Influence of the substrate type on CVD grown homoepitaxial diamond layer quality by cross sectional TEM and CL analysis

Abstract

a b s t r a c t a r t i c l e i n f oTo assess diamond-based semiconducting devices, a reduction of point defect levels and an accurate control of doping are required as well as the control of layer thickness. Among the analyses required to improve such parameters, cross sectional studies should take importance in the near future. The present contribution shows how FIB (focused ion beam) preparations followed by electron microscopy related techniques as TEM or CL allowed to perform analysis versus depth in the layer, doping and point defect levels. Three samples grown along the same week in the same machine with identical growth conditions but on different substrates (CVD-IIIa (110) oriented, CVD-optical grade (100) oriented and a HPHT-Ib (100) oriented) are studied. Even though A-band is observed by CL, no dislocation is observed by CTEM. Point defect type and level are shown to substantially change with respect to the substrate type as well as the boron doping levels that vary within an order of magnitude. H3 present in the epilayer grown on HPHT type of substrate is replaced by T1 and NE3 point defects for epilayers grown on the CVD type one. An increase of excitonic transitions through LO phonons is also shown to take place near the surface while only TO ones are detected deeper in the epilayer. Such results highlight the importance of choosing the correct substrate.

Identification of Dislocations in Synthetic Chemically Vapor Deposited Diamond Single Crystals

“…Dislocations can be revealed by plasma etching, , or the stress that surrounds them can be imaged in a birefringence microscope. , Low temperature cathodoluminescence (CL) can also allow visualizing dislocations, since they behave as efficient recombination centers for free excitons . Other heavier techniques can allow a direct identification of the type and Burgers vector of dislocations: such as X-ray topography with a synchrotron beam (XRT) , or transmission electron microscopy (TEM) . However, they rely on bulky equipments and cannot be carried out in a systematic fashion.…”

Boron concentration profiling by high angle annular dark field-scanning transmission electron microscopy in homoepitaxial δ-doped diamond layers

“…Lett. 103, 042104 (2013) in SIMS measurements, [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25] leading to systematic overestimates of the real width, similar to the case of silicon. 26 If this is the case, it is legitimate to ask how the layer thickness should be estimated from broadened SIMS profiles.…”

“…11 Among these techniques, cathodoluminescence (CL) 12,13 is clearly the most sensitive 14 while ensuring a high spatial resolution since cross sectional analysis can be also carried out on FIB preparations. 15,16 However, the signal is too weak in heavily doped diamond to image the spatial distribution of dopants. The need of an imaging method able to quantify boron content and layer thickness becomes obvious when d-doped devices are being developed.…”

Boron‐doped Homoepitaxial Diamond CVD from Microwave Plasma‐Activated Ethanol/Trimethyl Borate/Hydrogen Mixtures

“…The ratio between the intensities of the bound exciton emission related to the boron impurity BETO and the free exciton FETO peak I BETO / I FETO remains constant through the surface of the samples under study, which indicates the homogeneity of film properties. Using the relationship with empirical coefficient of 2.4 × 10 17 cm −3 between I BETO / I FETO in CL spectra measured at 80 K, and acceptor boron‐concentration,25 the doping level of sample #347 is estimated to be ∼1 × 10 19 cm −1 . Two broad mid‐gap emission peaks were observed throughout the surfaces of the films.…”

“…Two broad mid‐gap emission peaks were observed throughout the surfaces of the films. The more intense one at around 510 nm is typical for boron‐doped material and related to recombination via donor‐acceptor pairs,26, 27 and the less intense one at 410 nm is a band‐A emission attributed to extended defects such as dislocations28 and other defects containing sp 2 ‐phase 25, 29. The A‐band emission originates partly from the substrate due to secondary photoluminescence.…”