Etching Kinetics of (100) Single Crystal Diamond Surfaces in a Hydrogen Microwave Plasma, Studied with In Situ Low‐Coherence Interferometry

Abstract

A low‐coherence interferometry (LCI) was used to measure in situ the etch rate (ER) of synthetic single crystal (SC) diamonds in H2 microwave plasma, at substrate temperatures in the broad range of 800–1370 °C. The method allows the collection of the kinetic data on a single sample without switching off the plasma. (100)‐orientated SC plates of CVD and IIa type HPHT diamond were systematically etched in pure hydrogen at pressure p = 130 Torr and microwave power density of ≈300 W cm−3. The activation energies Ea of 42 ± 5 and 32 ± 4 kCal mol−1 have been determined for the CVD and HPHT substrates. An enhanced etching rate of a subsurface defected layer with thickness of ∼1 μm or less, formed upon polishing of the samples, is revealed. Surface morphology, roughness, and the shape of etch pits produced by a selective etching of defects, were characterized with optical profilometry. CH and dimer C2 radicals were detected in the H2 plasma with optical emission spectroscopy, as a result of the diamond etching, and the emission intensity of these species was linked to the substrate etch rate.

Quality Assessment of In Situ Plasma‐Etched Diamond Surfaces for Chemical Vapor Deposition Overgrowth

“…It should be noted that these measurement results are subject to large tolerances due to the limitation of thickness determination by weighing. Nevertheless, they show a clear trend and are in a good qualitative agreement to the observation of increased initial etching of a “defected layer” by Yurov et al [ 13 ] It implies that a near‐surface region of the diamond has a substantially higher defect density than the bulk diamond, which amplifies the etching. The equilibrium etching rate of the bulk diamond after longer etching times is below 100 nm h −1 , consistent with literature data.…”

“…The equilibrium etching rate of the bulk diamond after longer etching times is below 100 nm h −1 , consistent with literature data. [ 13 ]…”

“…[ 10,11 ] Unfortunately, only a few results of systematic approaches to better understand the influence of different parameters on the pretreatment process have been published so far: The prevalent findings are that the etching rate exponentially increases with rising substrate temperature and an additional incline in the misorientation angle as well leads to an increase of the etching rate. [ 12,13 ] The importance hereby is the actual evaluation of the effectiveness of the polishing‐induced subsurface damage removal. Usually, it is indirectly performed by subsequent epitaxial overgrowth; [ 6–8,14 ] however, this method implicates various collateral factors of influence, which cannot be accounted for.…”

“…Therefore, a direct method to assess the performance of the etching pretreatment would lead to a significant improvement and simplification on its optimization. In addition to in situ low‐coherence interferometry, [ 13 ] the only applied direct evaluation approach on in situ plasma etching of diamond substrates is to determine the saturation of the etch pit density, [ 4 ] which, however, implies using a method that produces a very rough surface. Thus, this work proposes omega scans by high‐resolution X‐ray diffraction (HRXD) as a nondestructive tool to effectively optimize the in situ plasma etching pretreatment process of diamond substrates on grounds of structural quality gain.…”

Chemical Vapor Deposition Single‐Crystal Diamond: A Review

“…A pretreatment to remove defects on the substrate surface is needed as they may induce further defects, such as threading dislocations, in the growing epitaxial layers. Polishing of the substrates results in formation of subsurface damage layer the thickness of which is of the order of a few micrometers in case of mechanical polishing with “scafe.” [ 22 ] Yurov et al [ 23 ] observed an enhanced etching rate in hydrogen microwave plasma of the first surface layer for both IIa type HPHT and CVD (100)‐oriented substrates compared with the bulk diamond etching. They assigned such kinetics to the presence of the polishing‐induced damaged layer, and estimated its thickness of ≈1 μm for a CVD substrate and ≈0.2 μm for the HPHT substrate.…”