In this letter, we report on the synthesis of monolayers of Mo S 2 via chemical vapor deposition directly on thin films of A l 2 O 3 grown by spatial atomic layer deposition. The synthesized monolayers are characterized by atomic force microscopy as well as confocal Raman and photoluminescence spectroscopies. Our data reveal that the morphology and properties of the 2D material differ strongly depending on its position on the substrate. Close to the material source, we find individual flakes with an edge length of several hundred microns exhibiting a tensile strain of 0.3 %, n-doping on the order of n e = 0.2 × 1013 cm−2, and a dominant trion contribution to the photoluminescence signal. In contrast to this, we identify a mm-sized region downstream, that is made up from densely packed, small Mo S 2 crystallites with an edge length of several microns down to the nanometer regime and a coverage of more than 70 %. This nano-crystalline layer shows a significantly reduced strain of only <0.02 %, photoluminescence emission at an energy of 1.86 eV with a reduced trion contribution, and appears to be p-doped with a carrier density of n h = 0.1 × 1013 cm−2. The unusual p-type doping achieved here in a standard chemical vapor deposition process without substitutional doping, post-processing, or the use of additional chemicals may prove useful for applications.

In recent years, there has been increasing interest in 2D materials for high performance electronic and optoelectronic applications,1–3 including flexible devices4–6 and field-effect transistors.7 An important class of 2D materials is transition metal dichalcogenides (TMDCs) with the chemical formula MX 2. The hexagonal structure is formed by a sheet of transition metal (M) atoms covalently bound in between two layers of chalcogen (X) atoms. Some TMDCs like molybdenum disulfide ( Mo S 2) have semiconductor properties like an indirect bandgap in the near infrared. When their spatial extension is confined in one dimension, this bandgap becomes a direct bandgap in the visible spectrum,8–10 making these materials even more attractive for (flexible) optoelectronics. This however requires the fabrication of heterostructures, i.e., the combination of the TMDC with at least one dielectric material. The current approach of layering Mo S 2 on top of high-k dielectrics utilizes growth via chemical vapor deposition (CVD) on selected substrates, typically Si O 2, and subsequent transfer onto the final dielectric materials. This process suffers from various limitations including the loss of structural integrity of the Mo S 2 layer, no control over the doping, and contamination through residuals from the transfer process.11,12 Growing Mo S 2 directly on top of the dielectric could overcome these severe limitations.7 

While this approach seems straightforward, it is still a challenge to obtain a continuous layer of Mo S 2 using a common fabrication method such as CVD. The resulting monolayers are typically highly crystalline, but their lateral dimensions rarely exceed a few hundred microns.7,13 Also, the large-scale fabrication of tailored dielectrics poses a major problem, in particular if the flexibility of the 2D material is to be maintained, requiring ultrathin dielectric films. To address both problems, it has recently been tried to combine CVD growth of Mo S 2 with the synthesis of the dielectric material via atomic layer deposition.7 The resulting flakes were on average rather small (edge length of a few microns), could be processed into a working field-effect device and showed a PL signal at room temperature (RT). Atomic layer deposition (ALD) has been a powerful tool for a wide variety of thin film applications in both industry and research.14 It is already in use worldwide for the fabrication of microelectronics and can deposit nanolaminated stacks of films.15–17 However, one of the biggest drawbacks of this technique is its low deposition rate and therefore it may not always be commercially viable considering the price and its limited throughput.18 

We therefore suggest to use the newly developed process of atmospheric-pressure spatial atomic layer deposition (AP-SALD) to overcome this drawback. This technique separates the precursor gases not temporally but spatially, making it a faster and more cost-effective way of creating a thin layer with sufficient control over the layer thickness.6,19,20 AP-SALD has already been successfully utilized for the growth of uniform films for different applications including solar cells,21,22 light emitting diodes,23–25 and quantum tunneling diodes.26 Even at 15 nm thickness and high aspect ratio, this method can archive a continuous and pinhole-free layer.27–29 Here, we report on the first results of growing a 2D material on a dielectric film deposited by AP-SALD. Monolayer MoS 2 is grown via CVD on thin Al 2O 3 films deposited by AP-SALD. By a detailed analysis, we show that the combination of those two scalable approaches may be used to fabricate mm-sized samples largely covered with nano-crystalline p-doped 2D- Mo S 2.

The growth of TMDCs via CVD is typically dominated by the material supply. This, in turn, is governed by the material sources, the TMDC itself, and the substrate. The most common substrate is thermally grown Si O 2 on top of a Si wafer, which allows the fabrication of field-effect transistors (FETs) based on the as grown-material without an additional transfer step. This may however not be an ideal substrate for the synthesis of large MoS 2 samples, as thermally grown Si O 2 is amorphous and exhibits a rather low roughness (root mean square) of z R M S < 0.2 nm. Thus, if the fabrication on a FET is not the final goal, better suited substrates can be used which may facilitate large-scale growth. For Mo S 2, epi-polished, single crystals of A l 2 O 3 have proven to be an almost ideal substrate with atomic scale roughness and a lattice constant of a A l 2 O 3 = 4.8 Å. The lattice is commensurable to the MoS 2 lattice ( a M o S 2 = 3.2 Å30) and mm-sized, epitaxial monolayers could be grown via CVD.31 As reference systems for our study, we have therefore grown MoS 2 via CVD on amorphous Si O 2 [see Fig. 1(a)] and on crystalline A l 2 O 3 [see Fig. 1(b); for details see Sec. IV]. On both substrates, individual flakes with the typical star-like triangular form have grown. On average, the flakes on Si O 2 are somewhat smaller than the ones on A l 2 O 3. Both samples exhibit the typical in- and out-of-plane Raman modes E 2 g 1 and A 1 g at wavenumbers 382.5 and 402.6 cm−1 for the Si O 2 substrate, and 384.5 and 402.5 cm−1 for the sapphire substrate, respectively, as well as a strong photoluminescence signal at an energy around E = 1.8 eV, the latter confirming the materials 2D nature. From the position of the Raman modes, we deduced the strain and doping of the grown flakes following the approaches presented in Refs. 12, 33, and 34. Pollmann et al. utilize a matrix T to quantify the strain- and doping-dependant Raman shifts based on unstrained, undoped Mo S 2 as the reference.32 The reference matrix is given as
T = ( 0.490 % / cm 1 0.073 % / cm 1 0.088 × 10 13 cm 2 / cm 1 0.464 × 10 13 cm 2 / cm 1 ) .
From this, strain and doping can be directly calculated as
T ( Δ E 2 g 1 Δ A 1 g ) = ( Δ S t r a i n Δ D o p i n g ) .
We find that Mo S 2 on Si O 2 exhibits a tensile strain of 0.43 % and is highly n-doped with n e = 1.1 × 1013 cm−2, and on sapphire it shows a compressive strain of 0.6 % and is less n-doped with n e = 1.4 × 1012 cm−2.
FIG. 1.

AFM/optical images and PL/Raman spectra of Mo S 2 monolayers grown via CVD on the two most common substrates. AFM of Mo S 2 on Si O 2 (a) and on sapphire (b), and the corresponding optical images (inset). The scale bar is 10 μm. The height of the layer was measured to be 0.654 nm (a, line, white), which is consistent with the nominal single layer height of Mo S 2. (b) clearly shows the atomic terraces of the single crystalline substrate. The PL spectra from both samples show a distinct RT photoluminescence at 1.76 eV ( Si O 2, c) and 1.8 eV ( A l 2 O 3, d), respectively, that is characteristic of the Mo S 2 monolayer. The Raman spectra show the fingerprint A 1 g and E 2 g 1 vibrational modes (inset, c, d).

FIG. 1.

AFM/optical images and PL/Raman spectra of Mo S 2 monolayers grown via CVD on the two most common substrates. AFM of Mo S 2 on Si O 2 (a) and on sapphire (b), and the corresponding optical images (inset). The scale bar is 10 μm. The height of the layer was measured to be 0.654 nm (a, line, white), which is consistent with the nominal single layer height of Mo S 2. (b) clearly shows the atomic terraces of the single crystalline substrate. The PL spectra from both samples show a distinct RT photoluminescence at 1.76 eV ( Si O 2, c) and 1.8 eV ( A l 2 O 3, d), respectively, that is characteristic of the Mo S 2 monolayer. The Raman spectra show the fingerprint A 1 g and E 2 g 1 vibrational modes (inset, c, d).

Close modal

From Fig. 1, it is obvious that, in principle, the amorphous structure of Si O 2 does not prevent the synthesis of large monolayer flakes and thus it seems promising to investigate the possibility to grow TMDCs on amorphous A l 2 O 3. To this end, a Si wafer with a 285 nm Si O 2 layer on top was used as a base substrate. Then, 20 nm of A l 2 O 3 were deposited on top of the wafer via AP-SALD (for details, see Sec. IV), resulting in a continuous layer. The thickness can be chosen freely within the limits of the method and in particular it can be chosen to be sufficiently thin to enable flexible devices. For the purpose of this study, we have chosen 20 nm which should be thick enough to prevent the underlying Si O 2 layer from influencing the Mo S 2 growth, e.g., by interface charges.34 The A l 2 O 3 layers deposited in this way have an average roughness of z R M S = 0.33 nm, as shown in Fig. 2(a), which is slightly larger than for a layer grown by conventional ALD ( z R M S = 0.22 nm, shown for comparison in Fig. 2(b), for details on layer generation see the supplementary material35), but comparable to the roughness of a clean, amorphous Si O 2 surface with z R M S = 0.15 nm (see Fig. S1 in the supplementary material35).

FIG. 2.

AFM images of the A l 2 O 3 substrate deposited by AP-SALD (a), conventional ALD (b). In both cases a uniform, continuous layer is formed. Statistical analysis of the height data yields a roughness of about z R M S = 0.33 nm for the AP-SALD layer and z R M S = 0.22 nm for the ALD layer.

FIG. 2.

AFM images of the A l 2 O 3 substrate deposited by AP-SALD (a), conventional ALD (b). In both cases a uniform, continuous layer is formed. Statistical analysis of the height data yields a roughness of about z R M S = 0.33 nm for the AP-SALD layer and z R M S = 0.22 nm for the ALD layer.

Close modal

The AP-SALD substrate was then used in a standard CVD process to deposit Mo S 2 (for details, see Sec. IV). In Fig. 3(a), an optical image from the as-grown Mo S 2 is shown. At the top left side of the sample, close to the material source, Mo S 2 has grown as individual flakes that have the typical triangular shape with a size of a few 10 μm. Moving to the bottom right, i.e., further downstream, one can see that the density of the flakes increases while their size decreases. This clearly shows that the amorphous AP-SALD substrate allows for very different flake morphologies, depending on the temperature and material supply. In particular, there is a large region, extending roughly from the middle of the first third of the substrate (position of the smaller, green square) to its lower left right edge, where the flake size appears to be too small to be resolved by optical microscopy and a seemingly continuous film has been synthesized. Note, that this region is on the order of several square millimeters. To the best of our knowledge, this has never been achieved with an amorphous Si O 2/ Si substrate and furthermore did also not occur on ALD-grown A l 2 O 3 substrates (see the supplementary material35). Further analysis with AFM reveals the true morphology in this region. As shown in Fig. 4, the seemingly uniform film is made up from densely packed nano-crystallites with an average height of 7 Å. The edge length of the individual flakes decreases from several microns [Figs. 4(a)4(c)] down to <100 nm [Figs. 4(d)4(f)] within a span of 120 μm. At the same time, the filling factor increases from 35 % for the individual flakes to about 75 % for the nano-crystallites.

FIG. 3.

Optical images, exemplary Raman, PL spectra and mapping of Mo S 2 monolayers grown via CVD on the AP-SALD-grown A l 2 O 3 layer. Optical image (a) of the substrate fabricated by AP-SALD after the Mo S 2 process. Downstream direction is marked with the arrow. Separated flakes (upstream) are visible as well as a continuous layer of flakes (downstream). The transition area in between (green square) is shown in (b). The individual flakes have a triangular structure; Raman and PL spectra shown in (c) are taken from the areas marked by the circles. The PL peaks for the nano-crystallites (b, red or light gray) have more intensity and are more narrow than for the individual flakes (b, black). The Raman spectra show the typical Mo S 2 modes for both (c, inset). The E 2 g 1 and A 1 g modes where measured at 384 and 403 cm−1 for the individual flakes and at 385 cm−1 and 404 cm−1 for the nano-crystallites. A PL map (d) shows the intensity at 1.86 eV and that the nano-crystalline region expands over several hundred microns. In the lower part, where no PL activity is measured, no a- A l 2 O 3 was deposited and no Mo S 2 growth was detected.

FIG. 3.

Optical images, exemplary Raman, PL spectra and mapping of Mo S 2 monolayers grown via CVD on the AP-SALD-grown A l 2 O 3 layer. Optical image (a) of the substrate fabricated by AP-SALD after the Mo S 2 process. Downstream direction is marked with the arrow. Separated flakes (upstream) are visible as well as a continuous layer of flakes (downstream). The transition area in between (green square) is shown in (b). The individual flakes have a triangular structure; Raman and PL spectra shown in (c) are taken from the areas marked by the circles. The PL peaks for the nano-crystallites (b, red or light gray) have more intensity and are more narrow than for the individual flakes (b, black). The Raman spectra show the typical Mo S 2 modes for both (c, inset). The E 2 g 1 and A 1 g modes where measured at 384 and 403 cm−1 for the individual flakes and at 385 cm−1 and 404 cm−1 for the nano-crystallites. A PL map (d) shows the intensity at 1.86 eV and that the nano-crystalline region expands over several hundred microns. In the lower part, where no PL activity is measured, no a- A l 2 O 3 was deposited and no Mo S 2 growth was detected.

Close modal
FIG. 4.

Peak-force-microscopy images taken in different regions of the sample starting from micron-sized individual flakes (a)–(c) in the upstream region to the nano-crystallites typically found in the downstream region (d)–(f). The filling factor increases from 49 % (a) to 73 % (f). The scale bar is 1 μm.

FIG. 4.

Peak-force-microscopy images taken in different regions of the sample starting from micron-sized individual flakes (a)–(c) in the upstream region to the nano-crystallites typically found in the downstream region (d)–(f). The filling factor increases from 49 % (a) to 73 % (f). The scale bar is 1 μm.

Close modal

To further verify that this nano-crystalline film region is made up from Mo S 2 monolayer flakes, Raman and PL spectroscopies were performed. For easy comparison, we chose the transition region (green square in Fig. 3) and start upstream with an analysis of the individual flakes [Fig. 3(b), black circle], where the Raman E 2 g 1 and A 1 g modes are found at 384.2 and 403.0 c m 1, i.e., exhibiting a difference of 19 c m 1 [Fig. 3(c), black curve in the inset]. The difference in wave numbers confirms that these flakes are single layer Mo S 2,36 and this is corroborated by a moderate PL at 1.85 e V  [Fig. 3(c), black curve]. The PL signal shows a strong trion contribution at 1.8 eV, as shown by the blue fits in Fig. 3(c), that points toward a significant density of excess charges. From the Raman mode positions, we again infer that Mo S 2 here is single layer,36 n-doped with n e = 0.3 × 1013 cm−2, and experiences a tensile strain of 0.4 %.32–34 The strain is comparable to that observed on the Si O 2 substrate (see Fig. 1), but the doping is reduced quite drastically when compared to both Si O 2 and sapphire as substrates.

Next, we will analyze the nano-crystalline region located further downstream [Fig. 3(b), red or light gray circle]. Note, that the overall intensity of the Mo S 2 Raman modes [Fig. 3(c), inset] is in general very low, compared to the Si signal at 520 c m 1, which is probably due to the different optical properties of the AP-SALD substrate.37 The different intensity ratio of the in-plane E 2 g 1 mode and the A 1 g mode maybe due to the different polarization dependence of the two modes.36 

Despite the comparatively poor signal-to-noise ratio, we were able to perform a quantitative analysis. We find that both the in- and out-of-plane modes have shifted in comparison to the modes observed from individual flakes [see red and black curves in the inset of Fig. 3(c)], exhibiting an even smaller difference of 19 cm 1. The E 2 g 1 mode shifts to 385.2 cm−1 which shows that the strain of the material is now negligible (<0.02 %) and thus comparable to the reference value of freestanding Mo S 2.33,38 This relaxation is probably connected to the reduced size of the flakes. The A 1 g mode red-shifts to 404.0 cm−1 and therefore the material here is p-doped with n h = 0.1 × 1013 cm−2.33,39 The observed change in doping is in striking contrast to values reported in the literature so far, where Mo S 2 usually shows n-type doping on A l 2 O 3 and SiO 2,33,40,41 while p-type doping has been observed so far only on metallic substrates, as for example, Au.33,42 P-type doping of Mo S 2 can be achieved in different ways,39,43 e.g., also by adding appropriate reactants in the CVD process but the resulting carrier density determined in a field-effect transistor was an order of magnitude lower.44 

Interestingly, in comparison, the PL in the nano-crystalline region has a significantly higher intensity at an energy of about 1.86 eV [Fig. 3(c), red or light gray curve]. This considerable PL signal again confirms that the nano-crystallites consist of single layer Mo S 2.9 The trion contribution is strongly reduced to I T / I E = 0.83 ( I T / I E = 0.44 at 77 K, see SM Fig. S235) as derived from the orange fits in Fig. 3(c). In comparison, the large flakes exhibit a ratio of I T / I E = 2.00. The reduced trion contribution indicates that less excess charge carriers are available and mirrors the red-shift of the Raman modes associated with p-type doping in the nano-crystalline region. Low-temperature PL measurements (see the supplementary material35) show that the peaks barely shift (around 20 meV) which is to be expected for a non-strained monolayer. Finally, a PL mapping of a large sample region [Fig. 3(d)] shows that indeed the whole nano-crystalline film is PL-active.

We have synthesized monolayers of Mo S 2 via CVD on thin films of A l 2 O 3 grown via AP-SALD. The resulting flakes show a large variation in properties depending on their position on the substrate. While this effect is in principle well-known from all CVD processes,45 we observe a strong and quite unusual difference in morphology, and related to this, unexpected changes of vibrational and optoelectronic properties. In particular, while the larger, individual flakes show only weak, trion-dominated PL and appear n-doped and strained, we identified a region with a very high (up to 73 %) coverage of nano-crystalline Mo S 2 with negligible strain, p-type doping and a significant PL emission dominated by the A exciton. Whether this unusual doping can be maintained upon further processing remains to be seen.

The unique morphology of Mo S 2 monolayer nano-crystallites combined with a high coverage appears exclusively on the AP-SALD-grown substrates and could be exploited in applications where a high density of reactive edges is advantageous, e.g., catalysis,46 sensing,47 or electrochemical48 applications. In terms of (flexible) devices, it would be beneficial to further increase the coverage up to a completely continuous, closed film. This would probably lead to an increase in strain as the nano-crystallites would no longer be isolated from each other. However, the p-doping, which is clearly the more intriguing property, should not be affected. One way to obtain a continuous film could be a second processing step, exploiting the fact that the growth is dominated by the transport of materials across the flake area (and not along the edges),49 which could lead to a preferential growth in-between the pre-existing flakes. Clearly, this is delicate balance and the simultaneous suppression of competing processes such as multilayer growth and decomposition of pre-existing flakes would require a high degree of control and an extended parameter study to establish the correct growth parameters, which is beyond the scope of this paper. An alternative way could be to combine the AP-SALD-grown substrate with another variant of CVD, i.e., metal organic CVD which is known to produce uniform and closed layers of Mo S 2 composed of many small crystallites.50 

AP-SALD is a modification of the conventional ALD approach, in which the different precursor exposure steps of conventional ALD are not temporally but instead spatially separated. In this work, a custom-built, linear “zone-separated” AP-SALD system was used. This arrangement works with a fixed reactor head through which the individual precursor gases pass in different channels. The gases exit in parallel streams at different locations on the bottom of the head creating precursor-filled zones. The individual zones are separated by an inert nitrogen gas. Furthermore, in between the individual gas channels, there are exhaust channels through which the gases are removed using a vacuum pump. The substrate was placed on a heated substrate stage approximately 100 μm below the reactor head and oscillated back and forth at 50 mm/s to deposit a film with the desired 20 nm thickness (80 cycles). By eliminating the vacuum chamber and purge steps, the throughput is higher and the process is more cost effective than that of a conventional ALD system.

Trimethylaluminum (TMA, Strem Chemicals) and distilled water were used as the precursors. To deliver the TMA to the reactor head, nitrogen was bubbled through the TMA at 38 SCCM and combined with a carrier nitrogen flow of 87 SCCM. Similarly, the distilled water was bubbled with a 125 SCCM nitrogen flow and combined with a carrier nitrogen flow of 125 SCCM. A flow of 2000 SCCM was delivered to six inert nitrogen gas channels separating the precursor channels. The heated substrate stage was held at 250 °C throughout deposition.

The custom-made synthesis protocol is based on the system by Lee et al.51 and is used to fabricate crystalline Mo S 2 monolayers on various substrates.32 The Si O 2 substrate was a p-doped silicon wafer in (100) orientation with a 285 nm oxide thickness from Graphene Supermarket. The Sapphire substrate was a c-plane (0001), epi-polished monocrystalline A l 2 O 3 wafer with a 0.2° offcut from Roditi. It was annealed at 1000 °C for 60 min prior to synthesis.

The molybdenum source was ammonium heptamolybdate (AHM, Sigma Aldrich), an inorganic ammonium salt soluble in water. The solution with 200 g L−1 was deposited onto the substrate in a single droplet of 1 mm in diameter at the upstream edge of the substrate. This substrate, including the precursor material, was heated to 300 °C for 30 min to form Mo O 3.52 In a 1 Vol% solution the seeding promoter, cholic acid sodium salt (Sigma Aldrich), was spin coated onto the substrate containing the Mo O 3 and put into a crucible. A second crucible was filled with 40 mg sulfur (S powder, Sigma Aldrich, 99.98 %). The chemical vapor deposition took place in a tube furnace (ThermConcept, ROK 70/750/12-3z) with three thermally separated heating zones. The crucible containing the substrates was placed into the second heating zone (downstream) and the crucible containing sulfur was placed into the first heating zone (upstream). The tube was purged with inert argon gas. For the synthesis, the Ar flow rate was set to 500 SCCM. The downstream zone was heated to 750 °C in 10 min and held at that temperature for 20 min. The upstream zone was heated to 150 °C within 10 min and held at that temperature for 20 min. The process was ended by rapidly cooling down the samples.

Atomic force microscopy images were taken with a Veeco Dimension 3100 AFM in a tapping-mode using Nanosensors NCHR-50 tips. The peak-force-microscopy measurements were performed with a Bruker Dimension Icon in the PeakForce Tapping Mode using ScanAsyst-Air tips. The AFM results were analyzed and visualized by Gwyddion 2.60. The filling factors were evaluated by increasing the contrast between flakes and substrate until a black and white image was obtained and then counting the respective pixels.

Raman and PL spectroscopy were performed with a WiTec alpha300 RA confocal Raman spectrometer. All measurements were performed with 532 nm excitation wavelength and an output power of 4 mW. For Raman measurements a 1800 g/mm grid, and for PL measurements a 300 g/mm grid was used, respectively.

A.M. and M.S. acknowledge financial support from the Bundesministerium für Bildung und Forschung (BMBF; Federal Ministry of Education and Research) (Project No. 05K16PG1), Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) and through project No. 429784087. Support by S. Franzka with the peak-force-microscopy measurements by the Interdisciplinary Center for Analytics on the Nanoscale (ICAN) of the University of Duisburg-Essen (DFG RIsources reference: RI_00313), a DFG-funded core facility (Project Nos. 233512597 and 324659309), is gratefully acknowledged. K.P.M. acknowledges funding from NSERC Discovery (Nos. RGPIN-2017-04212 and RGPAS-2017-507977), Canada Foundation For Innovation John R. Evans Leaders Fund (Project No. 35552), and Ontario Ministry of Research, Innovation and Science ORF-Small Infrastructure (Project No. 35552). The University of Waterloo’s QNFCF facility was used for this work. We further acknowledge financial support from the DFG through IRTG 2803 2D-MATURE (Project No. 461605777)

The authors have no conflicts to disclose.

André Maas: Data curation (equal); Investigation (equal); Methodology (equal); Visualization (lead); Writing – original draft (equal); Writing – review & editing (equal). Kissan Mistry: Investigation (equal). Stephan Sleziona: Investigation (equal). Abdullah H. Alshehri: Investigation (equal). Hatameh Asgarimoghaddam: Investigation (equal). Kevin P. Musselman: Conceptualization (equal); Funding acquisition (equal); Project administration (equal); Writing – review & editing (equal). Marika Schleberger: Funding acquisition (equal); Project administration (equal); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
C.
Chen
,
Z.
Feng
,
Y.
Feng
,
Y.
Yue
,
C.
Qin
,
D.
Zhang
, and
W.
Feng
,
ACS Appl. Mater. Interfaces
8
,
19004
(
2016
).
2.
W. J.
Yu
,
Y.
Liu
,
H.
Zhou
,
A.
Yin
,
Z.
Li
,
Y.
Huang
, and
X.
Duan
,
Nat. Nanotechnol.
8
,
952
(
2013
).
5.
M.
Amani
,
R. A.
Burke
,
R. M.
Proie
, and
M.
Dubey
,
Nanotechnology
26
,
115202
(
2015
).
6.
K. P.
Musselman
,
C. F.
Uzoma
, and
M. S.
Miller
,
Chem. Mater.
28
,
8443
(
2016
).
7.
H.
Bergeron
,
V. K.
Sangwan
,
J. J.
McMorrow
,
G. P.
Campbell
,
I.
Balla
,
X.
Liu
,
M. J.
Bedzyk
,
T. J.
Marks
, and
M. C.
Hersam
,
Appl. Phys. Lett.
110
,
053101
(
2017
).
8.
K. F.
Mak
,
C.
Lee
,
J.
Hone
,
J.
Shan
, and
T. F.
Heinz
,
Phys. Rev. Lett.
105
,
136805
(
2010
).
9.
A.
Kuc
,
N.
Zibouche
, and
T.
Heine
,
Phys. Rev. B
83
,
4059
(
2011
).
10.
A.
Splendiani
,
L.
Sun
,
Y.
Zhang
,
T.
Li
,
J.
Kim
,
C.-Y.
Chim
,
G.
Galli
, and
F.
Wang
,
Nano Lett.
10
,
1271
(
2010
).
11.
E.
Pollmann
,
J. M.
Morbec
,
L.
Madauß
,
L.
Bröckers
,
P.
Kratzer
, and
M.
Schleberger
,
J. Phys. Chem. C
124
,
2689
(
2020
).
12.
E.
Pollmann
,
L.
Madauß
,
S.
Schumacher
,
U.
Kumar
,
F.
Heuvel
,
C.
vom Ende
,
S.
Yilmaz
,
S.
Güngörmüs
, and
M.
Schleberger
,
Nanotechnology
31
,
505604
(
2020
).
14.
R. W.
Johnson
,
A.
Hultqvist
, and
S. F.
Bent
,
Mater. Today
17
,
236
(
2014
).
15.
H.
Kattelus
,
M.
Ylilammi
,
J.
Saarilahti
,
J.
Antson
, and
S.
Lindfors
,
Thin Solid Films
225
,
296
(
1993
).
16.
K.
Kukli
,
J. Electrochem. Soc.
144
,
300
(
1997
).
17.
K.
Kukli
,
J.
Ihanus
,
M.
Ritala
, and
M.
Leskela
,
Appl. Phys. Lett.
68
,
3737
(
1996
).
18.
J. A.
Liddle
and
G. M.
Gallatin
,
ACS Nano
10
,
2995
(
2016
).
19.
K.
Mistry
,
A.
Jones
,
M.
Kao
,
T. W.-K.
Yeow
,
M.
Yavuz
, and
K. P.
Musselman
,
Nano Express
1
,
010045
(
2020
).
20.
P.
Poodt
,
A.
Lankhorst
,
F.
Roozeboom
,
K.
Spee
,
D.
Maas
, and
A.
Vermeer
,
Adv. Mater.
22
,
3564
(
2010
).
21.
K. P.
Musselman
,
S.
Albert-Seifried
,
R. L. Z.
Hoye
,
A.
Sadhanala
,
D.
Muñoz-Rojas
,
J. L.
MacManus-Driscoll
, and
R. H.
Friend
,
Adv. Funct. Mater.
24
,
3562
(
2014
).
22.
Y.
Ievskaya
,
R. L. Z.
Hoye
,
A.
Sadhanala
,
K. P.
Musselman
, and
J. L.
MacManus-Driscoll
,
J. Vis. Exp.
113
,
e53501
(
2016
).
23.
R. L. Z.
Hoye
,
K. P.
Musselman
,
M. R.
Chua
,
A.
Sadhanala
,
R. D.
Raninga
,
J. L.
MacManus-Driscoll
,
R. H.
Friend
, and
D.
Credgington
,
J. Mater. Chem. C
3
,
9327
(
2015
).
24.
R. L. Z.
Hoye
et al.,
Adv. Mater.
27
,
1414
(
2015
).
25.
D.
Di
et al.,
Adv. Mater.
29
,
1605987
(
2017
).
26.
A. H.
Alshehri
,
K.
Mistry
,
V. H.
Nguyen
,
K. H.
Ibrahim
,
D.
Muñoz-Rojas
,
M.
Yavuz
, and
K. P.
Musselman
,
Adv. Funct. Mater.
29
,
1805533
(
2019
).
27.
K. P.
Musselman
,
D.
Muñoz-Rojas
,
R. L. Z.
Hoye
,
H.
Sun
,
S.-L.
Sahonta
,
E.
Croft
,
M. L.
Böhm
,
C.
Ducati
, and
J. L.
MacManus-Driscoll
,
Nanoscale Horiz.
2
,
110
(
2017
).
28.
H.
Asgarimoghaddam
,
Q.
Chen
,
F.
Ye
,
A.
Shahin
,
B.
Song
, and
K. P.
Musselman
,
Nanotechnology
35
,
095401
(
2023
).
29.
H.
Asgarimoghaddam
,
Q.
Chen
,
F.
Ye
,
A.
Shahin
,
B.
Song
, and
K. P.
Musselman
, “
Zinc aluminum oxide encapsulation layers for perovskite solar cells deposited using spatial atomic layer deposition
,”
Small Methods
(in press).
30.
N.
Wakabayashi
,
H. G.
Smith
, and
R. M.
Nicklow
,
Phys. Rev. B
12
(2),
659
(
1975
).
32.
E.
Pollmann
,
L.
Madauß
,
V.
Zeuner
, and
M.
Schleberger
, “Strain in single-layer Mo S 2 flakes grown by chemical vapor deposition,” in
Encyclopedia of Interfacial Chemistry
(Elsevier, Amsterdam, 2018), pp. 338–343.
33.
S. E.
Panasci
,
E.
Schilirò
,
G.
Greco
,
M.
Cannas
,
F. M.
Gelardi
,
S.
Agnello
,
F.
Roccaforte
, and
F.
Giannazzo
,
ACS Appl. Mater. Interfaces
13
,
31248
(
2021
).
34.
E.
Schilirò
,
R.
Lo Nigro
,
S. E.
Panasci
,
S.
Agnello
,
M.
Cannas
,
F. M.
Gelardi
,
F.
Roccaforte
, and
F.
Giannazzo
,
Adv. Mater. Interfaces
8
,
2101117
(
2021
).
35.
See the supplementary material online for the roughness of the untreated Si O 2 surface. The deposition of amorphous A l 2 O 3 on Si O 2 via conventional ALD as comparison with AP-SALD and the subsequent growth of Mo S 2. Furthermore, low temperature ( T = 77 K) PL measurements of the nano-crystallite area are shown.
36.
C.
Lee
,
H.
Yan
,
L. E.
Brus
,
T. F.
Heinz
,
J.
Hone
, and
S.
Ryu
,
ACS Nano
4
,
2695
(
2010
).
37.
H.
Zhang
,
Y.
Wan
,
Y.
Ma
,
W.
Wang
,
Y.
Wang
, and
L.
Dai
,
Appl. Phys. Lett.
107
,
101904
(
2015
).
38.
D.
Lloyd
,
X.
Liu
,
J. W.
Christopher
,
L.
Cantley
,
A.
Wadehra
,
B. L.
Kim
,
B. B.
Goldberg
,
A. K.
Swan
, and
J. S.
Bunch
,
Nano Lett.
16
,
5836
(
2016
).
39.
E. Z.
Xu
,
H. M.
Liu
,
K.
Park
,
Z.
Li
,
Y.
Losovyj
,
M.
Starr
,
M.
Werbianskyj
,
H. A.
Fertig
, and
S. X.
Zhang
,
Nanoscale
9
,
3576
(
2017
).
40.
K. F.
Mak
,
K.
He
,
C.
Lee
,
G. H.
Lee
,
J.
Hone
,
T. F.
Heinz
, and
J.
Shan
,
Nat. Mater.
12
,
207
(
2013
).
41.
B.
Radisavljevic
,
A.
Radenovic
,
J.
Brivio
,
V.
Giacometti
, and
A.
Kis
,
Nat. Nanotechnol.
6
,
147
(
2011
).
42.
E.
Pollmann
,
S.
Sleziona
,
T.
Foller
,
U.
Hagemann
,
C.
Gorynski
,
O.
Petri
,
L.
Madauß
,
L.
Breuer
, and
M.
Schleberger
,
ACS Omega
6
,
15929
(
2021
).
43.
X.
Xin
,
Y.
Song
,
S.
Guo
,
Y.
Zhang
,
B.
Wang
,
Y.
Wang
, and
X.
Li
,
J. Alloys Compd.
829
,
154635
(
2020
).
44.
J. S.
Lee
,
C.-S.
Park
,
T. Y.
Kim
,
Y. S.
Kim
, and
E. K.
Kim
,
Nanomaterials
9
,
1278
(
2019
).
45.
N.
Kumar
,
R.
Tomar
,
N.
Wadehra
,
M. M.
Devi
,
B.
Prakash
, and
S.
Chakraverty
,
Cryst. Res. Technol.
53
,
1800002
(
2018
).
46.
L.
Madauß
et al.,
Nanoscale
10
,
22908
(
2018
).
48.
G.
Li
et al.,
J. Am. Chem. Soc.
138
,
16632
(
2016
).
49.
E.
Pollmann
,
A.
Maas
,
D.
Marnold
,
A.
Hucht
,
R.-M.
Neubieser
,
M.
Stief
,
L.
Madauß
, and
M.
Schleberger
,
2D Mater.
9
,
035001
(
2022
).
50.
K.
Kang
,
S.
Xie
,
L.
Huang
,
Y.
Han
,
P. Y.
Huang
,
K. F.
Mak
,
C.-J.
Kim
,
D.
Muller
, and
J.
Park
,
Nature
520
,
656
(
2015
).
51.
52.
Z. M.
Hanafi
,
M. A.
Khilla
, and
M. H.
Askar
,
Thermochim. Acta
45
,
221
(
1981
).

Supplementary Material