High resolution SIMS depth profiling of nanolayers

TL;DR

This paper introduces a refined TOF SIMS depth profiling method called gentleDB, achieving near-atomic resolution of nanolayer structures and revealing detailed interface features.

Contribution

The paper presents a novel gentleDB approach combining low energy and pulsed ion beams, significantly improving depth resolution in SIMS analysis of nanolayers.

Findings

Achieved depth resolution confined to about 2 monolayers.

Demonstrated ability to resolve 6 nm thick nanolayers and interface roughness.

Validated the method on MgO/ZnO nanolayer stacks.

Abstract

We report results of high-resolution TOF SIMS (time of flight secondary ion mass spectrometry) depth profiling experiments on a nanolayered structure, a stack of 16 alternating MgO and ZnO ~5.5 nm layers grown on a Si substrate by atomic layer deposition. The measurements were performed using a newly developed approach implementing a low energy direct current normally incident Ar+ ion beam for sample material removal by sputtering (250 eV and 500 eV energy), in combination with a pulsed 5 keV Ar+ ion beam at 60{\deg} incidence for TOF SIMS analysis. By this optimized arrangement, a noticeably improved version of known dual-beam (DB) approach to TOF SIMS depth profiling is introduced, which can be called gentleDB. We apply the mixing-roughness-information model to detailed analysis of experimental results. It reveals that the gentleDB approach allows ultimate depth resolution by confining the ion beam mixing length to about 2 monolayers. This corresponds to the escape depth of secondary ions, the fundamental depth resolution limitation in SIMS. Other parameters deduced from the measured depth profiles indicate that a single layer thickness equals to 6 nm so that "flat" layer thickness d is of 3 nm and interfacial roughness {\sigma} is of 1.5 nm thus yielding d+2\bullet{\sigma}=6 nm. In essence, we have demonstrated that the gentleDB TOF SIMS depth profiling with noble gas ion beams is capable of revealing structural features of a stack of nanolayers, resolving its original surface and estimating the roughness of interlayer interfaces, which is difficult to obtain by traditional approaches.