Tissue Chip Imaging with Scanning Electron Microscopy

Abigail Mullen

University of Rochester, Department of Chemistry
Opt 407 Spring 2025

1. Introduction

Tissue chips simulating the human blood brain barrier and the immune system are studied in the McGrath Lab. In this project, I imaged samples provided the McGrath Lab to how the endothelial and pericytic cells interacted with the membrane of the chip.

The following methods were used used in this project: ethanol ladder dehydration, critical point dehydration, gold sputter coating, Scanning Electron Microscopy (Secondary Electron and InLens secondary electron), and focused ion beam (FIB)

2. Methods

The sample were fixed before I received them from the McGrath lab. Apon receiving them, they were rinsed with DI water then place in a solution of 50% ethanol water. The ethanol ladder dehydration sequenced continued with 60%, 70%, 80%, 90%, 100% ethanol solutions for an hour with each step then the sample was left in fresh ethanol overnight.

The sample were dried next with the Tousimis PVT-3B Critical Point Dryer. The sample chamber of the critical point dryer filled with 100% ethanol. The sample was placed in the sample holder, then transferred the into the ethanol. The chamber temperature was decreased from room temperature to about 0°C within a couple minutes. The ethanol was purged at the same rate as the CO₂ inlet rate. The excess ethanol was collected until only CO₂ gas was being expelled. The chamber was then filled with liquid CO₂. Afterwards. the temperature and pressure were increased to approximately 31°C and 1,070 psi respectively. This state was maintained for four minutes. The chamber was decompressed slowly at around 100 psi per minute. Once finished, the sample was removed.

The sample stub was placed on the stage inside the bell jar if the Denton Vacuum DESK-II DC Sputtering System. The chamber pressure was decreased to approximately 100 mTorr. The chamber was then backfilled with argon gas to raise the pressure to approximately 300 mTorr, then decreased to 150 mTorr. Repeat this gas cycling three times then pumped down about 50 mTorr. A current of 15 mA was run through the cathode for 60 seconds depositing gold layer of approximately 60 Å.

3. Results & Discussion

3.1 Surface of the chip

Although every sample was meticulously dried following procedures to preserve the bioligical stuctures, the images show the structure have degraded. This can be seen in the images as the extracellular matrix have peeled away from the chip and pulling away from the nucluei.

Figure 1 - Secondary electron images at an accerating volatage of 5kV at 30x magnification (left) and 149x magnification (right).
The sample is a nanoporous silicon nitride membrane,100nm thick, with a 1 μm layer of endothelial cells on one side and 1 μm layer of pericytic cells on the other side. The endothelial and pericytic cells touch through the openings in the microporous portions of the membrane. This chip replicates the barrier in the brain that permits small molecules in and out. Figure 1 shows the overall structure of the chip. The nuclei of the brain endothelial cells appear as the dark spots on the membrane.

Figure 2 - Secondary electron Images at an accerating volatage of 5kV at 28.20Kx magnification (left) and 18.66 Kx magnification (right).

Figure 3 - Secondary electron images at an accerating volatage of 5kV at 4.34Kx magnification (left) and 1.06Kx magnification (right).
The images with false color highlight the different features on the sample. The blue is the extra cellular membrane. The orange is the nucleus of brain endothelial cells. The pink is some type of bacteria cell.

Figure 4 - InLens images of bacteria cells at an accerating volatage of 15kV at 5.13Kx magnification (left) and 27.06Kx magnification (right).

3.1 Cross-Section

The chip was cracked with a pair of tweezers in an attempt to get a cross-section image of the sample. The resulting images shown in Figure 5 show that the extracellular membrane away from the membrane. A second attempt was made with using FIB. First, a layer of platinum was deposited then the sample was ablated. This attempt did not yield better results as seen in Figure 6. No endothelial or pericytic cells were seen interacting through the membrane.

Figure 5 - Secondary electron images at an accerating volatage of 5kV at 42.63Kx magnification (left), 22.84Kx magnification (middle), and 5.55Kx magnification (right).

Figure 6 - Secondary electron image and InLens images of the sample with a layer of platinum deposited (left) and after FIB (right).

4. Conclusion

The sample was not properly dehydrated making it difficult to see key features. In the future, HMDS drying could be used dry the sample instead of critical point drying

Acknowledgments

I would like to express my thanks to Michelle Trempel from the McGrath Lab for providing my samples. Additionally, Sean O’Neill for helping me with the FIB.