spectroscopy

[jordanc740]

My research question is: "What is the effect of different light sources on wavelength spectroscopy measured with homemade and commercial spectroscopes?" However, i'm not quite sure how i can measure this. I've been reading about how to read the lines of the spectrum but am confused by it. Please help.
Thanks!

[deleted-71588]

I'm not sure I understand your research question. I'm having trouble understanding how this differs from something like: "What is the spectrum produced by different light sources?" which can be answered by a literature search for many light sources. Have you formed a hypothesis? If you want to do a good science fair project, you need to first understand "The scientific method" and how to formulate "your question" so that it can be tested.

There are a lot of interesting things related to spectrums that can be great science fair projects; however, unless you come up with a testable hypothesis that is interesting, you will likely end up with a demonstration project that might take a lot of effort but isn't interesting from a scientific standpoint.

[jordanc740]

Thanks Craig..
I understand what your saying about my question. My original question was "What are the effects of different light sources on wavelenght spectroscopy?" However, my teacher reccomended I do a more hands on project. It is similar to the ideas on this site : http://www.arborsci.com/CoolStuff/cool22.htm
She thinks this project would be more worthhile. Any ideas on how to word this question? I was thinking I would look at the spectroscopy of fluorescent, incandescent, ane LED light
sources with the homemade spectroscopes. Then, my measurement would be how far off of the commercial spectroscopes measurement each was, to determine how effective they are. However, while completing my Experimental Research Plan, I ran into the problem of how to measure/graph this.. If you have any ideas or comments please reply to this post.
Thanks!

[deleted-71588]

If I'm interpreting your plan correctly, the plan involves building one or more homemade spectroscopes and then comparing the results obtained in measuring 3 different light sources with those homemade instruments and a commercial instrument.

The purpose of a spectroscope is to measure the amount energy in a small frequency range over some range of frequencies. From a calibration standpoint, you need to be concerned with three things, the ability to measure a frequency accurately, the ability to discriminate a narrow range of frequencies, and the ability to measure the incident energy within each narrow band of frequency accurately. From an operational standpoint, you also need to be able to measure over the entire range of desired frequencies and to measure a range of energy or power levels.

Your choice of light sources dictates the frequency range your instrument needs to handle and the peak power levels involved. The known spectrum of an LED source will be a very narrow band and fairly low power while the incandescent and flourescent sources will have a much wider band and probably have a lot more power.

Do you have some designs for homemade spectrographs you are considering? Their design will affect how you have to calibrate them. Calibration (or lack thereof) can easily be the biggest source of inaccuracy.

With prism and gradient spectrographs, calibrating the frequency is straight forward if you have a couple of known single frequency sources such as LED and or lasers at different ends of the desired spectrum.

Measuring intensity and Calibrating the measurement is a much more difficult task. Most simple instruments bypass most of the difficulties by only dealing with relative intensity judgements. These can be accomplished by utilizing a single light source and varying the distance from the source. Intensity should fall off as the square of the distance for omnidirectional sources.

How does one compare the accuracy between two instruments? You should set them up side by side and utilize the same light source and take a measurement with both instruments. For the LED light source measurement, comparing the center frequency (wavelength) measured and the intensity at the same distance will give you two pairs of numbers to compare. For the incandescent and flourescent sources, you will have to compare two sequences of number pairs. Doing this mathmatically is a bit beyond your grade level. You can simplify the comparison by comparing where the measurable spectrum starts and ends and picking a couple of peak intensities in between to compare.

BTW: most simple homemade spectrographs don't attempt to measure intensity. If the ones you choose to experiment with don't measure intensity your problem of comparision is simplified.

[jordanc740]

Yes it seems you are interpreting mw experiment plan correctly. However, I am a little confused about what your saying as to the three things to focus on with my measurement. See, all the research sources i've used fail to give me a direct answer. Once I am looking into the spectroscope, and i see the spectrum, what is it that I will record for my data? I was aware that the LED light is much less powerful that the other sources. For this reason, would this be a bad choice for this experiment?

Yes i do have some designs in mind for the homemade spectroscopes. They are basically the same as the ones on the link i said in my previous post.

Thanks for the advice on how to conduct the expermennt as well.

[deleted-71588]

I believe your confusion with my previous answer concerning more sophisticated spectroscopes used for analytical purposes comes from the lack of any means of measuring light intensity in a narrow band on the spectroscopes you are considering building

Yes i do have some designs in mind for the homemade spectroscopes. They are basically the same as the ones on the link i said in my previous post.

The spectrascopes made from using a CD as a reflective difraction gradient in your linked article does not have any provisions for measuring the light intensity in a given band.

This limits what can be measured to frequency (wavelength) properties only. In order to calibrate your homemade spectroscope frequency (wavelength) measurements, you will need to have a couple of known frequency (wavelength) light sources. A red and a green (or blue) LED with known wavelengths would be good choices to use for calibration. The typical low power red LED produces a 635 nm wavelength but there are some ultra red at 660 nm and super red at 633 nm. The most common green LEDs come in 555 nm, 560 nm, and 565 nm. The more rare blue LEDs come in 470 nm and 430 nm.

With the use of two calibration sources at different wavelengths, you can mark on your display the center of these two narrow band spectrums and indicate their numeric wavelength. From this, you can construct a scale to be able to predict what the wavelength of any other visible spectrum display is.

[jordanc740]

Thanks for the advice on the light sources. Are all of these LEDs available at a local hardware store? Also, by using the LEDs would i be elimeinating the oter ligth sources from my experiment?

How could I mark the center of the two readings on my display, and how would I create the scale to make these predictions?

Thanks.

[deleted-71588]

Are all of these LEDs available at a local hardware store?

Not likely. Places like Radio Shack are probably your best brick and mortar store front alternative. See http://www.radioshack.com/family/index. ... Id=2032298 for what their online selection is. There are lots of online electronics supply places where you can purchase LED's. Be sure to understand whether you need a current limiting resistor or if it is built into what ever ones you purchase.

Also, by using the LEDs would i be elimeinating the oter ligth sources from my experiment?

That is up to you and should depend on what hypothesis you are testing. Even if you don't utilize other light sources as part of your official experiment, you should try some out so that you can see what broad spectrums look like for yourself.

How could I mark the center of the two readings on my display, and how would I create the scale to make these predictions?

Once you have the spectroscope partially assembled, you can try it out using a single color LED (white is not a single color). In a dark room, you should see a fairly narrow band of color appear on whatever you are projecting it onto (maybe a piece of white paper attached to the end of a shoe box in the example). If you have one side off to allow access, you can use a pencil to mark where the center of say the red LED's band apears. Repeat for a green or blue LED. At this point, you have a simple geometry problem of "How do you divide a line into equal segments?". If you put this into a search engine like ask.com or look it up in a book on drafting (mechanical drawing), you should come up with some good methods.

[jordanc740]

Thanks for the advice. I understood everything you said except for the part about dividing the line into equal segments..
How would this be useful to me?

Thanks.

[deleted-71588]

If you want to figure out what wavelength the spectrom of some unknown is, then you need to be able to judge where it falls on the scale appropriate to your spectroscope.

If you wanted to measure how long something is in inches, you would use a ruler marked off in inches and parts of inches. You probably have done this in school with a ruler. Another name for a ruler is a scale.

For your spectroscope, you need to construct a scale that is marked off in wavelength units.

For example, if you used a known 635 nm Red LED and a known 470 nm Blue LED as your calibration sources, you would know precisely where 470 nm and 635 nm show up; however, you need some way to judge where things inbetween fall.

635 nm - 470 nm = 165 nm difference. If you divided this into 11 equal spaces, you would end up with a scale marked off every 15 nm because 165 nm / 11 = 15 nm. How did I come up with 11, I actually did this one backward and looked at all of the prime factors of 165 (3,5,11) and then chose which ones made the most sense to have a reasonable number of divisions.

Hope this helps.

[jordanc740]

Thanks.
The scale idea makes a lot more sense now as this solved many questions I had.
From the list of LEDs you sent me, would it be best to get the most intense red and least intense blue to create a wider scale?
If so, which ones would this be?
Lastly, how will i project the spectrum readings onto the paper to make this scale?

[deleted-71588]

From the list of LEDs you sent me, would it be best to get the most intense red and least intense blue to create a wider scale?

"intense" as in "intensity" of light sources refers to how much light power is produced. In the case of LED's, this is usually expressed in MCD (milli-candles) which compares it to 1/1000 th the amount of light produced by a single candle. For your project, the intensity of the light source is NOT that important to your project. What is important is that you know the wavelength of the light being produced and that you pick two different colors that are fairly far appart and that you can acquire easily at a reasonable price. Red is a great choice for one end because it is near the longest visible wavelengths. Likewise, a blue one is a great choice for the other end because it is toward the shortest visible wavelength (violet is shorter; however, I don't know of any way to produce purple LED's).

Lastly, how will i project the spectrum readings onto the paper to make this scale?

In the http://www.arborsci.com/CoolStuff/cool22.htm article, they show a "shoebox" reflective spectroscope where light comes in through a slit, reflects off the CD surface and projects to an end. If you leave the end of the shoebox in place and cover the inside of it with a piece of white paper and provide a viewing hole, you should be able to position the CD (e.g. your reflective gradient) such that light coming through the slit projects onto the paper.

[jordanc740]

Okay that all sounds great.
One thing though.. Is there a way with the other two spectroscopes on the Arbor Science site to project them onto a piece of paper?
Thanks.

[deleted-71588]

With a transmittance version spectroscope, I'd recommend using a piece of clear plastic as the viewing screen and some fine line markers to mark a scale on it. I'm probably missing it, but I only noticed two types on that site. If you could give me a paragraph number of figure number for the one I missed, I'll be glad to take a look.

[jordanc740]

The third spectroscope they sugggested is just positioning a CD to where the spectrum is visible on it. Could i just draw the scale onto the CD?
Thanks.

[deleted-71588]

The third spectroscope they sugggested is just positioning a CD to where the spectrum is visible on it. Could i just draw the scale onto the CD?

Personally I don't consider that a spectroscope because there is really no way to calibrate it. A drop of water, a prism, a piece of crystal, or anything that has fine equally spaced lines (e.g. a gradient) is capable of separating light into its component wavelengths. Without some means of controlling the incident angle such as a slit that is somehow held at a constant angle with respect whatever is acting like a prism or gradient, you can't control where the spectrum will appear (if at all).

You maybe able to take that idea and make it into a spectroscope by attaching the CD to something that can be used to hold it at a constant angle that also can act as a slit so that the spectrum is always displayed in the same place on the CD surface in which case you can draw the scale onto the CD.

[jordanc740]

Do you have an ideas for what to use to hold the CD in place?

[deleted-71588]

Maybe a small box with a slit in the end for light to enter, a diagonal cut on two sides to hold the CD at an appropriate angle to the slit (essentially match the angle you get to work just holding the CD and light source).