Subscriber access provided by UNIV OF NEWCASTLE
Surfaces, Interfaces, and Applications
Vapor Phase Cleaning and Corrosion Inhibition of Copper Films by Ethanol and Heterocyclic Amines Luis Fabián Peña, Jean-Francois Veyan, Michael A Todd, Agnes Derecskei-Kovacs, and Yves J. Chabal ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b13438 • Publication Date (Web): 18 Oct 2018 Downloaded from http://pubs.acs.org on October 21, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Vapor Phase Cleaning and Corrosion Inhibition of Copper Films by Ethanol and Heterocyclic Amines Luis Fabián Peña†*, Jean-Francois Veyan†, Michael A. Todd‡, Agnes Derecskei-Kovacs‡, Yves J. Chabal† †
Department of Materials Science & Engineering, The University of Texas at Dallas, Richardson, Texas 75080, United Sates ‡ Versum Materials, Inc., 8555 South River Parkway, Tempe, AZ 85284
Abstract Cleaning and passivation of metal surfaces are necessary steps for selective film deposition processes that are attractive for some microelectronic applications (e.g., fully aligned vias or selfaligned contacts). For copper, there is limited knowledge about the mechanisms of the copper oxide reduction process and subsequent passivation layer formation reactions. We have investigated the in-situ cleaning (i.e., oxidation and reduction by vapor-phase species) and passivation of chemical-mechanical polishing (CMP)-prepared Cu films in an effort to derive the mechanisms associated with selectively tailoring the surface chemistry. By monitoring the interaction of vapor-phase ethanol with surface species generated after ozone cleaning at 300 °C, we find that the optimum procedure to remove these species and avoid byproduct redeposition is to use atomic layer deposition (ALD)-like binary cycles of ethanol and N2, with active pumping. We have further explored passivation of clean Cu using benzotriazole and 2,2'-bipyridine in an ALD environment. Both molecules chemisorb on clean Cu in an upright orientation, with respect to the metal surface at temperatures higher than the melting point of the organic inhibitors (100 ≤ T < 300 °C). Both molecules desorb without decomposition from clean Cu above 300 °C but not from Cu2O. Previous studies related to the passivation of Cu surfaces using heterocyclic amines have focused on solution-based or ultra-high vacuum applications of the passivation molecules onto single crystalline Cu samples. The present work explores more industrially relevant vaporphase passivation of CMP-cleaned, electroplated Cu samples using ALD-like processing conditions and an in-situ vapor-phase clean. Keywords: copper, cuprous oxide, ethanol, ALD, benzotriazole, ozone, heterocyclic amines, corrosion inhibition 1. Introduction Copper is the most widely used material in semiconductor interconnect technology due to its superior conductivity and resistance to electromigration. Although several methods of depositing copper or copper oxide (e.g., conformally on nanostructured surfaces with high aspect ratio) have been developed using atomic layer deposition (ALD),1-6 a reduction step is required to convert deposited copper oxide into metallic copper (Cu(0)). Promising results have been demonstrated to selectively reduce copper oxide at ambient pressure conditions and at temperatures below 310 ˚C using carbon monoxide,7 forming gas,8 acetic acid,9 formic acid,10 and ethanol (EtOH)10-11 as reducing agents. Unfortunately, the reaction pathways and surface chemistry of reduction with
ACS Paragon Plus Environment
ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
alcohols and carboxylic acids are poorly understood due to the lack of in-situ experimental data.12 Interest in controlling Cu deposition, reduction and protection arises from the challenges associated with scaling sub-10 nm device dimensions in manufacturing that require ‘bottom-up’ patterning methods.13 For example, at these dimensions, lithography (a ‘top-down’ process) becomes prohibitively expensive due to the need for multiple mask passes per layer and the pattern placement errors that result from them.14 This has led the efforts to develop a ‘bottom-up’ approach, particularly for applications such as fully aligned vias and self-aligned contacts for future device generations.15-17 Area-selective deposition (ASD) of materials has been proposed as a way to circumvent this issue and reduce the cost of lithography.15 ASD requires selective surface functionalization after a patterning step to alter the reactivity of specific surfaces for further deposition.18-19 ASD in conjunction with ALD, known as area-selective atomic layer deposition (AS-ALD), is rapidly gaining interest because of its potential for eliminating edge placement errors by enabling self-aligned fabrication schemes for next-generation nanoelectronics.20-21 ASD can be achieved using inherent selectivity22 or organic inhibitors23-24 during the deposition process. However, a lack of fundamental understanding and control of the surface chemistry regarding the formation and stability of passivation layers under industrially relevant processing conditions is currently limiting the demonstration of manufacturable selective deposition processes. Specifically, control over the surface reactivity of Cu(0) is of interest to hinder corrosion and to provide selectivity for subsequent ALD processing. Such protection may be afforded by the chemisorption of a monolayer of selected molecules on Cu provided that such a passivation scheme can effectively prevent undesired film deposition on the passivated surface. Heterocyclic amines are particularly attractive as passivating agents owing to their stability at relatively high temperature (T > 100 °C) and the ability of the nitrogen heteroatoms to coordinate to transition metals. For example, since the 1950s, benzotriazole (BTAH) has been used as a copper corrosion inhibitor25-26 and recently has been introduced into chemical mechanical polishing (CMP) slurries to control etching at various pH levels.27-28 Extensive investigations using x-ray photoelectron spectroscopy (XPS),29-31 spectroscopic ellipsometry (SE),29 high resolution electron energy loss (HREELS),32 reflection absorption infrared spectroscopy (RAIRS),26, 31, 33-34 surface-enhanced Raman spectroscopy (SERS)35-36 and scanning tunneling microscopy (STM)33 have explored the adsorption and corrosion inhibition properties of BTAH on Cu. Density functional theory (DFT) calculations have enabled useful insights into mechanisms suggested by these experimental investigations.33-34, 37-41 Similarly, 2,2’-Bipyridyl (BPD) has been examined on Cu, Ag and Au metal electrodes, although not to the same extent as BTAH.4246 It is believed that the corrosion inhibition properties of heterocyclic amines correlate to the bonding and orientation ascertained on metal surfaces, i.e., adsorption can take place in a variety of configurations, on different facets and surfaces.39 Despite extensive work to validate this hypothesis, there are several contradicting accounts in literature that preclude a firm conclusion.47-48 The ambiguity is mostly due to the neglect of effects arising from solution
ACS Paragon Plus Environment
Page 2 of 23
Page 3 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
components and competing surface reactions in studies seeking to understand the interaction of BTAH with copper in solution. In contrast, vapor-phase BTAH interaction with Cu eliminates adverse effects arising from solution-based investigations. Furthermore, if BTAH undergoes dehydrogenation of the labile hydrogen (from N-H) on the surface of Cu, which provides a pathway for a Cu-N covalent bond, the chemisorption and orientation of BTA- can be determined and validated.32, 49-51 Moreover, studies on polycrystalline Cu films at ambient pressures, e.g., in an ALD environment, are most relevant for industrial applications, yet remains unexplored. For the semiconductor industry, vapor-phase deposition of organic monolayers on solid substrates, as passivating adlayers, is an area of particular interest to inhibit corrosion and selectively tailor ALD growth. In this work, we use in-situ reflection absorption infrared spectroscopy (RAIRS) to investigate each step of the vapor-phase oxidation (O3) and reduction (EtOH) processes, with the objective of optimizing the cleaning procedure of polycrystalline Cu metal surfaces in an ALD environment. Additionally, we characterize the chemisorption and corrosion inhibition properties of vapor-phase benzotriazole (BTAH) and 2,2’-Bipyridyl (BPD) for clean Cu metal surfaces. We demonstrate a process based on ALD-like binary cycles of ethanol and N2, with active pumping to remove surface impurities and avoid byproduct redeposition during EtOH vapor cleaning and selective passivation of clean Cu and Cu2O with BTAH and BPD in an ALD environment. 2. Experimental Procedures 2.1. Materials Rectangular samples (4.5×2 cm2) were cut from a 300 mm wafer (supplied by Versum Materials, Inc.) with 1.5 μm thick Cu film that had been chemically cleaned and polished to a mirror-like finish using a proprietary process. Benzotriazole (BTAH) [Product B11400], 2,2'-bipyridine (BPD) [Product D216305] and Ethanol (200 proof, anhydrous) [Product 459836] were obtained from Sigma-Aldrich; molecular structure drawings of the organic inhibitors are presented in Scheme 1. All chemicals were loaded separately into stainless-steel ampoules under an inert atmosphere to prevent oxygen contamination. The BTAH and BPD ampoules were heated slightly higher than the melting temperature (melting point of BTAH and BPD is ca. 97-99 °C and 70-73 °C, respectively) of the chemicals to obtain a higher vapor pressure; chemicals were vapor drawn into the reactor (base pressure of ALD reactor Pb = 5×10-5 Torr). Several precautionary procedures had to be taken to perform these studies to avoid crosscontamination between experiments since BTAH has a high melting point and ambient pressures were used. The primary one involved performing long N2 purges (~1 hour) post BTAH exposures; a series of ozone and baking treatments of the reactor were also completed prior to each experiment. Similar precautionary steps were implemented for the investigation with BPD. Despite these technical challenges, measurements were successfully collected and analyzed without concern of previous contamination.
ACS Paragon Plus Environment
ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Scheme 1: Structural formulae of Benzotriazole (BTAH), Benzotriazole anion (BTA-) and 2,2’Bipyridyl (BPD). BTAH
-
BTA
BPD
2.2. Sample preparation and Ozone (O3) cleaning Initial sample preparation included visual inspection of the Cu sample surfaces for particulate impurities and physical damage (i.e., scratches). Sample surfaces were then swept with compressed N2 gas to remove surface particulates prior to introduction into the ALD reactor. Samples, without any prior solvent cleaning, were then exposed to O3 (16% wt. O3/cm3 concentration supplied by an In-USA Ozone generator) at 100 °C for 10 min using a flow rate of 300 standard cubic centimeters per minute (cm3/min) and a pressure of 9.6 Torr. 2.3. In-Situ Characterization: Reflection Absorption Infrared Spectroscopy (RAIRS) Surface characterization was performed using a home built ALD reactor with in-situ FTIR capability performed in reflection using a grazing incidence geometry (85° incidence from the surface normal), as necessary for metallic surfaces, see Figure 1. Infrared absorption spectra are recorded with a Thermo Nicolet 6700 spectrometer (frequency range of KBr beamsplitter: 4000 400 cm-1) equipped with a Globar IR source and mercury cadmium telluride (MCT/B) detector. Three loops consisting of 500 scans each (at a resolution of 4 cm-1) are averaged for data processing. Note that reflectance spectra are plotted so that vibrational absorptions appear as “negative” peaks (dips); reflectance is defined as (ΔR/ Ro) with ΔR = Ro - Ra, where Ro and Ra denote the reflectivities of clean and adsorbate-covered surfaces, respectively.52-55
Figure 1: A simplified top view schematic of the ALD reactor with in-situ RAIRS at grazing incidence (85° from the surface normal) used for investigating surface reactions on reflective surfaces. Reactor comprises of a chamber, maintained at 110 °C, containing a heated substrate stage holder (T = 30 to 600 °C); a precursor delivery system for EtOH, vapor-drawn into a line that was purged with N2; and a pumping system with a turbo pump and an automated pressure controller valve to control pumping speed.
ACS Paragon Plus Environment
Page 4 of 23
Page 5 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Gases in
Baratron
IR beam
Manual gate valve -protect the KBr windows
MCT-B detector Sample holder (stage)
MCT/B IR detector
IR interferometer
2.4. Ex-Situ Characterization: X-ray photoelectron spectroscopy (XPS) XPS measurements were performed using a Physical Electronics Quantum 2000 ESCA Microprobe equipped with a hemispherical analyzer, a monochromatic Al Kα (1486.6 eV) radiation source, detection takeoff angle of 45° and a pass energy of 11.75 eV. The transfer time from the ALD reactor to the XPS chamber was between 5 and 10 minutes, during which time some surface contamination (mostly physisorbed hydrocarbons) took place on the surface. The oxygen (1s, 530.3 eV) for CuO and adventitious carbon (1s, 284.5 eV) were taken as internal standards for binding energy calibration.56 3. Results 3.1. Surface preparation for EtOH etching: O3 treatment and 300 °C anneal The spectra to characterize the cleaning process are obtained using in-situ IR spectroscopy by taking a reference of the starting Cu surface, then examining changes after each processing step, always performing the IR measurements at the same temperature as that of the reference surface. The results after O3 exposure at 100 °C are shown as black spectra in Figure 2(a) and (b), and as red spectra after subsequent annealing to 300 °C for 30 min, with all measurements performed with the sample at 100 °C. The formation of carbonates (-CO3), CO and CuO is clearly observed, with characteristic signatures at 1498, 1311 and 1030 cm-1 for carbonates, at 2231 and 1901 cm-1 for CO, and at 580 cm-1 for CuO. The mode assignments are summarized in Table 1. After 300 °C annealing, the blue spectra in Figure 2(c) referenced to the original surface (prior to any treatment) show that the carbonates created by O3 treatment are removed. In addition, CuO is reduced to Cu2O (645 cm–1) and the appearance CO and CO2 species is observed (modes at 2192 and 2335 cm–1). The relatively high vibrational frequency measured for CO is attributed to single coordination (i.e., linear bonding) adsorption;57 note that the CO stretch frequency is very sensitive to intermolecular repulsion and dipole-dipole coupling, as illustrated in inset in Figure 2(c). Figure 2: Differential IR reflectance measurements taken following O3 treatment at 100 °C (spectra in black, referenced to the as-introduced Cu sample) and after sample annealing to 300 °C (spectra in red,
ACS Paragon Plus Environment
ACS Applied Materials & Interfaces
referenced to the O3 treated Cu surface) with an emphasis on (a) carbonate species formed and desorbed from surface, (b) CuO transformation to Cu2O by thermal activation and (c) residual surface species post annealing to 300 °C (spectra in blue, referenced to the as-introduced Cu sample). All spectra collected at 100 °C, i.e., after treatment at 300 °C the sample was cooled down to 100 °C. The dashed line in (b) is to help visualize the path of the baseline. Processing conditions: 10 min O3 exposure (P = 9.6 Torr) and 30 min anneal in N2 atmosphere (P = 2 Torr) at 300 °C. Note: The spectra in the inset in Figure 2(c) illustrate the frequency shift of CO of four nominally similar experiments. The frequency variance greater than 2177 cm-1 indicate that the surface coverage ≥ half monolayer, in addition to CO2 species present. 1311
(a)
2231
(b)
1901 1030
% Reflectance %Reflectance
2
645
2
1498 2250
2000
1750
Cu2O
1500
1250
1000
750
700
650
580 CuO 600
550
500
2192
(c)
2337 1 00.0 2 1 00.0 0 9 9.98
645
9 9.96 9 9.94
2
%Refl ectance
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 23
9 9.92 9 9.90 9 9.88 9 9.86 9 9.84
CO
9 9.82 9 9.80 9 9.78 2 300
2 250
2 200
2 150
Cu2O
2 100
W ave num be rs (c m-1 )
2226 Blueshift 2177 2500
2250
2000
1750
1500
1250
1000
750
500
Wavenumber (cm Wavenumber (cm) -1) -1
Table 1. Assignments of Measured IR Vibrational Bands (in cm-1) of Surface Species on Cu Oxide from Literature Values (range) Mode CO2 (physisorbed)
Exptl. 2337
CO
2231
Single co-ordination (θ>.5)
CO
2192
Lit. 2380, 2320 2234, 2230 2194
ref 58
57, 59
57, 59
Single co-ordination (θ
