Why can't we frequency double 1208nm ?

I just finished reading goninanbl00d's DPSS primer. One of the more useful threads/primers I've read in a while actually - I found it quite well layed out and straightforward.

However, I'm still not entirely clear as to why we can't frequency double some other wavelengths apart from 1064. For example, why could we not use KTP to frequency double a 1206nm diode like this one to end up with a fairly unique 604nm orange colour?

This seems too simplistic, but I've heard that Corning had some green 530 (?) nm lasers that were really just 1060nm diodes, directly doubled (as opposed to traditional 808 pumped DPSS). They're used in some of the ShowWX pico projectors. The fact that I haven't seen this done anywhere, is enough to suggest to me that it isn't that simple. I'm presuming that KTP can't simply double anything you throw at it. However, I'm also somewhat left with the impression from that DPSS primer that KTP can double anything above 500nm.

Or perhaps there is some other kind of SHG crystal that could be used? May not be cheap though.

I'm sure it wouldn't be cheap. That diode alone is listed at $700 (although thorlabs seems to be about 2x the competitive price of elsewhere on diodes). But if people spend three digits (nearly four) on 589, I'm sure they would on 604?

rhd said:

I just finished reading goninanbl00d's DPSS primer. One of the more useful threads/primers I've read in a while actually - I found it quite well layed out and straightforward.

However, I'm still not entirely clear as to why we can't frequency double some other wavelengths apart from 1064. For example, why could we not use KTP to frequency double a 1206nm diode like this one to end up with a fairly unique 604nm orange colour?

This seems too simplistic, but I've heard that Corning had some green 530 (?) nm lasers that were really just 1060nm diodes, directly doubled (as opposed to traditional 808 pumped DPSS). They're used in some of the ShowWX pico projectors. The fact that I haven't seen this done anywhere, is enough to suggest to me that it isn't that simple. I'm presuming that KTP can't simply double anything you throw at it. However, I'm also somewhat left with the impression from that DPSS primer that KTP can double anything above 500nm.

Click to expand...

Thanks. :yh:

As I've said before about KTP and 'throwing photons KTP with a diode for them to shit out a stream of photons out the other end', it isn't a very straightforward process. Finding the correct wavelength is only the start of it.

First off, due to a property inherent to KTP, it cannot double well below 500nm, and not all in the blue, as it cannot be phase-matched for those wavelengths. This same property allows it to double 1064nm with such brutal efficiency.

KTP is also used in the SHG of 1319 and 1342 for 660 and 671nm respectively (both are Nd:YAG/YVO4 lines). It's also used in SFG (sum frequency generation) for both 589 and 594.5nm.

The second issue that comes to mind is power density- and although it may seem easy to get plenty of power (5W in a C-mount 808 is easy to get), cramming that power into something that is usable is not nearly as easy.

First off, although 3W (like in the diode you linked) may seem like a lot of power, bear a few things in mind:

1. Being a multimode, multi-emitter device, the output beam will be very messy, not to mention extremely astigmatic.
2. There'll be no easy way to focus that down to the spot size needed for effective SHG.

To put things into perspective, your dinky little 5mW green pen laser has watts of power circulating through the cavity at any given time. Only a small percentage ever makes it out as 532nm, of course, and most of that power remains in the cavity.

The other thing about said dinky pen laser is that the intracavity 1064nm beam is often micrometers wide. Without a FAC, the intracavity beam is usually slightly smaller than diode's active emitting area, while with a FAC, the intracavity 1064 is often narrower.

So, you have many watts of power in a tiny spot. SHG, being a non-linear process, works better almost exponentially as power levels rise.

Put those two together, and you'll see that sending a single diode for a pass through a doubling crystal won't work.

Case-in-point: An SSY-1 (flashlamp-pumped pulsed Nd:YAG laser with a nominal peak power of 200mJ) has trouble achieving green output when it is shot through a KTP crystal. The same laser with a passive Q-switch (which serves to increase the peak power) has no problem with extracavity doubling. Without the Q-switch, the power is on the order of kW, while with the Q-switch, it rises to hundreds of kW (or even into the MW range if pumped correctly).

Of course, with the diode being multimode and mult-emitter, getting it to the point size necessary is extremely difficult. And even if you do manage to break a handful of the laws of physics and pull it off, it'll be an even bigger challenge to contain it within the doubling crystal.

On top of that, the diode has to be phase-matched with the doubler crystal. Due to the beam profile and astigmatism involved often this can't be done with a diode.

When you have seen a directly-doubled diode (such as the Novalux Protera 488 series), they use a special external-cavity VESCEL diode. The VESCEL diode design eliminates the astigmatism and poor beam profile inherent to normal diodes, while the external-cavity design means that the doubling optic is also part of the cavity, giving the power densities needed. Phase-matching can then be achieved, and a TEC is used to stabilise the doubling crystal and the diode to ensure temperature fluctuations do not result in an un-phase-matching of the setup.

It's far from easy, however, it can be done. It's also very expensive, and often, it's easier to just go DPSS instead of messing around with doubling of a diode. Sure, laser diodes scale well in terms of power, but by the time you hit multi-mode multi-emitter devices, it's near damn impossible to use the output for anything but pumping another laser (or burning stuff, if you're into that).

Well, if you could somehow SFG 1319nm and 1112nm you would get 603.3nm...

-Trevor

Hallo,
I'm appologizing a bit for some advertisment, but would like to point out that we are offfering 7W @ 1210nm multimode and 350mW (100pm FWHM) ex fiber for single mode.
But I would agree that goninanbl00d that multimode is not a desired dource for SHG.

Anyhow, from our single mode BTF it is possible to get up to 70mW of doubled light.

We are providing any wavelength for visible SHG. Even some of them with more power.

So is it possible, just an idea, that the crystals react to light the same way a tube would on a chime? Like there are some materials that vibrate at certain sound frequencies and will vibrate at each harmonic.

Is it the crystals vibrating at a specific light frequency in reaction to the incoming light, and changing the input to match those higher harmonics also lowers the response, just like trying to get a higher note out of a tube?

PM sent to answer. Lets keep off the dead threads.