Large Sample Transmission Electron Microscopy

A project for OPT 407: Electron Beam Methods in Microscopy

By: Kilean Lucas

: Transmission electron microscopy (TEM) is a powerful imaging method that allows scientists and researchers to view structures that are not easily visible in techniques such as light microscopy or at higher magnification than scanning electron microscopy (SEM).

However, in order to transmit electrons through the sample, it must be around 100 nm thick. Additionally, the historical TEM sample preparations are 3 mm in diameter, which is a dimension that is too small for any practical microfluidic application.

Therefore, we purchased a used TEM sample holder and modified it such that it would accept a silicon chip that is commonly used in microfluidic applications. This gave us an unprecedented ability to image samples larger than 3 mm in diameter, which will allow us to analyze samples directly from a microlfuidic device.


: Exosomes are lipid vesicles 30 – 150 nm in diameter that are secreted by cells. These vesicles contain micro-RNA that are the ‘business cards’ of the cell of origin. By capturing these exosomes, it is possible to use them to screen for diseases such as cancer as an early detection mechanism. Using silicon nanomembranes in a microfluidic device allows for lab-on-a-chip device that can be used as an addition to a standard blood test, where we can separate blood components, isolate exosomes and then analyze them on chip to give a preliminary diagnosis. To develop a model system for the exosomes, we used 60 nm gold nanoparticles and captured them in the pores of a 100 nm thick silicon nitride nanomembrane.

Methods and Results

: A used sample holder was purchased for the FEI Tecnai G2 TEM housed in URnano and then modified from the format to fit 3 mm samples to 5.4 mm chips. The sample holder tongue is slightly wider than 5.4 mm, so it was only necessary to create a slot wide enough to fit the sample in the axial direction. The modification allows the sample to sit closer to the plane of the 3 mm sample position, which allows for easier imaging in the range of focus of the microscope.

Standard TEM imaging involves taking a single image at one sample tilt (typically zero degrees). However, due to the thinness of the sample, we can tilt the stage ± 15 degrees and generate a series of images at these various angles. Using an image processing program (FIJI) we stacked the tilted images and aligned the features in the image to generate an aligned stack. The stack was animated to create a gif that demonstrates the 3D structure of the pores. The information from the gif was rendered into an STL file, which allows us to generate a 3D printed reconstruction of the membrane.

One common method for generating topographical information in a microscopy lab is atomic force microscopy (AFM). We can use AFM to compare the depth information given by the tilt stack reconstruction which lets us see that the resolution of scanning probe microscopy for imaging small pores is lower than reconstructing them from the tilt stacked images. This is due to the tip geometry, which can’t go into the pores very far without interfering with the sides.

The most powerful ability of this technique is being able to take a sample directly from a microfluidic diagnostic device and put it into the TEM for imaging. The gold in the model system was a perfect demonstration of capture, which we expect to see with exosomes. Because gold is very dense, it does not transmit electrons very easily and thus shows up darker in TEM.


: Transmission electron microscopy is a powerful tool that allows for high resolution imaging of many different types of samples. However, due to historical size limitations, samples that are TEM compatible in terms of thickness and also practical for microfluidic applications are not viable for use in the TEM due to their larger dimensions. By modifying a TEM sample holder, we have made it possible to image samples that are larger than standard TEM systems are capable of handling, opening the doors to a new world of large area TEM imaging.


: 1. Striemer, et al. (2007) Size and charge based separation of macromolecules using ultrathin silicon membranes, Nature 445:749-53
2. DesOrmeaux, J. P., et al. (2014) Nanoporous silicon nitride membranes fabricated from porous nanocrystalline silicon templates, Nanoscale. 6:10798-10805


:I would like to thank the Center for Emerging and Innovative Sciences at the University of Rochester for their support in purchasing the TEM sample holder used in this project. I would also like to thank Brian McIntyre for his tremendous support and assistance. Without him, and his humor, this project wouldn't have been possible.

Please enter any comments, criticisms, questions, etc. below.

Your name:

Email address: