Get email delivery of the Cadence blog featured here
In researching and thinking about silicon photonics last week, I found myself going down the rabbit hole of quantum physics. Like ya do.
Since I didn’t take any physics classes in college, most of what I know about quantum physics is self-taught, which means that there are probably vast swathes of basic things that I don’t understand, like, for example, why we do consider light as a particle and a wave, depending on context. I know that we DO, I just don’t know WHY.
So I looked into it.
The double-slit experiment was first performed by Thomas Young in 1801. His experiment was part of classical physics, well before quantum mechanics and the concept of wave-particle duality. He believed it demonstrated that the wave theory of light was correct, and his experiment is sometimes referred to as Young’s experiment or Young’s slits.
The essence of the experiment is this: When shining light through a partition with two slits in it, the resulting pattern shows that there is wave interference, showing for sure that the light-as-a-wave theory is correct.
By the way, this also works for water waves and sound waves. This is why when a piano or violin is out of tune, the tone you hear isn’t a pure sound—what you’re hearing are the waves from the competing strings interfering with each other. (In a piano, there are three strings per note, and if they’re out of tune with each other, you get interfering waves of sound when you strike a key. In a violin, the musician tuning it is listening not necessarily for the actual pitch, but for the pattern of fourths or fifths that interfere with each other in a particular way.)
But light isn't just a wave, it can also be looked at as a particle, called a photon.
So what happens if you shoot a single photon at the double slits? Turns out, if you shoot a bajillion photons, one at a time, they still form an interference pattern, the same interference pattern as do the waves of light.
Basically, that means that all the possible paths of these particles can interfere with each other, even though only one of the path actually happens. Even if the photons are sent through the slits one at a time, there’s still a wave present to produce the interference pattern. The wave is a wave of probability, because the experiment is set up so that the scientists don’t know which of the two slits any individual photon will pass through. (The wave of probability is called the Copenhagen interpretation.)
Quantum physics, y’all. How is this possible? Mind blown yet?
It gets even weirder.
If scientists set up the experiment so they CAN figure out which slit the photon goes through the barrier and hits the screen, the interference pattern collapses. (This is called a “which way” experiment.) If you have ever heard at a cocktail party that observing quantum particles change their behavior and you never understood what they meant, this is it. Observation eliminates the wave of probability, because that wave is based on all the possible outcomes. Once you know the outcome, then there’s no wave.
Now get this. (And here is where my mind is just bent too far.) If the scientists set up the experiment such that the detection happens AFTER the photon has gone through the slit (called a “delayed choice” experiment), the interference pattern ALSO collapses. So basically, the observation affects the behavior of the particle after the fact. The photon has already passed through the barrier, so somehow observing the photon affects the behavior of the photon BEFORE it goes through the barrier.
Which happened in the past.
Time travel comes from this theory. Multiverse theory. Quantum tunneling. Quantum entanglement.
And Meera’s brain getting broken.
Now, how does this apply to Cadence? When we’re talking about using light to create the mask on the silicon wafer to create the die (that are packaged into chips that go into packages that are installed on circuit boards that go into devices…), there is a lot of weird geometry that goes on in this creation.
First, let’s think about nanometers. They are one billionth of a meter. A human hair is around 75 microns (abbreviated 75μm) or 75,000nm (nanometers) in diameter. A human red blood cell is 6,000-8,000nm across, and the Ebola virus is about 1,500nm long and 50nm wide. This is unthinkably small.
At this level, waves of light (193nm) pass through something a bit like Young’s Slits to create the mask and we can predict how it will interfere with itself to give just the pattern we want, which may be only 32nm across, or even smaller. (There’s an amazing photograph in a blog post on Forbes Magazine, but know that it was written seven years ago and is likely already out of date!) Light starts acting very strangely at this level of smallness, but instead of figuring it out from the beginning, we start from the pattern we want on the wafer, and then work out what we need to put on the mask to get it. This is called reverse lithography.
I maintain that the engineers who have figured this out are actually magicians.
P.S. An awesome video is on PBS’ Space Time, if you’re the kind of learner who likes to see this stuff explained in real time. This is worth every minute of the approximately 10-minute video.
And its follow-up, which goes into more detail about how a “quantum eraser” re-writes the past.