Challenges have arisen in selecting suitable candidates for interconnects and metal contacts due to the exponential increase in metal resistivity at scaled pitches. Molybdenum (Mo) has emerged as a promising alternative to the traditional metals such as copper or tungsten owing to its low electrical resistivity and electron mean free path. In this study, we investigated the formation of a molybdenum film grown by thermal atomic layer deposition (ALD) using a MoO2Cl2 solid precursor and H2 and NH3 gases as the reducing agents. A molybdenum nitride film served as the seed layer on a SiO2 substrate before molybdenum film deposition. The analysis focused on the film's phase, morphology, chemical bonding states, and resistivity across various thicknesses. X-ray diffraction (XRD) confirmed the presence of polycrystalline BCC planes. Our analyses confirmed the successful growth of the molybdenum metal thin film, which, at a thickness of 10 nm, exhibited a record-low resistivity of approximately 13 μΩ cm.

As device pitches continue to shrink, MOSFET devices have encountered significant challenges in selecting suitable materials for interconnects and contact metals. This is primarily due to the exponential increase in metal resistivity at scaled pitches, which adversely affects device performances and reliability by increasing signal delay and power consumptions. Traditionally, tungsten (W) and copper (Cu) have been the predominant choices for metal lines because of their low resistivity above their electron mean free path (EMFP) dimensions. While the bulk resistivities of tungsten and copper are as low as 5.28 and 1.67 μΩcm,1 respectively, their resistivity sharply increases as the required film thickness becomes ultrathin, falling below their EMFP.2–4 Therefore, the fabrication of devices requires alternative metals with short EMFP and lower resistivity at tight pitches to achieve low figures of merit, the product of resistivity, and EMFP(ρo × λ).5 This leads to the emergence of molybdenum,4,6,7 cobalt,8–12 and ruthenium13–16 as promising candidates. Among these next-generation metals, molybdenum (Mo) stands out due to its lowest bulk resistivity, high thermal stability, high conductivity, and low coefficient of thermal expansion.5 Compared to Cu and W, Mo metal possesses the advantage of maintaining low resistivity, particularly with thickness reduction. Table I compiles the resistivity data of traditional materials, such as Cu and W, and Mo at approximately 10 nm thicknesses using the ALD method. In examining the properties of polycrystalline metal films, key determinants of low resistivity include the film's thickness, the purity of the metal, and grain size.22–24 For refractory metals such as Mo, the impact of grain size on reducing resistivity is more significant than the effect of film thickness, especially when compared to Cu. This difference is attributed to the weaker grain boundary scattering and stronger surface scattering in thinner Cu films, alongside Mo's significantly shorter electron mean free path (EMFP) compared to that of copper.4,25

TABLE I.

Previously reported results for resistivity near a 10 nm thickness using ALD deposition methods.

MaterialThickness (nm)Resistivity (μΩ cm)Reference
Cu 12 20.1 17  
Cu 12 19.1 18  
15 19  
Mo 24 18.6 20  
Mo 17.5 16.5 21  
MaterialThickness (nm)Resistivity (μΩ cm)Reference
Cu 12 20.1 17  
Cu 12 19.1 18  
15 19  
Mo 24 18.6 20  
Mo 17.5 16.5 21  

Mo film deposition for metallization applications has been extensively studied using physical vapor deposition (PVD)6,7,24,26,27 and chemical vapor deposition (CVD).28,29 However, these processes are considered unsuitable for scaled-down devices since high conformality is required for high aspect ratio patterns. To resolve these issues, Mo film deposition by ALD process is strongly suggested due to its excellent step coverage and smooth surfaces.20,21,30 Yet, challenges remain in depositing high-Mo content films without oxygen and thus in achieving low-resistivity films. The properties of Mo films by the ALD have not met expectations due to the formation of byproducts and impurities within the films, resulting from most of the widely used precursors; using Mo precursors like MoCl5, MoOCl4, Mo(CO)6, and other Mo-based precursors, such as those with halogen or fluorine, led to insufficient reduction and resulted in producing Mo compound films. Mo compound films such as Mo–oxide31–33 and Mo–carbide films34,35 are often produced during the ALD process. In contrast, using MoO2Cl2 as a Mo source precursor enables the deposition of a high-Mo composition film with sufficient reduction with H2 reducing agent by forming HCl and H2O.20,21 It also makes it possible to deposit the molybdenum nitride (MoN) film by using a different reducing agent, NH3, in the same chamber. In this investigation, the MoN film served as an essential seed layer for Mo film deposition on SiO2 on the Si substrate, as its presence is critical for initiating the Mo deposition process,21 in addition to offering high thermal stability.36 

In this study, we investigate the properties of a Mo metal thin film deposited by the thermal ALD process with MoO2Cl2 as the Mo source precursor and H2 gas as the reducing agent. Successful ALD growth of high-purity Mo thin films was verified through impurity content analysis by x-ray photoelectron spectroscopy (XPS) and phase analysis by x-ray diffraction (XRD). In addition, the resistivity of the Mo metal film grown by ALD is demonstrated for films below 10 nm thickness.

A Mo thin film was deposited on a SiO2 100 nm-thick film on a 12-in. Si wafer at 650 °C chamber temperature using a thermal ALD instrument manufactured by Hanwha Corporation. The SiO2 layer was thermally produced by LPCVD on top of a Si substrate. During full pumping, the base pressure was approximately 0.01 Torr, and the working pressure was 11–20 Torr during Mo film deposition. The precursor used for depositing the Mo and MoN films was MoO2Cl2, and the reactants used to deposit them were H2 and NH3, respectively. Since MoO2Cl2 is a solid-phase precursor, maintaining the reproducibility of the process with consistent source usage was challenging.37 The canister was maintained at a temperature of 120 °C during the process. Ar gas was used as the carrier gas to deliver the source to the chamber and as the purge gas. Prior to the actual deposition, a dummy process was conducted for 1 h.

The film thickness and uniformity were measured using x-ray reflectometry (XRR), and the film thickness was confirmed by analyzing a high-resolution transmission electron microscopy (HR-TEM) image. The phase and the crystallite size of the film were examined through grazing incidence x-ray diffraction (GIXRD). The film surface roughness and morphology were measured using an atomic force microscope (AFM, NX-10 model from Park Systems) in the noncontact mode within a 1 × 1 μm2 size range and confirmed with a TEM image. The grain size of the film, as observed in a TEM image, was determined using image j software. The chemical binding energy and composition of the Mo film were measured through high-resolution x-ray photoelectron spectroscopy (HR-XPS, Thermo Fisher Scientific K-Alpha+) with sputtering. The analysis mode was in monoatomic, and the x-ray source was 200 μm diameters. The film sheet resistance was measured by a four-point probe (4PP).

In this study, the following four steps, comprising one deposition cycle, were repeated to deposit the Mo film under continuous Ar gas flow: Mo precursor dose with 150 SCCM Ar carrier gas for 0.1 s, a purge of extra precursor molecules and byproducts with 3000 SCCM Ar gas for 2 s, reactant dose with 15 000 SCCM H2 gas for 9 s, and a byproduct purge with 3000 SCCM Ar gas for 1 s. The MoN film was deposited using the same four steps in one deposition cycle, except that NH3 was used instead of H2 as the reducing agent. The final film stack was the Mo/MoN/SiO2/Si substrate.

Considering the essential role of the MoN film as a seed layer for subsequent Mo film deposition, the feasibility of its deposition was initially confirmed. Figure 1 depicts the well-deposited 30 nm MoN film, as evidenced by the analyses of thickness measurements from TEM images, planes observed by GIXRD, and atomic concentration examined through XPS depth profile. A MoN film was deposited using NH3 as the reducing agent [Eq. (1)]. The MoN film phase and planes were confirmed through the GIXRD analysis with reference to the International Centre for Diffractional Data Powder Diffraction File (ICDD PDF) for hexagonal Mo0.82N #01-075-1006 in blue solid lines and ICDD PDF cubic MoN0.506 #01-076-3075 in blue dashed lines, as shown in Fig. 1(b). As illustrated in Fig. 1(c), the MoN film composition showed that it comprised a Mo/N ratio of 5.5/1,
Mo O 2 C l 2 + 2 N H 3 MoN + 2 HCl + 3 H 2 O .
(1)
FIG. 1.

Characteristics of the 30 nm MoN film deposited by ALD, including (a) thickness, (b) crystalline planes, and (c) atomic concentration.

FIG. 1.

Characteristics of the 30 nm MoN film deposited by ALD, including (a) thickness, (b) crystalline planes, and (c) atomic concentration.

Close modal

After validating the feasibility of MoN film deposition, a 4 nm MoN film, which will be used as a seed layer for the Mo film, was deposited on the SiO2 substrate as shown in TEM image in Figs. 2(a) and 2(b). The thickness of the MoN film as a seed layer and the process temperature were adapted from Lee et al.21 The MoN seed layer is required due to its ability to moderate the electronegativity differences between O (3.44) and Mo (2.16), with Si (1.90) and N (3.04). Incorporating N into the film structure reduces the electronegativity disparity, thereby enhancing the film's chemical stability and making MoN an ideal seed layer. Figure 2(c) shows the smooth morphology of the 4 nm MoN film with root mean square (RMS) value of 0.3 nm. A 2 nm-thick Mo film was deposited on the MoN film to see whether the Mo film could maintain the surface roughness of the underlying MoN substrate; the RMS value of the 2 nm Mo film was 0.3 nm, similar to that of the 4 nm as-deposited MoN film.

FIG. 2.

TEM image of the 4 nm MoN film at (a) 100 and (b) 500 K resolutions, and (c) AFM image of the 4 nm MoN film.

FIG. 2.

TEM image of the 4 nm MoN film at (a) 100 and (b) 500 K resolutions, and (c) AFM image of the 4 nm MoN film.

Close modal
To confirm the self-limiting reaction of our ALD process, we measured the thickness of the film over the deposition cycle. As shown in Fig. 3, the thickness linearly increased as the cycle increased; the GPC of the Mo film was ∼0.64 Å/cycle and the GPC of the MoN film was ∼0.95 Å/cycle. The film was deposited with the following chemical reaction:
Mo O 2 C l 2 + H 2 Mo + 2 HCl + H 2 O .
(2)
FIG. 3.

Molybdenum film and molybdenum nitride film thicknesses as a function of the cycle.

FIG. 3.

Molybdenum film and molybdenum nitride film thicknesses as a function of the cycle.

Close modal

The targeted Mo film thicknesses were chosen to be 6, 10, and 20 nm since the Mo film resistivity was predicted to be lower than those of the Cu and W films below 10 nm thickness.4 The deposited film thicknesses were confirmed with a TEM analysis in Fig. 4, and the TEM images proved the continuous film deposition with a 100 K resolution [Figs. 4(a)4(c)], and the 500 K resolution TEM image in Figs. 4(d) and 4(e) confirmed the polycrystalline growth and film thicknesses. The film deposition uniformity was examined through the XRR thickness analysis on a 12-in. wafer at five points. The film thickness nonuniformity on the 12-in. wafer was below 6% for the 6 and 10 nm Mo films. A 16 nm-thick Mo film was used to observe the atomic concentration of the film confirming the deposition of the pure-Mo film as shown in Fig. 5. The MoN layer was deemed too thin to be detected by XPS, so in order to confirm the proper deposition of the MoN film for the seed layer before Mo film deposition, the previously discussed Figs. 1 and 2 displayed its MoN concentration and thickness. Furthermore, the observation of the MoN phase being detected in the subsequent GIXRD graph (Fig. 7) corroborates the existence of the MoN seed layer.

FIG. 4.

TEM images of 6 (a), 10 (b), and 16 nm (c) thick Mo films on the MoN seed layer. High-resolution image of each site for (d)–(f), respectively.

FIG. 4.

TEM images of 6 (a), 10 (b), and 16 nm (c) thick Mo films on the MoN seed layer. High-resolution image of each site for (d)–(f), respectively.

Close modal
FIG. 5.

Atomic concentration of the 16 nm-thick Mo film observed by XPS depth profile.

FIG. 5.

Atomic concentration of the 16 nm-thick Mo film observed by XPS depth profile.

Close modal
FIG. 6.

Chemical bonding states of the as-deposited 16 nm Mo films by XPS measurements: (a) Mo3d, (b) O1s, and (c) Cl2p.

FIG. 6.

Chemical bonding states of the as-deposited 16 nm Mo films by XPS measurements: (a) Mo3d, (b) O1s, and (c) Cl2p.

Close modal
FIG. 7.

XRD 2θ diffraction peaks for polycrystalline Mo film. 6 (a), 10 (b), and 16 nm Mo (c) on the 4 nm MoN on SiO2 substrate.

FIG. 7.

XRD 2θ diffraction peaks for polycrystalline Mo film. 6 (a), 10 (b), and 16 nm Mo (c) on the 4 nm MoN on SiO2 substrate.

Close modal

Figure 6 depicts the fitted chemical bonding energy of the Mo film as determined by the XPS surface analysis. The Mo3d, O1s, and Cl2p peaks were fitted by rearranging the C1s value to 284.8 eV. As shown in the Mo 3d XPS spectrum from Fig. 6(a), the Mo3d5/2 peak had a binding energy of 228.0 eV with a standard deviation of 3.08 eV. The result agreed with the standard sample splitting between Mo 3d3/2 and Mo 3d5/2 (3.1 eV) and the standard metallic Mo0 binding energy (Mo 3d5/2 at 227.9 eV). Also, O1s and Cl2p peaks were not detected in the XPS analysis. This indicates that the high-purity Mo film was successfully grown with ALD. Using MoN as the seed layer for the Mo film and depositing it at a high temperature induced a shorter nucleation delay and led to a film with a low impurity content.

As shown in Fig. 7, polycrystalline growth of the Mo film with standard BCC Mo planes was observed through GIXRD with the thickness variations of 6, 10, and 16 nm. As observed in the GIXRD analysis, the standard BCC Mo planes were detected, including Mo (110) at 40.7°, Mo (200) at 58.8°, and Mo (211) at 73.9°. The structure factor of the peak position was calculated using Bragg's law. When the molybdenum films were not thick enough, x-ray measurements penetrate through the Mo layer to the MoN seed layer, detecting the MoN planes alongside the standard Mo planes of Mo (110), Mo (200), Mo (211), and Mo (220). Thicker films have a higher number of atoms in the film, leading to increased x-ray scattering; the intensity measured was higher for the 16 nm Mo film than those for the other thinner films. The planes from the XRD measurement matched with the body-centered-cubic Mo ICDD PDF#00-004-0809 shown in orange solid lines. The crystallite sizes were calculated using the information from the XRD graphs. The full width at half maxima (FWHM) and the 2θ peak position from the graph were substituted into the Scherrer equation: FWHM = K λ L Cos θ. The positions of FWHM and the 2θ peak are depicted in Fig. 7(c), where L represents the average crystallite size, θ is the peak position (2θ/2) in radians, λ represents the wavelength of the x ray (Cu-K-alpha, set to 0.1548 Å), K is typically set to 0.94 for most cases involving spherical crystallites, and FWHM in radians. The comparison of the same planes extracted from the three film thickness samples showed that the crystallite size trend went upward as the Mo film thickness increased (Table II).

TABLE II.

Crystallite size of Mo films calculated from FWHM and 2θ peak positions; Mo with film thicknesses of 6, 10, and 16 for Mo (110), (200), and (211) planes.

Mo film thickness (nm)Crystallite size (nm)
Mo planes
(110)(200)(211)
8.3 4.3 8.3 
10 11.6 12.3 12.0 
16 14.1 14.0 12.6 
Mo film thickness (nm)Crystallite size (nm)
Mo planes
(110)(200)(211)
8.3 4.3 8.3 
10 11.6 12.3 12.0 
16 14.1 14.0 12.6 

To investigate the effect of the carrier Ar gas and H2 gas flow rates on crystallite size, films from Lee et al.21 were used for comparison. These were deposited using the same reactor under the same conditions as in this work, except for those certain specific parameters. To assess the impact of the carrier Ar gas, films of similar thickness but deposited under different flow rates were compared. A 32 nm-thick Mo film with a 100 SCCM Ar flow rate was juxtaposed with a 30 nm-thick Mo film deposited using a 50 SCCM carrier Ar gas flow; both films contained approximately 95% Mo. The crystallite size of the films increased as the carrier Ar flow increased from 50 to 100 SCCM, as organized in Table III. Additionally, a 16 nm-thick Mo film referenced from Lee et al.,21 which utilized 10 000 SCCM of the reactant H2 gas, was compared to the 16 nm-thick Mo film used in this study, which employed 15 000 SCCM H2 gas, to find the effect of H2 flow rate on the film properties. The crystallite size of the film grown under 10 000 SCCM H2 was smaller than that in the films using 15 000 SCCM H2, as can be compared by consulting Tables II and III. The ligands were removed by H2 because ligands themselves were hydrogenated, and the metal atoms self-agglomerated;38,39 Vos et al.11 showed that an increasing H2 ratio affects the O impurity content by causing it to decrease and, thus, decreasing the resistivity. The above results indicate that sufficient Ar carrier gas and H2 reactant gases are necessary to reduce ligands from the precursors, enabling the growth of the Mo metal in a favorable plane. Also, Lee et al.21 confirmed that the resistivity decreased as the H2 flow rate increased, even at a similar thickness, and that the resistivity decreased as the carrier Ar gas flow rate increased. Additionally, adequate thermal energy was provided for hydrogen and ligand migration to break the bond from its ligand.40 By depositing at a temperature higher than 450 °C reported in the study of van der Zouw et al.,20 the active atom diffusion was enhanced during the nucleation and fostered the grain growth from a thermodynamics perspective, leading to lower film resistivity, which will be discussed later in this study.

TABLE III.

Crystallite size of Mo films from Lee et al.21 with different carrier Ar gas flows and reactant H2 gas flows for the Mo (110), (200), and (211) planes.

Crystallite size (nm)
Mo planes
(110)(200)(211)
Carrier Ar 50 SCCM 16.4 17.3 15.6 
100 SCCM 17 17.4 17.4 
Reactant H2 10 000 SCCM 7.3 7.8 10.17 
Crystallite size (nm)
Mo planes
(110)(200)(211)
Carrier Ar 50 SCCM 16.4 17.3 15.6 
100 SCCM 17 17.4 17.4 
Reactant H2 10 000 SCCM 7.3 7.8 10.17 

Figure 8 shows the plan-view TEM image of (a) 6, (b) 10, and (c) 16 nm Mo films' surfaces and the grain size distributions. Considering the grain size, Fig. 8(a) is taken at 100 K resolution, while Figs. 8(b) and 8(c) are taken at 40 K resolution. The mean grain size of the films increased from 5.4 to 20.3 nm and then to 43.3 nm, respectively. This increasing trend in grain size with increasing thickness correlates with the patterns observed in crystallite size, while the differences in grain size relative to thickness exhibit a more pronounced divergence. The surface morphology and the RMS values of the Mo thin films were obtained using the three-dimensional image obtained from the AFM, as shown in Fig. 9. The surface roughness of the Mo film increased as the film thickness increases; RMS values were approximately 0.5, 0.6, and 0.7 for 6, 10, and 16 nm-thick Mo films, respectively. The increasing surface roughness of the film exhibits a notable correlation between the grain size as shown in Fig. 8 and the crystallite size previously calculated from FWHM in GIXRD measurements as listed in Table II.

FIG. 8.

Plan-view TEM image of 6 (a), 10 (b), and 16 nm Mo (c) on the 4 nm MoN on SiO2 substrate to measure the grain size.

FIG. 8.

Plan-view TEM image of 6 (a), 10 (b), and 16 nm Mo (c) on the 4 nm MoN on SiO2 substrate to measure the grain size.

Close modal
FIG. 9.

AFM surface roughness and morphologies of ALD Mo films: 6 (a), 10 (b), and 16 nm (c) thick Mo films on the MoN seed layer.

FIG. 9.

AFM surface roughness and morphologies of ALD Mo films: 6 (a), 10 (b), and 16 nm (c) thick Mo films on the MoN seed layer.

Close modal

Figure 10 shows the resistivity of the Mo films with varying thicknesses measured by 4PP, represented in error bars, and Table IV compares the average sheet resistances and resistivities of the Mo films. The resistivity was calculated by multiplying the sheet resistance measured by 4PP with the thickness measured by TEM. In accordance with previous research on metal resistivity, sheet resistance and resistivity decreased as the film thickness increased, approaching the bulk metal characteristics. The bulk Mo resistivity is 5.34 μΩ⋅cm.1 The 10 nm molybdenum thin film deposited in this study notably achieved the lowest resistivity of 13 μΩ cm at 10 nm. This accomplishment represents the most favorable value among the various ALD-deposited molybdenum metal thin films reported in published journals,20,21,41 and this was achieved without additional postprocessing procedures such as heat treatment.

FIG. 10.

Error bars of Mo films’ resistivity obtained from 4PP measurements on the SiO2 substrates.

FIG. 10.

Error bars of Mo films’ resistivity obtained from 4PP measurements on the SiO2 substrates.

Close modal
TABLE IV.

Average sheet resistance and resistivity of Mo films calculated from the 4PP measurement. Thicknesses of Mo films were 6, 10, and 16 nm.

Mo film thicknessAverage sheet resistanceResistivity
(nm)(Ω/□)(μΩ cm)
55.1 33.0 
10 13.0 13.0 
16 7.5 11.9 
Mo film thicknessAverage sheet resistanceResistivity
(nm)(Ω/□)(μΩ cm)
55.1 33.0 
10 13.0 13.0 
16 7.5 11.9 

In summary, low resistivity Mo films were deposited via ALD using MoO2Cl2 as a Mo source precursor and H2 as the reducing agent, achieving a noteworthy low resistivity of 13 μΩ cm at 10 nm thickness. The study revealed that films deposited with higher H2 and Ar flow rates exhibited larger grain sizes and lower resistivities, attributed to the effective reduction of ligands by the H2 reactant dose from the Mo precursor. And as the thickness increased, the resistivity decreased with increased grain sizes and surface roughness. Furthermore, the Mo film grown by ALD was confirmed to exhibit approximately 100% high purity Mo content based on XPS analysis and XRD analysis, by having polycrystalline planes of BCC structured Mo film without any impurities of Mo–O bonding. While the results obtained in this study indicate the potential applicability of the Mo film as an alternative metal to Cu or W and the potential usage of MoO2Cl2 precursor, maintaining the reproducibility of the solid precursor is still a challenge. Therefore, developing Mo liquid precursor and broadening the Mo precursor should be pursued in the future. This approach will enhance the versatility and applicability of Mo film deposition processes.

This work was supported by Hanwha Corporation (Vacuum Equipment R&D Division, Semiconductor Research Center) with film preparation using Hanwha thermal ALD equipment. This research was supported by the Next-Generation Intelligence Semiconductor Program (No. 2022M3F3A2A01072215) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT.

The authors have no conflicts to disclose.

So Young Kim: Conceptualization (equal); Data curation (equal); Investigation (equal); Methodology (equal); Writing – original draft (equal). Chunghee Jo: Formal analysis (lead); Methodology (supporting). Hyerin Shin: Software (supporting); Visualization (lead). Dongmin Yoon: Conceptualization (supporting); Software (lead). Donghyuk Shin: Investigation (equal). Min-ho Cheon: Formal analysis (supporting). Kyu-beom Lee: Visualization (supporting). Dong-won Seo: Data curation (supporting). Jae-wook Choi: Resources (supporting). Heungsoo Park: Resources (equal); Supervision (equal); Validation (lead); Writing – review & editing (equal). Dae-Hong Ko: Funding acquisition (lead); Project administration (lead); Supervision (equal); Writing – review & editing (equal).

The data that support the findings of this study are available within the article and its supplementary material

1.
W. M.
Haynes
,
D. R.
Lide
, and
T. J.
Bruno
,
CRC Handbook of Chemistry and Physics
, 95th ed. (
CRC
, Boca Raton,
2014
).
2.
J. F.
Guillaumond
,
L.
Arnaud
,
T.
Mourier
,
M.
Fayolle
,
O.
Pesci
, and
G.
Reimbold
,
2003 IEEE International Interconnect Technology Conference
, 4 June 2004, Burlingame, CA (
IEEE
,
2003
), pp.
132
134
.
3.
T.
Sun
,
B.
Yao
,
A.
Warren
,
V.
Kumar
,
K.
Barmak
, and
K. R.
Coffey
,
2003 IEEE International Interconnect Technology Conference
, 1–4 June 2008, Burlingame, CA (
IEEE
,
2008
), pp.
141
143
.
4.
5.
D.
Gall
,
J. Appl. Phys.
119
,
085101
(
2016
).
6.
D.
Tierno
,
K.
Croes
,
A.
Ajaykumar
,
S.
Ramesh
,
G.
Van Den Bosch
, and
M.
Rosmeulen
,
2021 IEEE International Reliability Physics Symposium
, 21–25 March 2001, Monterey, CA (
IEEE
,
2021
), pp.
1
6
.
7.
D.
Tierno
,
M.
Hosseini
,
M.
Van Der Veen
,
A.
Dangol
,
K.
Croes
,
S.
Demuynck
,
Z.
Tokei
,
E. D.
Litta
, and
N.
Horiguchi
,
2021 IEEE International Interconnect Technology Conference
, 6–9 July 2021, Kyoto, Japan (
IEEE
,
2021
), pp.
1
3
.
8.
J.
Yoon
,
H.-B.-R.
Lee
,
D.
Kim
,
T.
Cheon
,
S.-H.
Kim
, and
H.
Kim
,
J. Electrochem. Soc.
158
,
1179
(
2011
).
9.
J. P.
Klesko
,
M. M.
Kerrigan
, and
C. H.
Winter
,
Chem. Mater.
28
,
700
(
2016
).
10.
M. M.
Kerrigan
,
J. P.
Klesko
,
S. M.
Rupich
,
C. L.
Dezelah
,
R. K.
Kanjolia
,
Y. J.
Chabal
, and
C. H.
Winter
,
J. Chem. Phys.
146
,
052813
(
2017
).
11.
M. F. J.
Vos
,
G.
Van Straaten
,
W. M. M. E.
Kessels
, and
A. J. M.
Mackus
,
J. Phys. Chem. C
122
,
22519
(
2018
).
12.
B.
Zhu
,
Z. J.
Ding
,
X.
Wu
,
W. J.
Liu
,
D. W.
Zhang
, and
S. J.
Ding
,
Nanoscale Res. Lett.
14
,
76
(
2019
).
13.
T.
Aaltonen
,
P.
Alén
,
M.
Ritala
, and
M.
Leskelä
,
Chem. Vap. Depos.
9
,
45
(
2003
).
14.
S.
Cwik
,
K. N.
Woods
,
M. J.
Saly
,
T. J.
Knisley
, and
C. H.
Winter
,
J. Vac. Sci. Technol. A
38
,
012402
(
2020
).
15.
H. M.
Kim
,
J. H.
Lee
,
S. H.
Lee
,
R.
Harada
,
T.
Shigetomi
,
S.
Lee
,
T.
Tsugawa
,
B.
Shong
, and
J. S.
Park
,
Chem. Mater.
33
,
4353
(
2021
).
16.
O.
Varela Pedreira
,
M.
Stucchi
,
A.
Gupta
,
V. V.
Vega Gonzalez
,
M.
Van Der Veen
,
S.
Lariviere
,
C. J.
Wilson
,
Z.
Tokei
, and
K.
Croes
,
2020 IEEE International Reliability Physics Symposium
, 28 April–30 May 2020 (
Dallas, TX
,
2020
), pp.
1
7
.
17.
L.
Wu
and
E.
Eisenbraun
,
J. Vac. Sci. Technol. B
25
,
2581
(
2007
).
18.
L.
Wu
,
W.
Zeng
, and
E.
Eisenbraun
,
ECS Trans.
11
,
67
(
2007
).
19.
M.
Yang
,
A. A. I.
Aarnink
,
J.
Schmitz
, and
A. Y.
Kovalgin
,
Thin Solid Films
646
,
199
(
2018
).
20.
K.
van der Zouw
,
B. Y.
van der Wel
,
A. A. I.
Aarnink
,
R. A. M.
Wolters
,
D. J.
Gravesteijn
, and
A. Y.
Kovalgin
,
J. Vac. Sci. Technol. A
41
,
052402
(
2023
).
21.
B. J.
Lee
,
K. B.
Lee
,
M. H.
Cheon
,
D. W.
Seo
, and
J. W.
Choi
,
Coatings
13
,
1070
(
2023
).
22.
J. W.
Lim
and
M.
Isshiki
,
J. Appl. Phys.
99
,
094909
(
2006
).
23.
D.
Choi
et al,
J. Vac. Sci. Technol. A
29
,
051512
(
2011
).
24.
C. H.
Lee
,
F. G.
Guo
, and
C. C.
Chu
,
Chin. J. Phys.
50
,
311
(
2012
).
25.
S.
Dutta
,
K.
Sankaran
,
K.
Moors
,
G.
Pourtois
,
S.
Van Elshocht
,
J.
Bömmels
,
W.
Vandervorst
,
Z.
Tokei
, and
C.
Adelmann
,
J. Appl. Phys.
122
,
025107
(
2017
).
26.
H.
Ahn
,
D.
Lee
, and
Y.
Um
,
Appl. Sci. Converg. Technol.
26
,
11
(
2017
).
27.
D.
Choi
,
Microelectron. Eng.
216
,
111084
(
2019
).
28.
A. N.
Lubnin
,
R. V.
Lebedev
,
V. I.
Lad’yanov
,
B. E.
Pushkarev
,
R. R.
Faizullin
,
L. K.
Baldaev
, and
S. Y.
Treshchev
,
Prot. Met. Phys. Chem. Surf.
58
,
76
(
2022
).
29.
A. H.
El-Hoshy
,
J. Electrochem. Soc.
118
,
2028
(
1971
).
30.
B. S.
Lim
,
A.
Rahtu
, and
R. G.
Gordon
,
Nat. Mater.
2
,
749
(
2003
).
31.
J. A.
Drayton
,
D. D.
Williams
,
R. M.
Geisthardt
,
C. L.
Cramer
,
J. D.
Williams
, and
J. R.
Sites
,
J. Vac. Sci. Technol. A
33
,
041201
(
2015
).
32.
S.
Lee
,
H. L.
Yang
,
B.
Kim
,
J.
Lee
,
H.
Lim
, and
J. S.
Park
,
Ceram. Int.
49
,
5292
(
2023
).
33.
M.
Diskus
,
O.
Nilsen
, and
H.
Fjellvåg
,
J. Mater. Chem.
21
,
705
(
2011
).
34.
A.
Mameli
,
M. J. M.
Merkx
,
B.
Karasulu
,
F.
Roozeboom
,
W. E. M. M.
Kessels
, and
A. J. M.
MacKus
,
ACS Nano
11
,
9303
(
2017
).
35.
M. J.
Ha
,
H.
Kim
,
J. H.
Choi
,
M.
Kim
,
W. H.
Kim
,
T. J.
Park
,
B.
Shong
, and
J. H.
Ahn
,
Chem. Mater.
34
,
2576
(
2022
).
36.
L.
Stöber
,
J. P.
Konrath
,
V.
Haberl
,
F.
Patocka
,
M.
Schneider
, and
U.
Schmid
,
J. Vac. Sci. Technol. A
34
,
021513
(
2016
).
38.
L.
Alig
,
M.
Fritz
, and
S.
Schneider
,
Chem. Rev.
119
,
2681
(
2019
).
39.
M.
Dai
,
J.
Kwon
,
M. D.
Halls
,
R. G.
Gordon
, and
Y. J.
Chabal
,
Langmuir
26
,
3911
(
2010
).
40.
M.
Bergschneider
,
N.
Ashburn
,
X.
Lang
,
A. C.
Kummel
, and
K.
Cho
,
MRS Adv.
8
,
768
(
2023
).
41.
D.
Seghete
,
G. B.
Rayner
,
A. S.
Cavanagh
,
V. R.
Anderson
, and
S. M.
George
,
Chem. Mater.
23
,
1668
(
2011
).