Self-consistent absorption correction for quantitative energy- dispersive X-ray spectroscopy of InGaN layers in analytical transmission electron microscopy

Self‐consistent absorption correction for quantifying very noisy X‐ray maps: group III nitride nanowires as an example

“…For quantification, we first calculate the Ga K/L ratio for each nanowire and compare it to simulation for thin films (Walther and Wang, ). The Ga K/L ratio from the raw maps for the 217 nm thin nanowire is only 1.125, which is actually smaller than the minimum of 1.24 predicted for an only a few nanometres thin GaN specimen from Monte Carlo simulations (Walther and Wang, ).…”

“…For the established methods of X‐ray absorption correction in analytical transmission electron microscopy (TEM) estimates of specimen thickness and density (Cliff and Lorimer, ), mass‐thickness (August and Wernisch, ) or probe current (Watanabe et al ., ) are needed. We recently developed a self‐consistent absorption correction method based on effective sensitivity factors (denoted as k *‐factors) based on Monte Carlo simulations and K/L intensity ratios of at least one heavier element directly measured from the spectrum to be quantified (Walther and Wang, ; ). If maps have very low counts they are susceptible to strong shot noise and the background cannot be estimated from an integer‐valued map because it will be below a single count on average.…”

“…We recently developed a self-consistent absorption correction method based on Correspondence to: Thomas Walther, Department of Electronic & Electrical Engineering, University of Sheffield, Mappin Street, Sheffield, S1 3JD, South Yorkshire, UK. Tel: +44 114 222 5891; e-mail: t.walther@sheffield.ac.uk effective sensitivity factors (denoted as k*-factors) based on Monte Carlo simulations and K/L intensity ratios of at least one heavier element directly measured from the spectrum to be quantified (Walther and Wang, 2015;. If maps have very low counts they are susceptible to strong shot noise and the background cannot be estimated from an integer-valued map because it will be below a single count on average.…”

Effective absorption correction for energy dispersive X‐ray mapping in a scanning transmission electron microscope: analysing the local indium distribution in rough samples of InGaN alloy layers

“…The self‐consistent X‐ray absorption correction method employing k*‐ factors, as defined (Walther & Wang, ) and tested for individual spectra (Walther & Wang, ) in our previous InGaN work was applied to calculate the indium content for each pixel of the X‐ray elemental map. In L , Ga K and Ga L maps and several maps for background correction were recorded to construct background subtracted intensity maps of I InL , I GaK and I GaL .…”

“…Therefore, it is possible for a routine to automatically select a suitable k‐ factor based on the Ga K / L ratio measured in the same spectrum to interpolate the relative absorption of the indium L‐line and so to quantify the chemistry of a rough InGaN sample at each point without the need for a specimen geometry measurement. We have termed such absorption correction factors, if plotted versus measured K / L ratios, k *‐factors (Qiu et al ., ; Walther & Wang, ). It is worth pointing out that the k *‐factor method does also not need the electron beam current to be measured, in contrast to the zeta‐factor method, which recovers the mass‐thickness product indirectly from this by iteration (Watanabe et al ., ).…”