Selective growth of vanadium dioxide on patterned Al/SiO₂ substrates by metal–organic chemical vapor deposition†

E. K. Bagochus*^a, S. V. Mutilin^a, V. N. Kichay^b and L. V. Yakovkina^b
aRzhanov Institute of Semiconductor Physics SB RAS, 13 Lavrentiev Aven, Novosibirsk, 630090, Russia. E-mail: bagochus@isp.nsc.ru
^bNikolaev Institute of Inorganic Chemistry SB RAS, 3 Lavrentiev Aven, Novosibirsk, 630090, Russia

First published on 7th August 2024

---

---

Introduction

Vanadium dioxide (VO₂) has attracted significant attention in recent years because of its reversible insulator to metal phase transition (IMT) at near-room temperature.¹ The phase transition can be induced by external stimuli, such as an electric field, optical pumping or heat.^2,3 The proximity of the phase transition to room temperature, along with a change in resistivity of up to five orders of magnitude, makes VO₂ a particularly interesting material for practical applications, e.g. as a switch in optical waveguides,⁴ memristor devices,^5,6 thermal sensors and thermochromic coatings,^7,8 and Mott transistors⁹ and in neuromorphic circuits.¹⁰ There is a growing interest in integrating a functionally rich VO₂ material onto silicon substrates. VO₂ film synthesis on silicon substrates has been a topic of interest for several years due to its potential for incorporation of functional oxide electronic devices with silicon integrated circuits. This is due to the need for increased information processing speed and energy efficiency which are the current challenges in modern electronics and photonics. However, the direct synthesis of polycrystalline VO₂ films on the surface of various materials, followed by lithographic operation steps involving interactions with oxygen and water, as well as temperature treatment, can significantly degrade the film quality and uniformity.¹¹

Area selective deposition is a technique that has become increasingly popular in the micro- and nanoelectronic industry due to its ability to eliminate costly lithography steps and reduce the number of edge placement errors.¹² Selectivity is achieved by organizing the growth process in such a way as to promote deposition on the substrate and prevent deposition on the mask, or vice versa. Area-selective deposition can be achieved by chemical vapour deposition (CVD), atomic layer deposition (ALD) and molecular layer deposition (MLD).^13–20 One of the most promising methods for the selective synthesis of structures is chemical vapour deposition. Selectivity can be achieved through the catalytic effect of free electrons on the growth material surface,²¹ different adsorption/desorption rates²² and different surface diffusion rates.²³ Additional factors, such as mask shape and surface diffusion,²⁴ can also affect the nucleation process. Approaches involving the area-selective growth of VO₂ micro- and nanostructures, as well as single nanocrystals on silicon substrates, showed promising results.²⁵ However, the mechanisms leading to the effects of selective VO₂ structure synthesis on heterogeneous surfaces remain unexplored. In this regard, it is important to determine the processes responsible for the kinetics of selective VO₂ nucleation, which include characterizing the nucleation density on heterogeneous surfaces and determining the nucleation rate and the film growth rate at these sites. One of the principal challenges is to define the conditions necessary for area-selective synthesis, which should yield a minimal (or indeed zero) density of nuclei on the non-growing surface.

A method for the efficient area-selective CVD synthesis of VO₂ films on the surface of aluminium (Al) pads using silicon dioxide (SiO₂) as a masking material is described in this paper. Aluminium was chosen for selective CVD due to its wide application in complementary metal-oxide-semiconductor (CMOS) circuits and its potential in such fields like optics and electronics. The proposed method allows for the formation of micro- and nanostructures of VO₂ on a Si substrate during the synthesis process, without subsequent lithography and, accordingly, processing in chemically aggressive environments. The obtained results will be valuable for developing methods to form micro- and nanostructures using the area-selective deposition of vanadium oxide that is compatible with the CMOS technology.

Experimental

Vanadium oxide films were synthesized using metal–organic chemical vapour deposition (MOCVD) in a horizontal two-zone hot-wall tubular reactor (Fig. 1). The reactor pressure was kept constant at 2 torr. Argon was used both as a carrier and diluent gas, while oxygen served as an oxidizing agent. The gas flows were regulated using mass flow meters. The precursor employed was vanadyl acetylacetonate (VO(acac)₂) which was heated to 145 °C. The samples, measuring approximately 1 × 1 cm², were placed on an inclined (about 37°) metal substrate holder, in order to achieve a uniform film thickness along the substrate.^26–28 The temperature in the reaction zone was maintained between 450 and 530 °C, and the synthesis duration ranged from 5 to 60 min, depending on the experimental conditions.

The concentration of the precursor in the reaction region was altered by adjusting the ratio of F1 and F2 flows that represent the carrier gas and diluent gas, respectively (Fig. 1). It is important to note that changing only the carrier gas flow would result in a change in the reactor pressure and overall gas dynamics in the substrate region. Therefore, the total gas flow rate in all experiments was constant at 190 standard cubic centimetres per minute (sccm). In the following, to determine some experimental conditions, we will use the parameter “carrier gas fraction”, determined as:

	(1)

As the ratio n decreases, the volume concentration of the precursor in the reactor also decreases. Determining the exact relation between the carrier gas fraction and the precursor volume concentration is a challenging task. The precursor vapours under our experimental conditions are not saturated, and the presence of hot reactor walls makes it difficult to account for the effects of precipitation and evaporation of precursor reaction products. In this work, the fractions of carrier gas n presented in Table 1 were considered.

---

The VO₂ film synthesis was carried out on three types of substrates: a) silicon wafers (100) with a 300 nm thick thermal oxide layer, b) silicon wafers (100) with a 300 nm thick thermal oxide layer over which a 70 nm thick continuous aluminium layer was deposited by magnetron sputtering and c) silicon wafers (100) with a lithographic pattern consisting of a SiO₂ substrate and Al pads. The feature sizes in the photolithographic pattern ranged from 10 to 1000 μm. Type a and b substrates were used for the composition analysis by X-ray diffraction (XRD) and measurement of VO₂ film thickness and electric resistivity. Type c wafers were obtained from type b wafers by photolithography followed by selective mask etching of aluminium in 10% orthophosphoric acid. The effects of area-selective synthesis were investigated on the type c substrates.

The morphology of the formed structures was analysed using a scanning electron microscope (SEM) JEOL-ISM-6700F and Tescan Vega3 at electron beam energies ranging from 2 to 15 kV. Additionally, an atomic force microscope (AFM) Solver P47 Pro (NT-MDT, Russia) was used to study the surface morphology. AFM measurements were performed in the semi-contact mode using HA_HR silicon cantilevers (TipsNano, Russia) with a needle radius of approximately 10 nm. Energy-dispersive X-ray spectroscopy (EDS) was used to examine the formed structures using a Tescan Vega3 microscope equipped with an Oxford X-Max microanalysis system. Thickness measurements were performed by measuring the step height on the AFM and by analysing the intensity of the film/substrate peaks on the EDS spectra. For the multipoint film thickness measurement, the EDS spectrum analysis was used following the technique described in ref. 29. The surface was irradiated with an electron beam with an energy of 10 keV. At this energy, the X-ray generation region has a depth of approximately 1 μm. In this case, the film and substrate lines will be present in the spectrum. As the film thickness increases, the ratio of film/substrate peak intensities will increase. To calibrate the method, we measured several substrates with known thicknesses and interpolated the a curve of the film/substrate spectral ratio (k) versus VO₂ film thickness (d) (see Fig. S1–S4 in the ESI†). The material was irradiated using a defocused beam with a diameter of approximately 50 μm, allowing for the measurement of the average thickness over this area. This enabled the estimation of the relative thickness of the material in areas where the film was not continuous.

The crystal structure of films and nanocrystals was investigated by X-ray diffraction analysis (XRD) on a Bruker D8 Advance diffractometer (Bragg–Brentano geometry, CuKα radiation, Ni-filter, LYNXEYE XE-T linear detector, 5–65° 2θ range, 0.03° 2θ step, 5 s per step). Indexing the diffraction patterns was carried out using the data for the compounds reported in the PDF-22022 database (Powder Diffraction Files (PDF), International Centre for Diffraction Data, USA).

The electrical measurements of VO₂ nanostructures in the temperature range from 20 to 100 °C were carried out by the standard two-contact measurement method using an Agilent B1500 analyzer. Two ohmic contacts were formed using pressed tungsten needles. During the measurements, the temperature T was changed at the rate of 1 °C per 30 seconds in the forward and reverse directions.

Results

To investigate the area-selective synthesis processes, VO₂ was synthesized on all three types of substrates simultaneously in a single process. The synthesis temperature T (from 450 to 530 °C) and the carrier gas fraction n (from 0.23 to 1) were varied. In each case, the average thicknesses of VO₂ films on Al and SiO₂ were compared to determine the growth rate on each surface as a function of synthesis conditions.

Dependence of VO₂ film thickness on synthesis conditions

The film thicknesses were measured at 16 points across the sample, and the minimum, maximum, and average thicknesses (or effective thickness if the film is not continuous) are indicated in the plots. The dependence of film thickness on temperature and carrier gas fraction n is shown in Fig. 2. At the synthesis temperature of 450 °C, a monotonic film thickness growth with the increasing precursor concentration in the gas mixture is observed for both surface types. Such dependence is typical of the transport limited growth mode.³⁰ In this case, the growth rate on the surface of silicon and aluminium is the same (Fig. 2a). When increasing the synthesis temperature up to 530 °C, the VO₂ film growth rate decreases for both substrate types. Moreover, on the silicon oxide surface, the growth rate deceleration is more noticeable than on the aluminium surface. The decrease in the growth rate with the increasing synthesis temperature can be explained by the precursor decomposition on the reactor walls. However, the differences in the film growth on different substrates with increasing temperature are clearly due to differences in the rates of surface reactions for each substrate.

It is important to note that at the synthesis temperature of 530 °C, the VO₂ film on the silicon oxide surface was not uniform in thickness: with a solid film in the centre and individual crystals at the substrate edges. To quantitatively compare the VO₂ film thicknesses for different substrate regions, in the case of non-continuous films, we introduce the parameter effective thickness d*:

d* = 〈V〉ncr,	(2)

where 〈V〉 is the average crystallite volume and n_cr is the concentration of crystallites per unit area. The average crystallite volume was estimated from the AFM measurements using the Gwyddion software.³¹ The influence of tip effects in our case leads to an increase in the average crystallite volume of about 15%, which is insignificant. The EDS effective layer thickness measurements are in agreement with the AFM average crystallite volume measurements, taking into account the correction for tip effects.

The effective thickness of the VO₂ film synthesized at 530 °C on the silicon oxide surface was 3–5 times lower than the thickness of the VO₂ film on the aluminium surface under the same synthesis conditions.

In order to analyse in detail the thickness distribution of the VO₂ film on SiO₂ and Al surfaces, EDS measurements were carried out and film thickness profiles were plotted on the first and second types of substrates. Fig. 3a shows the thickness distribution of the vanadium dioxide film grown on the aluminium surface. Points A and B mark the locations with the maximum and minimum thickness. Fig. 3c and d show the AFM images of the film at points A and B. The thickness distribution of the vanadium dioxide film grown on the silicon oxide surface is shown in Fig. 3b. Points C and D mark the locations of maximum and minimum thickness. The AFM images of the film at points C and D are shown in Fig. 3e and f. It can be seen that on the aluminium surface, the film fills the entire surface of the sample and is relatively uniform in thickness, whereas on the silicon oxide surface there is an area of continuous film with a large thickness variation in the centre of the sample and the rest of the sample area has only scattered crystals. The SEM images of continuous films on the surface of aluminium at point A (Fig. S5†) and silicon oxide at point D (Fig. S6†) are shown in the ESI.† As can be seen in Fig. 3c and d, the average in-plane diameter of the grain of the film on the aluminium substrate is 300 nm. On the silicon dioxide substrate, the average in-plane diameter of the grain in the solid film region is also about 300 nm, but in the growth region of the single crystallites, the in-plane diameter is significantly larger, being 700–1000 nm.

Dependence of VO₂ film thickness on the synthesis time

To study the growth process kinetics, a series of syntheses lasting from 5 to 60 minutes were carried out at a temperature of 490 °C and a carrier gas fraction of n = 1 on the substrates of the first and second types. The dependence of the average film thickness on the synthesis time is shown in Fig. 4. The VO₂ film synthesis on the SiO₂ surface was first considered. After 5 minutes of the synthesis, the growth on the SiO₂ surface was almost absent and the density of crystallites was less than 1 mm⁻². After 10 minutes of the synthesis, free-standing crystals formed on the substrate. At a synthesis time of 30 minutes, there is an area of a continuous film 100 nm thick in the centre of the sample and single crystals at the substrate periphery. At a synthesis time of 60 minutes, a continuous ∼150 nm thick film is formed over the entire substrate surface. Thus, as the synthesis time increases, the crystallite concentration increases, and the area filled with a continuous film appears, and at a synthesis time of 60 minutes, the entire substrate is covered with a homogeneous film.

On the surface of aluminium, unlike on silicon dioxide, a continuous vanadium dioxide film with a uniform thickness of 30 nm was formed already at the synthesis duration of 5 minutes. With the increasing synthesis time, the film thickness increased, but, in all cases, the film was continuous and the thickness variation did not exceed 15 nm.

The average growth rate at the synthesis time of 60 min was 2.5 nm min⁻¹ on both aluminium and silicon oxide substrates. However, on the aluminium surface, the growth was uniform over the entire substrate surface and the rate did not change significantly during the synthesis; growth on the silicon oxide surface showed significant spatial non-uniformity. Evolution of the film morphology with the increasing synthesis time is shown in Fig. S7 in the ESI.†

Area-selective synthesis of VO₂

To determine the effect of area-selective synthesis of the structures on patterned Al/SiO₂ substrates of the third type, nine different synthesis parameters, i.e. temperature and carrier gas fraction, were investigated. The diagram of the selectivity effect as a function of synthesis conditions is shown in Fig. 5a. The synthesis conditions, under which partial and full growth selectivities are exhibited, are shown. Partial selectivity refers to the growth of individual crystals on the silicon dioxide surface, while full selectivity refers to the complete absence of growth on the silicon dioxide surface. Under certain conditions, there is no growth selectivity, i.e., both on the aluminium surface and on the silicon dioxide surface, a continuous film of equal thickness is formed. However, even in this case, there is a region around the aluminium sites with a width of about 0.5 μm in which no growth occurs. SEM images of these typical growth patterns are shown in Fig. 5b.

At the synthesis temperature of 450 °C and the carrier gas fraction n from 0.23 to 1, the uniform growth of the vanadium dioxide film is observed both on the aluminium surface and on the SiO₂ surface. When the synthesis temperature increases and/or the carrier gas fraction decreases, a decrease in the thickness of the vanadium dioxide film on the silicon dioxide surface is first observed, followed by a complete cessation of the nucleation process. At the same time, a continuous vanadium dioxide film is present on the aluminium surface in all cases.

Determination of the phase composition for the grown films

V–O compounds are known for their complex phase diagram and polymorphism of phases within the same stoichiometry.³² The XRD study of the obtained films showed the same phase composition of the film on the surface of aluminium and silicon oxide (under synthesis conditions that result in a continuous film growth on SiO₂) consisting mainly of the VO₂ M1-phase (Fig. 6).

Electric measurements of VO₂ films

Prior to the VO₂ synthesis, no additional measures were taken to remove the natural oxide layer from the aluminium surface and since the gas mixture contains oxygen, the native oxide layer is formed on the aluminium surface during the growth process. This fact allowed us to measure the resistances of the VO₂ films synthesised on Al in the lateral geometry. The resistance of the ∼100 nm thick VO₂ film on the surface of aluminium, measured by the two-contact method, is about 10⁵ Ω at room temperature. Given that the surface resistance of the 70 nm aluminium layer is less than 1 Ω, it can be concluded that the current flows along the film and not through the aluminium layer. At the same time, when the distance between the contacts is changed, the measured resistance increases, thus indicating that the current flows along the film and not through the substrate. Dependence of the measured resistance on the distance between the probes is presented in the ESI† (Fig. S8). We suggest that the native oxide on the aluminium surface is not only not destroyed in the process of film synthesis, but also retains its dielectric properties, thereby insulating the vanadium dioxide film from the substrate at applied voltages less than 0.5 V. A voltage of less than 0.5 V was employed to prevent Joule heating.

Temperature resistance measurements were made for all VO₂ films synthesised under different conditions on aluminium and silica surfaces. In the case of SiO₂, measurements were made only for those samples where the synthesis resulted in the formation of a continuous film. Fig. 7 shows the results of the resistance temperature dependence measurements for the VO₂ films synthesised on Al and SiO₂ substrates. The resistance of the VO₂ film on Al is higher due to the comparable surface roughness of the initial Al surface (approximately 23 nm) to the VO₂ film thickness, in contrast to the SiO₂ surface roughness of approximately 1 nm. Consequently, the effective thickness of the VO₂ film on Al is reduced, resulting in a higher resistance. An AFM image of the Al/SiO₂ interface is shown in the ESI† (Fig. S9). The summarized results of the resistance temperature measurements of VO₂ films synthesised under different growth conditions on Al and SiO₂ surfaces are presented in Table 2.

---

Table 2 shows the resistivity ratios at room temperature (20 °C) and at 100 °C. The measurements show that all samples synthesised at 490 °C or higher, regardless of the precursor concentration and substrate material, show a sharp resistance jump at 68 °C, which is characteristic of the phase transition in vanadium dioxide. On the aluminium surface, the resistance jump is generally smaller than on the silicon oxide surface. This may be due to the grain structure of the substrate (Al) itself. On thin films less than 100 nm thick on the aluminium surface, a change in the resistance at 68 °C is observed, but the magnitude of such a change is small because the vanadium dioxide film thickness is comparable to the roughness of the original substrate. Thus, by varying the precursor concentration at a given temperature, it is possible to selectively grow films with good electrical characteristics.

Discussion

In this work, significant differences in the nature of vanadium dioxide film growth on the surface of silicon oxide and aluminium were found. These differences are expressed in the growth rate at temperatures above 490 °C, growth uniformity and the phenomenon of selectivity. We attribute these differences to the different rates of precursor desorption from these surfaces. With the increasing synthesis temperature, a decrease in the growth rate and then a complete cessation of the growth on the silicon oxide surface is observed. In ref. 33, an increase in the growth rate of VO₂ was observed with the increasing temperature up to 400 °C, after which an increase in temperature led to a decrease in the growth rate due to an increase in the precursor desorption rate. The difference in the film growth rates on the surface of aluminium and silicon oxide at high temperatures may also be due to the fact that the desorption from the silicon oxide surface is more active than that from the surface of aluminium. The difference in the desorption rate can be explained by the exchange of ligands between VO(acac)₂ and aluminium; the possibility of such a reaction was demonstrated in ref. 34. In this case, the resulting molecule would have a lower volatility, which would lead to a decrease in the desorption rate. A schematic illustration of the interaction between the precursor and the patterned Al/SiO₂ substrate is shown in Fig. 8.

At a synthesis temperature about of 500 °C and synthesis duration of 30 min or less, a difference in the uniformity of the vanadium dioxide film on the surface of silicon oxide and aluminium is observed: on the SiO₂ surface, film formation starts with individual crystals and on the aluminium surface, it starts with a continuous film. This behaviour can be explained by different changes in the surface free energy (surface free energy ΔF_S) during the vanadium dioxide film growth on the surface of aluminium and on the surface of silicon oxide. For example, in ref. 35 the effect of surface spreading of vanadium pentoxide on the aluminium oxide surface and the absence of such an effect on the surface of silicon oxide were observed, which is explained by different ΔF_S. In the case of ΔF_S < 0, the film will be deposited in such a way as to maximize the film–substrate contact area, and the growth will start with the formation of a continuous film, which is observed for the growth on aluminium. In the case of ΔF_S > 0, the new material will be formed in such a way as to minimize the film–substrate contact area, and the growth follows an island mode.

At the silicon oxide surface, it appears that a minimum critical precursor concentration is required for the nucleation process. As the surface concentration decreases due to an increase in temperature and/or a decrease in pressure, the nucleation centre density decreases and single crystals are formed instead of a continuous film. Although the concentration of admolecules is insufficient to nucleate new crystallites, they can participate in the re-growth of existing crystallites. This explains the larger size of free-standing crystallites, compared to a solid film. In the presence of admolecule trapping centres, which can be aluminium sites or an existing vanadium dioxide film, the growth does not start at all, thus resulting in the phenomenon of selectivity.

For a qualitative description of the selectivity process, the case of the lateral aluminium–silicon oxide interface was considered. It is assumed that all precursor molecules striking the aluminium surface immediately participate in the reaction. Precursor molecules striking the silicon oxide surface can diffuse over the surface until they desorb or become trapped. In this case, there will be surface diffusion of admolecules in the direction from SiO₂ to Al, but there will be no diffusion in the opposite direction because Al captures all the admolecules coming to it. Then, the film formation on the SiO₂ surface will occur only at some distance r_ng from the interface with Al, where the admolecule concentration exceeds the critical concentration.

Let us evaluate the influence of synthesis parameters on the size of the no-growth region around the Al sites qualitatively. Assuming that the admolecules diffusion across the surface follows the Fick law, the distribution of the admolecule concentration around the Al sites will be as follows:

	(3)

where N is the admolecule concentration on the substrate surface, N₀ is the equilibrium concentration of admolecules and D is the surface diffusion coefficient.³⁶

Equating the left part of the equation to the critical nucleation concentration N_c, at which the film growth starts, it is possible to estimate the distance r_ng from the Al interface of the site, from which the stable film growth on the SiO₂ surface starts, as follows:

	(4)

In the range of x < r_ng, no growth will occur because the admolecule concentration will be insufficient for the nucleation process (Fig. 9).

These estimates are in qualitative agreement with the experiment. In the bottom picture of Fig. 5b, a region of about 0.5 μm width is observed around the aluminium pad, where no growth is observed. As the carrier gas fraction decreases, the partial pressure of the precursor decreases and, consequently, the radius of the no-growth region increases. The distance between the aluminium sites on our samples was up to 1 mm; so, in the case of full growth selectivity, we can say that the radius of the no-growth region was more than 500 μm. This effect also explains the character of the film thickness distribution on the silicon oxide surface: the sample boundaries contain many defects that act as precursor capture centres so that as we approach the sample edge, the film thickness decreases to almost zero, but the growth of vanadium dioxide does occur on the side faces of the substrate. In particular, it follows eqn (1) in which the radius of the no-growth region increases with the increasing temperature and decreasing precursor partial pressure which, in turn, depends on the carrier gas fraction (ESI† eqn (S8)). This is the dependence of the growth selectivity on the synthesis parameters observed in our experiments, as can be seen in Fig. 5a.

Despite the different natures of nucleation on the surface of aluminium and silicon oxide, the vanadium dioxide film properties and phase composition depend only on the synthesis conditions. For example, the phase transition of VO₂ films is observed on both aluminium and silicon surfaces. The XRD studies of the films also show a similar phase composition. Thus, by controlling the synthesis parameters such as temperature and precursor concentration, it is possible to perform the area-selective synthesis of high-quality VO₂ structures.

Conclusions

The general approach for the area-selective synthesis of vanadium dioxide on patterned Al/SiO₂ substrates using the CVD method was formulated. The nucleation and growth kinetics of VO₂ films on Al and SiO₂ surfaces are qualitatively investigated as a function of time, synthesis temperature and precursor concentration in the reaction zone. It was shown that by varying the precursor concentration, the nucleation kinetics of VO₂ crystals on the SiO₂ surface can be effectively controlled. Three VO₂ growth modes have been demonstrated on patterned Al/SiO₂ substrates depending on the synthesis conditions: the absence of area-selectivity, where a continuous film grows over the entire substrate surface; full area-selectivity, where the film grows only on the Al pads; and an intermediate state, where the nucleation is absent on SiO₂ only near the Al pads. The resistance ratio before and after the phase transition in the VO₂ films synthesized on SiO₂ and Al surfaces is significantly dependent on the synthesis temperature, but, practically, independent of the precursor concentration in the gas phase. The best structures exhibit a resistance ratio more than 10³ times. The obtained results provide an important step towards the controlled selective growth, which is essential for the development of high-performance nanophotonic and nanoelectronic devices based on VO₂ grown on a Si platform.

Data availability

The data supporting this article have been included as part of the ESI.†

Author contributions

Bagochus E.: writing – original draft, formal analysis, and investigation. Mutilin S.: writing – review & editing and conceptualization. Yakovkina L.: methodology and data curation. Kichay V.: investigation and resources.

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

This work was supported by the Russian Science Foundation (grant no. 21-19-00873) and the Ministry of Science and Higher Education of the Russian Federation.

References

1. F. J. Morin, Phys. Rev. Lett., 1959, 3, 34–36 CrossRef CAS .
2. Y. Ke, S. Wang, G. Liu, M. Li, T. J. White and Y. Long, Small, 2018, 14(39) DOI:10.1002/smll.201802025 .
3. S. Kabir, S. Nirantar, L. Zhu, C. Ton-That, S. K. Jain, A. B. A. Kayani, B. J. Murdoch, S. Sriram, S. Walia and M. Bhaskaran, Appl. Mater. Today, 2020, 21, 100833 CrossRef .
4. K. J. Miller, K. A. Hallman, R. F. Haglund and S. M. Weiss, Opt. Express, 2017, 25, 26527 CrossRef CAS .
5. T. Driscoll, H.-T. Kim, B.-G. Chae, B.-J. Kim, Y.-W. Lee, N. M. Jokerst, S. Palit, D. R. Smith, M. Di Ventra and D. N. Basov, Science, 2009, 325, 1518–1521 CrossRef CAS PubMed .
6. J. del Valle, Y. Kalcheim, J. Trastoy, A. Charnukha, D. N. Basov and I. K. Schuller, Phys. Rev. Appl., 2017, 8, 054041 CrossRef .
7. V. Melnik, I. Khatsevych, V. Kladko, A. Kuchuk, V. Nikirin and B. Romanyuk, Mater. Lett., 2012, 68, 215–217 CrossRef CAS .
8. Z. Shao, A. Huang, C. Cao, X. Ji, W. Hu, H. Luo, J. Bell, P. Jin, R. Yang and X. Cao, Nat. Sustain., 2024, 7, 796–803 CrossRef .
9. M. A. Belyaev, A. A. Velichko, P. P. Boriskov, N. A. Kuldin, V. V. Putrolaynen and G. B. Stefanovitch, J. Sel. Top. Nano Electron. Comput., 2014, 1, 26–30 CrossRef .
10. M. Ignatov, M. Ziegler, M. Hansen, A. Petraru and H. Kohlstedt, Front. Neurosci., 2015, 9 DOI:10.3389/fnins.2015.00376 .
11. T. Chang, X. Cao, N. Li, S. Long, Y. Zhu, J. Huang, H. Luo and P. Jin, Matter, 2019, 1, 734–744 CrossRef CAS .
12. J. Mulkens, M. Hanna, B. Slachter, W. Tel, M. Kubis, M. Maslow, C. Spence and V. Timoshkov, Metrology, Inspection, and Process Control for Microlithography XXXI, 2017, vol. 10145, p. 1014505 Search PubMed .
13. R. Clark, K. Tapily, K.-H. Yu, T. Hakamata, S. Consiglio, D. O'Meara, C. Wajda, J. Smith and G. Leusink, APL Mater., 2018, 6(5) DOI:10.1063/1.5026805 .
14. G. N. Parsons and R. D. Clark, Chem. Mater., 2020, 32, 4920–4953 CrossRef CAS .
15. S. K. Song, H. Saare and G. N. Parsons, Chem. Mater., 2019, 31, 4793–4804 CrossRef CAS .
16. M. F. J. Vos, S. N. Chopra, M. A. Verheijen, J. G. Ekerdt, S. Agarwal, W. M. M. Kessels and A. J. M. Mackus, Chem. Mater., 2019, 31, 3878–3882 CrossRef CAS .
17. J.-O. Carlsson, Crit. Rev. Solid State Mater. Sci., 1990, 16, 161–212 CrossRef CAS .
18. J. Soethoudt, H. Hody, V. Spampinato, A. Franquet, B. Briggs, B. T. Chan and A. Delabie, Adv. Mater. Interfaces, 2019, 6(20) DOI:10.1002/admi.201900896 .
19. M. J. M. Merkx, T. E. Sandoval, D. M. Hausmann, W. M. M. Kessels and A. J. M. Mackus, Chem. Mater., 2020, 32, 3335–3345 CrossRef CAS .
20. K. Cao, J. Cai and R. Chen, Chem. Mater., 2020, 32, 2195–2207 CrossRef CAS .
21. K. Tsubouchi and K. Masu, Thin Solid Films, 1993, 228, 312–318 CrossRef CAS .
22. X. Zhong, R. Jordan, J.-R. Chen, J. Raymond and J. Lahann, ACS Appl. Mater. Interfaces, 2023, 15, 21618–21628 CrossRef CAS PubMed .
23. J. E. Greenspan, C. Blaauw, B. Emmerstorfer, R. W. Glew and I. Shih, J. Cryst. Growth, 2003, 248, 405–410 CrossRef CAS .
24. J. A. Venables, Introduction to Surface and Thin Film Processes, Cambridge University Press, Cambridge, 2000 Search PubMed .
25. S. V. Mutilin, V. Y. Prinz, V. A. Seleznev and L. V. Yakovkina, Appl. Phys. Lett., 2018, 113(4) DOI:10.1063/1.5031075 .
26. L. V. Yakovkina, S. V. Mutilin, V. Y. Prinz, T. P. Smirnova, V. R. Shayapov, I. V. Korol'kov, E. A. Maksimovsky and N. D. Volchok, J. Mater. Sci., 2017, 52, 4061–4069 CrossRef CAS .
27. H.-Y. Pak and K.-W. Park, Numer. Heat Transfer, Part A, 2000, 37, 407–423 CrossRef .
28. F. C. Eversteyn, P. J. W. Severin, C. H. J. Van Den Brekel and H. L. Peek, J. Electrochem. Soc., 1970, 117, 925 CrossRef .
29. A. Franquet, T. Conard, M. Gilbert, T. Hantschel and W. Vandervorst, J. Phys.: Conf. Ser., 2013, 417, 012033 CrossRef CAS .
30. H. O. Pierson, in Handbook of Chemical Vapor Deposition (CVD), Elsevier, 1999, pp. 36–67 Search PubMed .
31. D. Nečas and P. Klapetek, Open Phys., 2012, 10, 181–188 CrossRef .
32. H. A. Wriedt, Bull. Alloy Phase Diagrams, 1989, 10, 271–277 CrossRef CAS .
33. M. B. Sahana, M. S. Dharmaprakash and S. A. Shivashankar, J. Mater. Chem., 2002, 12, 333–338 RSC .
34. P. Van Der Voort, M. G. White and E. F. Vansant, Langmuir, 1998, 14, 106–112 CrossRef CAS .
35. C.-B. Wang, Y. Cai and I. E. Wachs, Langmuir, 1999, 15, 1223–1235 CrossRef CAS .
36. K. Yamaguchi, M. Ogasawara and K. Okamoto, J. Appl. Phys., 1992, 72, 5919–5925 CrossRef CAS .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ce00315b

This journal is © The Royal Society of Chemistry 2024