Do you know the relevant conditions for in situ heating experiments in the TEM?

In many fields of materials science in situ experiments in the TEM have proven to be paramountcomplements to conventional static investigations. These types of experiments allow for areal-time observation of the sample’s response to a specific external stimulus. In terms of theimpact of thermal energy (i.e. heating) the advent of MEMS (microelectromechanical systems)based heating holders for the TEM has greatly enhanced the choice of experimental possibilities.Due to the small size of MEMS-based heating elements, rapid temperature changes and a highstability can be achieved. In addition, the amount of emitted infrared radiation is sufficiently smallso in situ nanoanalysis using X-ray spectrometry (EDXS) can be used up to 1000 °C [1].What remains a challenge, however, is the knowledge of the relevant experimental conditions. Forinstance, the sample temperature is often only estimated from the heating holder readout. Thistemperature might significantly deviate from the local temperature of the area of interest.Therefore, different methods are available for determining the sample’s temperature moreaccurately like electron diffraction or plasmon energies [e.g. 2, 3]. On the other hand, theperformance of the X-ray detector will change when exposed to a large amount of infraredradiation, which might render quantitative analysis results questionable [4].In this contribution we will focus on the performance of the EDXS detector and the quality of theobtained spectra when exposed to a heated sample. As shown in figure 1 the energy resolution ofan SDD EDXS detector degrades significantly for temperatures above 500 °C. This can beattributed to the increasing emission of infrared radiation from the hot parts of the sample and theheating holder. When plotting the measured infrared intensities in an Arrhenius plot, a goodcorrespondence to calculations can be found.As an application example, the in situ investigation of a vertical cavity surface emitting laser(VCSEL) device is also shown. In terms of reliability one critical part in oxide defined VCSELs isthe aperture, which might undergo secondary oxidation and furthermore cause stress in thestructure and degrade the device performance. Such an aperture was prepared in plan-view and isshown in fig. 2 [5]. During in situ heating experiments the generation of so-called dark line defectscould be observed. Preparations for in situ biasing experiments are currently ongoing and we hopeto be able to show first experimental results.