Buying 650 nm laser for Ramen spectroscopy

[akf32]

Hello!

I am a student working on a project to build a Raman spectrometer with a 650 nm laser. Previously we have used a Picotronic 650 nm 10 mW laser, but were having issues with it's stray light. (It basically had two dots). I am currently in the process of looking for a new laser and was wondering if anyone had any recommendations for where to look or had experience with this sort of thing in the past. So far I've looked at Odicforce but those lasers are all very cheap which makes me question their quality for a project like this. Any help or advice is very appreciated!

 

[Mattronium]

What are the basic requirements for the laser? It sounds like you don't need much power.
How important is linewidth, beam profile/quality?

Without knowing much about this topic myself, it sounds like the main need is a good beam shape and maybe 'clean' spectrum? So I would intuitively think a normal single mode 650nm diode might work. If not that, then spacial filtering or external cavity for better linewidth? But that is just my guess.

 

[akf32]

What are the basic requirements for the laser? It sounds like you don't need much power.
How important is linewidth, beam profile/quality?

Without knowing much about this topic myself, it sounds like the main need is a good beam shape and maybe 'clean' spectrum? So I would intuitively think a normal single mode 650nm diode might work. If not that, then spacial filtering or external cavity for better linewidth? But that is just my guess.

We want the laser's linewidth and beam to be high quality to reduce stray light as much as possible. For power we don't have super strong requirements, just not so high that it can burn anything.

 

[kecked]

Do you really need 650nm. You can go with a 671nm dpss and use a line filter. for raman Id recommend going a bit deeper to 785. You can cheat that to 808nm which is easy to come by. The real questions are the limits of you spectrophotometer and how close in you want to get. I’m assuming stokes line. You will also not be able to compare your results to standard results without calculating all the shifts.

few pointers. Fibers do fluorescence so watch out for that. If you use a gas laser you will find a LOT of stray light that will swamp all signal So plan on line filters, edge filters, and nd filters. You will also have slow phase issues.

The nd of your objective is critical. The simple beam splitter type is deceptionly alluring but beware. You give up 50% of your signal. Don’t think for a second you can ditch a line filter and edge filter. You need both. Hats why I asked how close in you want to go. Less than 100 wave units is difficult.

The shorter your source the more trouble you’ll get into.

hope this helps

ohif you go diode laser you will need strict temp control or the wavelength will drift. Sure there are 3xternal cavity lasers that can narrow up the center but temp still drifts. Point is raman lasers cost for a reason. Also power does matter. Your signal is 1x10-6 below your excitation not counting for losses In collection efficiency, filtering, quantum efficiency, matrix absorption…..I’m not saying watts but you need at least 50-100mw in reserve.

 

[liveforphysics]

I'm fussy about how narrow the excitation lasers mode happens to be. Many diodes give you a second and third frequency content very near to your excitation wavelength, and it makes all your Raman output data ambiguous to see real double Raman lines vs the excitation noise.

I experiment with Raman from 473nm to 880nm. Most off the shelf Raman's will use 785nm single mode laser excitation at 25-250mW. This is done so a single laser device can have pretty OK results measuring just about anything (because 785nm can barely trigger a PL emission). Depending on what you're looking at, sometimes 785nm has an awful signal to noise and melts your subject before you can get a decent signal to noise. Sometimes 473nm or 520nm can use just 1-10mW of excitation power to get clean and well defined Raman lines, and for other samples, those wavelengths stimulate so much PL (photonic luminescence) that you will never get useful signal and you must use 785nm and the additional heat and low Raman emissions efficiency. This is why more fancy Ramans now have multi-pump selection options, with popular wavelengths being 532nm and 633-636nm, as well as 785nm pumps.

Its worth mentioning you can dodge the PL emission noise problem by ducking it in shorter wavelength UV than efficiently pumps the material, and sometimes a short pump wavelength like 405nm can give excellent clean lines (but most often its just a white noise PL mess).

 

[liveforphysics]

Also, keep in mind, 650nm can definitely still drive heavy PL emissions that will block out your signal completely depending on the sample you're measuring.

 

[liveforphysics]

One more tip for Raman, the laser is the most easy part. The hard part and art is in your tricks to block the Raleigh scattered light from blowing out your Raman signal. I started out using dielectric notch filters, but the cost a fortune and never performed well for me. You want a series of dielectric mirrors at 45deg angles and Raleigh light beam traps behind each one. Lower cost and works much better. I now use a double monochromator as a filter, but it's a physical huge device and $$$.

 

[kecked]

I gave up and bought one.

 

[LSRFAQ]

LOOK for DIY ECDL articles in journals such as Review of Scientific Instruments or consider diodes with built in Bragg Gratings to fix the wavelength. Your average university engineering department will find it fairly easy to stabilize an LD with a little practice.

Then there is this website, one of my favorites:

Orange

Or just get a HENE with a tube length under say 14-16 inches. They tend to self stabilize.

Your time is more expensive then a good diode. Getting the wrong diode results in failure to publish. Tell your PI that.

Steve

 

[LSRFAQ]

One of many vendors:

VBG Archives

Volume Bragg Grating (VBG) Laser Diodes offer narrow linewidth in a variety of stabilized wavelengths, output powers, and free-space & fiber-coupled options

www.rpmclasers.com

Volume Bragg Gratings are good to between .01 to .1 nm. Usually far better if you stabilize the temperature and tune the drive current.

As others have said filters are a way to get far better results. Laser bandpass or Lasefr Cleanup filters for Confocal and Raman are all over Ebay.


Steve