question about radio detection

Is it possible to make a device that can detect single photons of radio waves with wavelengths of many meters even if they enter the device at a rate of no more than, say, one per second and, if so, how might it basically work?

I tried googling this but got nowhere.

I understand the main reason why radio detectors are not designed to detect single photons of radio waves is because each photon has such a small amount of energy that their energy is indistinguishable from typical thermal energy in the same detector thus there would be the problem of the detector seeing itself! So could, in theory at least, you make a camera (as opposed to a radio dish ) that can detect individual radio photons providing you made it cold enough; say, 0.1K?
I guess it would have to be a huge camera with an aperture many kilometers in diameter else, because of the diffraction limit, the image would be too blurred to be of much use. But perhaps it could have a radio-frequency lens made of a metamaterial to get around that?

Originally posted by humy
Is it possible to make a device that can detect single photons of radio waves with wavelengths of many meters even if they enter the device at a rate of no more than, say, one per second and, if so, how might it basically work?

I tried googling this but got nowhere.

I understand the main reason why radio detectors are not designed to detect single photon ...[text shortened]... use. But perhaps it could have a radio-frequency lens made of a metamaterial to get around that?

In a related field. I studied the effects of our local gravitational lensing from our sun and discovered an interesting effect: If you look at the long view of the sun, say go back a billion Km, and observe radiation being lensed by the sun, first thing you notice is that radiation focuses at around 1000 AU, roughly.

Now if you look at say, UV light, it comes to a nice crisp focus, at least on first look.

However, if you look at wavelengths say a million Km long, as that radiation goes by the sun, the entire gravitational lensing effect is very limited for such large wavelengths. So if you had a detector at the first focal point (it's really a focal line leading nearly to infinity, but that's another story) and you have a detector that can detect radiation say from wavelengths of 1 million Km down to UV or so, you will find a maximum wavelength that can be focused by the sun.

This leads to the idea that you could figure out the mass of a star by going to the first focal length, and just measuring the cutoff wavelength of radiation, assuming there IS million Km wavelength radiation floating around the universe🙂 but a measure of such can tell you the mass of the star directly with no need for any other measurements.

One problem with the single photon measurement is that a naturally occurring set of wavelengths of such a low frequency, say less than one megahertz, is it is almost impossible to actually generate a single photon of such a wavelength. So you would be dealing with some kind of multi-wavelength radiation train. So your detector would have to be the one that could parse out one photon, some kind of switch that could, say, capture a single wavelength, say 1000 meters long, and direct it to a reflective box that could store that photon, in a way similar to the optical wavelength reflective capture devices, like a perfect set of mirrors that would let that photon bounce back and forth with no degradation so you could work with that photon for whatever reason you are after. At very low frequencies that might require say, a superconductive box that would allow that photon to be captured, and of course it would have to be very large to accommodate such a beast of a photon!

The detector I don't think would have to be as cold as you suggest although that would help. The main thing you would need would be reactive circuitry that would be sensitive to the wavelength you are after, some kind of tuned circuit that would in effect amplify a single frequency of interest. It would exactly amplify directly but be much more responsive to a single frequency than others and so preferentially capture the wavelength of interest.
So that's my 2 cents worth on the issue.

I think this would be difficult. With near visible frequencies you can produce an atomic transition - so something quite definite has happened for circuits to detect. With normal radios the signal induces an oscillation in the free electrons in the metal in your antenna. The problem is that with a single photon only one electron would be affected and it would be hard to tell the difference between that and a collision between the electron and a lattice vibration, never mind whatever electro-magnetic background there is.

The only way I can think of is if the antenna were at absolute zero (or as near as practicable) and one could detect the effect of an electron being knocked from the Fermi surface. If there is a band gap then the gap would represent the minimum energy photon that could be detected. If there is no band gap then background effects couldn't easily be filtered out.

As long as you don't mind a large noise to signal ratio, then you could probably build a statistical picture of about how many photons had hit the antenna, but you'd have a job identifying a single photon. If you are expecting to identify single photon interactions then I doubt it's possible.

thanks guys. I will mull all this input over.

Originally posted by DeepThought
I think this would be difficult. With near visible frequencies you can produce an atomic transition - so something quite definite has happened for circuits to detect. With normal radios the signal induces an oscillation in the free electrons in the metal in your antenna. The problem is that with a single photon only one electron would be affected and ...[text shortened]... photon. If you are expecting to identify single photon interactions then I doubt it's possible.

I think you mean a low signal to noise ratio, say 3 db or so. One problem is generating a single wavelength with known circuitry. If you use switching to cut off a signal from a frequency source you don't just cut off the signal, you generate sidebands and the faster the switching the more sidebands you generate, so you could make harmonics hundreds of times the wavelength you are studying. That may not be a problem if you have sufficient filtering but it does introduce an annoyance factor🙂 I was thinking maybe you could have a resonant box that stores up some of that original wavelength although for low frequencies, not sure what frequency you are talking about but say one megahertz, that is a significant size! If you have a 1/2 wavelength resonant chamber, that would amount to a box 491 feet or about 150 meters on a side. Pretty significant size. So the idea there would be to have a switch that would somehow effectively open one face of the box to release exactly one wavelength and then close so that would have to happen in microsecond time scales and the higher the frequency, the faster that switch has to be, so 1000 megahertz, the switch now has to be in the nanosecond time frame. Of course at that frequency, the resonant chamber would be 1000 times smaller you get that at least.

Not sure of the physics of just quickly opening a port, even that action might generate sidebands, imagining a door that can open and close in a microsecond, of course it couldn't be an actual matter door, it would have to be some kind of interference effect, not sure how that would be done.

The problem with switching is the electromagnetic field itself. If you time the shut off or on say at minimum electric field strength, you are cutting into a higher magnetic field since the two fields are off by some set amount of phase angle. And if you cut off the switch at minimum magnetic field strength you run into a higher electric field so I don't know how you could ever generate a single wavelength in the first place. Magnetic fields and electric fields just don't like being cut off, they go TWANG and generate other sideband effects so getting a single wavelength at radio wavelengths is going to be a significant challenge no matter how you achieve it.

You have any ideas on how that could be done?

Originally posted by sonhouse
I think you mean a low signal to noise ratio, say 3 db or so. One problem is generating a single wavelength with known circuitry. If you use switching to cut off a signal from a frequency source you don't just cut off the signal, you generate sidebands and the faster the switching the more sidebands you generate, so you could make harmonics hundreds of time ...[text shortened]... ificant challenge no matter how you achieve it.

You have any ideas on how that could be done?

I wrote high noise to signal ratio to emphasise the noise - I realise it's normally expressed the other way round. You might be able to do the switching with something like a liquid crystal, but the switching would still take a finite amount of time, so it would be impossible to completely cut out effects from the switching. This kind of thing is why I studied Theoretical Physics...