• State Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 2 answers
  • Subscribers 27 subscribers
  • Views 1391 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Wavelength / Spectral Range Conversion Question

Tom Myers2
Intellectual 475 points
Other Parts Discussed in Thread: DLP4500NIR

Hi,


I am checking over the final specifications of the NIR scan before making a purchase and I just want to double check the spectral range.

If I take the following Raman Spectrum for methamphetamine, then it has a range approx. 400 through to 1800 cm-1. By my calculation this is a nanometer equivalent of 25000nm to 5555nm, in which case the spectral range of the NIR scan being 1350nm to 2450nm would be too limited.

I suspect an error in my conversion, so I would welcome any thoughts and comments.

Thanks,

Tom

  • 0
    TI__Mastermind 43185 points

    Hello Tom,

    Your numbers are correct.  400 - 1800 cm-1 translates to 5555  - 25000 nm in wavelength.

    The DMD window in NIRscan unit is optiminzed for the range from 700 - 2000 nm for greater than 96% transmission per window pass and greater than 90% for 2000 - 2500 nm.

    Beyond this the efficiency is considerably lower.  This is a fundamental diffraction grating property since the wavelenght of light is on the same order of size as the the µ-mirrors themselves.  Even if you switched out the grating to match the range you desire, the efficencly would be much lower. 

    Generally Raman Spectrometery is already working with very small amounts of light so that this unit may not be the best match for your application.

    Fizix

  • 0 in reply to Fizix
    Intellectual 475 points

    Hi,


    So the conversion from cm-1 to nm does not have to take into account the excitation source wavelength?


    Just for my clear understanding - the wavelength range that I wish to operate in then the wavelengths are actually larger than a single mirror? This may sound like a uneducated suggestion, but can more than one mirror not be switched at one time to be the full width of the wavelength at the sacrifice of resolution?

    I do not suppose you do a DMD with "larger" mirrors?

    Thanks,

    Tom

  • 0 in reply to Tom Myers2
    TI__Mastermind 43185 points

    Hello Tom,

    No the excitation does not play into the conversion. It will have it's own wavenumber.

    The measure "cm-1" can be thought of as the spatial frequency (as oppossed to temporal frequency) of a wave.  This is analogous to the time domain where you have frequency (cycles/sec) and period (sec/cycle).  The wavenumber in cm-1 tells you number of cycles/cm and the wavelength is in length/cycle (typically in nm/cycle for visible light and µm/cycle for IR).  Just as period is the reciprocal of frequency, wavelength is the reciprocal of wavenumber.

    Your concept of using multiple µ-mirrors has some merit, and there have been some SPIE articles over the last several years exploring this methodology.  The angular distance between primary orders is asin(l/d) where d is the pitch of the grating.  This makes it immediately obvious that if l > d then the only real order in the space is the 0th order.

    However, you can introduce a periodic pattern on the array that has a "super-period" greater than d.  For example, 2 on, 1 off, 2 on, 1 off, . . . .  This has a "super-period" of 3d which will introduce secondary orders (sub-orders) in the diffraction pattern spaced asin(l/3d) apart.  However the sub-orders will not be nearly as bright. To start with you are throwing away 1/3 of the light  (i.e. 1/3 of the µ-mirrors are off). 

    The brightest of these sub-orders will be the one nearest the blaze condition.  In addition the brightness of the sub-orders are affected by the on/off ratio, with the max somewhere near 50% on/off.  Moreover, the larger the "super-period" the greater then number of sub-orders to disperse the light into since they are more closely spaced.

    The question then becomes how much light can you get to your detection system from one of these "sub-orders" and will it be enough to work with? The assumption made was that since you are doing Raman, you may not be able to afford to throw away very many photons.

    This may have been a bit convoluted, but hopefully helps.

    Fizix

  • 0 in reply to Fizix
    Intellectual 475 points

    Hello Fizix,


    Thank you for a very comprehensive answer. I am very sad to learn that the DMD is not suitable for my application.

    Just for my reference what types of compounds is the DMD aimed at? There are some common bonds of many compounds outside the seemingly limited range of the DMD.

    Thanks,

    Tom

  • 0 in reply to Tom Myers2
    TI__Mastermind 43185 points

    Hello Tom,

    The range achievable by the DLP4500NIR is aimed at direct transmission/refection aborption spectra in the 700-2500 nm range.  Quality control of products such as olive oil, or distinction of different plastic compounds that have signatures in this region are just a few examples.

    We are actively looking at Raman using this system to determine over what range it can be effective.

    I suspect that the methodology that you suggest may work fairly far out (up to 10 µm possibly) for direct transmission reflectance absorption.

    Thank you for the feedback.  We definitely consider this type of feeback in our development cycles.

    Fizix

  • 0 in reply to Fizix
    Intellectual 475 points

    Hello,


    Please could you clarify/expand upon your statement:

    "I suspect that the methodology that you suggest may work fairly far out (up to 10 µm possibly) for direct transmission reflectance absorption."


    Thanks,

    Tom

  • 0 in reply to Tom Myers2
    Intellectual 475 points

    Hello,


    I think there has been a major over sight here! The cm-1 is a shift from the excitation laser and therefore is not directly converted to wavelength as I previously suggested and you confirmed.

    I did more learning and found this: - https://www.princeton.edu/cefrc/Files/2011%20Lecture%20Notes/Alden/Lecture-8-Raman.pdf

    Which suggests if I use a 785nm laser I can look at molecules ranging from 0cm-1 upto at least 4500cm-1 at this corresponds to an actual spectral range 785nm to 1213nm which falls within the working limits of the NIR DMD and therefore the NIR scan may be suitable for my application?


    Do you agree?


    Regards,

    Tom

  • 0 in reply to Tom Myers2
    TI__Mastermind 43185 points

    Tom,

    This is very good news indeed!  If this turns out to be the case this means that the Raman response wavelengths fall squarely in the range for which the NIRscan was designed. 

    Please let us know how it goes.

    Fizix

  • 0 in reply to Fizix
    Intellectual 475 points

    Hi,


    Do you agree with my expetctation of how Raman will work?

    I notice from your literature that the DMD window has good transmission from 700nm upto around 2500nm. However, from looking at the NIR scan documentation the spectral range is shown from around 1300nm to 2500nm, could I get it to operate below 1300nm if I modified the software?

    Thanks,

    Tom

  • 0 in reply to Tom Myers2
    TI__Mastermind 43185 points

    Hello Tom,

    I should slightly revise what I said previously.  The range of wavelengths you expect is squarely in the range for which the DMD is optimized.

    The NIRscan grating and optics cover the range it lists (1350 nm to  2450 nm).  You would have to change the grating and it's angle and  likely the two lenses also to work in a the lower wavelength range.  It is very difficult to make a spectrometer that cover a range from wavelenth X to 2X+ since oders will begin to overlap

    So the answer is that this is a limiitation of this particular implementation , but not a fundamental DMD limitation.

    Fizix