# Is it possible to generate light with an antenna?

An antenna (e.g. a dipole) is able to radiate at a certain frequency thanks to the EM field generated by a current provided by a signal generator at such a frequency.

So, for instance:

Voltage source at frequency f (representing an amplifier) + Antenna made of conductors = Radiation at frequency f

My question is: is it possible (or, maybe will it be possible) to generate light by using a voltage source at frequency of light (480-750THz)?

All light sources I've seen are realized by using mechanisms different from EM radiation, like LEDs, LASER etc.

• sure. pass enough amps through it, it will generate light ;-) – danmcb Mar 31 at 10:38
• Passing enough amps through the antenna will create an arc which isnt what the OP is asking. – Miss Mulan Mar 31 at 10:54
• @MissMulan: Pass enough current through an antenna and it will glow from heat, not an arc. Have you ever heard of jokes? – JRE Mar 31 at 12:50
• Is it polite to request @ThePhoton 's presence? – Marcus Müller Mar 31 at 12:55
• physics.stackexchange.com/questions/5046 “Can I use an antenna as a light source?” – Jacob Krall Mar 31 at 18:14

Photonics is a thing, and it very much treats light as electromagnetic waves in waveguides on silicon, with mixers, delay lines, matching circuits, amplifiers etc as you'd do for a RFIC.

Just as an antenna doesn't generate the signal to be transmitted, an oscillator is used to generate a carrier wave – just that the oscillator happens to be a laser diode, sometimes on the same substrate, instead of a discrete transistor, an L and a C. An antenna is just an impedance matcher between transmission line and free-space, and that's what you find on every laser diode die, standard LED die at the point where you try to convert the surface-bound or substrate-travelling lightwave into something emitted into the fiber or a lens or free space.

What you definitely don't find is the classical dipole being fed with a current coming from an oscillator. That doesn't work, as most such components are comparable in size as the antenna. But, there are photonic antennas. These convert a free-space EM wave to a guided wave, just as your phone's antenna converts free-space waves to waves in a coax cable (and vice versa), while simulataneously mixing things with a lightwave. This is not an antenna in the sense of a $$\\lambda/4\$$-monopole or something.

They are fiddly:

THz Antenna coupled to a high-speed silicon plasmonic photodetector enabling opto-electronic generation and detection of waves by photomixing within a THz-bandwidth. Source: Institute of Photonics and Quantum Electronics (IPQ), Karlsruhe Institute of Technology

• This is the correct answer. Optical antennas are within reach of photolithography. Oscillators, amplifiers, and mixers need different technology than we use for longer wavelengths, but that's normal: you can't make millimeter waves with DeForest's triodes either. However, if you'd asked about x-rays... – John Doty Mar 31 at 18:23
• It's worth noting that the THz antenna you're showing does not produce light in the visible spectrum - it's much longer than that at around 300μm (vs 700nm for the longest visible red), effectively the far infrared. Closer to microwaves than to visible light. Nature already provides us with excellent antennas for visible light - atoms. – J... Apr 2 at 8:48
• @J... exactly! I hoped to make it clear that it's more of a mixer / RF antenna hybrid device than a "light antenna" in the last sentence of my answer. – Marcus Müller Apr 2 at 20:18

The wavelength of 750 THz is 400 nm so, your dipole would need to be 200 nm long to make a radiating antenna but your biggest problem (in 2021 and probably beyond) is being able to produce a frequency of 750 THz. Leave it to the optical guys is my advice.

Another consideration (from classical Friis equations) is that the path loss (from a transmitting antenna to a receiving antenna) is proportional to frequency. This is because a receive antenna shrinks with frequency and therefore cannot capture the same magnitude E-field. In other words, as frequency rises, dipole length shortens and E-field "captured" reduces.

Friis path loss equation for distance (d) in metres and frequency (f) in MHz: -

$$20\log_{10}(d) + 20\log_{10}(f) - 27.55$$

• If distance is 1 metre and frequency is 1 GHz, the path loss is 32.45 dB

• At 1000 GHz (1 THz), the path loss is 92.45 dB

• At 750 THz, the path loss (over 1 metre) is 149.95 dB i.e. the path loss is getting quite high (and that's just for a distance of 1 metre)

• Remark: the usual caveat about frequency-dependent path loss applies: By the same factor that the path loss for a fixed distance increases, the maximum antenna gain you can achieve with an antenna of fixed dimensions increases. Physics is non-discriminatory in that way: While you certainly have higher path loss, it also becomes significantly easier to build antennas with very high gain (an 1 GHz antenna that has a beam with an opening angle as narrow as a 750 THz laser pointer for 8€ will be quite a sight to behold). – Marcus Müller Mar 31 at 12:53

The other answers are correct, the frequency of light (i.e., visible light) is so fast that it's really hard to do. People do make THz light (even longer wavelength than infrared) using antennas, and can even make mid-infrared light using nanometer scale antennas - but this is highly technical, requires teams of scientists, and millions of dollars.

There is another way to generate light from an antenna. Drive it with an incredibly highly current such that it heats up. As it gets hotter it will generate blackbody radiation in the infrared (and a little red). It will probably melt shortly after that, but you created visible light from an antenna!

• "It will probably melt shortly after that" - the trick is to use tungsten wire and an inert atmosphere. Argon is a popular choice, though nitrogen also works well and might be easier to find. – John Dvorak Apr 1 at 6:16
• Good idea, someone should put that into a glass orb, or under a vacuum, and patent that! No more candles and gas lamps! – KD9PDP Apr 1 at 9:40
• @KD9PDP And someday perhaps we'll find this has LED to something better. :) – Don Branson Apr 1 at 13:50
• @JohnDvorak I'm not sure, but nitrogen might not be so inert at incandescent temperatures. en.wikipedia.org/wiki/Tungsten_nitride Anyway, argon is readily available at welder's supply stores. – Solomon Slow Apr 2 at 1:17

Here is an open access paper describing an Yagi-Uda antenna emitting light at 830 nm.

Kullock, R., Ochs, M., Grimm, P. et al. Electrically-driven Yagi-Uda antennas for light. Nat Commun 11, 115 (2020). https://doi.org/10.1038/s41467-019-14011-6

No, it isn't possible. The properties of the materials are very different at such frequencies.

Consider the materials typically used in chip manufacturing. We use aluminum as a conductor, and silicon dioxide (glass) as an insulator. But at 750 THz, aluminum blocks the EM field, either reflecting it or absorbing it as random thermal energy. And glass passes the field quite readily — that's why we use glass fibers for long-distance telecommunications today.

So, with that in mind, how would you actually construct an "antenna"?

• upvoted this although I don't fully agree: photonics is a thing, and it very much treats light as electromagnetic waves in waveguides on silicon, with mixers, delay lines, matching circuits, amplifiers etc as you'd do for a RFIC. Just as an antenna doesn't generate the signal to be transmitted, an oscillator is used to generate a carrier wave – just that the oscillator happens to be a laser diode, sometimes on the same substrate, instead of a discrete transistor, an L and a C. An antenna is just an impedance matcher between transmission line and free-space, and that's what you find on every – Marcus Müller Mar 31 at 12:37
• laser diode die, standard LED die at the point where you try to convert the surface-bound or substrate-travelling lightwave into something emitted into the fiber or a lens or free space. – Marcus Müller Mar 31 at 12:37
• @MarcusMüller: But that's precisely my point -- at the very least, the roles of the materials are reversed, and that's the distinction the OP is asking about. – Dave Tweed Mar 31 at 12:43
• And that's why I upvoted - you're right, OP has the wrong impression! – Marcus Müller Mar 31 at 12:44
• Aluminum acts as a conductor at optical (and lower) frequencies. That's why it's a good reflector of light. You can make an antenna with it. – John Doty Mar 31 at 20:31

Per KD9PDP above, about 40 years ago I encountered another ham with a VERY strong signal on 75 meters single side band. It turns out he was running 3kW peak IN HIS CAR! To accomplish this, he had a specially wound 3 phase generator under the hood that generated high voltage and reduced the complexity of his linear amp power supply. When he spoke, a totally awesome visible ball of ionization formed around the tip of his center loaded antenna and a sound, like AM demodulated SSB was clearly audible. He had a sacrificial piece of wire on the end of his antenna that had to be replaced periodically due to the arcing.

Of course such a setup was illegal for exceeding the power limitation, and would now be further illegal due to modern limitations on human exposure to electromagnetic fields and other safety requirements in amateur transmitting equipment.

But, yes, he produced light at the end of his antenna.

Now, back to my high power 30 watt rig... . . . K6YVL

Both radio waves and light are the same thing:EM radiation . So if you could create a RF signal with a frequency of 750THz we would receive this as light.