During heating experiments in the electron microscope, thermal drift and settle times pose a major problem for in situ imaging and analysis. The area of interest can drift considerably when a conventional bulk heater is used during temperature ramp.
This requires lower microscope magnifications so that the amount of drift can be visualized by operators, who can then adjust the position of the stage to account for it. Even slight changes in temperature can cause the areas of interest to rapidly drift out of view, making it challenging to acquire images and analytical data from the same area.
The Fusion heating and electrical biasing platform has been designed to significantly reduce settle time and thermal drift. Semiconductor devices called E-chips are used by Fusion that have a monolithic ceramic membrane. This membrane serves as the heater and sample support, and it has an extensively small thermal budget at just 700 x 700 µm, enabling heating rates of 1,000 °C per ms and up to 1200 °C with cooling at about the same rate.
The ceramic membrane is made of materials that have matched coefficients of thermal expansion (CTEs), which helps to reduce drift. The heating capability of Fusion is attractive for in situ heating experiments, where high stability and low drift is essential for image acquisition and analytical analysis of the same areas at atomic resolution.
In addition, the combination of fast ramp rates and settle times allow the user to obtain data quickly and more accurately, saving considerable time without affecting the data quality.
Two sets of experiments were carried out to quantify the Fusion’s drift and settle time characteristics. In the first set, drift and settle time were measured as the temperature was increased. The temperature was increased from room temperature (RT) to 500 °C using a double tilt holder, 500 °C to 1000 °C and RT to 1200 °C at 40 kX magnification in a JEOL 2010F operating at 200 kV.
This magnification was chosen so the area of interest remains in the field of view across the whole temperature excursion.
The settle time is defined as the instant the temperature was applied to when the sample was found to stop drifting. The table given below shows the amount of drift and settle time determined for each temperature excursion. Images at RT (top), and 350 °C (bottom) are shown in Figure 1.
The image given at the top was taken after the sample was seen to stop drifting at ~1 minute.
Table 1. Fusion displacement as a function of time and temperature
|RT to 500 °C
|RT to 1200 °C
|500 °C to 1200 °C
Figure 1A-B. A represents brightfield TEM image at RT. B represents bright-field TEM images at 350 °C
Fusion’s HRTEM imaging capabilities at high temperatures is described in the second set of experiments. HRTM images were acquired while the temperature was increased to 900 °C in a JEOL 2010F. Upon immediate ramping to 350 °C, resolutions less than 1.2 Å were consistently obtained.
Figure 2 depicts a high-resolution image that was taken immediately after a 4 °C/second ramp. 1.2 Å resolution is indicated by the FFT of this image (inset).
Figure 2. HRTEM image immediately after 4 °C/sec ramp.
The heating membrane of the Fusion can rapidly change temperature (millisecond timescale). This rapid change in temperature mainly promotes the drift once the temperature change is completed. Figure 2 illustrates high-resolution images that were taken after temperature was immediately changed from RT to 350 °C.
These images were taken to demonstrate how rapidly the heating membrane settles following a temperature ramp. The sample drift is opposite in direction but equal in magnitude for cooling and heating. Imaging the same area of interest was done before and after the temperature excursion, without actually changing the position of the microscope stage.
High resolution can be maintained by Fusion while heating. Fusion can also ramp at user-defined rates, in addition to stepping immediately to a specific temperature. High-resolution will not be affected if the ramp rate, or temperature jump, is kept at reasonable rates.
When the membrane heats, it expands. During expansion, the membrane will bow and the 2D projection of bowing is reflected as drift. Also, sample type can contribute to the drift and settle time characteristics.
Thermal drift and settle time can be applied to all heating experiments in the electron microscope. Images and analytical data can be obtained faster and more accurately by reducing drift and settle time during and after temperature changes.
The temperature step experiments were done at Sandia National Laboratories in Livermore, CA. The temperature ramp experiments were done at the Karlsruhe Institute of Technology in Karlsruhe, Germany.
This information has been sourced, reviewed and adapted from materials provided by Protochips.
For more information on this source, please visit Protochips.