Vacuum Ultraviolet-Enhanced Oxidation—A Route to the Atomic Layer Etching of Palladium Metal

Abstract

Low-temperature, plasma-free atomic layer etching (ALE) of Pd 0 is explored. A vacuum ultraviolet (VUV) light source (115 < λ < 400 nm) is used in conjunction with a controlled O 2 gas exposure to produce PdO x at 100 °C. The amount of PdO x that forms is dependent on the duration of coexposure of O 2 at 1 Torr and VUV irradiation. A minimum coexposure time of 1 min is required to partially oxidize 2 nm Pd while 3 min is required for 20 nm Pd films, which is verified in situ using X-ray photoelectron spectroscopy (XPS). Formic acid vapor is used to complete the etch cycle, which does not etch Pd 0 and only removes the PdO x that forms. Repeated etch cycles on a ∼20 nm Pd thin film yield an etch rate of 2.81 Å/cycle, which is characterized in situ using XPS and ex situ using X-ray reflectivity. The morphology of the Pd 0 surface does not change appreciably with etching. The only observed effect of repeated ALE cycles is a reduction in the roughness of the Pd film, which is measured using atomic force microscopy.

A Vacuum Ultraviolet-Enhanced Oxidation Mechanism for Pd: Near-Surface Oxidation for Atomic Layer Etching

“…Therefore, the surface consisting of Pd(111) and Pd(100) is a reasonable representation of the Pd surface. The use of Pd(111) and Pd(100) also allows for comparison between simulations reported herein and literature. − However, it is important to note that the surfaces investigated herein are not atomically smooth, and as we have shown, Pd surface roughness decreases as ALE is performed, and the rate of oxidation depends on surface area . This implies that high-order facets are likely present for the initial ALE cycles and oxidation occurs more rapidly.…”

“…The Pd surface that results after repeated etch cycles is also investigated, as VUV-enhanced oxidation is studied with ALE in mind. ALE is performed as described elsewhere . After 5, 10, and 20 ALE cycles, Pd(111) and Pd(200) are the only features detectable by XRD (not shown).…”

“…ALE of Pd has been elusive thus far as it is difficult to limit oxidation to the topmost layer of Pd (∼2–4 Å) using conventional O 2 plasma exposures. We have demonstrated that co-exposure of vacuum ultraviolet (VUV) light and O 2 gas leads to PdO x formation wherein the Pd 3d X-ray photoelectron (XP) feature for Pd 2+ at 336.7 eV develops as oxidation proceeds . After sufficiently short exposures to VUV at low enough temperatures, PdO x formation is limited to the near-surface region.…”

“…VUV irradiation of O 2 yields singlet atomic oxygen, O­( 1 D), triplet atomic oxygen, O­( 3 P), and O 3 , according to eqs – where production of O 3 proceeds via recombination in the presence of a third body, M . Thermal treatment of Pd with O 2 does not form PdO x for substrate temperatures up to 200 °C . Therefore, the formation of PdO x at temperatures T < 200 °C under VUV and O 2 co-exposure is attributed to the presence of O 3 , O­( 1 D), and O­( 3 P).…”

“…Gas-phase reactions in the stratosphere at comparable pressures to the experimental setup used herein (1 Torr) predict O 3 to have a lifetime ( viz ., the abundance/total loss rate) of approximately 1000 s, while atomic O has a lifetime of about 0.002 s. , Measurements of O 3 (96% pure) decomposition to O 2 (2O 3 → 3O 2 ) indicate a half-life ranging between 9 × 10 3 and 1.3 × 10 5 s at 300 K and 4 Torr; the half-life increases due to passivation of surfaces from continued O 3 exposure . The lifetime of atomic O produced by VUV/O 2 co-exposure is unknown, as is the relative abundance of O and O 3 ; however, it is reasonable to assume that O is short-lived, relative to O 3 . We hypothesize that a gas composition consisting of mainly atomic O, O 3 , and O 2 is incident on a Pd sample when the surface is illuminated by the VUV light source, and a gas composed of O 3 and O 2 is incident on a Pd sample shadowed from the VUV light source.…”