High yield and wide lateral size growth of α-Mo2C: exploring the boundaries of CVD growth of bare MXene analogues

Synthesis of Mo2C bare MXenes, without surface terminations groups, via chemical vapor deposition (CVD) on metal foils is scientifically a very intriguing crystal growth process, and there are still challenges and limited fundamental understanding to overcome to obtain high yield and wide crystal size lateral growth. Achieving large area coverage via direct growth is scientifically vital to utilize the full potential of their unique properties in different applications. In this study, we sought to expand the boundaries of the current CVD growth approach for Mo2C MXenes and gain insights into the possibilities and limitations of large area growth, with a particular focus on controlling Mo concentration. We report a facile modification of their typical CVD growth protocol and show its influence on the Mo2C synthesis, with growth times spanning up to 3 h. Specifically, prior to initiating the CVD growth process, we introduced a holding step in temperature at 1095 °C. This proved to be beneficial in increasing the Mo concentration on the liquid Cu growth surface. We achieved an average Mo2C crystals coverage of approximately 50% of the growth substrate area, increased tendency of coalescence and merging of individual flakes, and lateral flake sizes up to 170 μm wide. To gain deeper understanding into their CVD growth behavior, we conducted a systematic investigation of the effect of several factors, including (i) a holding step time on Mo diffusion rate through molten Cu, (ii) the Cu foil thickness over the Mo foil, and (iii) the CVD growth time. Phase, chemical and microstructural characterization by x-ray diffraction, x-ray photon spectroscopy, SEM and scanning/transmission electron microscopy revealed that the grown crystals are single phase α-Mo2C. Furthermore, insights gained from this study sheds light on crucial factors and inherent limitations that are essential to consider and may help guide future research progress in CVD growth of bare MXenes.

In the field of MXenes research, typical synthesis relies commonly on top-down wet chemical etching routes.This involves selectively etching A-group elements in a 3D bulk MAX phase to derive the corresponding 2D MXene material [5,[21][22][23][24][25][26][27][28].As a result, the surfaces of the synthesized MXenes are inevitably terminated with functional groups such as O, OH, and F [10,[29][30][31], which affect their surface chemistry and deteriorate their properties.Theoretical predictions have shown that bottom-up synthesized MXenes may exhibit superior electrochemical performance compared to their wet-chemistry route derived counterparts [32][33][34][35][36].To overcome this limitation, researchers are exploring bottom-up synthesis approaches, particularly chemical vapor deposition (CVD) and solid-state reaction synthesis, for the direct growth of MXenes with functional groups-free surfaces [37,38].
The advancements made in CVD growth of MXenes are still in their early stages and with limited progress.The first successful CVD synthesis of molybdenum carbide (Mo 2 C), a M 2 X type MXene analogue, was reported by Xu et al in 2015 [39].Here, it is important to acknowledge that this approach do not directly result in monolayers of Mo 2 C sheets as in the case of MAX phase etching route but rather ultrathin crystals of Mo 2 C [40,41].For this reason, here we term these ultrathin crystals of Mo 2 C as bare MXene-analogues.More recently, Wang et al reported direct synthesis of layered stacks of Ti 2 CCl 2 MXenes, i.e.Ti 2 C monolayers with Cl surface termination groups on their both surfaces through solid state reaction and CVD methods [38,42].However, direct bottom-up synthesis of bare M 2 X type MXenes devoid of any surface terminations is yet to be achieved.
To this day, Mo 2 C remains the only CVD-grown M 2 X type bare MXene analogue that has been reported.However, the achievement of high yield, large crystal size, and extensive area coverage growth of Mo 2 C, similar to what has been achieved in the case of other 2D materials such as graphene, boron nitride or transition metal dichalcogenides (TMDCs) [43][44][45][46][47][48][49], is yet to be realized.
It is agreed that [50][51][52][53][54], during CVD of Mo 2 C based on metal foils stack composed of a Mo foil at the bottom and a Cu foil on top, the growth temperature typically surpasses the melting point of Cu (1086 °C).Moreover, the Mo atoms from the Mo foil either diffuse through the liquid Cu (i.e.bulk diffusion) and/or from the edges of the Cu droplets on Mo foil (i.e.surface diffusion) towards the top surface.Growth occurs on the surface of liquid Cu, where Mo atoms combine with C to form Mo 2 C. The carbon is sourced through dehydrogenation of CH 4 , a common carbon precursor gas in CVD methods, on the catalytic Cu surface.
To date, research efforts in Mo 2 C growth have mostly focused on studying CVD parameters such as gas flow rates (i.e. carrier gases H 2 and N 2 , carbon precursor gas CH 4 ) and Cu metal foil thickness in the growth substrate stack of Cu/ Mo [50][51][52][54][55][56].Young et al studied the combination of Ag and Cu foils in the growth substrate and observed that the CVD growth temperature required to synthesize Mo 2 C can be lowered to below 1000 °C due to Ag-Cu alloying [57].They also show that the diffusion of Mo through the liquid Cu layer above is a crucial Mo diffusion pathway in this CVD approach.However, apart from studies on different copper thicknesses [53,54,56], to the best of our knowledge, no reports have explored the other possibilities that may regulate the Mo concentration on the Cu growth surface.
It is commonly observed that the CVD growth of Mo 2 C occurs as discrete individual crystals, which have lateral sizes of just a few tens of micrometers, scattered over the growth surface with island-like growth behavior.Previous studies have shown that vertical stacking growth is preferrable at low CH 4 flow rates (0.3 sccm), while lateral crystal growth is more favored at high CH 4 flow rates (3 sccm) [50,52,56,58].It is important to note here that the use of CH 4 flow rates of 3 sccm or higher has been found to result in the growth of graphene on the top Cu surface in tandem with Mo 2 C formation [50,59,60].There is no clear reported evidence showing that Mo 2 C can grow laterally as a continuous film covering the entire growth substrate, and lateral merging between individual crystals to form large size crystals is not commonly observed.These observations highlight the need to explore other possible CVD parameters and gain a deeper understanding of the fundamental pathways and factors that affect the growth of Mo 2 C during CVD.Achieving continuous large area growth or at least full coverage over the entire substrate area would be highly beneficial for potential energy storage applications, such as using it as an anode electrode material in rechargeable Li-ion batteries or supercapacitors [61,62].
In this study, we propose a modified protocol to the conventional CVD growth approach for Mo 2 C on Cu/Mo metal foil stacks.We investigated the impact of this modified protocol in regulating the concentration of Mo during CVD growth.We introduced a holding step at the growth temperature (similar to an annealing step) prior to the CVD growth step.In addition, we studied the impact of growth time and Cu foil thickness on the growth, centered around this holding step parameter.Our observations reveal that the holding step significantly increases the concentration of Mo on the Cu melt surface during CVD growth.Furthermore, we show that this is also a controllable parameter for increasing bulk diffusion of Mo through the Cu and surface mobility of Mo atoms on liquid Cu.The optimum CVD growth conditions for achieving high yield uniform growth, over about 1 cm 2 area, and large lateral size Mo 2 C flakes growth were found to be as following: a holding step time of 2 h at growth temperature of 1095 °C, and CVD growth time of 3 h.The optimum growth substrate stack comprised of two Cu foils (each 25 μm thick) wrapped over a Mo foil (100 μm thick).Our results show that under these conditions, at least 50% Mo 2 C coverage area of the growth substrate is achieved.We also report clear observations of lateral merging between individual crystals, and the MXene flakes reaching a lateral size of about 150 μm wide.
The paper is structured as follows: in section 2.1, we first describe the modified CVD growth protocol explored and present the main results, including Mo 2 C growth yield (coverage area over growth surface) and microstructural characterization of as-grown Mo 2 C crystals under optimum CVD growth conditions and determined from sections 2.2 and 2.3.Section 2.2 discusses the influence of different holding step times at a constant growth time and Cu thickness.In section 2.3, we present the results of the influence of different Cu foil thicknesses and growth times on Mo 2 C growth at the optimum holding step time of 2 h.The conclusions and outlook are presented in section 3. The description of the materials and methods used is provided in section 4.

Facile modification of CVD growth protocol
The CVD growth of Mo 2 C was carried out in a horizontal quartz tube furnace, as illustrated in figure 1.The growth substrate consists of a two metal foils stack, Cu and Mo, with the Cu foil placed above the Mo foil, as shown in the schematic in figure 1(a).Figure 1(b) illustrates the setup of a typical horizontal quartz tube CVD furnace used in this work.Hydrogen gas, at a flow rate of 200 sccm, serves as a process gas to provide reducing environment in the tube, while CH 4 , at a flow rate of 3 sccm, serves as the carbon precursor gas.The CVD growth temperature used in this study was 1095 °C, which is above the melting point of copper (1086 °C).  the Cu/Mo interface and the top of the liquid Cu surface, Mo atoms undergo bulk diffusion through liquid copper towards its surface.The formation of graphene and Mo 2 C MXene crystals on the liquid copper is illustrated in stages III and IV, respectively.These are two competing growth processes known to occur [43,45].Figure 1(e) portrays the time and temperature profile of the CVD growth approach studied in this work.'H' denotes the holding step where the growth temperature has been reached but no CH 4 gas flow is present.'G' represents the CVD growth step with a CH 4 flow.The modification in the protocol consists in the introduction of a holding step at the growth temperature before introducing the methane flow, i.e. before the CVD growth process step.
Figure 2 shows overview images of Mo 2 C crystals grown on the foil at the optimal CVD conditions found in this study for high yield and large area coverage growth; namely a Cu thickness of 50 μm, a holding step time of 120 min at a growth temperature of 1095 °C, and a growth time of 180 min.Finally, small amounts of Mo +6 are detected, with peaks at 232.6 and 235.8 eV BE for the 3d 5/2 and 3d 3/2 peaks, respectively.
In the case of the C 1s spectrum presented in figure 3(b), the deconvolution indicates that carbon sp 2 bonds (graphene), at 284.3 eV BE, and Mo-C bond peak, at 283.6 eV BE, are the dominant components; minor peaks at 284.8 eV BE and 286.3 eV BE can be assigned to C-C sp 3 bonds and C-O bonds, respectively [65,66].The O 1s spectrum, shown in figure 3(c), is dominated by the signal from SiO 2 at 532.7 eV BE with an additional peak at 530.2 eV BE.This latter peak has a binding energy typical for oxygen in Mo oxides [65][66][67].Surface oxygen species on Mo 2 C have been reported in the energy region 531-533 eV BE; 66 their presence cannot be ascertained in our experiment due to the overlapping signal from SiO 2 in this energy range.
The Mo 2 C crystals were further characterized by scanning transmission electron microscopy (STEM) and associated energy dispersive x-ray spectroscopy (EDX) and electron diffraction (ED) techniques.The orthorhombic phase is the ordered counterpart of the hexagonal phase and it is energetically favorable with respect to the other molybdenum carbide structures.Figures 4(e

Influence of holding step on the nucleation density and lateral crystal growth of Mo 2 C
To investigate the effect of holding time on Mo 2 C growth, we conducted experiments with three different holding step durations: 30 min, 60 min, and 120 min, while maintaining all other CVD parameters constant.The growth time for all experiments in this series was 3 h, and copper foil thickness was 50 μm.The as grown samples were investigated by light microscopy and backscattered electron (BSE) imaging in SEM.The average coverage area and particle size distribution analysis on the recorded images were performed using the ImageJ software (see methods section).
Figures 5(a) and (c) show the light microscopy images of the 30 min and 120 min holding time cases, respectively.The corresponding particle size analysis maps are shown in figures 5(b) and (d).These images clearly illustrate that the coverage area of Mo 2 C over the growth substrate significantly improves from 21% to approximately 50% as the holding step time increases from 30 to 120 min.
However, in the case of 60 min holding time, we observed that the coverage area decreased to 11%.This will be further discussed in the following part of the section.
The Mo 2 C particle size distribution and surface area coverage of the as grown samples under three different holding times was, furthermore, characterized by BSE-SEM imaging.The atomic number sensitive image contrast mechanism in BSE-SEM imaging mode is more beneficial over light microscopy to obtain high contrast difference images, with Mo and Mo 2 C areas appearing relatively brighter and Cu and graphene areas appearing relatively darker.Representative BSE-SEM images of regions from the edge and center of the as-grown foils, in each case, are shown in figures 6(a)-(f) and the corresponding results from the particle size analysis in figures 6(g)-(h).In the case of a 30 min holding step, the nucleation density of Mo 2 C growth was observed to be homogeneous both at the edges and center regions of the foils, as seen in figures 6(a) and (b).It was observed that the lateral size of the crystals formed in the center regions are slightly larger than at the edges.The minimum, mean and maximum particle size of the Mo 2 C flakes were observed to be 5 μm, 19 μm, 90 μm, respectively, as shown in figures 6(g) and (h).The Mo 2 C flakes average coverage area, with respect to the total area of the image, was determined to be 20% (see suppl.info.figure S4).
When the holding step time was increased to 60 min, we observed that the nucleation density of the Mo 2 C decreased compared to the previous case, though homogenous both at the center and edges, as shown in figures 6(c) and (d).On the other hand, the lateral size of the Mo 2 C flakes increased relative to the 30 min case, both at the center and edge of the growth foils.The minimum, mean and maximum particle size of the Mo 2 C flakes were observed to be 8 μm, 43 μm, 127 μm, respectively, as shown in figures 6(g) and (h).However, the observed flakes coverage area, relative to 30 min case, decreased down to 11% (see suppl.info.figure S4).
With further increase of the holding step time to 120 min, we observed that the nucleation density and lateral size of the flakes increased, in comparison to the previous two holding step times.The lateral size of the flakes in the center of the growth substrate was relatively larger compared to the edge regions, as shown in in figures 6(e) and (f).The minimum, mean and maximum particle size of the Mo 2 C flakes were observed to be 8 μm, 44 μm, 174 μm, respectively, as shown in figures 6(g) and (h).The Mo 2 C flakes average coverage area significantly increased to about 50% (see suppl.info.figure S4).
In summary, based on these SEM observations and Mo 2 C particle size distribution analysis, the following trends can be delineated: (a) with increase of the holding step time, from 30 min, 60 min to 120 min, the maximum lateral size of the Mo 2 C flakes doubled i.e. an increase from 90 to 127 μm and further to 174 μm, respectively.(b) In regard to Mo 2 C coverage area on Cu surface, we observed that the coverage area decreases from 20% to 11% when the holding step time is increased from 30 to 60 min.It rises again to about 50% as the holding time is increased to 120 min.(c) In all the cases, the SEM observations show that relatively large lateral size flakes were more common at the centre of the growth foils compared to the edges.
The increase in the mean and maximum lateral size of Mo 2 C flakes with holding step time was observed to occur due to both lateral crystal growth and coalescence merging between them (see suppl.info.figure S3).This is similar to lateral size growth through coalescence between individual flakes reported by Geng D et al in their work on CVD growth of graphene flakes on liquid copper [68].The observed minimum in the coverage area of Mo 2 C when holding time was increased to 60 min was unexpected and not fully understood (see suppl.info.figure S4).However, it can be explained by considering different processes that take place, in terms of mobility of solute Mo atoms and Cu metal liquid, during the Mo 2 C growth, viz.(a) Mo atoms bulk diffusion through liquid copper layer towards the surface, (b) Mo atoms surface diffusion on the liquid copper surface and (c) local segregation of Mo and Cu atoms on the liquid copper surface [68][69][70].We hypothesize that between 30 and 60 min of holding time the concentration of Mo on the copper surface may reach an interim saturation point slowing down the kinetics of Mo bulk diffusion, while the mobility of Mo atoms which are already on the Cu surface may increase through diffusion leading to more Mo segregation clusters during this time.This can lead to fewer nucleation sites and in combination with the low-lateral crystal growth rate of Mo 2 C flakes, during the following CVD growth step, the sample cannot reach the same coverage area achieved with a 30 min holding time.In the case of 120 min holding step, the time window is longer to overcome the bulk diffusion saturation point and allowing for higher Mo concentration in Cu surface prior to the next growth step, thereby increasing the potential nucleation sites density on liquid copper surface.This coupled with coalescence merging between the crystals in the following growth step may contribute to further increase in Mo 2 C coverage area on Cu, as observed for the 120 min case.This is in line with the observations of Young et al in their study on the synthesis mechanism of Mo 2 C on Ag-Cu alloy, where they clearly observed that the liquid Cu layer acts both as a diffusion pathway and barrier for Mo [57].We further compared our CVD-grown Mo 2 C crystals with the typical Mo 2 CT x MXene obtained from MAX phase etching.For SEM investigations, the commercially procured MXene powder was subjected to mild ultrasonication in IPA before drop casting it onto blank Cu foil.Our SEM observations show that the Mo 2 C crystals derived from chemical etching typically have a lateral size on the order of 1 μm (see suppl.info.figure S6).The observation that, in all cases, larger flakes form at the center of the growth foil compared to the edges indicates that lateral crystal growth rates are relatively higher in the central regions of the foils.We suggest that this is probably the effect of de-wetting between the liquid Cu layer and Mo during the CVD growth.When the Cu is present in the liquid state on Mo it takes a positive surface curvature (convex shape).Owing to this, the Cu layer is relatively flat in the middle of the foil as compared to edges and corners, where the Cu surface curvature is most likely steeper with respect to the underlying Mo surface.The flat Cu layer regions are energetically more favorable for lateral crystal growth compared to edges with higher slope of the curvature as observed by Meng et al studying crystal growth on curved surfaces [71].We clearly observed the presence of such de-wetting morphology and fragmentation of liquid Cu into smaller islands at the edges and corners of our growth foils in the case of 120 min holding time case (see suppl.info.figure S1) .
These results point towards the conclusion that the introduction of a holding step in CVD protocol is a significant growth parameter effecting the Mo concentration on the top surface of the Cu by providing more time for the Mo bulk diffusion through liquid copper towards its surface.In other words, the holding step facilitates the increase in Mo concentration on the liquid copper surface prior to the CVD growth step.

Influence of copper thickness and growth time on Mo 2 C growth at optimum holding time
We have studied the impact of two different copper thicknesses (25 and 50 μm) and growth times (120 and 180 min) while maintaining a constant holding step time of 120 min (labeled '2H').The reason for studying these specific Cu thicknesses and growth times is that for Cu foil thickness higher than 50 μm, we observed that the growth density proportionally reduces due to increase of Mo bulk diffusion path length, i.e. through liquid Cu layer.A growth time beyond 180 min was may not significantly improve the Mo 2 C coverage as it will be limited by graphene formed on the Cu surface.Figure 7 shows the BSE-SEM characterization results, while figure 8 presents the corresponding analysis of particle size distribution and coverage area fraction plots.Figures 7(a), (b) showcases the BSE-SEM images of the samples grown with one copper foil (indicated in the SEM images label as '1 Cu') of 25 μm thickness on Mo foil, with CVD growth time of 2 h (indicated in the SEM images label as '2 G') and 3 h (indicated in the SEM images label as '3 G'), respectively.Upon 2 h growth, Mo 2 C crystals were formed with a coverage area of approximately 35%, and the predominant lateral size of the crystals was found to be in the range of 1-10 μm.
In the case of 3 h growth time, significant portions of the growth foils, particularly in the middle area, lacked Cu, exposing bare Mo.This is shown in the supplementary information figure S5, where the bright contrast region shows the bare Mo foil uncovered by Cu.This most likely occurred due to migration of liquid Cu from the Mo foil edges towards its underside, owing to poor Cu wettability on Mo and to some extent possible Cu evaporation during CVD at temperatures above its melting point [72,73].In the regions of the foils where copper was present, the coverage area was observed to be poor, as shown in figure 7  thickness may be too thin for such long exposures above its melting point i.e. a holding step time of 120 min and CVD growth time of 180 min.
Figures 7(c) and (d) show the findings for a copper thickness of 2 foils (total thickness of 50 μm), after 2 and 3 h of CVD growth, respectively.We observed that as the CVD growth time increased, both the flake size and corresponding coverage area increased.After 3 h of CVD growth, we noted significantly large lateral size flakes of up to about 170 μm and the merging of individual crystals, as depicted in figures 7(e) and (f). Figure 8(a) presents the surface coverage area fraction of Mo 2 C when the copper layer thickness was 25 μm and 50 μm, and the CVD growth time for each case was 120 min and 180 min.The observed trend indicates that the surface coverage area drops significantly for the 25 μm case from about 35%-5% when the growth time is increased from 120 to 180 min.However, for the 50 μm thick copper layer, the coverage area fraction increases with the growth time from 20% to about 50%.This study highlights that for a holding step time of 120 min, the optimum thickness of Cu is 50 μm for high-yield growth.Figures 8(b) and (c) correspond to Mo 2 C particle size distribution histograms for 25 μm copper thickness for growth time of 120 min and 180 min, respectively.Here, it is evident that the average particle size for the 120 min case is about 10 μm, and the maximum size of flakes observed is 30 μm.Whereas, in the case of 180 min growth time, the average size drops to 5 μm, and the maximum size of the crystals does not exceed 7 μm.As explained before, this may be due to de-wetting of Cu on Mo foil and/or Cu evaporation during prolonged exposure to 1095 °C temperature.Figures 8(d) and (e) show the corresponding histograms for a 50 μm thick copper layer and the CVD growth time of 120 and 180 min.Here, we clearly observed an increasing trend in the average particle size from 10 to 50 μm, and the maximum flake size increasing from about 50-170 μm with the increase in growth time.
The results of this study clearly point that a minimum Cu foil thickness of 50 μm is optimal compared to 25 μm thick Cu foils for our CVD protocol involving a holding step time of 120 min followed by growth time of 180 min.

Conclusions and outlook
In conclusion, this work presents a facile modification to the CVD approach for synthesizing α-Mo 2 C, a bare MXene analogue, which can result in the growth of high-quality single phase Mo 2 C crystals with large lateral size width and high surface area coverage.This is accomplished by incorporating a holding step, i.e. pre-annealing at the growth temperature before the CVD growth.By increasing the holding time and growth time, we observed lateral crystal size of Mo 2 C through the merging of crystals at the edges.We achieved high surface area coverage Mo 2 C growth, up to 50%, and large single phase Mo 2 C flakes with a lateral size width reaching close to 200 μm.Our results clearly suggest that the holding step facilitates bulk diffusion dynamics of Mo and enhances the Mo concentration on the CVD growth surface, which is the molten copper.The optimum CVD conditions, for high yield growth were found to be pre-growth holding time of 120 min, copper thickness of 50 μm, and growth time of 180 min, corresponding to the thickest Cu layer and longest growth time explored.
The as-grown Mo 2 C crystals were analyzed using light microscopy and backscattered SEM imaging to examine their particle size distribution and surface coverage area, respectively.Moreover, a comprehensive suite of techniques including XRD, Raman spectroscopy, XPS, SEM, and S/ TEM was employed to characterize the obtained crystals.Our analysis revealed that the Mo 2 C phase is orthorhombic, with [100] crystallographic direction perpendicular to the surface.This work provides valuable insights into possibilities and limitations of large area coverage and wide lateral size growth of Mo 2 C in metal foils based CVD approach.Despite an explored total process time of 5 h, while continuously maintaining the growth substrate consisting of Cu foil on top of Mo foil at a temperature of 1095 °C, this study clearly shows that the growth kinetics of Mo 2 C crystals is significantly slower than other 2D materials such as graphene, BN, and TMDCs, particularly in terms of lateral growth.The formation of graphene on the molten copper growth surface is another key factor acting as a barrier for large area full coverage growth of Mo 2 C. Turker F et al has reported in their work that formation of graphene acts as a key rate-limiting step in the Mo 2 C growth by acting as an additional Mo diffusion barrier layer [55].However, its presence is beneficial in promoting more uniform thickness Mo 2 C crystals growth as found by the work of Geng et al [51].This they propose is likely due to the lower energy barrier for surface diffusion of Mo and Mo 2 C over graphene.The formation of graphene is thus likely undesired with respect to growth rate but advantageous for Mo 2 C crystal thickness control in our CVD approach.Another limiting factor, which is challenging to control, is the non-homogenous surface curvature and wetting properties of liquid Cu layer on Mo surface during the CVD growth process, which hinders uniform nucleation and full coverage growth of Mo 2 C. We intend to highlight here that CVD growth of Mo 2 C is scientifically a fascinating process, combining the elements of both conventional CVD growth and liquid metal solvent inorganic crystals growth [43,44,74,75].It is a complex mechanism and to fully unravel its understanding the initial growth stages need to be studied.This research direction may prove beneficial to make progress in large area coverage CVD growth of bare MXenes.On the other hand, it is also essential to explore alternative CVD growth approaches, such as metal organic chemical vapor deposition (MOCVD).In MOCVD, a metal organic precursor source of Mo can be used instead of Mo metal foil, allowing more direct control of Mo and C concentration during growth.This needs to be explored to investigate if it may lead to the growth of covered large area of continuous film of Mo 2 C growth.

Mo 2 C CVD Growth
In this study, commercially procured 0.1 mm (0.004 in) thick 99.95% molybdenum foils (metal basis) from Alfa Aesar and 0.025 mm thick 99.98% copper foils (metal basis) from Sigma-Aldrich were used.The CVD growth experiments were conducted in a custom-built horizontal quartz tube furnace with an inner diameter of 22 mm, an outer diameter of 25 mm, and a length of 1200 mm.The CVD growth was performed at atmospheric pressure in an H 2 /CH 4 environment, with a process gas (H 2 ) flow rate of 200 sccm and a carbon precursor gas (CH 4 ) flow rate of 3 sccm, and the furnace temperature was set to 1095 °C.
To prepare the typical samples for CVD growth, molybdenum and copper metal foils were cut into sizes of about 1 cm × 1 cm and 1.5 cm × 1.5 cm, respectively.The foils were then cleaned with 2% HCl for 2 min, followed by ultrasonic treatment for 10 min in a 1:1:1 ratio of acetone, isopropyl alcohol (IPA) and ethanol mixture, and finally with DI water for 10 min.The cleaned foils were dried using N 2 gas.
We prepared the metal foils stack in such a way that the Cu foil covers the entire surface area of the Mo foil below and wraps around its edges, extending slightly over to the backside of the Mo foil.Such foils preparation was adopted to enable more suitable conditions for bulk diffusion of Mo through the liquid copper during growth, and to strongly limit surface diffusion of Mo from the edges of the Cu melt droplets.However, it does not avoid de-wetting of liquid copper except at the edges of the Mo foil.The wrapped Mo/Cu foils were then placed on an Al 2 O 3 ceramic support and loaded into the quartz tube.
The CVD growth of Mo 2 C crystals was performed according to the following protocol: first, the quartz tube was flushed with Ar (100 sccm) for 10 min, followed by 20 min of flushing with an Ar/H 2 (100/200 sccm) mixture to remove any residual air from the tube.The Ar flow was then stopped, and only the flow of H 2 was present.The furnace was heated to 600 °C under an Mo 2 C atmosphere with a ramping rate of 18 °C min −1 and held at this temperature for 10 min to remove any surface native oxides from the Cu foils.The furnace was then ramped up to the growth temperature of 1095 °C, above the melting point of copper.The samples were held at this temperature in just H 2 flow for different time periods before the CVD growth process was initiated by allowing the additional flow of CH 4 gas (3 sccm) for different CVD growth durations.

Mo 2 C transfer
To transfer the as-grown Mo 2 C crystals from the growth substrate foils to other target substrates, typically Mo 2 C/Si wafers, the PMMA-assisted transfer method was used.The process involved spin coating E-beam resist PMMA 950 K (Allresist GmbH) on the Cu/Mo foil with as-grown Mo 2 C flakes using the KLM SCC 200 spin coater (Schaefer Technologie GmbH) with spinning steps of 10 rps for 5 s and 30 rps for 60 s.The spin-coated foil was cured on a hot plate for 10 min at 50 °C.Subsequently, the sample was immersed in 1 M ammonium persulfate solution, maintained at 80 °C, to etch the Cu layer and detach the PMMA support with the Mo 2 C crystals.The PMMA support was then rinsed in DI water for a few minutes.Following this, the PMMA support was transferred onto a clean SiO 2 /Si wafer or TEM grid and dried at 120 °C for 10 min to enhance the adhesion of the PMMA support with the surface of the target substrate.Finally, PMMA was removed by exposing it to hot acetone vapors.The Witec alpha 300RA correlative microscope was used for light microscopy and Raman characterization.The studies were conducted on both the as-synthesized Mo 2 C on the growth foils and on Mo 2 C after they were transferred from the growth foils to blank SiO 2 /Si wafers.The 10× and 100× objectives were used for this purpose.The Witec suite software was employed to carry out large area scans and imaging of the as-grown foils, using the image stitching feature.
Raman vibrational studies were conducted using a green laser with a wavelength of 532 nm, with laser power ∼2 mW, integration time 20 s and accumulations set to 30.
X-ray diffraction (XRD) was performed for phase and structural characterization on an Empyrean (PANalytical) diffractometer equipped with a Pixcel3D detector.The measurements were taken using a Cu radiation (Kα 1 = 1.540 597 Å) with goniometer a step size of 2θ = 0.0002°u nder normal conditions at RT, to cut out the Cu Kα 2 radiation a 4-bounce monochromator was used on incident and a double crystal analyzer in the diffracted beam optics.The analysis was conducted on the Mo 2 C after they were transferred to SiO 2 /Si substrates.The position of the optics was optimized on the (200) planes of the Mo 2 C crystals.
X-ray photoelectron spectroscopy (XPS) on the Mo 2 C crystals was performed using a Thermo Fisher Scientific Thetaprobe Angle-Resolved x-ray Photoelectron Spectrometer.The spectra were acquired using a monochromatic Al Kα x-ray source and a spot size of 100 μm, with spectra acquisition in standard, non-angle-resolved mode.No charge compensation was required for these samples.At least 8 different locations were probed on each of the samples investigated, both on the as-grown metal foils and after they were transferred to SiO 2 /Si wafers.Peak fitting of the spectra was done using the CasaXPS software (www.casaxps.com).Synthetic components with a Voigt profile (70% Gaussian, 30% Lorentzian) were used in the peak fitting process after subtraction of a Shirley type background.A spin-orbit splitting of 3.18 eV was set for the Mo 3d components, with a 3-2 area ratio between the 3d 5/2 and 3d 3/2 components and with an identical peak width for both components of a chemical state.

SEM
SEM investigations were conducted using the TFS Helios NanoLab 450 HP and TFS Qemscan 650 F equipped with two Bruker x-ray energy-dispersive spectrometer (EDX) detectors.To study the morphology, size, and coverage of the asgrown Mo 2 C crystals over the growth foil surface area, both secondary electron and backscattered electron SEM imaging and EDX measurements were performed.The typical SEM parameters used were 2 kV and 1.6 nA on Helios, and 10 kV, 15 kV spot size 3.5 on Qemscan.

S/TEM
The scanning/transmission electron microscopy (S/TEM) investigations were conducted on the Mo 2 C crystals after their transfer onto lacey-carbon support film coated gold 300mesh TEM grids (SPI Supplies).These studies were performed using Atomic-scale investigations were conducted on an TFS Titan G2 60-300 kV microscope, combining (S)TEM imaging, EDX spectroscopy and electron diffraction.The microscope is equipped with a CEOS DCOR probe-corrector and Super-X EDX detectors.Observations were performed at 300 kV with a probe convergence angle of 24 mrad.The camera length was set to 60 mm and simultaneous STEM imaging was conducted with 3 detectors: high-angle annular dark-field (HAADF) (collection angles 101.7-200 mrad), ADF (collection angles 22.4-101.7 mrad), and annular brightfield (ABF) (collection angles 8.5-22.4mrad).The resulting spatial resolution achieved was approximately 0.08 nm.

Particle size and surface area coverage analysis
To determine the particle size and area fraction coverage of the Mo 2 C crystals on the growth foils, both light microscopy and BSE-SEM images from each growth experiment were analyzed using the color-threshold and analyze particles features in the ImageJ image processing software [76].To ensure more accurate statistics, multiple areas from each sample were imaged under similar imaging conditions and used for analysis.The particle size of individual particles was determined from the corresponding area, and the surface coverage area was taken from the area fraction measurement result generated by the ImageJ software.

Figure 1 (
c) shows that under these flow rate conditions of H 2 and CH 4 , both graphene and Mo 2 C growth occur as the result of CVD process.The inset in figure 1(c) depicts the significant stages involved during the CVD growth process of Mo 2 C. Stage I represents the initial state of two metal foils stack, while stage II schematically illustrates the state during the holding time step after reaching the growth temperature of 1095 °C.At stage II, copper is in liquid state above the Mo foil.Due to the concentration gradient of Mo atoms between

Figure 1 .
Figure 1.Concept illustration of the CVD growth of molybdenum carbide.Schematic representation of (a) Cu, Mo growth foils stacking order, (b) horizontal quartz tube CVD furnace setup, (c) final Mo 2 C MXene crystals and graphene grown in the CVD process (the inset shows light microscopy images of as grown Mo 2 C crystals after transfer to a SiO 2 /Si wafer).(d) Schemes showing the four significant stages involved in the growth of Mo 2 C MXene.(e) The time temperature profile of the modified CVD growth protocol investigated in this work (here H refers to the holding time at 1095 °C and G refers to growth time where the CH 4 gas flow is present).
Figure 2(a) shows a large area scan image of the substrate after growth under light microscopy at 10× magnification.It clearly shows the high coverage growth of the MXene crystals (blue contrast) over the entire Cu surface.

Figure 2 (
b) shows the low magnification backscattered SEM (BSE-SEM) of the same sample, where Mo 2 C crystals appear as bright contrast and the underlying copper surface appear in dark contrast.This further validates the large area coverage of grown Mo 2 C, and one can clearly see the presence of coalesced large lateral size crystals scattered randomly (also see supplementary information figures S1, S3). Figure 2(c) shows the x-ray diffraction (XRD) pattern of CVD-grown Mo 2 C crystals after transferring from the growth substrate to blank SiO 2 /Si wafers.The XRD pattern matches well with the orthorhombic crystal structure of Mo 2 C (i.e.α-Mo 2 C) with space group Pbcn 60 (PDF-4+ 2022-ICDD card no 00-031-0871).This indicates the single-phase structure of the asgrown Mo 2 C crystals, with their [100] crystallographic direction perpendicular to basal facet surface, i.e. the vertical growth direction of the Mo 2 C crystals is along the [100] direction.Our XRD observations match with the previously reported work by Kang et al [63] and Caylan et al [64] on CVD grown α-Mo 2 C phase.Full wide range XRD scans of from α-Mo2C transferred on SiO 2 /Si wafer and the blank SiO 2 /Si wafer, which serves as target transfer substrate, is shown in the supplementary information figure S9.

Figure 2 (
d) shows a Raman spectrum taken from the asgrown Mo 2 C after transferring it to a Si wafer, where the

Figure 2 .
Figure 2. (a) Light microscope large area scan image of as grown foil with Mo 2 C crystals; inset, at lower right corner, shows the photograph of two Cu foils, each with a thickness of 25 μm, wrapped over Mo foil prior to CVD, the rectangular marker on the foil represents the region from which image (a) is recorded.(b) Low magnification BSE-SEM image showing the overview of as grown Mo 2 C crystals (bright contrast).(c) X-ray diffractogram (indexed with PDF-4+ 2022-ICDD card no 00-031-0871) and (d) Raman spectrum of the as grown molybdenum carbide after their transfer to blank SiO 2 /Si wafer.

Figure 3 (
a) illustrates the deconvoluted Mo 3d spectrum, displaying three chemical states.The Mo 3d spin-orbit pair at 228.1 eV and 231.3 eV binding energy (BE) is attributed to Mo-C bond of Mo 2 C.This binding energy fits well with the results obtained by Velusamy et al for unterminated Mo bound to C[65]; metallic Mo has similar binding energies, but its presence can be excluded in the sample transferred to Si.The spin-orbit pair at approximately 228.9 and 232.1 eV BE is in the lower range typical for Mo +4 as well as for surface oxidized Mo 2 C species (oxygen terminated Mo 2 C) [65-67].
Figure 4(a) shows a lowmagnification high-angle annular dark-field (HAADF)-STEM image of a hexagonal 2D micro-sized crystal.The corresponding EDX elemental maps (figures 4(b) and (c)) indicate that the flake is composed of Mo and C that are homogeneously distributed across the whole crystal (see suppl.info.figure S8 for corresponding EDS spectrum).Selected area electron diffraction (SAED) patterns were acquired by tilting the crystal at a series of angles.
Figure 4(d) illustrates a SAED pattern along the [100] zone axis of the orthorhombic

Figure 3 .
Figure 3. XPS characterization of the Mo 2 C crystals grown under optimum CVD growth conditions; the plots show the high-resolution spectra of (a) Mo 3d, (b) C 1s, and (c) O 1s, including the peak fitting components.
) and (f) show atomically-resolved HAADF and (annular brightfield) ABF-STEM images, respectively.HAADF-STEM imaging is sensitive only to heavy Mo atoms, it shows them closely-packed into a hexagonal arrangement in the crystal structure.ABF-STEM imaging is sensitive to both light and heavy element atoms, where atoms appear in dark contrast and the background appears bright.Hence in 4(f) we can see the crystal structure arrangement of both C and Mo atoms simultaneously.The ordered nature of the orthorhombic structure is evident, with the C atoms located at the center being surrounded by six Mo atoms along the [100] crystallographic direction.

Figure 4 .
Figure 4. (a)-(c) HAADF image of a hexagonal-shaped micro-sized Mo 2 C flake and the corresponding EDX elemental maps.(d) SAED pattern along the [100] zone axis of the orthorhombic a-Mo 2 C phase with the characteristic 020 superlattice reflection denoted by the blue circle.(e) High-resolution HAADF image showing the atomic positions of Mo atoms (denoted by red dots).(f) High-resolution ABF image and the corresponding intensity profile (inset) revealing the arrangement of the Mo and C atoms (red dots/arrows annotate Mo atom positions, green dots/arrows annotate C atom positions, cyan line marks the area in the image from which the overlaid intensity line profile is extracted).

Figure 5 .
Figure 5.Light microscopy images of the top surface of the as-grown foils after (a) 30 min holding step and (b) corresponding color threshold image showing the coverage area fraction of Mo 2 C crystals (red color), on the growth substrate.(c) The light microscopy image after 120 min holding step, (d) corresponding color threshold image showing the coverage area fraction of Mo 2 C crystals (red color).

Figure 6 .
Figure 6.BSE-SEM characterization and particle size analysis of growth experiments with different holding step time.The low magnification BSE-SEM images from growth foil center and edge regions for (a), (b) 30 min; (c), (d) 60 min and (e), (f) 120 min holding step time.The CVD growth time for all these samples was 180 min.(g) and (h) show the corresponding particle size analysis histograms performed on images (a)-(f).The scale bar in (a)-(f) is 100 μm.
(b), with only 4.61% coverage.Furthermore, the Mo 2 C crystals size was lower, predominantly in the range of 1-5 μm, as shown in figures 8(a) and (c).These findings indicate that the 25 μm Cu

Figure 7 .
Figure 7. BSE-SEM characterization of growth experiments with different copper foil thickness and CVD growth time.(a) and (b) show the representative BSE-SEM images of regions of the as grown Cu/Mo foil for the case of 1 Cu foil (thickness 25 μm) with CVD growth time of 2 h and 3 h, respectively.(c) and (d) show the as grown Cu/Mo foils for the case of 2 Cu foils (thickness 50 μm) with CVD growth time of 2 h and 3 h, respectively.(e) and (f) show two individual flakes from sample in image (d), where one can clearly see the lateral growth merging of individual Mo 2 C crystals to form wide large size flakes.(g)-(j) show SEM-EDX elemental mapping of as grown Mo 2 C flakes from sample in image (d).
Figures 7(g)-(j) show the representative SEM EDX mapping of Mo 2 C flakes in this sample.

Figure 8 .
Figure 8. Particle size and coverage area fraction plots of Mo 2 C growth with different CVD growth time and copper foil thickness.(a) shows the coverage area fraction plot and (b)-(e) show the particle size distribution histograms.In the labels, 2H refers to holding time of 2 h, 2 G and 3 G refers to CVD growth time of 2 h and 3 h respectively, 1Cu and 2Cu refers to copper foil thickness of 25 μm and 50 μm, respectively.