Nitride Materials: Synthesis, Crystal Structures, and Optical Properties

Nitride Materials: Synthesis, Crystal Structures, and Optical Properties

Djahida KerdoudFaouzia Benkafada Nora Boussouf Chahrazed Benhamideche 

University 20 Août 1955, BP 26, Route d’El Hadaiek Skikda 21000, Algeria

University Frères Mentouri Constantine 1, BP 325, Route Ain El Bey, Constantine 25017, Algeria

University Centre Abdel Hafid Boussouf, BP 26 RP, Mila 43000, Algeria

Corresponding Author Email: 
d.kerdoud@centre-univ-mila.dz
Page: 
103-108
|
DOI: 
https://doi.org/10.18280/acsm.460206
Received: 
6 February 2022
|
Revised: 
25 March 2022
|
Accepted: 
3 April 2022
|
Available online: 
30 April 2022
| Citation

© 2022 IIETA. This article is published by IIETA and is licensed under the CC BY 4.0 license (http://creativecommons.org/licenses/by/4.0/).

OPEN ACCESS

Abstract: 

Our research involves the preparation of transition metal nitrides of the composition Mn4N, NbN, Mo2N, TaN and ZrN. The synthesis of Li3N binary alkali metal nitride was also part of this work. Simple and cost-effective methods with relatively low impact on the environment have been privileged in the selection. The experimental work has focused on determining the optimum conditions of synthesis and the convenient high yield route to the desired nitrides, and ultimately improvement of the properties of the final materials. All samples were characterised by X-ray powder diffraction. Their structures will be discussed in more detail here. Optical band gap has been calculated from diffuse reflectance measurements. The air sensitivity of the nitrides was also probed.

Keywords: 

nitrides, X-ray diffraction, sub-stoichiometric nitrogen, Rietveld refinement, band gap

1. Introduction

Inorganic nitride materials have received considerable interest during the past decade because of their technological importance. Compared to metal oxides or metals, the properties of the corresponding nitrides are superior in some respects. Unfortunately, the development of these materials is still limited [1] by providing many interesting challenges to the synthetic solid-state chemist. Due to nitrogen’s high bond energy (941 KJ/mol), the formation of nitrogen gas requires high temperatures for activation; however, at these temperatures the M-N bonds in ionic/covalent nitrides become less stable. This is exacerbated by entropic effects, which favor lower nitrogen-to-metal ratios with increasing temperature.

Many nitrides are air and moisture sensitive, and rapidly form oxides, hydroxides and ammonia upon contact with oxygen or moisture. They must be prepared in oxygen-and water-free atmospheres to achieve high purity [2-5]. Synthetic approaches using more activated forms of nitrogen can be used to overcome the inertness of N2, but increased exothermicity can also result in diminished stoichiometric control and the activation of deleterious competing pathways.

These stringent synthesis constraints, coupled with the poor intrinsic stabilities of nitrides, impose significant risk on the exploratory synthesis of novel nitride materials. Recent developments in the handling methods for air sensitive samples and improved diffraction techniques have led to moving forward in the area of nitride chemistry, making a large increase in research [1, 6, 7].

In this work, there have been attempts to synthesise the alkali metal nitride Li3N and materials based on the nitrides of d-block transition metals mainly, Mn4N, NbN, Mo2N, TaN and ZrN. This contribution describes how nitride materials are produced using different methods. Powder X-ray diffraction with emphasis placed on the structural studies and optical band gap using UV-Vis reflectance spectroscopy are main focus in this paper.

2. Materials and Methods

All procedures were performed in an inert atmosphere inside a glovebox equipped with a circulative purification system to prevent moisture and oxygen contaminations, in which the H2O and O2 concentrations were kept below 1 ppm.

Manganese nitride Mn4N was prepared by heating Mn powder (Aldrich, 99.9%) in a flow of super pure NH3 at 600℃ for 3 h. An annealing at 500℃ has been followed for 20 h in the same atmosphere.

Niobium nitride NbN nitride was synthesised by the vapor-phase reaction of gaseous niobium pentachloride (NbCl5) with ammonia (NH3) at 200-1300℃ [8]. The formation process of NbN can be followed by changing heating temperature. The reaction of NbCl5 with ammonia to form NbCl5 and 5NH3 has been occurred first at around 200℃. The Nb4N5 was obtained at 250-950℃ is considered to be due the reaction of NbCl5-5NH3 with ammonia above about 250℃ leading to Nb4N5 product. Nb4N5 decomposes and can be converted to NbN at above 1000℃.

Molybdenum nitride Mo2N was produced by the temperature programmed reaction method previously described [9]. Approximately 0.1 g of molybdenum (VI) oxide powder (MoO3, Aldrich) was loaded onto a quartz wool plug in a quartz tube reactor. The oxide was first nitridated using a mixture of nitrogen/hydrogen with a predetermined temperature. After the material was cooled to room temperature, the material was passivated in a mixture of 1% O2 in He flowing for 24 h. This step was necessary to prevent pyrophoric oxidation upon contact with air.

Tantalum nitride TaN was prepared by the reduction in liquid ammonia. The reaction of solutions of tantalum pentachloride (TaCl5) with sodium was previously reported [10]. The obtained powder was heated under vacuum at 900℃ for 24 h.

Zirconium nitride ZrN was produced from Zirconium tetrachloride (ZrCl4, Alfa Aesar) and ammonia (NH3) injection using nonthermal plasma reactor [11]. The as-obtained powder was annealed at 250℃ and 0.3 Torr in 90 SCCM of argon flow for 1 h to remove the ammonium salt [12].

Lithium metal foil (99.9 purity, 0.75 mm thickness) was purchased from Alfa Aesar and was used for the preparation of Li3N. During the synthesis, about 1 g of Li metal foil was filed firstly until most of the metal oxide coating removed and a shiny surface appeared. In a recirculating glovebox filled with nitrogen, the Li metal was further filed until no oxide layer remained. After 3 days, the cleaned Li metal foil transformed spontaneously into a Li3N flake, which was further heat-treated outside the glovebox using a furnace to ensure the metal has fully reacted with nitrogen. The direct reaction took place under N2 gas at 200℃ for 24 h. Li3N powder was obtained after grinding using sandpaper or file.

All obtained samples Mn4N, NbN, Mo2N, TaN, ZrN and Li3N, were transferred directly to the glovebox after cooling in each case, which reduce the risk of impurity phases forming. All the ground samples were stored in the glovebox for further characterisation.

3. Materials Characterisation

All product powders were characterised by the standard powder X-ray diffraction (PXD) technique with a PANalytical X’Pert Pro MPD powder diffractometer (Cu Kα source, 40 kV, 40 mA). Samples were loaded in the glovebox into a bespoke aluminum sample holder consisting of a sealed chamber equipped with a Mylar window. All phase analyses were performed using the HighScore plus software. For the phase-pure samples, Rietveld refinements in diffraction patterns were conducted using the GSAS and EXPGUI packages [13, 14]. The scale factor, zero point and background were refined in initial cycles. A shifted Chebyschev polynomial function (background function 1 in GSAS) was employed to model the background. The unit cell parameters, peak profile parameters and atomic parameters were refined subsequently. The peak shape was modelled using the pseudo-Voigt function (profile function 2 in GSAS).

Powdered samples were irradiated by UV-Vis spectrophotometer. The powder was placed in the sample holder inserted into diffuse reflectance spectrometer. The praying mantis mirror was placed over the top of the sample and the sample compartment sealed for light. Scans were run in reflectance mode over the operative range 190-1300 nm. Band gap values were calculated by extrapolation of the steepest part of the curve towards the baseline, reading off at 0% in nm and converting to eV using the Kubelka-Munk approach proposed by Tauc et al. [15], and David and Mott [16].

4. Results and Discussion

Analysis of PXD data for the full series of samples served to confirm their purity. Figure 1 shows the X-ray diffraction pattern of manganese nitride derived by the conventional heating method under ammonia gas. The peaks were indexed with the cubic phase of Mn4N (ICSD#01-089-4804) and the hexagonal phase of Mn2N0.86 (ICSD#01-071-0200).

Figure 1. Powder X-ray diffraction pattern of Mn4N. The Mn2N0.86 phase is indicated by the asterisks

Figure 2. Powder X-ray diffraction pattern of NbN. The Nb8N6.8 phase is indicated by the asterisks

Figure 3. Powder X-ray diffraction pattern of Mo2N. The Mo phase is indicated by the asterisks

The typical X-ray diffraction pattern of chemically prepared NbN powder, in the range of $30^{\circ} \leq 2 \theta \leq 120^{\circ}$, is shown in Figure 2. The major feature can be indexed as a hexagonal phase of NbN (PDF#03-065-3417) with the appearance of small and moderate quantity of tetragonal Nb8N6.8 (PDF#03-065-2252) as second phase.

Figure 4. Powder X-ray diffraction pattern of TaN. The Ta2N0.86 phase is indicated by the asterisks

The diffraction pattern of synthesised Mo2N in Figure 3 shows peaks corresponding to the planes which match with the pattern of Powder Diffraction File Database having the card No. 024-0768 with the most intense peaks occurring at diffraction angles, $2 \theta=37.6^{\circ}$ and $43.05^{\circ}$ which correspond to the (112) and (200) reflection planes, respectively. The diffraction at $2 \theta=40.5^{\circ}, 58.6^{\circ}, 73.6^{\circ}, 87.6^{\circ}$ and $101.4^{\circ}$  can be identified as the $(111),(200),(211),(220) $ and $(310)$crystal planes of Mo (PDF#00-004-0809).

The peaks recorded in Figure 4 match those of hexagonal TaN (PDF#00-039-1485) with the presence of slight impurity of hexagonal Ta2N0.86 (PDF#01-089-5199).

It was observed that the routes used for the synthesis of Mn4N, NbN, Mo2N and TaN produced a nitrogen deficient nitrides of formula Mn2N0.86, Nb8N6.8, and Ta2N0.86, as second phases, which is likely these materials formed are sub-stoichiometric nitrogen. This low value may be due to the refractory nature of these metal nitrides. This may also correspond to some nitrogen sub-stoichiometry as many transition metal nitrides are interstitial compounds and are metal rich, which they are known to tolerate wide ranges in composition [17]. It should be noted that some nitrogen loss may occur prior to analysis.

ZrN and Li3N have been synthesised and studied. Structural refinements performed against powder X-ray diffraction data were conducted using General Structure Analysis System (GSAS). The least-square refinements were carried out until the best fit was obtained between the entire observed powder diffraction patterns taken as a whole and the entire calculated patterns based on the published structures as starting models. The major parameters refined were scale factors, zero shifts, unit cells, Pseudo-Voigt peaks, shape, and profile parameters.

As-prepared ZrN was a fine, yellow/brown powder. In a typical synthetic method with an appropriate choice of conditions, the nonthermal plasma reaction generates crystalline phase-pure ZrN (PDF#03-065-9412). Most importantly, no oxide or hydroxide impurities have been detected.

Manually fitting the background using 12 coefficients gave the best fit. The experimental pattern fits well with the theoretical pattern as displayed in Figure 5. The main phase was indexed to a cubic cell (Space group No.225: $F m \overline{3} m$). Rietveld refinement yielded lattice parameters of a=4.58093(4) Å and a cell volume of 96.131(2) Å3.

Figure 5. Observed-calculated-difference (OCD) profile plot from the Rietveld refinement for ZrN. Observed data are shown in red (crosses), calculated (green solid line) and the difference between the two curves (pink). Black tickmarks correspond to peak positions of cubic ZrN

Selected Rietveld refinement data and atomic parameters are presented in Tables 1 and 2. Estimated standard deviation (esd) values are shown in parentheses.

Table 1. Crystallographic data obtained from the Rietveld refinement for ZrN powder

Chemical formula

ZrN

Crystal system

Space group

Lattice parameter, a /Å

Volume / Å3

Formula units, Z

Formula weight/ g mol-1

Calculated density/ g cm-3

No. of observations, parameters

Rwp, Rp

$\chi^{2}$

Cubic

$\operatorname{Fm} \overline{3} m(225)$

4.58093(4)

96.131(2)

4

356.154

6.152

5385, 25

0.2933, 0.2379

12.93

Table 2. Atomic parameters for ZrN powder

Atom

Wyckoff symbol

x

y

z

Occupancy

Uiso×100/Å2

N

4a

0

0

0

1

0.38(23)

Zr

4b

0.5

0.5

0.5

1

1.21(2)

The reaction between Li metal and nitrogen gas to form Li3N occurred spontaneously. Li metal foil was left in a nitrogen-filled glovebox, and after 3 days of exposure, the color of the Li foil changed from white to black. The as-obtained powder was heated at 200℃ for around 24 h to ensure a complete reaction. During the course of the experiment, N2 gas was flowed through the furnace and paraffin bubbler heating the crucible allowed for reaction temperature 200℃ to be reached. Once cooled, the crucible and its content were removed from the furnace and transferred to a glovebox where they were opened using pipe cutter, which allowed the crucible lid to be removed cleanly without polluting the sample. The black powder of Li3N turned purple/red with lustrous effect after annealing process in the same atmosphere. The final powder was thoroughly mixed and ground together using an agate pestle and mortar. The crystallinity and phase purity of the annealing powder after grinding was then determined by powder X-ray diffraction.

All the diffraction peaks of the resulting powder can be readily indexed to alpha-Li3N exhibiting strong intensity, indicating the high crystallinity of the powder sample. Weak peaks of Beta-Li3N were additionally observed. This may be due to the pressure induced phase transformation during the grinding process [18]. No other impurities, such LiOH and Li2O, were detected by PXD due to its limited sensitivity.

Figure 6. Observed-calculated-difference (OCD) profile plot from the Rietveld refinement for Li3N powder. Observed data are shown in red (crosses), calculated in green (solid line) and the difference between the two curves (pink). Black tickmarks are reflections from α-Li3N. The red tickmarks correspond to peak positions of β-Li3N

The experimental powder X-ray diffraction data for synthesised Li3N was refined via the Rietveld method in order to obtain weight percentage values for alpha and beta Li3N phases. The models for each of the calculations were obtained from the online Inorganic Crystal Structure Database (ICSD) and PowderCell 2.4 was used to obtain a visual comparison between the experimental and theoretical models.

Table 3. Selected Rietveld refinement data from PXD data of synthesised Li3N powder

Empirical formula

α-Li3N

β-Li3N

Crystal system

Space group

Lattice parameters/ Å        a

                                           c

Volume/Å3

Formula units, Z

Formula weight/ g mol-1

Calculated density/ g cm-3

Phase fraction/ wt.%

No. of observations

No. of variables

Rwp

Rp

$\chi^{2}$

Hexagonal

P6/mmm(191)

3.6447(1)

3.8737(1)

44.566(3)

2

24.301

0.905

70.37(15)

3589

44

0.1086

0.0801

1.685

Hexagonal

P63/mmc(194)

3.5724(3)

6.3417(6)

70.091(13)

4

55.423

1.313

29.63(29)

The refinement was carried out using structural parameters of alpha-Li3N in hexagonal symmetry $[$ Space group:No. 191: $P 6 / m m m, I C S D \# 34779]$, and beta-Li3N in hexagonal symmetry $[$ Space group $N o .194: P 63 / m m c, I C S D \# 156889]$. The observed-calculated-difference (OCD) profile plot is shown in Figure 6. Selected Rietveld refinement data and atomic parameters are presented below in Tables 3, 4 and 5. Estimated standard deviation (esd) values are shown in parentheses.

Rietveld refinement of the diffraction data yielded lattice parameters of a=3.6447(1), c=3.8737(1) for α-Li3N, and a=3.5724(3), c=6.3417(6) for β-Li3N. A cell volume of 44.566(3) and 70.091(13) were obtained, respectively. The agreement indices in the simulated pattern gives the best fit value of weighed Rp-Rwp and goodness of fit-$\chi^{2}$=1.685 with 70.37(15) % of α-Li3N and 29.63(29) % of β-Li3N phase fractions.

Table 4. Atomic parameters for α-Li3N

Atom

Wyckoff symbol

x

y

z

Li1

Li2

N

1b

2c

1a

0

0.3333

0

0

0.6667

0

0.5

0

0

Occupancy

Uiso×100/ Å2

0.573(9)

0.812(5)

0.647(3)

16.03(57)

11.15(19)

10.30(13)

Table 5. Atomic parameters for β-Li3N

Atom

Wyckoff symbol

x

y

z

Li1

Li2

N

2b

4f

2c

0

0.3333

0.3333

0

0.6667

0.6667

0.25

0.5915(15)

0.25

Occupancy

Uiso×100/ Å2

0.879(16)

0.651(10)

0.898(6)

6.63(52)

7.80(33)

10.26(20)

The diffuse reflectance is measured as a function of the wavelength of the incident radiation as in normal UV-Vis spectrometry within the operative range of 190-1300 nm.

The determination of the optical band gap is easily extrapolated from its absorption spectrum according to the process originally proposed by Tauc et al. [15], and David and Mott [16]. Tauc’s plot proposes the proportionality between the band gap (Eg) and the absorption coefficient ($\alpha$) of the material.

$(\alpha h v)^{1 / n}=A\left(h v-E_{q}\right)$       (1)

where, $h$ is the Plank constant, $\mathrm{v}$ is the light frequency and $A$ is a proportionality constant depending on the properties of the material. $n$ defined as power factor and its value depends on the nature of the optical transition probability [19]. The value of $n$ derive from the description of the electronic bands through their wavelength in function of the imaginary part of the dielectric constant of the material [15].

According to the Kubelka-Munk theory [20, 21] and Tauc plot of $(F(R) h v)^{1 / n}$ as a function of the incident radiation energy, the characteristic slop is fitted as a linear interpolation. The intersection of the slop with the x-axis ($h v$) directly provides the value of the band gap [22].

Optical band gaps obtained from the reflectance data on synthesised Mn4N, NbN, Mo2N, TaN, ZrN and Li3N nitride materials were estimated to be 2.18 eV, 3 eV, 3.1 eV, 1.32 eV, 2.8 eV, and 1.95 eV, respectively, as shown in Figures 7, 8, 9, 10, 11 and 12.

Figure 7. UV-Vis spectrum of Mn4N with Tauc plot for band gap determination

Figure 8. UV-Vis spectrum of NbN with Tauc plot for band gap determination

Figure 9. UV-Vis spectrum of Mo2N with Tauc plot for band gap determination

Figure 10. UV-Vis spectrum of TaN with Tauc plot for band gap determination

Figure 11. UV-Vis spectrum of ZrN with Tauc plot for band gap determination

Figure 12. UV-Vis spectrum of Li3N with Tauc plot for band gap determination

5. Conclusions

In the present study, it was demonstrated that the methods outlined, as under which nitride materials produced with high purity, were successfully applied. Some of the following important observations are:

(i) Additional of minor sub-stoichiometric nitrogen phases of formula Mn2N0.86, Nb8N6.8 and Ta2N0.86 have been detected by X-ray powder diffraction, which were probably formed prior to analysis. However, no evidence of any oxidation of the synthesised samples has been seen.

(ii) The Rietveld analysis based on the structure refinement has been adopted for precise determination of structural parameters. Nonthermal plasma reactor is considered the most convenient method leading to successful preparation of ZrN single phase with cubic structure. The Rietveld refinement analysis revealed also the high purity of synthesised Li3N powder, with the formation of 70.37(15)% alpha-Li3N and 29.63(29)% beta-Li3N, suggesting phase transformation during the grinding process.

(iii) Band gaps of all samples were obtained by analysing their UV-Vis reflectance spectra.

  References

[1] Brese, N.E., O’Keeffe, M. (1992). Crystal chemistry of inorganic nitrides. In: Complexes, Clusters and Crystal Chemistry. Structure and Bonding, vol 79. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0036504

[2] Sun, W., Holder, A., Orvañanos, B., Arca, E., Zakutayev, A., Lany, S., Ceder, G. (2017). Thermodynamic routes to novel metastable nitrogen-rich nitrides. Journal of Chemistry of Materials, 29(16): 6936-6946. https://doi.org/10.1021/acs.chemmater.7b02399

[3] Thomas, A.M. (2017). Introducing DDEC 6 atomic population analysis: Part 3. Comprehensive method to compute bond orders. Journal of RSC Advances, 7(72): 45552-45581. https://doi.org/10.1039/C7RA07400J

[4] Thomas, A.M., Nidia, G.L. (2016). Introducing DDEC 6 atomic population analysis: Part 1. Charge portioning theory and methodology. Journal of RSC Advances, 6(53): 47771-47801. https://doi.org/10.1039/C6RA04656H

[5] Maintz, S., Deringer, V.L., Tchougréeff, A.L., Dronskowski, R. (2016). LOBSTER: A tool to extract chemical bonding from plane-wave based DFT. Journal of Computational Chemistry, 37(11): 1030-1035. https://doi.org/10.1002/jcc.24300

[6] Niewa, R., Jacobs, H. (1996). Journal of Chemical Reviews, 96: 2053. 

[7] Niewa. R., DiSalvo. F.J. (1998). Journal of Chemical Reviews. Mater, 10: 2733.

[8] Akimasa, Y., Takashi, A., Ryoko, M., Yuzo, S. (1984). Journal of Bulletin of the Chemical Society of Japan, 57(6): 1582-1585.

[9] Choi, J.G., Brenner, J.R., Thompson, L.T. (1995). Journal of Catalysis, 154: 33.

[10] Ma, C.H., Zhang, W.F., He, J.L., Zhu, H.M. (2007). Synthesis and characterization of tantalum nitride nanopowder prepared through homogeneous reaction. Journal of Transactions of Nonferrous Metals Society of China, 17: s556-s559.

[11] Palko, A.A., Ryon, A.D., Kuhn, D.W. (1958). The vapor pressures of zirconium tetrachloride and hafnium tetrachloride. Journal of Physical Chemistry, 62(3): 319-322. https://doi.org/10.1021/j150561a017

[12] Yajima, A., Segawa, Y., Matsuzaki, R., Saeki, Y. (1983). Reaction process of zirconium tetrachloride with ammonia in the vapor phase and properties of the zirconium nitride formed. Journal of Bulletin of the Chemical Society of Japan, 56(9): 2638-2642. https://doi.org/10.1246/bcsj.56.2638

[13] Larson, A.C., Von Dreele, R.B. (2000). The General Structure Analysis System. Los Alamos National Laboratories. Los Alamos. NM.

[14] Toby, B.H. (2001). EXPGUI, Graphical User Interfaces for GSAS. Journal of Applied Crystallography, 34: 210-213. https://doi.org/10.1107/S0021889801002242

[15] Tauc, J., Grigorovici, A., Vancu, A. (1966). Optical properties and electronic structure of amorphous germanium. Journal of Physica Status Solidi (b), 15(2): 627-637. https://dx.doi.org/10.1002/pssb.19660150224

[16] Davis, E.A., Mott, N.F. (1970). Conduction in non-crystalline systems V. Conductivity, optical absorption and photoconductivity in amorphous semiconductors. Journal of Philosophical Magazine, 22(179): 903-922. https://doi.org/10.1080/14786437008221061

[17] Parkin, I.P., Nartowski, A.M. (1998). Solid state metathesis routes to group IIIa nitrides: Comparison of Li3N and Mg3N2 as nitriding agents. Journal of Polyhedron, 17(16): 2617-2622. https://doi.org/10.1016/S0277-5387(97)00454-3

[18] Sun, Y., Li, Y., Sun, J., Li, Y., Pei, A., Cui, Y. (2017). Stabilized Li3N for efficient battery cathode prelithiation. Journal of Energy Storage Materials, 6: 119-124. https://dx.doi.org/10.1016/j.ensm.2016.10.004

[19] Zhao, Y., Li, C., Liu, X., Gu, F., Jiang, H., Shao, W., Zhang, Li., He, Y. (2007). Synthesis and optical properties of TiO2 nanoparticles. Journal of Materials Letters, 61(1): 79-83. https://doi.org/10.1016/j.matlet.2006.04.010

[20] Kubelka, P., Munk, F. (1931). Ein Beitrag Zur Optik Der Farbanstriche. Journal of Zeitschrift für Technische Physik, 12: 593-601. 

[21] Kubelka, P. (1948). New contributions to the optics of intensely light-scattering materials. Part I. Journal of the Optical Society of America, 38(5): 448-457. https://doi.org/10.1364/JOSA.38.000448

[22] Nowak, M., Kauch, B., Szperlich, P. (2009). Determination of energy band gap of nanocrystalline SbSI using diffuse reflectance spectroscopy. Journal of Scientific Instruments, 80(4): 046107. https://doi.org/10.1063/1.3103603