Matthew Hamblin and Aaron Hawkins, Electrical Engineering
Mass spectrometry is an important tool for analytical chemistry that allows the chemical composition of a compound to be determined. In order to do so, it separates the compound into ions, and then detects the number of ions of different masses.
The charge on a single ion is very small, making it difficult to measure. In order to overcome this, current mass spectrometers must use methods such as electron multipliers to create a cascading effect of electrons until a measurable charge is reached or cryodetectors that measure the thermal change of an ion hit at low temperatures.1 Using methods such as these require the use of expensive, bulky machinery that can provide either vacuum conditions or extreme low temperature conditions, meaning that samples must be brought back to a lab for analysis.
MOSFET (metal-oxide semiconductor field-effect transistor) devices are commonly used in memory circuitry, in which a very small charge can be placed on the device and later read. If a similar MOSFET device was created with a very low capacitance, it could be sensitive enough that a single ion or electron hitting the device would create a measurable change in the device output.
My research focuses fabricating low capacitance MOSFET devices in order to create an ion detector sensitive enough to measure a single ion without dependence on cascading or low temperatures. This would allow the development of smaller, cheaper and more portable mass spectrometers.
The most critical part of this project is obtaining MOSFET devices with low enough capacitance to detect hits from a small number of ions. Although MOSFET devices are very common and commercially available, devices with a sufficiently low capacitance for single ion detection are not readily available. Because of this, our first step was to fabricate low capacitance MOSFET devices in the BYU cleanroom.
Many MOSFET devices have been fabricated at BYU, but making the devices sensitive enough for our design required fabricating devices smaller than had ever been successfully made at BYU. Several devices could be fabricated simultaneously on a four inch silicon wafer, and each wafer was made with devices of ten, five, two, and one micron gate lengths. The smallest devices previously made had ten micron gate lengths.
As we completed several wafers and tested the devices, we found that for the first several attempts the ten micron devices occasionally worked as we had hoped, but none of the smaller devices were able to work. Eventually we determined that one of the steps used in the process, in which the wafer was placed in a furnace to grow a layer of silicon dioxide, was causing previously implanted ions to diffuse further than desired. Although this additional diffusion had not been a problem for larger devices, the margin of error for the smaller devices was narrow enough that the smaller devices no longer functioned. To resolve this, a new recipe was developed in which all furnace steps occurred before ion implantation, eliminating the extra diffusion.
In addition to this, as we worked to resolve the problems with our own devices, we had the opportunity to prepare and simulate an amplifier circuit for professional fabrication as an alternative option.
In order to test our low capacitance MOSFET devices and our professionally fabricated amplifier circuit, we have begun developing a circuit board, and have been running simulations to determine the type of shielding needed for the circuit to ensure that the incoming ions only effect the correct part of the circuit.
Results & Discussion
Although we have not yet been able to test a completed circuit with an ion stream, we have been able to use our new recipe to fabricate functional low capacitance MOSFET devices with gate lengths as small as two microns. These devices are much smaller and more sensitive than any devices previously developed at BYU and should function well for ion detection.
We have also extensively simulated and tested the circuit that we have sent out for professional fabrication, and expect it to work as desired.
Using these devices and the circuit board and shielding we are developing, we will be able to perform tests with an ion stream as the project moves forward, and anticipate that our devices will be sensitive enough to measure a small number of ions.
MOSFET devices currently used in memory technology can be used to revolutionize mass spectrometry by making it smaller, more portable, and less dependent on factors like extremely low temperatures and pressures.
- D. W. Koppenaal, C. J. Barinaga, M. B. Denton, R. P. Sperline, G. M. Hieftje, G. D. Schilling, F. J. Andrade, and J. H. Barnes, “MS detectors.,” Analytical chemistry, vol. 77, no. 21, p. 418A–427A, Nov. 2005.