For years, your company has been making ICs using materials including an element, Hafnium, as a replacement for the silicon dioxide (SiO2) previously used. You did this because the SiO2 layer was getting too thin (1.2 nanometers), which led to manufacturability problems and quantum tunneling problems. 1.2 nanometers is only six diameters of a silicon atom. And if the insulator gets too thin, the electrons go right through it. Not good if you are trying to minimize a chip's power consumption. And that's what the customers want because they are often running memory and microprocessors using limited-capacity batteries.

Some smart guy figured out that if the dielectric was made employing compounds made with Hafnium, the dielectric constant (also called the "relative permittivity") could be increased from SiO2's 3.9 to about 24, a major improvement. This allowed you to increase the thickness of the gate insulator, reducing the quantum tunneling. It also allowed you to maintain, and even increase gate authority over the current in the semiconductor channel: This means that changes in voltage of the gate have a bigger effect, changing the current through the semiconductor channel. That might allow a lower supply voltage, too. Might this drop power consumption even further?

It worked very well. Over time, minimum feature sizes went from about 60 nanometers to today's 10 nanometers. But you want to go further. Yet, you are wondering where the next big material breakthrough is going to come from. Will you be able to maintain a semiconductor process down to 3 nanometers? How are you going to be able to "out-Hafnium" Hafnium?

All the chemical elements that nature provides are shown in a Periodic Table of the Elements. There are 92 elements up to Uranium, and only 81 of them are stable. (The rest are radioactive, so we don't use them.) You might imagine that each and every one of them has been considered, at one time or another, for use in semiconductor and dielectric devices. And probably that's true. If an element existed that would help, somebody would have found it by now. You think. But they haven't.

I believe you will succeed but will do so with my help. I have just patented an improvement to dielectrics based on an easy-to-do substitution to the atoms within a Hafnium-containing dielectric, and quite possibly others as well. One that I believe will greatly increase the possible dielectric constant. It requires a simple, easy to describe change in the makeup of the chemicals used in a deposition machine. And a bit of redesign of your IC process, to take advantage of the better characteristics of the Hafnium-containing dielectric layer.

I found something that your best scientists and engineers didn't even consider. Nobody expected to find an improvement, so nobody looked. The material I found was always there, but just well-hidden. Hidden in plain sight, you might say. Like a few trees, hidden in a forest. And your people didn't think they needed to enter that forest.


By and large, your people didn't consider isotopes. Well, neither did the rest of the semiconductor industry. Most engineers and even scientists usually think of isotopes as being atoms with different numbers of particles called "neutrons", and that is quite true. But that's not the only difference there is with some isotopes. A few special ones.

I would like to discuss this matter with some of your Semiconductor Process Scientists and Engineers, under an NDA (Non-Disclosure Agreement). Please talk to them, and have them contact us.

They and you need the improvements this invention can provide. Because if you don't use it, maybe your competitors will.

Jim Bell

Scroll to Top