• Members of the previous forum can retrieve their temporary password here, (login and check your PM).

Raman spectroscopy

Migrated topic.

Loveall

Established member
Donator
Senior Member
Recent years have seen a lot of activity and instrumentation development dedicated to Raman spectroscopy.

Thanks to advances and accessibility to economical solid state photodetectors and optical fibers, individuals are starting to build their own Raman systems.

Is there is any interest from some nexians to collaborate on this?

Here is a setup on witch I'm thinking to pull the trigger on:

Laser
- 532nm green laser with beam focused on optic fiber ($50)


Light transmission:
-3-way optical fused (FBT) fiber to connect the laser/sample/spectrometer. On the way in the fiber is singlemode to select 532nm. On the way back from the sample the fiber ends in a collimated connection that is sent to the spectrometer ($200).

Spectrometer:​
-Sony IMX385 based camera ($300). Supposedly can detect a handful of photons (!)
-Camera lens ($150)
-600 lines/mm refraction diffraction grating ($150)
-Long pass edge filter ($150)




So looking at a $1K investment. Any thoughts?


Edit: components are continually being changed. Updates in the thread.
 
Taking first baby Raman steps.

I ended up finding a cheap surplus spectrometer and got one of the many cheap 532nm green lasers that are on the market. Both together where <$250. Tested the laser source without any optics and it is giving a narrow peak (~10 pixels FWHM, and encouragingly one can identify the single peak pixel reliably).

A small step, but it is encouraging how narrow and clean the unfiltered raw laser line is. The cheap spectrometer does seem to have a couple hot pixels near ~700 pixel index (roughly the wavelength in nm but it is not fully calibrated yet), but that is not a show stopper.

Next, need to install some optics to remove the laser (edge filter) and help collect Raman emission (converging lenses). The spectrometer needs calibration and maybe an internal alignment adjustment. I've found a good edge filter for $70 and that's coming soon. I'll post more if I make more progress or anyone has questions.
 

Attachments

  • greenlaser.jpg
    greenlaser.jpg
    147.3 KB · Views: 0
yo, I’m gonna follow your lead on this project, bought the same components, and two 532nm bandpass filters ($18 ). For comparison, I will include ftir scans on the same samples. 😁
This is a learning opportunity for me, as I know very little about raman.
reading Raman Spectroscopy for Chemical Analysis (2000), I found out that lasers with outputs from 5mW to 5 watts may be used, so 200mW is quite suitable.
 
I used an old 12V netgear router power cord and spliced it into the open connection that comes with the laser (this open board connection was not pictured at the vendor site, but fortunately it came with the laser). Only one power polarity works since it is a diode (I got it wrong the first time, nothing was damaged). See picture below.

I'm working now on collecting optics after the laser goes through the sample. After some time thinking of this is the best I can come up with:

12.7mm (aka half inch or aka SM05) optical tube to easily put everything together (~$50)
Fiber optic SMA attachment for the optical tube ($25)
12.5mm edge filter, these seem to be good quality based on info in the science madness forum (~$45)
12.5mm AR coated converging lens (~$35). The kind designed to project a point source into a point projector which is double convex. I may get a couple lenses, don't know. I plan to play around with a free optical ray trace simulation to see what makes more sense as far as focal length (e.g., optical cable is 200um and accepts .22 NA rays I believe). Maybe I'm just splitting hairs at this point. This is the last piece I need to decide to order.
I may need a couple extra retaining rings to hold the filter and lens(es) in place inside the tube (only $4 each).

There is also the option of simply getting a pre-aligned SMA collector lens for which there is also a larger more expensive version. It should work but that is really designed to collect a collimated bean and the cheaper version has a clear aperture of 5mm which is a smaller than one gets with the optical tube ~10mm, meaning one can collect scattered Raman light over an angle 4x bigger. I'm going ahead down the optical tube path, but this other option is worth mentioning. For a similar price I prefer the optical tube because:
1) It is flexible/up-gradable (e.g. if one gets a working Raman, would swapping in achromatic lenses help put more light into the focal point?, or Could one could attach an Iris in the optical tube see if that cleans-up the signal?, etc.)
2) Attaching the bare edge filter should be very straight forward the optical tube (simple slide it in in front of the collecting lens(es)).

Last step would be to align the collector tube and the laser. I hope this is not too difficult, but have not thought about it much yet.

Where did you find the $18 band pass filters? I still need to order one or two of these. I'm not 100% they are needed since the green laser looks so clean, but folks say it is needed so I'm probably wrong and will need to get them.
 

Attachments

  • IMG_20180623_123225120_LL.jpg
    IMG_20180623_123225120_LL.jpg
    3.6 MB · Views: 0
e.bay 😁

I need to read much more into it, since you're already employing other optics. From what I've read, the bandpass filter is
typically used at the source, before the focusing lens. It's suggested, especially with DPSS lasers like 532, to filter out the IR wavelengths. Makes sense, as it would reduce noise.

The DIY scheme being proposed here is very attractive, as it is modular, and various optics/lasers could be swapped out depending on the application.
 
Right, the excitation Neodymium wavelength around 1064nm which I think is what most 532nm lasers use. That would be outside the spectrometer range (we would not see it), but it would heat up everything and generate noise. Some lasers filter that out internally better than others, not sure of the cheap laser we are using. I put my hand in front of the laser and could not sense any heat (for whatever that is worth). I think I'll test in both filtered and non filtered configuration.

I also got green laser safety glasses ($10). It is a must for anything beyond 5mW, and there have already been a couple times the laser hit the glasses as I was moving things around. Safety first :) .
 
Loveall said:
I also got green laser safety glasses ($10). It is a must for anything beyond 5mW, and there have already been a couple times the laser hit the glasses as I was moving things around. Safety first :) .
Safety first indeed!

Best to invent in some real glasses:

[YOUTUBE]

Kind regards,

The Traveler
 

this is the other module I got. Not exactly cheap, but should be suitable for Raman. I'll definitely need to get some goggles. I've read that response is a bell curve across the wavelength spectrum, dropping significantly at higher wavelengths (>650 nm) when using Si-CCD as a detector. Hence, I went with high-output.

What made you choose 1/2" lenses over 1"? portability, I suppose.
See pg. 96 on Raman Spectroscopy for Chemical Analysis (McCreery, 2000). Fiber optic sampling seems most
conducive/simplified to this CCD detector. Here's a general schematic for a probe

how about something like this, for solution analysis?

also, here's a great page on Raman from Thorlabs.
 
Wow, 800mW, you are not kidding around benzyme. :d

The links and attachments are very helpful and interesting, many thanks. :thumb_up:

The light tight cubet holder is really interesting. Also the cuvetes look really good. Honestly it looks a lot better than my current approach optical tube up against a clear glass vial being shot at by a laser. Are you going to have the laser output be by optical fiber?

The optical fiber (immersion?) sensor is teesting too. How cool would it be to analyze an extract by simply dipping a Raman probe into it? And I thought digital pH meters were cool... :p

I went with 1/2" because of price and what I could find on eBay and what I knew at the time. Each lens/filter is a little cheaper. Also thinking (guessing) that abberatioms will be smaller with lower aperture, but of course one captures less scatteres light so the net effect on signal is not clear (and one could allways add an iris or fancier lenses on the lareger system to optimize the signal). A 1" system could very well be better, what committed me was the edge filter (as I found a hopefully good cheap one which was 12.5mm).

For sure, all this is very upgradable and expandable. If the basic setup works upgrades are many. Could become an expensive hobby.
 
I was thinking of going the all-fiber route for simplicity, but the tube route (i.e. probe) makes sense for scanning surfaces .
and you aren’t kidding, the price adds up quick. a mounted beamsplitter from thorlabs is over $500 :shock:
Trying to determine whether or not I should get an optics breadboard for all the mounts.

I love browsing their site. so many cool toys, so little expendable funds (I’d hear about it from the wifey). I’ve already dumped a bit
into the analytical instruments I have, and am trying to justify Raman. As if I wasn’t already considered a mad scientist, seeing lasers in the lab would make me certifiable.
 
benzyme said:
I love browsing their site. so many cool toys, so little expendable funds (I’d hear about it from the wifey). I’ve already dumped a bit
into the analytical instruments I have, and am trying to justify Raman. As if I wasn’t already considered a mad scientist, seeing lasers in the lab would make me certifiable.

:lol:
 
This has to be one of the most exciting threads here at the moment. It may even inspire me to improve my income stream as this would be near the top of the list of fun toys to add to my collection.

Do we have any other Europe-based participants for this?
 
heh, yeah I got bit by the spec bug real bad.
After a bit of thought, this is very enticing due to the lack of sample prep needed, and no requirement
for high vacuum. Only one thing irks me about this analytical technique: there is no standardized sampling method.

I emailed sales@thorlabs to get a parts list for a probe like this
Components of a Raman Spectrometer - B&W Tek which would be suitable for scanning surfaces and glass containers, even amber vials.
 
If you're going to go the fiber route, I suggest getting a couple of these

Trying to get spectrum-studio running in Virtualbox on a macbook pro. I'm envisioning a setup where I can put all components in a backpack or man-purse, and analyze on the go.
*edit* I successfully got Spec studio running on the mb pro. Thorlabs never got back to me,
so I’m just going to wing it with that schematic.
 
This is fascinating. I'm going down the optical tube path, which may have been a mistake, but I'm committed now since parts are in order. Setup will be different than the fiber probe, so we can compare how they do and maybe learn moar about different options (we have the same spectrometer so comparisons would be meaningful I think).

Update on where I'm at:

1) Done: Laser and spectrometer up and running. Got the long-pass filter from ebay

2) In progress: I have ordered the 1/2" optical tube and parts to fill it up: laser cleanup filter, focusing lenses, SMA adapter (thorlab) and small cuvettes (ebay, since standard 12.5mm cuvette will not fit in the tube). Playing around with the optical ray tracer program led me to go for the aspheric collector lenses with 8mm of focal length. Setup description is below. One downside is that when the light hits the edge pass filter it is not collimated, so the long filter performance will not be optimal (worst case I need to change up the lens setup to collimate the beam before the edegpass filter and then add 3rd convergent lenses). Note that collimating the beam before the edgepass filter would also allow a 2x focused laser path through the cuvette since the edge pass filter is dichroic and should reflect the laser light).

3) Future work: I still need a way to align the fully loaded optical tube and the laser. After poking around I may get some stuff from the gun industry :surprised . They have very cheap systems to align optics on gun scopes using what they call a "picatinny" system. I think that would be sufficient and work, but not sure. Seems like the cheap route so worth a try I think.


Here is the simple optical tube setup modeled in optical ray tracer (open the spoiler to see all the details, it can be copied/pasted into the program and parts can be moved around - if anyone finds a better simulation let me know. Interested in a simple setup where collimated light hits the edge filter).

# OpticalRayTracer 9.6
# * OpticalRayTracer Home Page
# 2018.07.01 14:23:00 EDT
# Raman optical tube V0.0
# Notes:
# Laser coming from the left would be smaller (~2mm)
# Laser has been made wide to simulate the Raman emission in the cuvette
# If the bandpass filter is dichroic, Raman emission moving left should be reflected back
# Raman emission occurs everywhere in the cuvette the laser hits
# Only showing emission where the laser is focused in the cuvete
# Collimating the light hitting the edge pass filter could be a improvement (2x laser pass)


program {
antialias = false
askBeforeDeleting = true
beamAngle = 0.000000e+00
beamCount = 20
beamWidth = 1
clipboardGraphicXSize = 1280
colorArrow = -2147483393
colorBaseline = -16760832
colorBeam = -4194304
colorGrid = 1086374080
colorHighBackground = -1
colorLensOutline = 276865279
colorLensSelected = 268484608
colorLightSource = -16776961
colorLowBackground = -16777216
colorTerminator = -16777216
decimalPlaces = 4
defaultWindowHeight = 600
defaultWindowWidth = 900
dispScale = 0.015251
dispersionBeams = 8
divergingSource = false
helpScrollPos = 0
interLensEpsilon = 1.000000e-06
intersectionArrowSize = 0.000000e+00
inverse = false
maxIntersections = 64
selectedComponent = 6
selectedTab = 0
showControls = true
showGrid = true
snapValue = 0.000000e+00
surfEpsilon = 5.000000e-04
tableLineLimit = 500
virtualSpaceSize = 100.000000
windowX = 100
windowY = 100
xBeamRotationPlane = 0.000000e+00
xBeamSourceRefPlane = -30.000000
xOffset = 12.189982
yEndBeamPos = 3.500000
yOffset = 0.061811
yStartBeamPos = -3.500000
}

object {
active = true
angle = 0.000000e+00
centerThickness = 0.100000
dispersion = 59.000000
function = 2
ior = 1.520000
leftCurvature = 3
leftSphereRadius = 10.000000
leftZValue = 20.000000
lensRadius = 0.100000
name = Terminal Plane
rightCurvature = 3
rightSphereRadius = 10.000000
rightZValue = 20.000000
symmetrical = true
thickness = 0.100000
xPos = 33.962500
yPos = 0.005000
}

object {
active = true
angle = 0.000000e+00
centerThickness = 4.000000
dispersion = 59.000000
function = 0
ior = 1.520000
leftCurvature = 3
leftSphereRadius = 28.100000
leftZValue = 20.000000
lensRadius = 6.250000
name = LongpassFilter 540AELP
rightCurvature = 3
rightSphereRadius = 28.100000
rightZValue = 20.000000
symmetrical = true
thickness = 4.000000
xPos = 18.420800
yPos = 0.000000e+00
}

object {
active = true
angle = 0.000000e+00
centerThickness = 7.499915
dispersion = 59.000000
function = 0
ior = 1.499000
leftCurvature = 2
leftSphereRadius = 15.600000
leftZValue = 20.000000
lensRadius = 6.350000
name = ACL12708U-A
rightCurvature = 2
rightSphereRadius = 6.706500
rightZValue = 100.000000
symmetrical = false
thickness = 1.600000
xPos = -6.578100
yPos = 0.000000e+00
}

object {
active = true
angle = 0.000000e+00
centerThickness = 7.499915
dispersion = 59.000000
function = 0
ior = 1.499000
leftCurvature = 2
leftSphereRadius = 6.706500
leftZValue = 100.000000
lensRadius = 6.350000
name = ACL12708U-A Inverted
rightCurvature = 2
rightSphereRadius = 15.600000
rightZValue = 20.000000
symmetrical = false
thickness = 1.600000
xPos = 10.412000
yPos = 0.000000e+00
}

object {
active = true
angle = 0.000000e+00
centerThickness = 6.000000
dispersion = 59.000000
function = 0
ior = 1.520000
leftCurvature = 3
leftSphereRadius = 28.100000
leftZValue = 20.000000
lensRadius = 6.350000
name = FL05532-10 Bandpass
rightCurvature = 3
rightSphereRadius = 28.100000
rightZValue = 20.000000
symmetrical = true
thickness = 6.000000
xPos = -15.640300
yPos = 0.000000e+00
}

object {
active = true
angle = 0.000000e+00
centerThickness = 0.500000
dispersion = 59.000000
function = 0
ior = 1.520000
leftCurvature = 3
leftSphereRadius = 28.100000
leftZValue = 20.000000
lensRadius = 15.000000
name = Beckman 758407 wall
rightCurvature = 3
rightSphereRadius = 28.100000
rightZValue = 20.000000
symmetrical = true
thickness = 0.500000
xPos = -2.251000
yPos = 0.000000e+00
}

object {
active = true
angle = 0.000000e+00
centerThickness = 0.500000
dispersion = 59.000000
function = 0
ior = 1.520000
leftCurvature = 3
leftSphereRadius = 28.100000
leftZValue = 20.000000
lensRadius = 15.000000
name = Beckman 758407 wall
rightCurvature = 3
rightSphereRadius = 28.100000
rightZValue = 20.000000
symmetrical = true
thickness = 0.500000
xPos = 2.251000
yPos = 0.000000e+00
}

object {
active = true
angle = 0.000000e+00
centerThickness = 4.000000
dispersion = 59.000000
function = 0
ior = 1.330000
leftCurvature = 3
leftSphereRadius = 28.100000
leftZValue = 20.000000
lensRadius = 15.000000
name = Liquid Sample
rightCurvature = 3
rightSphereRadius = 28.100000
rightZValue = 20.000000
symmetrical = true
thickness = 4.000000
xPos = 0.000000e+00
yPos = 0.000000e+00
}
 

Attachments

  • raman+v0.jpg
    raman+v0.jpg
    195.3 KB · Views: 0
Back
Top Bottom