Growth of non-layered 2D transition metal nitrides enabled by transient chloride templates
Liqiong He, Jingwei Wang, Zhengyang Cai, Ruiting Liu, Shengnan Li, Yunhao Zhang, Zhi-Yuan Zhang, Jiarong Liu, Bilu Liu
TL;DR
Scientists developed a new method to create 2D transition metal nitrides with tunable magnetic properties using chloride templates.
Contribution
A universal synthesis method for non-layered 2D TMNs using metastable metal chlorides as transient templates.
Findings
Fifteen types of 2D TMNs and their alloys were successfully synthesized.
The TMNs exhibit tunable magnetic properties from antiferromagnet to hard magnet based on composition.
Abstract
2D transition metal nitrides (TMNs) have attracted significant attention due to their magnetic, electrical, and chemical properties at atomic thickness. However, the synthesis of 2D TMNs is still challenging, due to their strong isotropic metal-nitrogen bonding networks. Here, we report a universal synthesis of non-layered 2D TMN family by using corresponding metastable metal chlorides as transient templates. This approach takes advantage of the layered structures and low conversion energy barriers of transition metal chlorides (TMCls) to grow 2D TMNs. Fifteen types of 2D TMNs and their alloys were synthesized, demonstrating the versatility of this method. The 2D TMN family exhibits tunable magnetic characteristics ranging from antiferromagnet to hard magnet, which can be modulated by their composition. This work overcomes previous synthesis limitations, thus offering a pathway to…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
Topics2D Materials and Applications · MXene and MAX Phase Materials · Machine Learning in Materials Science
Introduction
Transition metal nitrides (TMNs) represent a class of materials with impressive properties, including ultrahigh mechanical strength, remarkable thermal stability, tunable magnetism, and superconductivity^1–4^. Recent advances in two-dimensional (2D) materials have spurred interest in reshaping TMNs into ultrathin 2D forms to unlock enhanced functionalities^5–8^. For instance, owing to unsaturated dangling bonds and low-coordination atoms, the exposed surfaces endowing 2D TMNs with excellent catalytic performance and chemisorption capabilities in energy-related applications^4,9–13^. Notably, under electron and phonon confinement at the 2D limit, these ultrathin TMNs exhibit properties distinct from their bulk counterparts, offering opportunities for fundamental studies in magnetism^5,14–16^, electronic transport^5,8^, optical properties^8^, etc. Additionally, due to their atomic-level thickness, 2D TMNs could exhibit high flexibility, making them promising building blocks for future 2D electronic, spintronics, and optoelectronic devices^17–21^. To achieve these property explorations and applications, the prerequisite is the universal and controllable synthesis of 2D TMNs.
Different from other nonlayered materials such as transition metal oxides (TMOs, e.g., Fe_2_O_3_) and transition metal chalcogenides (e.g., FeS), TMNs shows special isotropic hybrid metallic-covalent interactions between metal and nitrogen atoms in three dimensions^22–26^. This isotropy impedes the formation of thin 2D TMNs by using the established method in synthesizing other non-layered 2D materials^2,3,27–29^. Currently, approaches such as selective etching of MAX phases^17,30,31^, salt-templated growth^32,33^, and 2D template conversion^34,35^, have been developed to produce certain free-standing 2D TMN materials. For example, Urbankowski et al have fabricated 2D Ti_4_N_3_-based MXene by using the molten fluoride salts to etch Al atoms from Ti_4_AlN_3_ powder precursor^30^. Note that the produced materials contain unavoidable surface terminal groups, which influence the understanding of the intrinsic properties of pure 2D Ti_4_N_3_. In another work, Jin et al have obtained layered W_2_N_3_ nanosheets by ammonization of 2D Na_2_W_4_O_13_ salt-templates with similar lattice symmetry, while this method remains limited to a few 2D TMNs since the lattice matching requirement between salt and target materials^10^. Although the 2D template conversion approach shows promise for the universal growth of high-quality pristine 2D TMNs, selecting suitable template materials to avoid the long nitriding time caused by etching is critical^34,35^. Transition metal chlorides (TMCls) therefore is a potential choice since their low conversion energy, while their high-temperature instability makes this process practically unachievable^36,37^. How to stabilize and nitrogenize these metastable templates is critical for the successful growth of 2D TMNs.
Here, we develop a transient chloride template assisted reverse-thermal-field (RTF) strategy for the universal synthesis of a library of 2D TMNs and their alloys. We use TMCls as templates, taking advantage of their layered structure and low conversion energy barriers to corresponding nitrides. By applying a spatially RTF method during nitridation in a short time, we directly transform metastable TMCls templates into ultrathin TMNs with preserved integrity and single-crystallinity. Through changing the type of precursor and growth temperature, we have grown 7 types of 2D TMNs and 8 types of 2D TMN alloys. The high-resolution transmission electron microscopy (HRTEM) images confirm the structure and composition of materials. Magnetic measurements show that the 2D TMNs exhibit magnetic behaviors different from their bulk counterparts, and alloying is proven to be an effective way to modulate their magnetism.
Results
Growth of 2D TMNs by the reverse-thermal-field CVD method
As shown in Fig. 1a, compared to common transition metal compounds like oxides and sulfides, TMCls possess two key advantages which make them promising templates to grow 2D TMN. First, TMCls show the lowest solid-to-solid conversion energy to TMNs at high temperatures, which reduces the conversion time and thus inhibits crystal etching caused by NH_3_ and H_2_ during the conversion process. Second, unlike the transition metal oxides and sulfides which mainly show nonlayered structures, most of the TMCls exhibit a layered vdW structure (Supplementary Fig. 1). However, due to their high volatility and hygroscopicity, the 2D TMCl templates cannot exist stably at high temperatures, making the direct use of their chloride template unapplicable. Indeed, Supplementary Fig. 2 shows the optical microscopy (OM) images of a 2D TMCl flake exposed in air for 10 s, 20 s, 30 s, 40 s, and 50 s, and Supplementary Fig. 3 shows the OM images of deposited TMCl heated at 650 °C in an Ar atmosphere. The quick fading of the flake demonstrates its poor stability. In addition, as shown in the thermal gravimetric analysis (TGA) curves (Supplementary Fig. 4a), all the sublimation temperatures of TMCls (i.e., VCl_3_, CrCl_3_, MnCl_2_, FeCl_3_, CoCl_2_, and NiCl_2_) are below 200 °C, revealing that the 2D TMCl templates cannot maintain as solid crystal at high temperature. However, the thermodynamically supported conversion temperature from chlorides to nitrides is above their sublimation temperature, which indicates that the solid-to-solid conversion from 2D TMCls to 2D TMNs is difficult (Supplementary Fig. 4b), making the traditional CVD method unapplicable (Fig. 1b, c). Interestingly, we found that the sublimation rate of TMCl in NH_3_ was greater than its growth rate to nitride when the growth temperature was <700 °C, while it was smaller when the temperature was >700 °C (Supplementary Fig. 5). Therefore, the key is how to heat the metastable TMCl templates to the required high temperature in a short time.Fig. 1. Design and growth process of the reverse-thermal-field (RTF) chemical vapor deposition (CVD) method.a Calculations of the Gibbs energy barriers of the solid-to-solid conversion from transition metal chlorides, oxides, and sulfides to the transition metal nitrides (TMNs) counterparts at 1000 K. The positive energy means the conversion is theoretically unfavorable, while the negative energy means it is theoretically favorable. b Schematic of the metastable 2D transition metal chloride (TMCl) template. c Scheme to illustrate the temperature program and growth result of the traditional CVD method. The nitridation process starts when the temperature reaches the thermodynamically supported conversion temperature. And the TMCl is sublimated as MCl_x_ molecules before the nitridation process. d Scheme to illustrate the temperature program and growth result of the RTF CVD method. The dashed and solid lines represent the temperature of the precursor and substrate, respectively. The nitridation process starts when the thermal field is reversed. The detailed process is shown in the Methods and Supplementary Information.
Based on this, we develop an RTF method in which the fast switching of the thermal field makes it possible to satisfy the above requirements. Fig. 1d and Supplementary Figs. 6–9 illustrate the setup of the furnace and the growth process of RTF method and traditional CVD method (see details in Methods and Supplementary Information). The step 1 stands for the growth of 2D TMCl templates, and the step 2 represents their conversion to the corresponding 2D TMNs. In the step 1, the TMCl precursors were placed at the upstream of the growth tube (Field Ⅰ), where the growth temperature was 600-800 °C. At the same time, the mica substrates were placed at the downstream of the tube (Field Ⅱ), where the temperature was about 100-200 °C to grow a 2D TMCl template under Ar. Subsequently, in the step 2, the temperature of the Field Ⅰ and Field Ⅱ was reversed in a short time ( < 2 s) by moving the furnace while keeping the position of the quartz tube fixed, together with the introduction of NH_3_. In this step, the 2D TMCl templates were in situ converted to corresponding 2D TMNs when the temperature of Field Ⅱ quickly reached up to 600-800 °C. At the same time, the temperature of Field Ⅰ was quickly cooled down to 100-200 °C, inhibiting the extra evaporation of precursors induced contamination. After growth, the temperature of the furnace was cooled down to room temperature (R.T.). Notably, since the TMCls show similar physical and chemical properties, various 2D TMNs were fabricated by changing the type of precursor and tuning the growth temperature. Moreover, through simply mixing the precursors, a variety of 2D TMN alloys have been fabricated, further confirming the universality of this method.
Characterizations of 2D TMNs and their alloys
By using the above RTF method, 7 types of mono-metal non-layered 2D TMNs (i.e., CrN, MnN, VN, FeN, CoN, h-NiN, r-NiN, WN, and ZrN) as well as 8 types of 2D TMN alloys (i.e., r-Co_x_Ni_y_N, h-Co_0.6_Ni_0.4_N, Co_0.3_Fe_0.7_N, Cr_x_Fe_y_N, Co_0.2_Ni_0.1_Fe_0.7_N, Cr_x_Fe_y_Co_z_N, Cr_x_Fe_y_Co_z_Mn_w_N, Cr_x_Fe_y_Co_z_Ni_w_N) are grown (Fig. 2a, b). Taking CoN as an example, the lateral size and thickness of the 2D flakes are determined by the growth temperature and duration of step 1 (Supplementary Fig. 10 and Supplementary Fig 11). The lateral size exhibits a volcano-like relationship with both increasing temperature and time, whereas the thickness generally increases. The largest CoN flake obtained thus far has a lateral size of 51 μm (Supplementary Fig. 12a), achieved at a growth temperature of 750 °C with a 5 min growth time, while the thinnest flakes ( ~ 1.03 nm, Supplementary Fig. 12b) are obtained at 600 °C with a 2 min growth time. We performed systematic studies and found that the morphology of the samples is decided by both their crystal symmetry and growth temperature. Most 2D TMNs and alloys show hexagonal morphology, which is similar to their corresponding TMCls with six- or three-fold symmetry. Note that MnN, r-NiN, and r-Co_x_Ni_y_N show rectangular morphology, which is decided by the morphology of the template. For instance, as shown in Supplementary Fig. 13, 2D NiN exhibits a hexagonal morphology when synthesized at 650 °C, which corresponds to the hexagonal shape of the 2D template formed at that temperature. Conversely, when the step 1 temperature is increased to 800 °C, a rectangular shape 2D template forms, resulting in rectangular 2D NiN flakes. And this result may be attributed to their different composition of nickel chloride grown under different temperature. Taken together, we have grown 15 types of non-layered 2D TMNs and their alloys by using the RTF method.Fig. 2. Universal growth of 2D TMNs and their alloys by the RTF method.a The periodic table of elements shows the transition metal elements involved in this work. b Optical microscopy images of the 15 types of 2D TMNs and alloys. All the scale bars are 5 μm. The materials based on one, two, three, and four types of metal elements are highlighted in yellow, orange, green, and blue, respectively.
We then investigated the crystal structure and composition of the 2D TMNs. Fig. 3a shows the crystal structure of Fe_2_N (PDF#97-002-0390). The HRTEM image and selected area electron diffraction (SAED) pattern (Fig. 3b and its inset) reveals that the 2D Fe_2_N has a lattice spacing of 0.138 nm, which is assigned to the (110) plane. Low magnification image and the corresponding energy-dispersive spectroscopy (EDS) mapping indicate that Fe and N atoms are uniformly distributed (Fig. 3c). The corresponding EDS result shows the characteristic peaks of Fe and N atoms (Fig. 3d) with an atomic ratio of ~ 2:1, suggesting the composition of Fe_2_N. Similarly, we obtained HRTEM images of 2D VN and 2D MnN and their corresponding SAED patterns (Fig. 3f, j). The 2D VN shows a lattice spacing of 0.146 nm, and the 2D MnN shows a lattice spacing of 0.247 nm, which are consistent with their atomic structures (VN: PDF#97-016-9819; MnN: PDF#97-023-6828). The corresponding EDS mapping (Fig. 3g,k) and EDS results (Fig. 3h, l) show that all the elements are uniformly distributed and both the V: N and Mn: N atomic ratios are 1:1, which are in consistent with their corresponding crystal structures (Fig. 3e, i). The detailed HRTEM images and SAED patterns of the other 2D TMNs are shown in Supplementary Figs. 14–17, and the h-NiN and r-NiN belong to the same atomic structure (PDF#97-016-1755). These results suggest the high quality of the obtained 2D TMNs.Fig. 3. Structural and composition characterization of 2D TMNs.a Top-view atomic structure of Fe_2_N. b High-resolution transmission electron microscopy (HRTEM) image and selected area electron diffraction (SAED) pattern (inset) of the synthesized 2D Fe_2_N flake. c Low magnification TEM image, energy-dispersive spectroscopy (EDS) element mapping of Fe and N atoms, and (d) EDS of the 2D Fe_2_N flake. e Atomic structure of VN from the top view. f HRTEM image of the synthesized 2D VN flake. The inset is the corresponding SAED pattern. g Low magnification TEM image, elemental EDS mapping of V and N atoms, and h EDS of the 2D VN flake. i Atomic structure of MnN from the top view. j HRTEM image of the synthesized 2D MnN flake. The inset is the corresponding SAED pattern. k Low magnification TEM image, elemental EDS mapping of Mn and N atoms, and (l) EDS of the 2D MnN flake.
We also characterized the structure and composition of the 2D TMN alloys. HRTEM image and SAED pattern of the 2D Co_0.6_Ni_0.4_N alloy are shown in Fig. 4a, revealing a lattice spacing of 0.151 nm, which is similar to the CoN (d_220_ = 0.151 nm) and NiN (d_220_ = 0.152 nm). Highly uniform elemental distribution is also confirmed by EDS mapping in Fig. 4b, and the corresponding EDS result (Fig. 4c) shows the characteristic peaks of Co, Ni and N atoms, with the atomic ratio of Co: Ni: N is about 0.6: 0.4: 1. For Co_0.3_Fe_0.7_N, the HRTEM and SAED results in Fig. 4d reveal a lattice spacing of 0.142 nm, which is similar to the CoN (d_220_ = 0.151 nm) and FeN (d_220_ = 0.141 nm, PDF#97-023-6810). Interestingly, the FeN here belongs to the cubic structure rather than the previously fabricated 2D Fe_2_N. The EDS mapping (Fig. 4e) exhibits the highly uniform elemental distribution of Co_0.3_Fe_0.7_N alloy. The corresponding EDS result in Fig. 4f shows the characteristic peaks of Fe, Ni, and N atoms, with the atomic ratio of Fe: Co: N is about 0.7: 0.3: 1. Moreover, the HRTEM image of 2D Co_0.2_Ni_0.1_Fe_0.7_N and its corresponding SAED pattern (Fig. 4g) show a lattice spacing of 0.072 nm, which is almost half of NiN (d_220_ = 0.152 nm), CoN (d_220_ = 0.151 nm), and FeN (d_220_ = 0.141 nm, PDF#97-023-6810). This should be attributed to the lattice distortion caused by the different lattice constants of the three materials. Additionally, the corresponding EDS mapping (Fig. 4h) and EDS result (Fig. 4i) demonstrate the uniformly distribution of Ni, Co and Fe in the flakes with an approximate atomic ratio of Co: Ni: Fe: N = 0.2: 0.1: 0.7: 1. The detailed HRTEM image and EDS mapping of 2D r-Co_x_Ni_y_N are shown in Supplementary Fig. 18. The above results suggest that the RTF synthesized 2D TMN alloys exhibit high crystallinity and uniformity.Fig. 4. Structural and composition characterizations of 2D TMN alloys.a HRTEM image and SAED pattern of the synthesized 2D Co_0.6_Ni_0.4_N flake. b Low magnification TEM image, elemental EDS mapping of Co, Ni, and N atoms, and (c) EDS of the 2D Co_0.6_Ni_0.4_N flake. d HRTEM image and SAED pattern of the synthesized 2D Co_0.3_Fe_0.7_N flake. e Low magnification TEM image, elemental EDS mapping of Co, Fe and N atoms, and (f) EDS of the HRTEM image and SAED pattern of the synthesized 2D Co_0.3_Fe_0.7_N flake. g HRTEM image and SAED pattern of the 2D Co_0.2_ Ni_0.1_Fe_0.7_N flake. h Elemental EDS mapping of Co, Ni, Fe and N atoms and (i) EDS of the 2D Co_0.2_ Ni_0.1_Fe_0.7_N flake.
Magnetic properties of 2D TMNs and their alloys
Finally, we investigated the magnetic properties of 2D TMNs and their alloys by magnetic force microscopy (MFM) and superconducting quantum interference device (SQUID). Fig. 5a-f show the MFM phase contrasts of the 2D TMNs (i.e., VN, CrN, MnN, Fe_2_N, CoN, and NiN) at R.T. The results show that CoN has the largest saturation magnetization among these materials, which is consistent with their M-H loops measured by SQUID (Fig. 5g). The M-H loops of 2D TMN alloys were also characterized and summarized in Fig. 5h. In detail, the 2D Co_x_Fe_y_N alloy shows an antiferromagnetic behavior, with lower saturation magnetization (M_s_) and coercive force (H_c_) than both 2D CoN and 2D Fe_2_N. While 2D Cr_x_Fe_y_N alloy exhibits a ferromagnetic behavior (2D Fe_2_N: antiferromagnetic, 2D CrN: ferromagnetic) and moderate magnetism (M_s = 0.208 emu, H_c = 250 Oe) compared with 2D CrN (M_s = 0.068 emu, H_c = 188 Oe) and 2D Fe_2_N (M_s_= 0.377 emu, H_c_ = 252 Oe). The magnetic properties of all the 2D TMNs and their alloys are summarized in Fig. 5i and Supplementary Table 1. Taken together, we can either enhance or weaken the magnetic response of 2D TMNs by selecting appropriate metals or alloying, thereby expanding the performance range of 2D magnets to meet diverse application requirements.Fig. 5. Magnetism of 2D TMNs and their alloys.Atomic force microscopy (AFM) and magnetic force microscopy (MFM) images of individual (a)2D VN, (b) 2D CrN, (c) 2D MnN, (d) 2D Fe_2_N, (e) 2D CoN, and (f) 2D NiN flakes measured at R.T. The scale bar is 2 μm. g Magnetization versus magnetic field (M-H) loops of the 2D TMNs and (h) their alloys with the magnetic field applied out of plane. The results are measured by superconducting quantum interference device (SQUID) at 10 K. The effect of mica substrate has been eliminated (Supplementary Fig. 19). i Statistics of coercive force (H_c_) and saturation magnetization (M_s_) of the 2D TMNs and their alloys in (g) and (h). The materials with one, two, three and four types of metal elements are highlighted in yellow, pink, green and blue, respectively. And increasing H_c_ drives the transition from soft to hard magnetic materials.
Discussion
In this work, we developed a RTF CVD method to convert 2D TMCls and obtained fifteen types of non-layered 2D TMNs and their alloys. The key of the method is using metastable TMCls as transient templates, with features of their 2D layered structure and low conversion energy barrier to TMNs. Combined HRTEM images, SAED patterns and EDS spectra results reveal the single-crystalline structure and composition uniformity of these 2D materials. Magnetism characterizations show rich and tunable magnetic behaviors of 2D TMN by alloying. The general growth method for non-layered materials could enrich the 2D material family with exotic properties and applications.
Methods
Synthesis of 2D TMNs and their alloys
The 2D TMNs and their alloys were fabricated using the RTF method. The typical fabrication process was carried out in a movable furnace, as shown in Supplementary Fig. 6. The precursors were placed in a quartz boat located in the upstream region of the quartz tube (Field I), and the mica substrates were positioned in the downstream region (Field II). Argon (100 sccm) was used as the carrier gas throughout the process. The growth involved two distinct steps: Step 1 involved the deposition of the 2D TMCl templates. In this step, Field I (precursor zone) was moved to the heating center and heated to the target temperature (600-800 °C) from room temperature in 30 min and maintained for 2 min. During this step, Field II (substrate zone), situated approximately 10 cm away from the heating center, remained at a lower temperature (100-200 °C). At the end of Step 1, the furnace was swiftly translated along the rail to position Field II (substrate zone) at the heating center, initiating Step 2 (the conversion step). Simultaneously with this movement, a mixture of NH_3_ and H_2_ was introduced and maintained for 10 min. Consequently, the temperature of Field II was rapidly increased to 600-800 °C and held for 5 min, while Field I was cooled down to approximately 100-200 °C. Finally, the furnace was cooled to room temperature under an Ar atmosphere.
Traditional CVD growth of 2D TMN: the precursor was placed in the upstream section of a one-zone furnace (Field I), while the substrate was positioned at the furnace center (Field II, the heating center). In step 1, the precursor was heated to 300-400 °C (above its sublimation temperature) and carried by 100 sccm Ar to deposit onto the substrate. In step 2, the substrate was heated to 600-800 °C under a flow of H_2_ and NH_3_ to facilitate the nitridation of the deposited TMCl. The corresponding temperature profile is illustrated in Supplementary Fig. 9.
Transfer of 2D TMNs and their alloys: The 2D flakes grown on the mica substrates were transferred onto the TEM grid by a common method using Polystyrene (PS) as a sacrificial transfer film. PS was spin-coated on the sample at 2000 rpm for 60 s. The sample was then heated on a hot plate at 90 °C for 5 min. After baking, the sample was loaded in deionized water and then the peeled-off samples were transferred onto TEM grids. Finally, the PS was removed with methylbenzene.
Material characterizations and magnetic measurements: The morphology of the samples was observed by an optical microscope (Carl Zeiss Microscopy, Germany). The microstructure of the samples was investigated by high-resolution TEM at an acceleration voltage of 300 kV (FEI Tecnai F30, USA). SQUID (MPMS3, Quantum Design, USA) was used to measure the M-H hysteresis loops and M-T curves of the samples. AFM (Cypher ES, Asylum Research, USA) was used to measure the thickness of the samples. MFM (Cypher ES, Asylum Research, USA) was used to measure the magnetic properties of the nanoflakes.
Calculations of temperature-dependent Gibbs formation energy of compounds: Enthalpy data of compounds in this work were extracted from Lange’s handbook of chemistry^38^ (CrN, CrCl_3_, Cr_2_O_3_, FeCl_3_, Fe_2_O_3_, FeS, MnCl_2_, MnO_2_, NiCl_2_, NiO, NiS, CoO), FactSage^39^ (CrS, CoCl_2_, CoS, Fe_2_N, MnS, VCl_3_, V_2_O_5_, VS_2_), and DFT calculation (CoN, NiN, MnN, VN). All the compound data were extracted only at 298 K. The chemical potentials of the elements were obtained with the minimum Gibbs energy at a given temperature.
Supplementary information
Supplementary Information Transparent Peer Review file
Source data
Source Data
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Rumble, J. CRC Handbook of Chemistry and Physics. 98th edn, (CRC Press, 2017).
- 2Dean, J. A. Lange’s Handbook of Chemistry. (1999).