Public Lab Research note

Achieving a high quality spectrum

by Ankur_Bodhe | July 11, 2015 12:56 11 Jul 12:56 | #12048 | #12048

Hey Guys, So I wanted to use the spectrometer to Characterize Carbon Nanotubes. I used a 50 W halogen lamp and follwed the procedure as given in the manual of the DIY spectrometer by Public Labs. However, I am not able to achieve a sharp spectrum. I have tried: 1) Changing the angle of inclination of the gumstick camera 2) changing the diffraction grating 3) changing the distance between the grating and the slit 4) changing the focus of the gumstick camera Despite changing the bove mentioned parameters, I havent achieved satisfactory results. I have attached a picture of the spectrum I've got so far. Would like to know where is the problem and how to rectify it.



Hi, Ankur - nice spectrum. What would be an example of "ideal" results for you -- i just want to better understand exactly what you mean by sharp. Also, if you could link to your spectrum that would be helpful.

Reply to this comment...

Well There is a lot of blurring in the diffraction pattern that is formed.Also I have used the 0.12 mm collimated acetone slit. There is also a lot of overexposure of light and hence not giving an accurate spectrum. Since I am analysing Carbon Nanotubes, I need a well resloved spectrum inorder to find out absorbtion of specific wavelength. By Sharp I mean, that : A) Resolution of the pattern is better. B) No overexposure in certain regions of the spectrum C) No Blurring of the spectrum image I have attached a picture of absorbance graph of Carbon Nanotubes analysed by the Beckman Coulter DelsaMax PRO and one taken by Public lab spectrometer The picture shown below is by DelsaMax PRO CNT_1.jpg The picture shown below is by Public Lab spectrometer

SWNT_PL_spectrum.jpg the graph for the one taken by PL was arrived at downloading the CSV file of the sample spectrum and tthe blank spectrum and then using Beer lamberts law, the absorbance values were achieved. This was done using the spreadsheet. The Beer Lambert's Law given by A = log(Io/I) where Io - Intensity value of the blank (Cuvette with DI water) spectrum(ie, the average value got from the CSV file) I - intensity of the sample spectrum. Here the base 10 logarithm is used. The solution concentrations have been the same as the one shown by Beckman Coulter. As you can see, there is a lot of noise that is present in the absorbance graph. Hence I would like to know what can I do in order to achieve a more accurate graph.

Reply to this comment...

Hi, Ankur -

A) Resolution of the pattern is better

What is your hoped-for resolution? Can you link to your actual data on so we can take a look? One way to determine what resolution you're getting is to look at a fluorescent (CFL) lightbulb spectrum. @cfastie has published a couple notes on very high resolution mods of the basic PL spectrometer design -- and his suggestions were used to dramatically improve the v3.0 spectrometer:

B) No overexposure in certain regions of the spectrum

You can use an attenuator, as @stoft has described: - and also see his other research for gain correction and a proposal for HDR exposure correction.

C) No Blurring of the spectrum image

I think this is related to the first issue, but if there's a separate issue, can you elaborate?

Thanks for your input, Ankur! As with Chris Fastie's and Dave Stoft's suggestions, any improvements you can propose will be considered in upcoming revisions of the spectrometer kit design. We need all the input we can get -- that's how open source design works!

Is this a question? Click here to post it to the Questions page.

Reply to this comment...

Login to comment.