Could we use optical heterodyning to boost UV bands into VIS?

[Nathan McCorkle]

I came across this idea called optical heterodyning, it's the same
concept as RF heterodyning (commonly used for radio) where you have
your antenna signal (SIG) being mixed with a local oscillator (LO)
resulting in two new signals being created with frequencies LO+SIG and
LO-SIG or SIG-LO (whichever is non-negative) called the intermediate
frequencies (IF). With RF the mixer is a dual-input amplifier (I'm not
sure exactly what that is, it would be cool to see a part on mousr or
digikey) but in light it's just a 50/50 beam splitter with a
photodiode on the LO (i.e. laser) to keep the brightness stable.

Doing some algebra with and playing with a wavelength to frequency
calculator, I found that the 650nm red lasers on dealextreme.com would
boost 260nm light to 433.36nm and 280nm to 491.88nm. (The additive
heterodyne signals would be in the 180nm region of the UV)

So does anyone other than me think this could work? I already have a
red laser and a 50/50 beam splitter... I guess I would need at least a
UV low-pass filter for the SIG. Anyone know where to get one of those
cheap?

http://en.wikipedia.org/wiki/Heterodyne


HETERODYNE CHARACTERIZATION OF HIGH-SPEED PHOTOMIXERS FOR THE ULTRAVIOLET
BILLY WAYNE MULLINS - PhD Dissertation - 1989

(experimental setup starts on page 86)
http://www.dtic.mil/dtic/tr/fulltext/u2/a220914.pdf



--
-Nathan

[Simon Quellen Field]

In radio, you multiply the two signals. That can be done with a diode mixer.

Generally you get the sum and difference frequencies, and filter for the one you want. Using the same frequency for both inputs gets you a frequency doubler, and you don't have to filter, since the difference is DC. Digikey has lots of RF mixers. The one I have used the most is the NE602.

A beam splitter will not multiply. Otherwise you could get UV light by putting two beams of green light into a beam splitter. What you need is a medium that multiplies. The common green laser pointers use a frequency doubling crystal to double an infrared laser at 1064 nm into the green at 532 nm.

The photodiode in the article is not there to keep the brightness stable. It is there to detect the Doppler shift in the lidar apparatus he is working with. Light goes out, and is reflected back, and the signal and reflection are mixed, and the difference frequency is in the 10 GHz range for fast aircraft, which is why they need fast diodes.

Almost any non-linear response will work. You can disassemble a cheap green laser pointer to get the lithium niobate or potassium titanyl phosphate crystal out of it, and send any wavelength you want into it to get a doubling. The laser has to have a fairly high power in order to trigger the non-linear response, and the output will be much lower than the input, typically a tenth or less in optical power. There are losses, but the most obvious is the two photons in, one photon out limitation posed by simple physics.

The crystal is inside an optical cavity (a pair of partial mirrors) since you'll want many passes through the crystal to get the most out of it.

Check to see if the crystal and other optics in the green laser pointer operate in the UV ranges you are interested in. They may be opaque at those wavelengths.

Doubling a 200 mw 405 nm violet laser gets you into the 202 nm range, and you still have enough power to be useful (10 to 20 milliwatts). Don't bother with expensive low pass filters, just aim it at a DVD and pick off the high frequency beam -- it will be going in quite a different direction than the low frequency (405 nm) beam, since it has half the wavelength. Filters are for when you need a compact device and you don't care much about price.

Personally, I don't buy laser pointers above 5 mw. It is just too easy to accidentally damage your vision by aiming it at a window or doorknob. You won't even notice you have damaged yourself, because your brain corrects for so much. But you will gradually find that reading gets harder and harder. If you do get one of those lasers, epoxy it to a concrete block, and always wear low-pass goggles.

-----

Get a free science project every week! "http://scitoys.com/newsletter.html"

--
-Nathan

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To post to this group, send email to diy...@googlegroups.com.
To unsubscribe from this group, send email to diybio+un...@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

[Ben Hunt]

I think your goal is achievable, however I do not think that heterodyning is the solution. I wish I had a better grasp of why; my gut explanation is that they light waves are too small: only molecular antennas (crystalline structures on nm range down to the valence bands of gaseous atoms) can consume or produce them.

My understanding of light, esp. in the UV range, is it is fairly easy to go up in wavelength by losing power, but much harder to go down in wavelength by increasing power. 

I have heard about what Simon talks about, and it is definitely possible to use a medium which uses photon doubling to push out a ray of higher energy photons, although there is a lot of energy lost.

The easiest method to get visible light from the UV is through flourescence; this is how fluorescent lights work. All you need is a phosphor coating. Unfortunately both photon-doubling and flourescence cause a significant strain on the material. Phosphors have a tendency to oxidize when fluorescing, which is why they are put under vacuum or in a noble gas atmosphere.

I would be very interested to see if Simon's experiment works, however! Cheap and relatively high power (10 mW) UV light sources are hard to find. I might have to open up some lasers and try this out for myself. Thanks DIY BIO!

Ben

[Nathan McCorkle]

On Sun, Jan 6, 2013 at 10:33 AM, Simon Quellen Field <sfi...@scitoys.com> wrote:
> In radio, you multiply the two signals. That can be done with a diode mixer.

Wikipedia and a professor I had for a radio class said the signals get
added or subtracted, not multiplied... unless by multiplication you
mean amplification...

>
> Generally you get the sum and difference frequencies, and filter for the one
> you want. Using the same frequency for both inputs gets you a frequency
> doubler, and you don't have to filter, since the difference is DC. Digikey
> has lots of RF mixers. The one I have used the most is the NE602.
>
> A beam splitter will not multiply. Otherwise you could get UV light by
> putting two beams of green light into a beam splitter. What you need is a
> medium that multiplies. The common green laser pointers use a frequency
> doubling crystal to double an infrared laser at 1064 nm into the green at
> 532 nm.

No I'm interested in wavelength doubling, not frequency doubling.

From that PhD thesis:
"A 50/50 beam splitter combines the LO beam with the signal beam
and splits part of the combined beams off to the detector and part to a
NIST traceable surface-absorbing calorimeter. "

>
> The photodiode in the article is not there to keep the brightness stable. It
> is there to detect the Doppler shift in the lidar apparatus he is working
> with. Light goes out, and is reflected back, and the signal and reflection
> are mixed, and the difference frequency is in the 10 GHz range for fast
> aircraft, which is why they need fast diodes.
>

I understand that the thesis is using 334.5nm combined with a tunable
334.5nm LO, and feeding the difference (in the low GHz) into a power
analyzer... seems just like AM radio where the difference is audio
frequency and the data, is, well audio

> Almost any non-linear response will work. You can disassemble a cheap green
> laser pointer to get the lithium niobate or potassium titanyl phosphate
> crystal out of it, and send any wavelength you want into it to get a
> doubling. The laser has to have a fairly high power in order to trigger the
> non-linear response, and the output will be much lower than the input,
> typically a tenth or less in optical power. There are losses, but the most
> obvious is the two photons in, one photon out limitation posed by simple
> physics.
>
> The crystal is inside an optical cavity (a pair of partial mirrors) since
> you'll want many passes through the crystal to get the most out of it.
>
> Check to see if the crystal and other optics in the green laser pointer
> operate in the UV ranges you are interested in. They may be opaque at those
> wavelengths.
>
> Doubling a 200 mw 405 nm violet laser gets you into the 202 nm range, and
> you still have enough power to be useful (10 to 20 milliwatts). Don't bother
> with expensive low pass filters, just aim it at a DVD and pick off the high
> frequency beam -- it will be going in quite a different direction than the
> low frequency (405 nm) beam, since it has half the wavelength. Filters are
> for when you need a compact device and you don't care much about price.
>

Again I want to double the wavelength... heterodyning 650nm with 260nm
or 280nm to increase the wavelength to a range that can make it past
the CCD window. Here are my calculations:
650nm == 4.6122*10^5 GHz
260nm == 1.1530*10^6 GHz
280nm == 1.0707*10^6 GHz

(260nm) - (650nm) == X nm
1.1530*10^6 GHz - 4.6122*10^5 GHz == 691780 GHz
691780 GHz == 433.36 nm

(260nm) - (650nm) == X nm
1.0707*10^6 GHz - 4.6122*10^5 GHz == 691780 GHz
609480 GHz == 491.88 nm

--
-Nathan

[Simon Quellen Field]

On Mon, Jan 7, 2013 at 3:34 PM, Nathan McCorkle <nmz...@gmail.com> wrote:

On Sun, Jan 6, 2013 at 10:33 AM, Simon Quellen Field <sfi...@scitoys.com> wrote:
> In radio, you multiply the two signals. That can be done with a diode mixer.

Wikipedia and a professor I had for a radio class said the signals get
added or subtracted, not multiplied... unless by multiplication you
mean amplification...

Your professor was wrong. You need a non-linear device. Adding and subtracting are linear.

Wikipedia agrees. Read the third sentence. Also read the section "Mathematical principle" which explains what happens when you multiply two sine waves.

 

Again I want to double the wavelength... heterodyning 650nm with 260nm
or 280nm to increase the wavelength to a range that can make it past
the CCD window. Here are my calculations:
650nm ==  4.6122*10^5 GHz
260nm ==  1.1530*10^6 GHz
280nm ==  1.0707*10^6 GHz

(260nm) - (650nm) == X nm
 1.1530*10^6 GHz - 4.6122*10^5 GHz == 691780 GHz
691780 GHz == 433.36 nm

(260nm) - (650nm) == X nm
1.0707*10^6 GHz - 4.6122*10^5 GHz == 691780 GHz
609480 GHz == 491.88 nm

 I don't believe the author of the paper was talking about getting the difference frequency out as light or radio waves. He was seeing the difference frequency as an electrical signal in his photodiode. To do what you want, you would still use the potassium titanyl phosphate crystal. Note that multiplication of sine waves gives you four output frequencies -- the originals, plus the sum and difference frequencies. If your detector is only sensitive to the difference frequency, and the local oscillator is not modulated you won't need any additional filters. The local oscillator frequency will just add a DC bias which is always the same, and easily subtracted out.

You didn't mention your application. It sounds like you have a signal at 260nm and 280nm, and you don't have a detector that can see those wavelengths. It might be simpler to use a detector that can, such as a back-biased UV LED and op-amp. At a couple dollars, that would also be cheaper.

[Patrik D'haeseleer]

On Tuesday, January 8, 2013 2:48:24 PM UTC-8, Simon Field wrote:

On Mon, Jan 7, 2013 at 3:34 PM, Nathan McCorkle <nmz...@gmail.com> wrote:

On Sun, Jan 6, 2013 at 10:33 AM, Simon Quellen Field <sfi...@scitoys.com> wrote:
> In radio, you multiply the two signals. That can be done with a diode mixer.

Wikipedia and a professor I had for a radio class said the signals get
added or subtracted, not multiplied... unless by multiplication you
mean amplification... 

Your professor was wrong. You need a non-linear device.

I suspect his professor meant that the *frequencies* get added and subtracted. Which you achieve by multiplying the signals.

[John Griessen]

On 01/09/2013 12:19 AM, Patrik D'haeseleer wrote:
> I suspect his professor meant that the *frequencies* get added and subtracted. Which you achieve by multiplying the signals.

Probably. That kind of treatment "transforms" the nonlinear into the linear for ease of calculations.

[Nathan McCorkle]

Hmm, I am still thinking about this sort of thing, and just came across an article that seems to confirm that this sort of technique works/exists... but for some reason (dogma/history/technical-language-barriers??? or something else I'm failing to understand) they don't call it optical heterodyning, but rather 'sum frequency generation':

https://en.wikipedia.org/wiki/Sum_frequency_generation_spectroscopy

https://www.chemistryworld.com/news/dna-helix-has-chiral-water-spine/3007545.article

[John Griessen]

On 06/12/2017 08:15 PM, Nathan McCorkle wrote:
> Hmm, I am still thinking about this sort of thing, and just came across an article that seems to confirm that this sort of
> technique works/exists... but for some reason (dogma/history/technical-language-barriers??? or something else I'm failing to
> understand) they don't call it optical heterodyning, but rather 'sum frequency generation':

If you start with UV it could be a down-convert to visible, if you had a difference detector for light, (in radio tech talk).
Heterodyning is about a signal difference, not an addition.

The wikipedia sum-frequency article says the third light beam is tiny amplitude and the sum of the two input light beam
frequencies because of surface effects.

[John Ladasky]

I'm not sure exactly what this has to do with biology, but...

Non-linear optical phenomena do in fact exist, but they are hard to achieve.  Under the right conditions (high optical flux, special crystals or absorbent materials), you can get two photons to combine into one.  

The standard 532 nm laser is actually a 1064 nm laser with a frequency-doubling crystal.  Trying to make this relevant to biology, there's a technique called two-photon excitation microscopy which is sometimes a nice way to do fluorescence work.  The excitation source wavelength is far from the fluorescence emission wavelength, so working with fluorescent dyes with short Stokes shifts becomes much easier.  You don't need to design a high-performance optical rejection filter for the excitation source. 

There are other, more complex nonlinear optical phenomena.  Some introductory reading:

https://en.wikipedia.org/wiki/Nonlinear_optics

https://en.wikipedia.org/wiki/Two-photon_excitation_microscopy

[Nathan McCorkle]

On Wed, Jun 14, 2017 at 1:47 AM, John Ladasky <john.l...@gmail.com> wrote:
> I'm not sure exactly what this has to do with biology, but...

Mostly regarding the detection of biomolecules. I was thinking about
how to use cheaper light sources to substitute for less common
wavelengths, i.e. 260 and 280 nm.
The article I just posted is specifically about DNA research, and
implications of surface hydration effects on DNA interactions with
proteins, drugs, etc. It sounds pretty much like two-photon

> The standard 532 nm laser is actually a 1064 nm laser with a
> frequency-doubling crystal. Trying to make this relevant to biology,
> there's a technique called two-photon excitation microscopy

I'm familiar with both those concepts, I guess SFG is just a different
term for detecting the two-photon emissions that are characteristic of
certain molecules, which seem much weaker than the common two-photon
dyes. I wonder if it is a similar order of magnitude difference
between fluorescense and Raman scattering (reading now).

Very cool!