1 Introduction

Aluminium nitride (AlN) along with indium nitride and gallium nitride is a well-known wide energy gap semiconductor. However, due to the high energy of the surface optical phonons, AlN has become a prominent substrate for atomic layered composites [1]. The material is known to reduce remote interfacial phonon scattering, thereby improving the devices performance [1] to provide high charge mobility and dielectric constant [2]. The high thermal conductivity of AlN enables effective heat dissipation, which limit self-heating [3,4,5]. Its films can be reliably integrated [6] in large scale, for example 6″ and 8″ Si wafers [7, 8], making it both commercially and technologically significant for layered material production. In addition, de Almeida et al. [9] investigated the likelihood of apprehending two-dimensional (2D) hexagonal AlN by first-principles density functional theory (DFT) calculations [9]. Furthermore, de Almeida et al.’s [9] first-principles calculations revealed stability of hexagonal-shaped AlN sheet. The simulations model system consisted of 96 atoms per supercell of a hexagonal AlN. De Almeida et al. predicted that AlN has an indirect gap of 2.81 eV and a cohesive energy, which is 6% lower than the bulk AlN.

Nickel diamminebis[m-(cyano-C:N)]bis(cyano-C)di (C4H6N6Ni2) clathrate is as a depiction of phonon glass–electron crystal materials; the clathrate structures have been studied for possible applications in thermoelectricity generation [10,11,12,13,14,15]. Type-I clathrate crystal structure is covalently bonded with the lattice composed of elements primarily from groups 13 and 14 of the periodic table. Its structure has cages, where metal atoms can fill in [13, 16]. Clathrate compounds have been found with high figure of merit (ZT) values of 1.35 at 900 K but contain expensive elements such as Ge and Ga in Ba8Ga16Ge30 [17] crystal. About 46-mm-long crystal of the Ba8Ga16Ge30 clathrate has been grown using Czochralski method [17, 18]. Saramat et al. [17] revealed electrical, thermal conductivities, as well as Seebeck coefficient transiting from extrinsic to intrinsic behaviour in the 600–900 K temperature range with ZT equal to 1.63 at 1100 K. Thus, Ba8Ga16Ge30 clathrate is a strong candidate for high temperatures, as well as medium thermoelectric application. Efforts have also been made to include cheap elements such as Si and Al for thermoelectric application [19]. Bobnar et al. [19] developed and structurally studied Ba8AlxSi46−x clathrate-I phase, which revealed aluminium and silicon atoms: 5.7 aluminium atoms and 0.3 silicon atoms occupying the crystallographic site 6c, while 1.2 aluminium atoms and 22.8 Si atoms occupy site 24k. Further, atomic distribution was established using wavelength-dispersive X-ray spectroscopy [19], X-ray single-crystal diffraction [20], as well as NMR [19, 21] experiments. Condron et al. [22] generated Ba8Al14Si31 crystal using an aluminium flux method. Electron microprobe, 27Al MAS NMR and single-crystal neutron diffraction were used to determine the aluminium site occupancies. However, Si-based clathrates show low ZT values. A cross-substitution improves the properties by introducing more scattering centres and changing the electronic band structure, thus bringing value to the basic investigative process of chemical composition, such as structural variations and stability [23,24,25,26,27,28].

Defects play a significant role in influencing the physical properties, as well as the applicability of materials [29,30,31]. Clathrate type-I structure has been observed for compounds containing defects, such as Rb8Sn442 and Ba8Ge433 [32, 33]. In these compounds, the defects can order in the crystal structure and give a superstructure belonging to Ia\(\bar{3}\)d space group. At elevated temperatures, the defects randomise and a structural phase transition into the Pm\(\bar{3}\)n structure is observed. From the experimental optical spectra study, schematic energy band for SrTiO3 preformed by Mishra et al. [34] revealed that the detection of oxygen defect is important from a technological point of view, as these vacancies are known to improve the photo-catalytic activities and magneto-optical properties. The idea of disorder and tunnelling states in the Sr and Eu systems but not in the Ba systems, and the corroborative nature of the difference Fourier nuclear density and lattice thermal conductivity has been established in the literature [35,36,37]. Chakoumakos et al. [38] revealed that Sr (2) in the large cage should have displacement parameters of one order of magnitude larger than that of the framework atoms and Sr (1) atom in the small cage. This addressed the guest atom disorder and dynamics in multi-temperature single-crystal and powder neutron diffraction study of Sr8Ga16Ge30. Difference Fourier maps with the Sr (2) atom removed from the reference model showed non-ellipsoidal features in the nuclear density with low density at the centre of the cage. These features in the nuclear density suggest a four-site disorder model. Refinements of single-crystal and the neutron powder diffraction data resulted in similar residual factors for models with four-site disorder on 24k or 24j, and this suggests that the disorder could be more complex than just four sites [39]. The metal atoms in cages, however, introduce low lattice thermal conductivity in clathrates [40,41,42]; the doped material results in high thermoelectric performance [43,44,45] as determined by \({\text{ZT}} = S^{2} \sigma T/(k_{e} + k_{ph} )\). Where S, T and σ represent thermal power or Seebeck coefficient, the absolute temperature and electrical conductivity. Nevertheless, because of the inversion relationships between these parameters [46], it is challenging to improve thermoelectric performance by decreasing k or increasing S and σ. In this study, we investigate the traditional Clathrate C4H6N6Ni2 thermoelectric material with 2D AlN to form AlN/C4H6N6Ni2 heterostructures. Using ab initio methods, we study the geometry, electronic properties and optical properties to demonstrate that heterostructures can result in lowering lattice thermal conductivity, and can have improved electronic properties and reduced energy gap of composites, thus resulting in electrical conductivity increase. The study reveals the importance of heterostructures in electronic analysis, and demonstrates that improved electronic properties would lead to improving the charge generation and result in a better thermoelectric material.

2 Computational details

The ab initio calculations were made using the Cambridge Sequential Total Energy Package [47] program and employing Ultrasoft pseudopotentials as implemented in Biovia Materials Studio 2019 [48]. In this study, geometry optimisation, energy, electronic structure optical properties and charge generation and separation are computed. Additional functionals have been used to correct the underestimation of density functional theory. These include the generalised gradient approximation functional [49] with Perdue–Burke–Ernzerhof [50] to account for electron correlation effects. Geometry optimisation was done on ultra-fine quality with energy convergence tolerance of 5.0 × 10−6 eV/atom. The electronic cut-off energy was set at 400 eV with medium SCF tolerance. Discrete Fourier-transform algorithm [51] grid density was set at 42 × 48 × 160. The convergence window was set mixing at three electronic densities with charge mixing amplitude of 0.5 having a charge mixing cut-off. The Monkhorst–Pack k-points to define the accuracy of the Brillouin zone sampling were set at 8 × 8 × 1 for energy, and Monkhorst–park grid 16 × 16 × 1 k-point defined density of state (DOS) giving the actual spacing of (0.00704 × 0.000704 × 0.05604) 1/Å. Spin mixing amplitude was set at 2.0 having a cut-off of 1.51/Å. To prevent the heterostructure layers from collapsing on each other, a vacuum slab was set at 20 Å and this allowed geometric relaxation.

3 Results and discussion

3.1 Computed models geometry

Many physical properties of crystals are associated with their point groups. Some of these properties are piezoelectricity, pyroelectricity, external morphology, ferroelectricity, optical activity or shape of the crystal, indices of refraction, as well as etch figures. A point group is a group of symmetry operations that leave at least one point of an object or pattern unmoved. In 2D, the elements of the point group include the identity and may include one or more mirrors and/or one or more n-fold rotations. The symmetry operations are represented by matrices. C4H6N6Ni2 is a point group 4 with an order of four and four symmetry operations (1, 2, 4+ and 4−). Figure 1a shows an example of an abstract motif of n(C4H6N6Ni2) with 3D tetragonal lattice length parameters of 7.235 (Å), 8.310 (Å) along (a = b, c), with a 83 P4/M spacegroup. The primitive cells angle symmetry constraints of α = β = γ = 90°, respectively, with the symmetry point group 4 and a projection of clathrate. (C6D6)2n, catena-(bis(hexadeuterobenzene) tetrakis(m2-cyano)-tetrakis(ammonio)-di-nickel clathrate) are also members having the same symmetry. In this study, we generated models of AlN with lattice parameters of a = 0.311, b = 0.9385 and c = 0.498 nm, with a Fm3m space group. The primitive cells atomistic position was set at 90°, 90°, 120° for the α, β and γ, respectively [37], and built unit cells, which were magnified into supercells (Fig. 1b). Surfaces for reaction were cut along (101) (Fig. 1c) for C4H6N6Ni2 and layered heterostructures (Fig. 1d–f).

Fig. 1
figure 1

Model structures of C4H6N6Ni2 (a), AlN (b), C4H6N6Ni2 surface (101) (c), C4H6N6Ni2–AlN layer (d), C4H6N6Ni2–AlN–AlN layer (e), C4H6N6Ni2–AlN–AlN–AlN layer (f)

Full size image

3.2 Electronic properties

3.2.1 Band structure

The electronic band energy structures were generated in order to find out the influence 2D AlN had on the formation of heterostructures. Since lowering heat conductivity requires scattering of acoustic phonons, this involves anisotropic thermal vibrations, inclusion of heavy atoms with complex structure, filling lattice cages with atoms for disorder, superlattice structures (heterostructures), as well as defects/grain boundary scattering. C4H6N6Ni2 had a zero band gap, and this can be ascribed to the high aromaticity of C4H6N6Ni2 structure, while AlN model had a direct band gap energy of 6.013 eV near the G point. This was much higher than the theoretically determined value by de Almeida et al. [9] (2.81 eV) and an experimental energy gap of 5.0 eV by Mattila et al. [52]. However, Li et al. [53] revealed a direct band gap of 6.12 eV near the G point of AlN from ab initio calculations. The experimental optical reflectance, as well as absorption/transmission cathode luminescence bandgap measurements, reported a 6.03 eV by Silveira et al. [54]. Furthermore, their study ascribes exciton at 6.03 eV and a shoulder intense line at 6.01 eV ascribed to the neutral donor-bound-exciton recombination. It was observed from our plot (Fig. 2) that the number of energy bands increased and induction of energy gap was observed in our heterostructures, i.e. C4H6N6Ni2 surface (2c), C4H6N6Ni2–AlN layer (0.015 eV) (2d), C4H6N6Ni2–AlN–AlN layer (0.060) (2e) and C4H6N6Ni2–AlN–AlN–AlN layer (0.022 eV) (2f). An energy gap is important in that the thermoelectric materials should not be a perfect conductor of heat. Temperature is dependent on band gap, and the optical band gap of semiconductors tends to decrease with the increase in temperature as found out by Mishra et al. [55]. However, the material requires being a good electrical conductor and the small energy gap created is enough to allow the electrons to move from valence band (VB) to conduction band (CB) with minimum energy.

Fig. 2
figure 2

Model structures of AlN (6.013 eV) (a), C4H6N6Ni2 (b), C4H6N6Ni2 surface (c), C4H6N6Ni2–AlN layer (0.015 eV) (d), C4H6N6Ni2–AlN–AlN layer (0.060) (e), C4H6N6Ni2–AlN–AlN–AlN layer (0.022 eV) (f) represents the band structure, respectively

Full size image

3.2.2 Density of state (DOS)

Increasing Seebeck (α) can be archived by bandgap engineering, large DOS at E = 0 and large effective mass. To understand the materials nature at the Fermi level, the DOS was calculated (Fig. 3a–e), i.e. C4H6N6Ni2, AlN, C4H6N6Ni2–AlN layer, C4H6N6Ni2–AlN–AlN layer and C4H6N6Ni2–AlN–AlN–AlN layer. As shown in Fig. 3a, C4H6N6Ni2 material had increased degeneracy of band extrema. Furthermore, there was a clear luck of energy gap observed in C4H6N6Ni2. An intense decrease in the electron density of state at the Fermi level, i.e. zero energy AlN (Fig. 3b), was seen. The increase in extrema of heterostructures and high intensity in band structure can be ascribed to improve thermoelectric ZT as the Seebeck for the material is derived from the additions of contributions from each extremum [56, 57].

Fig. 3
figure 3

Calculated C4H6N6Ni2 (a), AlN (b), C4H6N6Ni2–AlN layer (c), C4H6N6Ni2–AlN–AlN layer (d) and C4H6N6Ni2–AlN–AlN–AlN layer (e) total densities of state (E = 0 eV is set as the Fermi level (Ef), and this is represented by the perpendicular grey-dashed line)

Full size image

3.2.3 Projected density of state (PDOS)

Individual element constituting our heterostructures was computed for their contribution to the interface projected as a plot of density of state (PDOS) (Fig. 4). Ni, C, N, H and Al projected density of state (E = 0 eV is set as the Fermi level). Ni has an atomic number of 28 with [Ar] 3d8 4s2 or [Ar] 3d9 4s1 electronic configuration; it has 2, 8, 16, 2 or 2, 8, 17, 1 electron per shell. The 3d8 or 3d9 contribution was dominant in the VB, as well as conduction band past the Fermi level (Fig. 4a). However, the s orbital showed dominance past 3 eV into the conduction band. Carbon (C) has an electron configuration of [He] 2s2 2p2 (Fig. 4b) with 2, 4 electrons per shell, and its 2p2 in p orbital occupies the VB and CB with its highest electron density in the VB. N has an atomic number of 7 with [He] 2s2 2p3 electronic configuration, and it has 2, 5 electrons per shell.

Fig. 4
figure 4

Calculated individual element Ni, C, N, H and Al projected density of state (E = 0 eV is set as the Fermi level (Ef), and this is represented by perpendicular grey-dashed line)

Full size image

The 2p3 contribution dominated in the VB, as well as in the conduction band past the Fermi level (Fig. 4c). However, it had the greatest contribution to the VB. Hydrogen has a symbol of H and atomic number 1. It has 1 electron per shell with electronic configuration of 1s1. It had an equal band contribution at the Fermi level. However, aluminium (Al) had the dominance in the conduction band with the [Ne] 3s2 3p1 electronic configuration, where the 3p1 orbital had the highest contribution at the Fermi level.

It was further observed that the p orbital contribution dominated the heterostructures generated from the individual elements C4H6N6Ni2, AlN, C4H6N6Ni2-AlN, C4H6N6Ni2–AlN–AlN and C4H6N6Ni2–AlN–AlN–AlN layer partial density of state, respectively (Fig. 5), in both the valence and conduction bands. The influence of adding the different Al layers was observed by the increase in the DOS (electrons/eV), with Fig. 6e being the densest.

Fig. 5
figure 5

Calculated C4H6N6Ni2 (a), AlN (b), C4H6N6Ni2–AlN layer (c), C4H6N6Ni2–AlN–AlN layer (d) and C4H6N6Ni2–AlN–AlN–AlN layer (e). Partial density of state, respectively (E = 0 eV is set as the Fermi level (Ef), and this is represented by perpendicular grey-dashed line)

Full size image
Fig. 6
figure 6

Calculated C4H6N6Ni2–AlN layer (a) and individual element Ni, N, C, Al, and H spin density of state, respectively (E = 0 eV is set as the Fermi level (Ef), and this is represented by perpendicular grey-dashed line)

Full size image

To explore the electron contribution and the orbitals near the Fermi level (Ef) at the interfaces of C4H6N6Ni2–AlN, C4H6N6Ni2–AlN–AlN and C4H6N6Ni2–AlN–AlN–AlN heterostructures, the PDOS of bulk C4H6N6Ni2, AlN and their heterostructures were plotted (Fig. 5). The absence of bandgap in C4H6N6Ni2 is observed (Fig. 5a) with 4s2 or 4s1 electrons dominating both the VB and CB before and past Ef mainly from Ni (Fig. 5a). At the Fermi level, AlN VB is stuck at the E = 0 eV of VB states with a clear separation (band gap) purely dominated by p orbital (Fig. 5b). N contributes 2p3 at VB, while the CB is made up of 3p1 from Al; this signifies upward band bending and p-type feature. The PDOS analysis of heterostructures from C4H6N6Ni2 and AlN shows that the VBM is largely contributed by 2p3, 4s2 or 4s1 states, respectively, while 3p1 makes the greatest contribution in CB. The electron transition from the VB to CB, i.e. C4H6N6Ni2 states to AlN states of in the composite reveal the orbitals involved and their influence in thermal electron generation or light absorption [15]. Therefore, from Fig. 5c–e, the VB of composites is dominated by p orbital states, as well as the CBM in layer. There is a distinct increase in the density of state of the layer build up, resulting in an increase with both VB and CB, respectively. Under heat (infrared) or visible light, it is possible for the electrons in the C4H6N6Ni2 states to be easily transferred into the AlN states.

3.3 Spin density of state

The C4H6N6Ni2–AlN monolayer and Ni, N, C, Al, and H spin density of state of starting materials were calculated and analysed using DFT against individual components and plotted in the state graph density. In the spin-polarised calculation, individual states contributions from alpha and beta were summarised. The spin-up contributions formed only self-states alpha, and only self-states spin-down made beta. The results and the difference in beta and alpha contributions were shown in a butterfly plot (Fig. 6). Ni, N, C, Al, and H had equal spin-down and spin-up energies. Despite the fact that all the elements had equal alpha and beta spins as observed (Fig. 6a), the composite shows slight shift at the Fermi level and this can be ascribed to the doping effect as seen in the emergence of a bandgap in the composites in the band structures.

3.4 The optical properties

New material gives a wealth of information about the low-energy excitations that govern the physics of material. Some of these excitations indicate the part of the solar spectrum, where they might be expected to appear. A number of quantities, often called optical constants, describe the response of materials to light/heat. These optical constants, such as the dielectric constant (\(\epsilon\)), the electrical conductivity (\(\sigma\)), the susceptibility (\(\chi\)) and the refractive index (n) among others, are neither constant nor independent but are functions of the pressure, external magnetic field, frequency, as well as temperature. By knowing two of these, one that describes the absorption in the solid and one that describes dispersion, all of the others may be calculated using Maxwell’s [58, 59] equations for macroscopic media as expressed using Gaussian units [60].

The value of reflectivity as a function of an incident angle is calculated using Fresnel’s formulas [61] from absorption index of sample material and refractive index. Our calculated reflectivity against wavelength plot reveals that at low wavelengths, i.e. below 200 nm (Fig. 7a), AlN is the highest. Single-layered heterostructure resulted in a drop in the reflectance, and subsequent addition of the layers show improvement in reflectivity with the C4H6N6Ni2–AlN–AlN–AlN layer having the highest reflectivity from 240 nm into the redshift.

Fig. 7
figure 7

Calculated optical properties of the C4H6N6Ni2 and AlN as well as the heterostructures

Full size image

Absorption is the process of converting radiant energy into internal energy of the medium. This occurs by two: mainly by Rayleigh scattering [62] and Compton scattering [63]. Thermoelectric effect occurs when photon/heat energy is consumed to release an electron from atom nucleus [64]. This effect arises from the fact that the potential energy barrier for electrons is finite at the surface of the material. AlN (Fig. 7b) has the least absorption, which mainly occurs in the ultraviolent region of the solar spectrum. Improvement was observed when the layered heterostructures were tested with the three layers being the highest at all wavelengths tested, i.e. 200 to 2500 nm.

The calculated optical conductivity (Fig. 7c) showed C4H6N6Ni2 exhibits a high conductivity in the high-frequency region of the solar spectrum; thus, as the wavelength increases the conductivity decreased in a uniform manner. However, a significant improvement in the conductivity with increase in layers was noted particularly at higher wavelengths with C4H6N6Ni2–AlN–AlN–AlN layer dominating. Thermoelectric materials require high electrical conductivity. Increasing electrical conductivity (σ) entails having a material with high charge carrier mobility, narrow energy gap semiconductors. To lower electron scattering, our calculated electrical conductivity (Table 1) using the Boltzmann constant relation [65] is as follows:

$$\sigma = \sigma_{0} \exp \left( { - {\raise0.7ex\hbox{${E_{\text{a}} }$} \!\mathord{\left/ {\vphantom {{E_{\text{a}} } {k_{T} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${k_{T} }$}}} \right)$$

where \(\sigma_{0}\) is a constant, \(E_{\text{a}}\) is the activation energy and k is Boltzmann constant. This show a gradual decrease in conductivity from the thermoelectric material C4H6N6Ni2 surface (101) with single layer to double layers; however, the three-layer system indicated an increase in conductivity and this can be ascribed to a reduction in electron affinity and ionisation energy as observed.

Table 1 Calculated electrical conductivity of AlN, C4H6N6Ni2 and the composites
Full size table

Our study of the permittivity of a material as function of frequency (dielectric function) (Fig. 7d) is derived from Maxwell’s equations, which state that a dielectric-metallic interface can support surface plasmon polaritons. Unique properties of interface waves arise and are the result of dielectric materials optical frequency-dependent dispersion characteristics. At higher wavelength above 1250 nm, C4H6N6Ni2–AlN–AlN–AlN layer showed a high dielectric function. However, the absence of negative permittivity means that the electric displacement vector and the electric field vector point in the same direction. The high wavelengths are appropriate for thermoelectric application indicating improvement in the generated composites having probable infrared wavelength of red heat.

3.5 Mechanism of charge transfer

To understand the charge transfer mechanism between the layers within heterostructures generated, electron density difference is necessary [66]. This will explain the observation on what happens when two surfaces interact. The interfacial interactions are expected to change the charge distribution between surfaces, which result in charge transfer across the heterostructures.

In order to explore the charge transfer process in the C4H6N6Ni2–AlN and C4H6N6Ni2–AlN–AlN heterostructures, three-dimensional charge density difference (Fig. 8) is plotted with the yellow iso-surface representing electron accumulation, while the green iso-surface represents the electron depletion with an iso-surface value of 0.003 Å−3. Figure 8 reveals that the charge densities are redistributed by forming hole- and electron-rich regions within the bulk C4H6N6Ni2, C4H6N6Ni2 surface (101) and AlN sheet interface. This is attributed to the large separation between the AlN sheet and C4H6N6Ni2 surface (101). In addition, a high charge depletion occurred on the bulk C4H6N6Ni2 accumulating on C4H6N6Ni2 surface (101), resulting in hole-rich C4H6N6Ni2 sites. Thus, holes accumulated in the C4H6N6Ni2 bulk region, while electrons accumulated in the region close to the surfaces; therefore, electrons are expected to be excited from the C4H6N6Ni2 to the surface (101) and crossover to AlN. C4H6N6Ni2 acts as electron donor and conduction medium to transport electrons to the AlN surface.

Fig. 8
figure 8

Calculated charge density difference of C4H6N6Ni2–AlN layer and C4H6N6Ni2–AlN–AlN layer with an iso-surface of 0.003 Å−3 (the yellow iso-surface represents electron accumulation, while the green iso-surface represents the electron depletion)

Full size image

The C4H6N6Ni2–AlN (Fig. 8a) and C4H6N6Ni2–AlN–AlN layer (Fig. 8b) differences in electron density show the Mulliken population analysis [67, 68], and the three-dimensional charge density difference [66] confirms the transfer of interlayer charge between the C4H6N6Ni2–AlN monolayers as previously observed in PDOS analysis. The charge density difference in the two-layer system (Fig. 8b) shows a strong charge relocation at the AlN interface region, which can lead to substantial quantities of charge generation in C4H6N6Ni2 and subsequent transfer increase in the layers, resulting in more charge accumulation in the AlN. This study, therefore, reveals that the interaction at the interface causes not only strong charge distribution fluctuation in each component, but also significant interfacial charge transfer and separation of heterostructures, which determine the mechanism of the enhanced thermos-catalytic performance of the heterostructures.

3.6 Electron affinity and ionisation energy

C4H6N6Ni2, AlN and their heterostructures were calculated for their electron-transporting properties (Fig. 9). The ionisation energy was obtained from work function + EVBM [69], where EVBM = − Work function + 0.5 energy gap [70] and Electron affinity = ionisation potential-band gap [71]. Energies of the band edges and vacuum level were used in calculations of electron affinity and ionisation energy of the individual monolayers as well as their corresponding heterostructures.

Fig. 9
figure 9

Calculated ionisation energy and electron affinity of AlN, C4H6N6Ni2, C4H6N6Ni2–AlN layer (one layer), C4H6N6Ni2–AlN–AlN layer (two layers) and C4H6N6Ni2–AlN–AlN–AlN layer (three layers) along the x-axis

Full size image

The calculated ionisation energy and electron affinity of AlN, C4H6N6Ni2, C4H6N6Ni2–AlN layer, C4H6N6Ni2–AlN–AlN layer and C4H6N6Ni2–AlN–AlN–AlN layer along the x-axis are shown in Fig. 9. AlN had the least ionisation energy of − 1.135 eV, which was followed by C4H6N6Ni2 with 2.814 eV. Generating a single layer resulted in a further increase in the ionisation energy to a maximum value of 4.102 eV in the two layers. However, after generating two layers, the three layers of AlN resulted in a reduction in the ionisation energy, where a similar trend was observed with the least being − 7.145 eV and the maximum electron affinity was 4.042 eV. The reduction in the three-layer system can be ascribed to the shielding effect and the distance from the thermoelectric material. Thus, it will be easy to move an electron from the thermoelectric material into AlN in the composite generated.

4 Conclusion

Using first-principles calculations, we demonstrate that it is possible to generate layered C4H6N6Ni2–AlN heterostructures with increased electrical conductivity and reduced thermal conductivity because of superlative structures. The new heterostructures resulted in improved properties, such as large effective mass, new energy gaps and large density of states at the Fermi level. Increasing electrical conductivity requires high-mobility charge carriers, narrow-gap semiconductors and lower electron scattering. Optical and electronic properties revealed improvement in the electron density of state, conductivity and dielectric function. The benefit arising from the improved electronic properties would lead to improving the charge generation and separation, thus resulting in a better thermoelectric material.