvetat, J.-M. Bonard and L. Forro, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 59, 5945–5949.
41. A. I. Shames, E. A. Katz, A. M. Panich, D. Mogilyansky, E. Mogilko, J. Grinblat, V. P. Belousov, I. M. Belousova and A. N. Ponomarev, Diamond Relat. Mater., 2009, 18, 505–510.
42. G. Singh, A. Choudhary, D. Haranath, A. G. Joshi, N. Singh, S. Singh and R. Pasricha, Carbon, 2012, 50, 385–394.
43. C. T. Chien, S. S. Li, W. J. Lai, Y. C. Yeh and H. A. Chen, Angew. Chem., Int. Ed., 2012, 51, 6662–6666.
44. F. Liu, Y. Cao, M. Yi, L. Xie, W. Huang, N. Tang, W. Zhong and Y. Du, Crystals, 2013, 3, 28–37.
45. T. Torchynska, B. Perez Millan, E. Velazquez Lozada, M. Kakazey and M. Vlasova, ECS Trans., 2015, 66, 297–303.
46. E. Velázquez Lozada, T. Torchynska, J. L. Casas Espinola, A. Vivas Hernandez, M. Kakazey, M. Vlasova, L. Shcherbyna and L. Castañeda, Mater. Sci. Semicond. Process., 2015, 237, 82–86.
47. T. Torchynska, Phys. E, 2011, 44, 56–61.
48. M. Kakazey, M. Vlasova and E. A. Juarez-Arellano, J. Nanopart. Res., 2015, 17, 118.
49. M. Kakazey, M. Vlasova, E. A. Juarez-Arellano and T. Torchynska, Mater. Sci. Semicond. Process., 2015, 39, 775–780.
50. A. B. Palotas, L. C. Rainey, C. J. Felderman and J. B. Vander Sande, Microsc. Res. Tech., 1996, 33, 266–278.
51. T. V. Torchynska, M. Dybiec and S. Ostapenko, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 72, 195341.
52. T. Ruf, S. Repp, J. Urban, R. Thomann and E. Erdem, J. Nanopart. Res., 2016, 18, 109.
53. M. Kakazey, M. Vlasova, M. Dominguez-Patiño, G. Dominguez-Patiño, Y. Enríquez-Méndez, M. C. Reséndiz-Gonzalez, T. Tomila, M. Ristic and M. V. Nikolic, Powder Technol., 2009, 93, 130–136.
54. D. M. Hoffmann, A. Hofstaetter, F. Leiter, H. Zhou, F. Henecker and B. K. Meyer, Phys. Rev. Lett., 2002, 88, 045504.
55. E. Erdem, J. Alloys Compd., 2014, 605, 34–44.
56. B. Schallenberger and A. Hausmann, Z. Phys. B: Condens. Matter Quanta, 1976, 23, 177–181.
57. D. Galland and A. Herve, Solid State Commun., 1974, 14, 953–956.
58. M. G. Kakazey, M. Vlasova, M. Dominguez-Patiño, G. Dominguez-Patiño, G. Gonzalez-Rodriguez and B. Salazar-Hernandez, J. Appl. Phys., 2002, 92, 5566–5568.
59. M. Kakazey, M. Vlasova, M. Dominguez-Patiño, I. Leon and M. Ristic, J. Mater. Sci., 2007, 42, 7116–7122.
60. L. E. Halliburton, N. C. Giles, N. Y. Garces, M. Luo, C. Xu, L. Bai and L. A. Boatner, Appl. Phys. Lett., 2005, 87, 172108.
61. V. A. Nikitenko, K. É. Tarkpea, I. V. Pykanov and S. G. Stoyukhin, J. Appl. Spectrosc., 2001, 68, 502–507.
62. H. G. van Bueren, Imperfections in crystals, North-Holland Pub. Co., Amsterdam, 2nd edn, 1961.
63. D. Arcon, Z. Jaglicic, A. Zorko, A. V. Rode, A. G. Christy, N. R. Madsen, E. G. Gamaly and B. Luther-Davies, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 74, 014438.
64. A. Hausmann, Solid State Commun., 1968, 6, 457–459.
65. S. Kunii, S. Tobita and E. Yirahara, J. Phys. Soc. Jpn., 1966, 21, 479–484.
66. G. H. Lee, T. Kawazoe and M. Ohtsu, Appl. Surf. Sci., 2005, 239, 394–397.
67. Y. Z. Yoo, Y. Osaka, T. Fukumura, M. Kawaski, H. Koinuma, T. Chikyow, P. Ahmet, A. Setoguchi and S. F. Chichibu, Appl. Phys. Lett., 2001, 78, 616–618.
68. A. B. M. A. Ashrafi, I. Suemune and H. Kumano, Jpn. J. Appl. Phys., 2002, 41, 2851–2854.
69. L. Xu, Y. Su, Y. Chen, H. Xiao, L. A. Zhu, Q. Zhou and S. Li, J. Phys. Chem. B, 2006, 110, 6637–6642.
70. A. B. M. A. Ashrafi and C. Jagadish, J. Appl. Phys., 2007, 102, 071101.
71. M. Kakazey, M. Vlasova, M. Dominguez-Patiño and J. Munguia-Diaz, Solid State Commun., 2008, 145, 122–125.
72. T. V. Torchynska and B. El Filali, J. Lumin., 2014, 149, 54–60.
73. C. Ton-That, L. Weston and M. R. Phillips, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 86, 115205.
74. X. Xu, C. Xu, G. Chen, J. Wu and J. Hu, Europhys. Lett., 2013, 101, 27009.
75. M. A. Reshchikov, J. Q. Xie, B. Hertog and A. J. Osinsky, J. Appl. Phys., 2008, 103, 103514.
76. S. Dutta and B. N. Ganguly, J. Nanobiotechnol., 2012, 10, 29.
77. Y. S. Choi, J. W. Kang, D. K. Hwang and S. J. Park, IEEE Trans. Electron Devices, 2010, 57, 26–41.
78. J. Lv, C. Liu, W. Gong, Z. Zi, X. Chen, K. Huang, T. Wang, G. He, S. Shi, X. Song and Z. Sun, Opt. Mater., 2012, 34, 1917–1920.
79. M. A. Reshchikov, H. Morkoc, B. Nemeth, J. Nause, J. Xie, B. Hertog and A. Osinsky, Phys. B, 2007, 401–402, 358–361.
80. M. K. Patra, K. Manzoor, M. Manoth, S. P. Vadera and N. Kumar, J. Lumin., 2008, 128, 267–272.
81. D. H. Zhang, Z. Y. Xue and Q. P. Wang, J. Phys. D: Appl. Phys., 2002, 35, 2837–2840.
82. D. I. Son, B. W. Kwon, J. D. Yang, D. H. Park, B. Angadi and W. K. Choi, J. Mater. Chem., 2012, 22, 816–819.
83. R. L. Liu, D. Q. Wu, S. H. Liu, K. Koynov, W. Knoll and G. Li, Angew. Chem., 2009, 121, 4668–4671.
84. H. Li, Z. Kang and Y. Liu, J. Mater. Chem., 2012, 22, 24230–24253.
85. E. Velázquez Lozada, T. V. Torchynska, J. L. Casas Espinola, B. Perez Millan and L. Castañeda, Phys. B, 2014, 453, 111–115.

---