This isn’t a book review, per se. I don’t even know if this “textbook” is still available, as I found at my university’s library book sale stuffed with out-of-date textbooks. But I wanted to highlight some of the technologies written in the book.
Note: I just got done with an interview with Sandia National Labs, and this book actually helped a lot with understanding more of the fabrication capabilities and equipment they possess.
Microfabrication, for me personally, is a very fascinating topic. By manipulating atoms, electrons, and photons in such a controlled way, one can create extremely practical devices that power the electronic needs of our everyday lifestyles. There isn’t much in terms of design theory in this book (that’s what Volume I-IV are for). However, despite being quite small (~150 pages or so), this paperback gives a precise overview of each possible process in microfabrication and the practical limitations for each. Additionally, the book is littered with (extremely beneficial) images from concept visualizations to experimental graphs to explain both procedures and parameter controls respectively.
To the fabrication methods!
This is the general term for the method that creates a desired pattern on the wafer. By applying a thin film of radiation-sensitive polymer, one can expose the material and change its properties. Exposure is typically done with UV light, but it can also be conducted through alternative means including electron and atom beams. When exposed to an appropriate liquid (solvent), one part is removed (washed away) while the rest of the material stays. If the exposed material is dissolved, the material is known as a “positive” resist. In contrast, a “negative” resist becomes resistant to solvents when it’s exposed to radiation.
Of course, these patterns are never useful on their own. However, they create “windows” that additional processes now have access to the wafer below.
Etching is when you want to remove material in the lithography windows. Etching can either be done either using wet (liquid) or dry (gas/plasma) methods. The majority of etching methods are driven through chemical reactions, which allows for chemical selectivity during the etching process. Alternatively, one can create an ion beam for a pure “physical” etch that removes material through atomic bombardment, though this method is typically slower than preferred chemical etching methods.
It’s interesting to note that some etching methods (wet or dry) are directional (anisotropic) and can be used to create novel or deep trenches in your design. For example, a directional beam of atoms will remove material in the beam’s path. Other methods will selectively attack the crystal lattice row-by-row and allow unique shapes in your design.
To add material, numerous methods are utilized to apply layers either on the atomic scale to create crystalline (epitaxial growth) layers or in “bulk” (poly-crystalline or amorphous) films (the latter being the easier and faster method). These methods include chemical vapor deposition (CVD), material sputtering, e-beam evaporation, and many others that result in thin film coatings to be applied the entire wafer.
Ion Diffusion & Implantation
In the making of microelectronic circuits, one wants to change the conductive properties of the Si wafer underneath all these films. That is where doping comes in, which allows for the creation of p and n doped materials necessary for diodes and transistor technology.
The easier method is to heat up the wafer environment and allow for material (vapor) to come into contact in your “window” regions of interest. Material accumulates on the surface and slowly makes it way into the wafer beneath the surface. Smaller and less interactive atoms will, of course, diffuse into the material at a faster rate.
The one thing with dopants is that they will still move around whenever the wafer is heated up in processing steps farther down the manufacturing line. Thus, one has to take into account ALL the high temperature manufacturing steps to make sure material diffusion does not get out of hand.
Sometimes, all you want to do is just change the chemical composition of the surface. The most common method is the oxidation of silicon (Si) to silicon dioxide (SiO2). SiO2 is an insulator and is a simple, yet robust barrier to many manufacturing methods. When photoresists used in lithography are not resilient enough for the required microfabrication processes, a layer of SiO2 can be grown underneath and etched to create more chemically “inert” windows.
The visually interesting aspect of growing SiO2 on Si wafers is that the wafer will change color based on the final SiO2 thickness. Thus, one can easily verify if the process went smoothly just by comparing the wafer color to a “look-up table” (but precision measurements are still used to understand your fabrication precision and consistency).
This is one of the easiest methods in a microfabrication setup, as it only involves heating up the wafers in an oven. No plasmas, no fancy chemicals. Of course, the gases present in the chamber are highly controlled as undesired chemicals can fuse to the surface and diffuse inwards when heated at such high temperatures.
Contacts & Packaging:
Finally, semiconductor chips have to be connected to the outside world and easily handled through macroscopic manufacturing processes (like being placed on a circuit board). One is typically familiar with standard processors and integrated circuits (ICs) being a black plastic box with metal leads coming out the sides or underneath. The semiconductor chip is connected to these leads typically through wire bonds “stitched” to both surfaces before the final device is completely confined in black plastic.
This is just a basic overview of the processes in microfabrication in this book. There are also a few additional topics on specific methods and insights for building specific designs, including BJTs (current-controlled switches) and MOSFETs (voltage-controlled switches). The last chapter details various methods on the design for MEMs (MicroElectroMechanical Systems). This is a fascinating area of research where microscopic gears, levers, springs, bridges, and many more unique shapes can be created using standard microfabrication capabilities.
After completing my 10 hour long on-site interview with Sandia, I realize that I do have a soft-spot for microfabrication. It’s a career area that I should pursue to fulfill my long-term career desires. But until then, someone has to be the sales engineer for LEDs!