Raman Spectrometer and Software, is it possible?

[Hive 8]

How hard would it be to make a raman spectrometer with software, sorry i am a complete noob in terms of building Spectrometers for chemical/biological analysis. I would like to build one based on a laser with a great resolution, i have some money set aside to make prototypes. Please let me know if that is possible and if someone is interested in guiding/helping me. I also have a desktop CNC mill and a CNC lathe.

[Jack Summers]

I will not claim to be an expert on raman spectrometry, but here are a couple of things to consider;

1.  You can buy off the shelf portable raman spectrometers these days at reasonably low cost.  Stellernet has a system for $3500.   (see their raman offerings at http://www.stellarnet-inc.com/PopularConfigurations_ramansystems.htm)   B&WTech and Ocean Optics offers spectrometers as well.

2.  For those who dont know what raman is, here is my rudimentry description (possibly some misconceptions in my understanding) ...   Raman spectroscopy involves measuring light scattered by molecules where the frequency of the scattered light is different than that of the light input to the sample. In the absense of an enhancement mechanism, raman spectra are very weak and unlikely to be detected by home built spectrometers.  The data can be greatly enhanced if the frequency of the input radiation overlaps the absorbance band of your molecules.  This is called resonance raman.  Resonance raman spectroscopy involves scattering of light by molecules via a mechanism where a photon is absorbed and then emitted at a different wavelength.  The energy difference between the absorbed and emitted photons is (usually) distributed into molecular vibrations.  Thus, if you excite with a laser (single wavelength) and measure the wavelengths of the scattered photons, you can determine some of the vibrational energies that are available to the molecule in question.  This, of course means that the molecule you are interested in must absorb light at the wavelength that your laser emits (a visible laser will give data on molecules that are colored).  The information from raman spectra and IR spectra are complementary in that they both involve the vibrational energies of the molecules.  They do have selection rules where some vibrational modes show up in IR and others are only observed in raman.  Both IR and raman give sharp peaks that are much more informative (and more specific to your analyte) than the broad absorbances seen with UV-vis. Another benefit of raman is that it uses either UV or vis excitation and detection.  That means you can use glass / plastic optics and there is not a problem with water absorbing your radiation.  

3.  Raman bands are never going to be as intense as Rayleigh scattering (scattering at the same wavelegth as excitation) so it is necessary to completely resolve the raman bands from the laser frequency.  

4.  The difference between the wavelengths of the raman bands and the laser will be relatively small (perhaps a few nm. Separation from the laser energy equals the energies of the molecular vibrations).      This means that you will have to have high wavelength resolution to make this work.

5.  I think the required wavelength resolution means that you will need your grating to be further from the camera. 

6.  I thinkk you will also need to have a slit betweeb the grating and detector so you can remove the laser frequency.  If you dont remove it, I would expect stray light problems.

7.  I think the software would not be a special problem.

8.  I have a textbook called Instrumental Methods of Analysis by Willard, Merritt, Dean, and Settle (I have the sixth ed, copyright 1981).  There is a lot of good  information in there about optics and laying out a spectrometer.  I will see what I can find out about making some of this available (copyright problems, right?).

 

Good luck,

Jack

[Spacecaptain]

Raman would be so incredible to crack for DIY analysis.

I do belive that we would need to increase manufacturing accuracy by an order, as probably also the complexity of the device would increase. Wouldn't the biggest problem be to have acces to a tunable IR Laser, as I understand you need to scan through a wavelength window to hit the right vibrational resonating frequency?
Probably you couldn't do a <100$ kit. What does such a Laser cost? Is it actually economically feasable?
As for eliminating the Laser nominal wavelength, I guess you could shade (eclipse) the part of the grating that covers that precise wavelength, so that you don't sense this input on your camera.

As for sensitivity, maybe a low Lumen spycam would do a better job at reading the lines than a webcam.

Just thinking out loud....

-- Post to this group at publicla...@googlegroups.com. To unsubscribe, email publiclaborato...@googlegroups.com. Options at https://groups.google.com/d/forum/publiclaboratory?hl=en

[Amirber]

Hi Hive8 and Jack,

I am currently using Stimulated Raman Spectroscopy (SRS) in my
studies. In our setup we use the SRS to prevent ionization so we
don't have a problem with blocking Rayleigh scattering which is a
major problem for optical uses, you could look at laboratory equipment
vendors for 532 nm blocking filter, but I was told those are not
cheap.

It is customary to use wave-numbers (WN[cm^-1]) rather than wavelength
(WL[nm] etc). If you use a laser source you could find the vibration
energy from the wave-number difference.
Most small biological molecules we have used have fundamental
vibrations in the region of 0-4000 wave-numbers. A pump beam at 532 nm
and a 4000 wave-numbers vibration would result at a ~640 nm photon
emitted. If you wish to calculate a specific vibration the mathematics
is quite simple:
WN[cm^-1] = 1E7/(WL[nm])
WN[Vibration] = WN[pump]-WN[stokes]

You can ignore the rotational bands, those will not show without
intense cooling and very high resolution. As for increasing
resolution, getting the grating further from the camera will improve
it just until you reach the diffraction limit (or less for less than
perfect optics), once again - it's evil Rayleigh, Raleigh's criteria
sets the relation between the maximal resolution to the wavelength
and aperture.

Hope I helped a bit,

Amir

[Jeffrey Warren]

Hey all - what would you be trying to identify with a Raman spectrocope? Something different than the uv-vis-ir ones we're already working on? 

It sounds like a neat experiment but personally I worry about getting sidetracked or trying too many things all at once... I'll probably spend most of my time working towards a really rigorous proof of concept contaminant identification with our existing prototypes. 

My sense is that if that can be demonstrated and reliably performed, that would be a huge step forward. I'd *love* to be able to tell people with confidence that they can build/buy one of these things and test soil or water in their neighborhood -- something i've been really careful not to promise... yet.

Jeff

[Ron Huber]

Jeff, We too look forward to the DIY spectrometer as we track down and mitigate the waste pits and other pollution deposits   around Penobscot Bay, Maine, that the 19th and 20th centuries left behind.  We'll also be using the balloon mapping kit to examine the many intertidal flats for waste discoloration. 

Keep up the good work, Jeff et al!
Ron Huber

On Wed, Sep 19, 2012 at 12:31 PM, Jeffrey Warren <je...@publiclaboratory.org> wrote:

Hey all - what would you be trying to identify with a Raman spectrocope? Something different than the uv-vis-ir ones we're already working on? 

It sounds like a neat experiment but personally I worry about getting sidetracked or trying too many things all at once... I'll probably spend most of my time working towards a really rigorous proof of concept contaminant identification with our existing prototypes. 

My sense is that if that can be demonstrated and reliably performed, that would be a huge step forward. I'd *love* to be able to tell people with confidence that they can build/buy one of these things and test soil or water in their neighborhood -- something i've been really careful not to promise... yet.

Jeff

On Wed, Sep 19, 2012 at 12:15 PM, Amirber <ami...@gmail.com> wrote:

Hi Hive8 and Jack,

I am currently using Stimulated Raman Spectroscopy (SRS) in my
studies.  In our setup we use the SRS to prevent ionization so we
don't have a problem with blocking Rayleigh scattering which is a
major problem for optical uses, you could look at laboratory equipment
vendors for 532 nm blocking filter, but I was told those are not
cheap.

--
Ron Huber
Friends of Penobscot Bay
POB 1871
Rockland Maine 04841

e: coast...@gmail.com
tel: 207-691-7485

cell 207-593-2744
web: www.penbay.net

[Mathew Lippincott]

Here is an extreme long-shot for Raman spectroscopy that could be very low cost-- I wonder if this team at the University of Cambridge could print a sheet of low-power lasers of different frequencies?

I don't feel competent enough to e-mail the lead authors, Professors H.J. Coles and Ian Hutchings, maybe someone else wants to take this on?

This suggestion is thanks to my dad, who wants to be helpful without butting into my work ;-)

[Amirber]

Hi Mathew, I doubt if this "paper laser" source will have the
necessary qualities, such as low bandwidth, wide spectral range, high
power and so on. I have found a paper dealing with SERS (http://
connection.ebscohost.com/c/articles/15382273/surface-enhanced-raman-
spectroscopy-laser-pointer-light-source-miniature-spectrometer) which
seems quite promising, but would require a notch filter. I don't know
if I can post the full paper here due to copyright stuff...

Amir

On Sep 19, 9:01 pm, Mathew Lippincott <mat...@publiclaboratory.org>
wrote:

> Here is an extreme long-shot for Raman spectroscopy that could be very low
> cost-- I wonder if this team at the University of

> Cambridge<http://pubs.rsc.org/en/Content/ArticleLanding/2012/SM/C2SM26479J>

> could
> print a sheet of low-power lasers of different

> frequencies<http://www.bbc.co.uk/news/science-environment-19641112>
> ?

>
> I don't feel competent enough to e-mail the lead authors, Professors H.J.

> Coles <http://www.eng.cam.ac.uk/~hjc37/> and Ian
> Hutchings<http://www.ifm.eng.cam.ac.uk/people/imh2/>,

> maybe someone else wants to take this on?
>
> This suggestion is thanks to my

> dad<http://www.linkedin.com/profile/view?id=6760895>,

> who wants to be helpful without butting into my work ;-)
>
>
>
>
>
>
>
> On Wed, Sep 19, 2012 at 10:59 AM, Ron Huber <coastwa...@gmail.com> wrote:
> > Jeff, We too look forward to the DIY spectrometer as we track down and
> > mitigate the waste pits and other pollution deposits   around Penobscot
> > Bay, Maine, that the 19th and 20th centuries left behind.  We'll also be
> > using the balloon mapping kit to examine the many intertidal flats for
> > waste discoloration.
> > Keep up the good work, Jeff et al!
> > Ron Huber
>
> > On Wed, Sep 19, 2012 at 12:31 PM, Jeffrey Warren <
> > j...@publiclaboratory.org> wrote:
>
> >> Hey all - what would you be trying to identify with a Raman spectrocope?
> >> Something different than the uv-vis-ir ones we're already working on?
>
> >> It sounds like a neat experiment but personally I worry about getting
> >> sidetracked or trying too many things all at once... I'll probably spend
> >> most of my time working towards a really rigorous proof of concept
> >> contaminant identification with our existing prototypes.
>
> >> My sense is that if that can be demonstrated and reliably performed, that
> >> would be a huge step forward. I'd *love* to be able to tell people with
> >> confidence that they can build/buy one of these things and test soil or
> >> water in their neighborhood -- something i've been really careful not to
> >> promise... yet.
>
> >> Jeff
>

> >> On Wed, Sep 19, 2012 at 12:15 PM, Amirber <amir...@gmail.com> wrote:
>
> >>> Hi Hive8 and Jack,
>
> >>> I am currently using Stimulated Raman Spectroscopy (SRS) in my
> >>> studies.  In our setup we use the SRS to prevent ionization so we
> >>> don't have a problem with blocking Rayleigh scattering which is a
> >>> major problem for optical uses, you could look at laboratory equipment
> >>> vendors for 532 nm blocking filter, but I was told those are not
> >>> cheap.
>
> >>> --
> > Ron Huber
> > Friends of Penobscot Bay
> > POB 1871
> > Rockland Maine 04841
>

> > e: coastwa...@gmail.com

[Spacecaptain]

Found a similar article using a green Laser pointer and an Ocean Optics
spectrometer:

http://academic.reed.edu/physics/courses/Physics332.s10/pdf/Raman_Spectroscopy.pdf

Seems that the setup would be very similar to the idea for UV
fluorescence spectrometer already investigated in the wiki:
http://publiclaboratory.org/notes/warren/7-18-2012/fluorescence-oil-spill-residue-diverse-spectrometer-use
http://publiclaboratory.org/notes/joshmc/4-28-2012/setup-uv-testing-specrtrometer

The question is: can we substitute the Ocean Optics spectrometer with
our own kitted machine? We would need to filter out the source
wavelength from the Laser. I worry about the glare and we will have to
phisically block this wavelength from arriving at the detector with a
filter (they discuss this in the linked article)

Answering to a previous mail from Jeffrey, where you asked what would we
be achieving with these spectrometers in comparison to the actual
design, I drew back on my experience with FTIR. Bothe FTIR and I guess
Raman habe libraries of spectra of a wide range of organic compounds to
which to compare your sample to. When going to collect samples in the
field, there will be a need for a collection and measurement protocol,
but even like this, you will end up with samples that will contain
hundreds or thousands of compounds. This will make the analysis very
complicated and you may only be able to identify metals in the visible
spectrum with the regular kit and maybe some high concentration organic
compound families, but I have some doubts on this.

What Raman and FTIR does for you is to produce very discrete and
accurate peaks that identify functional groups, such as radicals and
cyclic structures. You may not be able to pinpoint a single compound
against a database unless it is present in a pure form, but you will be
able to find out PAHs and polar molecules, hidroxile groups, amines,
ketones, aldehydes. Raman allows you to access vibrational modes of a
molecule, which can be very varied and I have less experience on how you
read these, but I would bet that you can get very telling fingerprints
(see Raman database links below)

So in a nutshell, Raman and FTIR are excellent analytical methods for
fingerprinting pure organic molecules against a database of known
compounds. FTIR would need the building of a Michelson interferometer,
which is rather tricky. If Raman is easier to implement, it would be a
worthwhile development that would allow us to tap into a wealth of
compound databases. The sample collection protocol could be refined to
be able to separate organic molecules, via simple chormatography for
example, in case you would need further charatcerisation of the
molecules in presence.

Some Raman and IR databases:
http://riodb01.ibase.aist.go.jp/sdbs/cgi-bin/cre_index.cgi
http://minerals.gps.caltech.edu/files/raman/
http://www.knowitall.com/literature/default.asp?gclid=CKiyzZ7sw7ICFY5TfAodXFEAaQ#Databases

[Jeffrey Warren]

Very cool, esp the article on the green laser pointer. Did you see these notes I posted?

http://publiclaboratory.org/notes/warren/7-18-2012/fluorescence-oil-spill-residue-diverse-spectrometer-use

Green laser making oil spill sample fluoresce: http://spectralworkbench.org/spectra/show/377

that last one i had to use a specially modified spectrometer with a long-exposure shot on my Canon S90, because the light was so relatively weak.

also http://publiclaboratory.org/notes/warren/7-26-2012/oil-residue-preparation-spectroscopy

Jeff

  -- Post to this group at publiclaboratory@googlegroups.com. To
unsubscribe, email publiclaboratory+unsubscribe@googlegroups.com. Options
athttps://groups.google.com/d/forum/publiclaboratory?hl=en

-- Post to this group at publiclaboratory@googlegroups.com. To unsubscribe, email publiclaboratory+unsubscribe@googlegroups.com. Options at https://groups.google.com/d/forum/publiclaboratory?hl=en

-- Post to this group at publiclaboratory@googlegroups.com. To unsubscribe, email publiclaboratory+unsubscribe@googlegroups.com. Options at https://groups.google.com/d/forum/publiclaboratory?hl=en

[Jack Summers]

Thanks to Amir for input.  Amir, what kind of enhancement are you using?

I spoke with a colleague who is also a 'real' raman spectroscopist.  He recommended considering surface enhanced raman (as in the Can J Chem Article Amir sent the ref for).  In this technique, light from the laser is bounced off a roughened silver surface  My colleague tells me that raman bands of molecules adsorbed on the silver are enhanced significantly when compared to other methods.  He also brought up the issue of fluorescence as a problem.  He said that the sharp raman bands would be superimposed over the broader fluorescence bands complicating the quantification of the material present.  Since the raman bands are sharp, it should be easy to determine what is in the sample qualitatively.  So, it sounds like fluorescence would not be too big of a problem (for determining what is present) as long as you do not saturate your detector.  (I did not get whether the fluorescence is an issue iwth the surface enhanced raman or just with resonance raman). 

 

Jack

 

  
On Wednesday, September 19, 2012 5:54:23 AM UTC-4, Hive 8 wrote:

[Amirber]

I'm always happy to brag and link to a paper I helped write (right now
we are working hard on the followup): http://pubs.acs.org/doi/abs/10.1021/jz300026a
The method we use is called ILSRS, where a UV laser can be used to
selectively ionize different species of the same molecule (in pure
state), two other lasers are used to scan for a vibration resulting in
a depletion of ionization. Thou I must say I have never used CARS or
SERS.

The big bonus for using Raman is that the area of fundamental
vibrations can be observed in visible light, the 0-4000 WN corresponds
to an IR laser at ~2500-3500 nm, so you could access a hole other area
of the spectrum, one which is much more interesting (to me at the
least). As for the paper by Etchegoin et.el. I think this could be
reproduced, I will try it out on a sample of R6G with a simple
spectrometer we have here, this is used in dye lasers and has a very
strong fluorescence as you can see in this photo:
http://en.wikipedia.org/wiki/File:Beer%E2%80%93Lambert_law_in_solution.JPG
This was used in the papare by Van Duyne, and it seemd thay had good
results.

As for the solution databases, take in consideration that Raman
activity is different for gas, liquids and solids, it is also
dependent on the temperature and pressure, so it is important to
consider how the spectra presented at the DB were made. A good source
for FTIR / FTRaman is Sigma Aldrich which produces quite a lot of
them, for example Tryptamine on which we are working now:
http://www.sigmaaldrich.com/catalog/product/sigma/t2891?lang=en&region=IL

Spectrum: http://www.sigmaaldrich.com/spectra/rair/RAIR003328.PDF

Mathew - do you think it would be possible to add a sort of a
monochromator with a spectral filtering capability? if we could
mechanically block the 532nm area maybe there won't be a need for an
expensive notch filter. By the way - how long do your long exposures
last? would you expect the samples to be analysed on the field or back
at the "home base"?

Amir

[Spacecaptain]

@Jeffrey, yes I had seen the page with the fluorescence post, but did not throw a closer look to that spectrum until now. A green pointer laser was used, am I right? I do not see the 532nm peak, has it been removed a posteriori or did the perpendicular setup take care of that one. Otherwise I would expect it to have overwhelmed your sensor with it's glare? I wonder what the peaks that are shown hint at...
An interesting experiment with that the setup would be to try the same R6G dye (Rhodamine 6G) Amir has mentioned, as it seems to be a good calibration dye. Here it's florescence spectra: http://www.springerimages.com/Images/Biomedicine/1-10.1007_s10895-009-0489-4-8

Have a look at this paper that talks about a methodology for testing oil spills with raman fluorescence spectrometry: http://www.earsel.org/Advances/3-3-1995/3-3_17_Patsayeva.pdf

@Jack: So according to your collegue, there is no unsurmountable problem in trying to build a cheap SERS?
SERS (surface enhanced raman spectrometry) has the advantage that it's signal is a lot stronger (for some lines, depending on the experiment and surface structure), but reading up on it I have noted that creating the sample to be measured is not straightforward and particularly, systematizing this process could be problematic, as the shape and size of the roughness is an important factor for each experiment. Also, doesn't the sample need to be adsorbed on the roughened surface prior to analysis? It seems like an intricate process that has to be recreated for each experiment, am I wrong?
Does your collegue have some input on how these roughened silver surfaces (or other metals, copper would be nice) could be reproduced consistently and cheaply?

This paper walks you through the creation of roughened silver surfaces and measuring of human serum albumin:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2737399/?tool=pubmed

Here a nice, short video explaing what the SERS does and why it works:
http://www.youtube.com/watch?v=JgGXgaxtao4

@Amir, you seem to know your way around this Raman thing. For simplicity's sake would it be easier to prepare samples in liquid form, be it pure or diluted in some sort of solvent (water/ethanol/acetone)? Temperature is a parameter we would have to factor in. Maybe Raman is to finicky to turn into a field equipment. Making a desktop version would be more controllable. All those Raman and FTIR databases that are avilable online are a very powerful resource that we should try to use.

-- Post to this group at publicla...@googlegroups.com. To unsubscribe, email publiclaborato...@googlegroups.com. Options at https://groups.google.com/d/forum/publiclaboratory?hl=en

[Jeffrey Warren]

Hmm. I've read about Rhodamine 6G but i've heard it's pretty toxic. I've also read somewhere that it will bind with lead and allow for lead identification with spectroscopy though. 

Hey all -- we've been talking about making a dedicated spectroscopy mailing list. If I created one, would folks be OK moving some of the deeper spectrometry discussions over there? The list is just starting to get quite busy on this topic. We can keep CC'ing in the main publiclaboratory list on the first post of a given thread, perhaps. 

Jeff

[Spacecaptain]

Fine with me.
I think this discussion is getting a little dense on the subject and non-interested people may get swamped.
I also have my doubts on how easy it would be to make one of these sensitive devices on a budget. Notch filters are expensive and monochromators are tricky to build, similar to interferometers for FTIR. But I am not an optics guy, I'd love to be proven wrong!

[Jack Summers]

My colleague did not appear to have serious objections but he was busy with a lab, so he did not have a lot of opportunity to thnk about possible problems.  He was concerned about the dynamic range of the camera (when compared to teh ccd's in a commercial instrument).  He also suggested buying a "real" difraction grating (about $60 from Edmund optics). He told me that the way to make the roughened silver (at least initially) was with electrochemically roughened silver mirror (as described in the pubmed abstract you provide).  I believe that silver mirrors are produced by mixing aqueous solutions of formaldehyde with silver nitrate.  I think this is done commercially using a two-nozzle spray gun, but could be done on small scale using medicine droppers and a microscope slide.  This should be a pretty cheap process.  I dont know about the surface to surface reproducibility of spectra.  I did note that the Can J Chem paper used silver produced by chemical vapor deposition (not something the average person will have in their garage).  I am not sure about the notch filter.  It seems to me that you might be able to get away from the notch filter using either a high pass filter (and just get the Stokes lines) or a low pass filter (for the anti-Stokes lines).

Jack

[Spacecaptain]

I think I will give it a try by using, as you recommmend, a high pass filter to capture the Stokes emissions, as these are going to be the stronger pnes and also are not (as much?) temperature dependant. Also I will probably be using a 405nm Laser pointer (what power would be fine?) as the Raman emission intensity is a funcion of λ^-4 (http://www.horiba.com/scientific/products/raman-spectroscopy/tutorial-faqs/raman-faqs/what-laser-wavelengths-are-used-for-raman-spectroscopy/) so keeping as low as possible in the excitation wavelength is the good choice both for chosing the highpass filter as well as getting the best emission intensity.

If Jeffrey doesn't mind, I have created a Raman specific page in the wiki, so that we can gather all the good information we have been discussing here in one accessible place: http://publiclaboratory.org/wiki/raman-spectrometry

[Jack Summers]

Hi,

I think the Can J Chem paper will probably be a good source to see what is possible.  It is nice that they list manufacturers, model numbers and prices for their components (Table 1 page 1437). I notice that Figure 1 indicates they are using a notch filter in the minitaturized system while Table 1 indicates a high pass filter.  Since the model numnber starts with LP, I suspect it is, in fact, a long pass filter.  The laser pointer they use emits at 532 nm and thier long pass filter has to cut that out and allow longer frequencies through.  I looked through my Edmond Optical catalog for an appropriate filter for the 405 nm wavelength you are considering.  The short answer is that I would consider a TechSpec OD4 notch filter.  Specs on this are; CWL; 405.0 nm, full width at half max; 20 +/- 2 nm, Transmission range: 325-540 nm.  A 12.5 mm dia filter (stock no A67-107) will set you back $150.  If you have $395 to put into the filter, consider the TechSpec OD-6 notch filter which has a fwhm of 9 nm.  That fwhm variable basically tells you how close you are going to be able to get to the laser wavelength.  According to my calc's, having a 20 nm fwhm means you can observe wavelengths outside of a 390 to 415 nm window.  Using Amir's formulae for raman shift, exciting at 405 and observing at 415 gives a raman shift of ~600 cm-1.  That would be the weakest vibration that you would be able to detect, and I think that will be plenty good enough.  As far as long pass filters go, I did not see any in my catalog that would work nearly as well (at least on paper) for the cost. Yoiu may be able to find something from another source.  The thing to look at is how close you can get to the laser line.

Another notable thing in the paper is the use of fiber optics.  I suspect their system is a whole lot simpler to set up than what you envision simply because they are starting with a fiber optic spectrometer.

Good luck,

Jack

[Spacecaptain]

Hi Jack,

For a long pass filter I found this one that starts transmitting at 420nm (25%) upwards:
http://www.asahi-spectra.com/opticalfilters/syousaik2_ver2.asp?key=XUL0422
But it still costs 200$, so not a cheap solution and it would cut off a bit more the observable range.
On the ther hand, did I get you right on the transmission range of your filters? 325-540 nm, that would cut quite a big swath to be observed from green upwards, did I understand you right? Unless one would not expect any Raman emissions further in the higher wavelengths, wouldn't that be throwing away a lot of information?
In the video demonstration that I linked to, the builder just obfuscated the blue 443 Laser line by mechanically eclipsing it.

I think I will start with the most basic setup and continue adding higher quality elements, documenting improvements at every step. I think the first improvement will be a holographic grating, of which I don't know if I will chose a transmission or reflection style. Transmission would probably make the eclipsing of the lasing emission unpractical:
http://www.edmundoptics.com/optics/gratings/reflective-holographic-gratings/1621

The next improvement will be replacing the webcam with one of these:
http://www.eureca.de/pdf/optoelectronic/spectronic-devices/EURECA_USB-Boards.pdf

It may be doable with both of the above improvements to aim for rather nice, little Raman spectrometer around 200$

Cheers!
Fernando

On 09/21/2012 04:51 PM, Jack Summers wrote:

Hi,

I think the Can J Chem paper will probably be a good source to see what is possible.οΏ½ It is nice that they list manufacturers, model numbers and prices for their components (Table 1 page 1437). I notice that Figure 1 indicates they are using a notch filter in the minitaturized system while Table 1 indicates a high pass filter.οΏ½ Since the model numnber starts with LP, I suspect it is, in fact, a long pass filter.οΏ½ The laser pointer they use emits at 532 nm and thier long pass filter has to cut that out and allow longer frequencies through.οΏ½ I looked through my Edmond Optical catalog for an appropriate filter for the 405 nm wavelength you are considering.οΏ½ The short answer is that I would considerοΏ½a TechSpecοΏ½OD4οΏ½notch filter.οΏ½ Specs on this are; CWL; 405.0 nm, full width at half max; 20 +/- 2 nm, Transmission range: 325-540 nm.οΏ½οΏ½A 12.5 mm dia filter (stock no A67-107) will set you back $150.οΏ½ If you have $395 to put into the filter, consider the TechSpec OD-6 notch filter which has a fwhm of 9 nm.οΏ½ ThatοΏ½fwhm variable basically tells you how close you are going to be able to get to the laser wavelength.οΏ½ According to my calc's, having a 20 nm fwhm means you can observeοΏ½wavelengthsοΏ½outside of a 390 to 415 nm window.οΏ½οΏ½Using Amir's formulae forοΏ½raman shift, exciting at 405 and observing at 415 gives a raman shift of ~600 cm-1.οΏ½ That would be the weakest vibration that you would be able to detect, and I think that will be plenty good enough.οΏ½ As far as long pass filters go, I did not see any in my catalog that would work nearly as well (at least on paper) for the cost. Yoiu may be able to find something from another source.οΏ½ The thing to look at is how close you can get to the laser line.

Another notable thing in the paper is the use of fiber optics.οΏ½ I suspect their system is a whole lot simpler to set up than what you envision simply because they are starting with a fiber optic spectrometer.

Good luck,

Jack

On Friday, September 21, 2012 3:49:49 AM UTC-4, spota wrote:

I think I will give it a try by using, as you recommmend, a high pass filter to capture the Stokes emissions, as these are going to be the stronger pnes and also are not (as much?) temperature dependant. Also I will probably be using a 405nm Laser pointer (what power would be fine?) as the Raman emission intensity is a funcion of οΏ½^-4 (http://www.horiba.com/scientific/products/raman-spectroscopy/tutorial-faqs/raman-faqs/what-laser-wavelengths-are-used-for-raman-spectroscopy/) so keeping as low as possible in the excitation wavelength is the good choice both for chosing the highpass filter as well as getting the best emission intensity.

If Jeffrey doesn't mind, I have created a Raman specific page in the wiki, so that we can gather all the good information we have been discussing here in one accessible place: http://publiclaboratory.org/wiki/raman-spectrometry

On 09/20/2012 11:32 PM, Jack Summers wrote:

My colleague did not appear to have serious objections but he was busy with a lab, so he did not have a lot of opportunity to thnk about possible problems.οΏ½ He was concerned about the dynamic range of the camera (when compared to teh ccd's in a commercial instrument).οΏ½ He also suggested buying a "real" difraction grating (about $60 from Edmund optics). He told me that the way toοΏ½make the roughened silver (at leastοΏ½initially) was with electrochemically roughened silver mirror (as described in the pubmed abstract you provide).οΏ½ I believe that silver mirrors are produced byοΏ½mixing aqueous solutionsοΏ½of formaldehyde with silver nitrate.οΏ½ I think this is done commercially using a two-nozzleοΏ½spray gun, but could be done on small scale using medicine droppers and a microscope slide.οΏ½ This should be a pretty cheap process.οΏ½ I dont know about the surface to surface reproducibility of spectra.οΏ½ I did note that the Can J Chem paper used silver produced by chemical vapor deposition (not something the average person will have in their garage).οΏ½ I am not sure about the notch filter.οΏ½ It seems to me that you might be able to get away from the notch filterοΏ½using either a high pass filter (and just get the Stokes lines) or a low pass filter (for the anti-Stokes lines).

Jack

οΏ½

On Thursday, September 20, 2012 3:16:09 PM UTC-4, spota wrote:

@Jeffrey, yes I had seen the page with the fluorescence post, but did not throw a closer look to that spectrum until now. A green pointer laser was used, am I right? I do not see the 532nm peak, has it been removed a posteriori or did the perpendicular setup take care of that one. Otherwise I would expect it to have overwhelmed your sensor with it's glare? I wonder what the peaks that are shown hint at...
An interesting experiment with that the setup would be to try the same R6G dye (Rhodamine 6G) Amir has mentioned, as it seems to be a good calibration dye. Here it's florescence spectra: http://www.springerimages.com/Images/Biomedicine/1-10.1007_s10895-009-0489-4-8

Have a look at this paper that talks about a methodology for testing oil spills with raman fluorescence spectrometry: http://www.earsel.org/Advances/3-3-1995/3-3_17_Patsayeva.pdf

@Jack: So according to your collegue, there is no unsurmountable problem in trying to build a cheap SERS?
SERS (surface enhanced raman spectrometry) has the advantage that it's signal is a lot stronger (for some lines, depending on the experiment and surface structure), but reading up on it I have noted that creating the sample to be measured is not straightforward and particularly, systematizing this process could be problematic, as the shape and size of the roughness is an important factor for each experiment. Also, doesn't the sample need to be adsorbed on the roughened surface prior to analysis? It seems like an intricate process that has to be recreated for each experiment, am I wrong?
Does your collegue have some input on how these roughened silver surfaces (or other metals, copper would be nice) could be reproduced consistently and cheaply?

This paper walks you through the creation of roughened silver surfaces and measuring of human serum albumin:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2737399/?tool=pubmed

Here a nice, short video explaing what the SERS does and why it works:
http://www.youtube.com/watch?v=JgGXgaxtao4

@Amir, you seem to know your way around this Raman thing. For simplicity's sake would it be easier to prepare samples in liquid form, be it pure or diluted in some sort of solvent (water/ethanol/acetone)? Temperature is a parameter we would have to factor in. Maybe Raman is to finicky to turn into a field equipment. Making a desktop version would be more controllable. All those Raman and FTIR databases that are avilable online are a very powerful resource that we should try to use.

On 09/20/2012 05:54 PM, Jack Summers wrote:

Thanks to Amir for input.οΏ½ Amir, what kind of enhancement are you using?

I spoke with a colleague who is also aοΏ½'real' raman spectroscopist.οΏ½ He recommended considering surface enhanced raman (as in the Can J Chem Article Amir sent the ref for).οΏ½ In this technique, light from the laser is bounced offοΏ½a roughened silverοΏ½surfaceοΏ½οΏ½My colleagueοΏ½tells me that raman bands of molecules adsorbed on the silver are enhanced significantly when compared to other methods.οΏ½ He also brought up the issue of fluorescence as a problem.οΏ½ He said that the sharp raman bands would be superimposed over the broader fluorescence bandsοΏ½complicating theοΏ½quantification of the material present.οΏ½ Since the raman bands are sharp, it should be easy to determine what is in the sample qualitatively.οΏ½ So, it sounds like fluorescence would not be too big ofοΏ½a problemοΏ½(for determining what is present) as long as youοΏ½do not saturate your detector.οΏ½οΏ½(I did not get whether the fluorescence is an issue iwth the surface enhanced raman or just with resonance raman).οΏ½

οΏ½

Jack

οΏ½

οΏ½οΏ½

On Wednesday, September 19, 2012 5:54:23 AM UTC-4, Hive 8 wrote:

How hard would it be to make a raman spectrometer with software, sorry i am a complete noob in terms of building Spectrometers for chemical/biologicalοΏ½analysis. I would like to build one based on a laser with a great resolution, i have some money set aside to make prototypes. Please let me know if that is possible and if someone is interested in guiding/helping me. I also have a desktop CNC mill and a CNC lathe.

-- Post to this group at publicla...@googlegroups.com. To unsubscribe, email publiclaborato...@googlegroups.com. Options at https://groups.google.com/d/forum/publiclaboratory?hl=en

-- Post to this group at publicla...@googlegroups.com. To unsubscribe, email publiclaborato...@googlegroups.com. Options at https://groups.google.com/d/forum/publiclaboratory?hl=en

[Jeffrey Warren]

On Fri, Sep 21, 2012 at 3:49 AM, Spacecaptain <spacec...@gmail.com> wrote:

If Jeffrey doesn't mind, I have created a Raman specific page in the wiki, so that we can gather all the good information we have been discussing here in one accessible place: http://publiclaboratory.org/wiki/raman-spectrometry

The wiki is there for everyone in the community to use for documenting research. Please use it! :-)

I created a new list and will be a little annoying fwding things back and forth for the next week or so, and I'll also send specific invitations to folks who have been very spectro-active.

https://groups.google.com/group/plots-spectrometry

It's also now a default-on option on the registration form: http://publiclaboratory.org/user/register (you have to be logged out to see)

Jeff