NIR absorbance spectra of tourmaline
Page 1 of 3

Author:  Brian [ Thu Apr 29, 2010 5:00 pm ]
Post subject:  NIR absorbance spectra of tourmaline

It is end of school year, so I have a couple hours to play around. What to do, what to do?

Bruce Fry alerted me to this G&G article, which describes how near-infrared (NIR) absorbance spectra can be used to identify copper-bearing tourmalines. The article describes how copper content can be inferred quantitatively using reflection spectroscopy. But I just have a transmission spectroscopy setup, so that my spectra can only provide qualitative results. That is, absorbance NIR spectrum from transmitted light can be used to determine if copper is likely present in amounts above the detection limits of a LA-ICP-MS measurement, but you'll still have to get a LA-ICP-MS measurement done if you want to know how much copper is present. Or send the tourmaline to someone with a reflection spectroscopy setup, and they can tell within somewhat large error bars what is the copper content.

Examining the article, I picked up the following points concerning the NIR absorbance spectrum of tourmaline.

*It seems both iron and copper can produce wide peaks at around 700nm (the border between red and infrared). The copper peak is usually (but not always) shifted slightly lower than 700nm, whereas the iron peak is usually shifted above 700nm. Also the iron peaks tends to be wider than copper peaks. But overall, it is really difficult to look at one spectrum and decide if the peak you see at 700nm is due to copper or due to iron.

*It seems that copper alone produces a wide peak at around 900nm. Larger amounts of copper produce a higher peak at 900nm than at 700nm.

*There is also some evidence in the article that manganese produces a wide peak at around 975nm. At least there is a very wide peak occuring at 975nm, that is not associated with copper, in the spectra of manganese-bearing tourmalines.

Well, so ammyguy has sent me some really small tourmaline chips... a yellow chip (ampa) and a blue chip. Below are (relative) absorbance spectra for the two, covering the red to NIR range from 600-1000nm. Oops, in the graph I labelled the yellow tourmaline as red... confusing myself with the red colored curve.


Note that both have peaks around 700nm: the yellow chip has a very wide peak located slightly above 700nm, and the blue chip has a somewhat narrower peak located slightly below 700nm

Note that only the blue chip has a significant peak around 900nm. Also note that after 900nm the blue chip's signal descends as wavelength increases (so that there is no peak at 975nm... in fact it is a minimum).

Also, in this blue chip's spectrum, the 900nm peak is about as high as the 700nm peak. As seen below in the spectrum of another blue chip he sent, the 900nm peak is higher than the 700nm peak.


So my conclusion is that it is not worthwhile trying find copper in the yellow chip... the 700nm peak is most likely due to iron. On the other hand, if the blue chips were of any significant size, it might be worthwhile to quantify the copper in them.

Finally, I am not encouraging anyone to send me tourmaline. Bruce Fry has offered to spectralize tourmalines for some nominal price. I like to try these things out once, but I get bored making measurements on the same things.

Author:  bruce_tourm [ Thu Apr 29, 2010 7:58 pm ]
Post subject:  Relative size of iron and copper peaks in tourmaline

I spoke with Paul the principle author of the G and G article. We discussed the relative size of the about 700 nm peak and the about 900 peak with copper/iron. The relative size depends not only on copper/iron absorption, but also the crystallographic orientation of the crystal when you take the measurements. When I was working with Dr. Rossman investigating Laurellite, the cuprian Elbaite that is a reverse Alexandrite color changer, he wanted a sample that was oriented with the c axis so the copper absorption measurement of Laurellite could be quantified by using a micro beam along with the spectrometer. I was unable to get such a sample and that ended his interest in the research. I have high quality absorption curves for Laurellite, both the c axis and the a/b axis that were taken at the SSEF lab in Switzerland. I would be happy to send you a copy if your interested Brian, when I get my principle computer repaired.


Author:  Brian [ Thu Apr 29, 2010 8:09 pm ]
Post subject: 

Thanks for that information Bruce.

I was thinking that the size of the peaks was kind of sketchy, as I noticed that I could change that peak height a bit by nudging the chip a bit one way or the other.

So gem orientation with respect to the light is another factor that has to be accounted for when trying to quantify the absorption.

Author:  Brian [ Tue May 11, 2010 1:01 pm ]
Post subject: 

Here are some more spectra of little pieces of tourmaline distinguished by their color.


The purple, turquoise, and blue pieces all show show a distinct peak at 900 nm. Of the several green pieces, only one (labeled "green 1") shows a distinct peak at 900 nm. The one labeled "green 2" shows a spectrum typical of those seen in the other green pieces.

Just a few interesting details to add. Everyone on these boards knows of the "if I heat this purple it'll turn turquoise" phenomenon. So I found it quite interesting that the purple and turquoise NIR spectra look essentially identical. In fact, I could have laid one curve directly on the other.

After testing the purple, I thought that the blue pieces would probably show a similar NIR spectrum, because there were some mixed-blue-purple pieces... blue at one end and purple at the other. Then I found it interesting to see that the blue NIR spectrum wasn't exactly the same as the purple's.

The green was a real surprise... after looking at a few greens and seeing the iron peak, I decided to try one last piece that seemed to glow a bit more than the others... and voila! it had a NIR spectrum nearly identical to the blues! Makes me wonder if the blues would heat and turn to green?

Author:  bruce_tourm [ Tue May 11, 2010 4:51 pm ]
Post subject:  Heating cuprian absorption changes come from Manganese

Hi Brian,

I have see the same flattening of the 900nm absorption curved with some cuprian tourmaline I have tested. I don't think that it is dependant on color, but on the richness of the gem's tone. The flattened curves also tend to get ragged and I was beginning to think that I not have enough light to analysis darker stones. The spectrometer's manufacture tested some of my gems and feels that I have enough light, but I will be upgrading my system with an integrating sphere and different light fibers.
The difference between the tourquois and purple cuprian tourmaline is not in the NIR as you have seen, but depends on the level of absorption at about the 520nm and comes from Mn+3. When the Mn+3 is reduced by heating to Mn+2, the purple will turn to tourquois, unless there is enough Ti+4 to form an IVCT reaction with the Mn+2(absorption in the lower 400nm). Then it turns green. At least this is what Dr. Rossman has proposed. If the cuprian tourmaline is blue and has very little titanium/manganese, I don't think that it will turn green by heating, but the whole question of why cuprian tourmaline is green, outside of being partially colored by iron is a question in my mind. Cuprian tourmaline that is green naturally has not been found to have it's color changed significantly by heat. One researcher told me that if there is enough Mn+2 in the cuprian tourmaline the gem will be green without the titanium. Whichever way the cuprian is caused to be green, copper alone can not cause cuprian tourmaline to be green. I have found that absorption curve in tourmaline between the long ultrviolet and about 450nm to be both complex and interesting. I have found two absorption peaks merging into one peak, which may be caused by the absorption of Mn+2, Mn+2-Ti+4 IVCT and or a color center caused by radiation. I would very much like to know the chemistry of these absorption peaks. In particular I want to know why some cuprian tourmalines are green threw heating or naturally.


Author:  Brian [ Tue May 11, 2010 7:32 pm ]
Post subject: 

Great information, Bruce.

I agree with your supposition that the flattening out of the 900nm peak in a couple spectra is likely due to the fact that there is no more light there to absorb.

The details about green color are interesting. Makes sense to me that the copper does not create the green, but it seems to add some je ne sais quoi to the stone. It was very easy to spot among the iron-colored stones.

Author:  Jason Barrett [ Wed May 12, 2010 12:31 am ]
Post subject: 

thanks for the report Brian..this kind of info adds greatly to the GO forum!!!

Author:  Alberto [ Wed May 12, 2010 7:47 am ]
Post subject: 

Neat report Dr. Brian, thank you! :D
i hope to have my HR4000 online (sooner or later) to start playing with spectras too....

Author:  bruce_tourm [ Wed May 12, 2010 9:14 am ]
Post subject:  A case history of a green tourmaline

Thanks Brian,

Sometimes a case history is interesting and I have a good one with a bright grass green round that I have prized for many years.

Color has always been important in my hobbies and I was drawn to tourmaline by it. I got back into lapidary after many years with the expressed intent of completing as much of the color wheel with tourmaline colors as I could. Now green is not the hardest color to get in tourmaline, but when it began to come out of a new deposit in Mozambique with other rarer colors, like lavender, I jumped on it. All I could finded out about the deposit was that it was an old dried up river bed that was hard to mine. Its location was hidden because the owner didn't want rough dealers tempting his men to steal and selling the best material to them.

The piece of green rough I received was heavily water worn and was filled with a web of rather uniform feathers. Still the color seemed good and bright, so I was pleased with the purchase. It hadn't cost that much because of the inclusions. It cut well and yielded a standard ten mm round without any really distracting flaws. I immediately loved the color, but the brightness of the gem was the real distraction. It had to be somehow different. This happened before the GIA discovered copper, as a chromophore in gem quality tourmaline from Mozambique, in samples that I supplied them. With the discovery of copper in the blue/purple reverse Alexandrite color changers, I was sure that my bright green gem was also cuprian, but I could not affort to prove it.

While the tourmaline were being tested by various groups for the GIA, I made contact with a laboratory that has done a lot of work with tourmaline using a micro probe and would be doing the GIA work. Later, when I went to visit them, I brought a selection of interesting tourmaline which included good old green. As the researcher screened my gems with an electron microscope, to get a qualitative idea of which gems to quantify, old green failed to show any copper. I was disappointed, but I was still amazed by the green which was now certainly an Elbaite without any exceptional chromophores. I looked at my top quality sea foam from Afghanistan, that practically glows in the dark, and told myself that iron alone can make fantastic tourmaline. Still old green wouldn't let me go.

I made a trip to LA to look over a specially designed spectrometer and discuss color. One of the stones I had looked at on the spectrometer was good old green. The gemstone produced a wonderfully clean smooth absorption curve in the visible range, that was a classic pattern for geen tourmaline. Another strike agains good old green being cuprian.

Next I was introduced to a researcher that was an expert in the use of spectrometry for chemical analysis. We discussed the influence of iron as a chromophore in some cuprian tourmaline and his idea that the 900nm absorption curve in the NIR was definitive for copper, while the 700nm absorption curve was common to both iron and copper. With the publication of his paper in Gems and Gemology, I decided to purchase a spectrometer like he used in his work. My collection had to be tested for copper and I also wanted to work with color beyond what my old eyes can see.

Now the end game is upon us and I put good old green under the fiber optic probe of my new spectrometer. The 900nm peak rose in stately spendor. Green was cuprian! My eyes had not deceived me for most of 10 years. Why had it failed the previous tests and kept me wondering all these years. It turned out that all the copper concentrations were very low on my gemstones that were tested with the electron microscope and I don't think that it takes very much copper to make a bright yellow green gem. The spectrometer in LA was designed for only the visible range of light and therefore could not separate iron from copper absorption at the 700nm wavelength peak.

Looking closely at the cuprian tourmaline, I think that brightness and instability in color under different white lights is more an indication of copper content than pure color. It certainly is with good old green, that still leaps out of the tray to greet you.


Author:  Brian [ Wed May 12, 2010 2:49 pm ]
Post subject:  Re: A case history of a green tourmaline

brucetourm wrote:
Looking closely at the cuprian tourmaline, I think that brightness and instability in color under different white lights is more an indication of copper content than pure color. It certainly is with good old green, that still leaps out of the tray to greet you.

Yep, that's what I meant with my French.

Now I'd like to ask you about any times that you've had the opposite experience. That is, have you found two stones whose colors and brightness look identical, where you find a 900nm peak in the spectrum of one but not in the other? If so, when you compare the VIS portion of the spectra, do you note any differences there?

Author:  bruce_tourm [ Wed May 12, 2010 4:36 pm ]
Post subject:  lets look at two yellow greens

It is nice to agree in two languages, Brian.

In my search I found two yellow greens that appear, to my eye, to be identical in color. One is round and the other one is oval. The 700nm absorption peak is present in both, but only the round shows the 900nm copper peak. The only other VIS absorption peaks are between say 400nm and 450nm and vary quite a bit in both yellow greens and oranges. I don't have the graphs in front of me, but I think all the absorption peaks were of similar intensities with the two yellow greens. I also found reds , some with and some without indications of copper. As we have spoken before you have to be careful with tourmaline that has a lot of Mn+3 because it has some absorption around 900 nm. I only claim the red tourmaline has copper if the slope change, between a gentle ramping up toward 900nm that is probably from Mn+3 and the 900nm copper peak, is distinct.

My sample of gemstones is limited, but I still strongly feel that copper makes a nicer gemstone, when properly cut, because generally it makes a brighter presence in tourmaline than iron. The biggest possible exception to this generalization is cyan. The paraiba like Afghanistan tourmaline can be outstanding, but I still feel that better quality cuprian has more visual punch.

Finally an example of another unusal tourmaline. It has a nice fruity orange color with a couple of absorption peaks in the 400nm to 450nm range. It has no absorption at 520nm which leaves out Mn+3 as a chromophore. The absorption graph bottoms out at about 600 nm, like many tourmaline that are not colored by chrome or vanadium. Then the graph indicates absorption that basically follows about a 20 degree slope off my chart at 850nm. The angle seems unusually high to me and I had a friend check its IR and it is correct for tourmaline. It doesn't look like either copper or iron absorption to me and I wish I could get it analyzed.


Author:  MK [ Wed May 12, 2010 4:49 pm ]
Post subject: 

Brian wrote:
The green was a real surprise... after looking at a few greens and seeing the iron peak, I decided to try one last piece that seemed to glow a bit more than the others... and voila! it had a NIR spectrum nearly identical to the blues! Makes me wonder if the blues would heat and turn to green?

funny, makes ME wonder if the green will irradiate to blue 8)

Author:  Brian [ Thu May 13, 2010 1:36 am ]
Post subject: 

MK wrote:
funny, makes ME wonder if the green will irradiate to blue 8)

Martin, one can always hope.

Bruce, I was just thinking before I posted last how I would be wary of any cuprian claim for a red tourmaline because of the 700nm peak that is usually shifted more into the visible, and because of the manganese possibility.

Alberto, you really need to clean up that workspace so you can set up your spectrometer.

Jason, you have got to stop sending me stuff that piques my curiosity. :wink:

Author:  Alberto [ Thu May 13, 2010 3:14 am ]
Post subject: 

Brian wrote:
Alberto, you really need to clean up that workspace so you can set up your spectrometer.

you really don't know HOW much i would do RIGHT NOW, neverthless it's not only a matter of space, still have no bench for samples but hope i'll have it soon... :wink: (read subliminal message here..LOL)

ciao and a BIG thank you to you and Bruce for your work!!!

Author:  bruce_tourm [ Tue Oct 19, 2010 12:33 pm ]
Post subject:  Re: NIR absorbance spectra of tourmaline

I have finally been able to purchase a one inch in diameter integrating sphere with culminating optics to focus on a small spot on the gemstone. I also changed my fiber optic arrangement to accomodate separate locations for the light to enter and leave the integrating sphere. I now place the gemstone table down on a fused silica microscope slide with high IR oil and attempt to determine the gemstones hue angle in CIELAB color space. The CIELAB is a graphic presentation of the visible area of the spectrum. It was developed years ago and uses polar coordinates to graph the hues/colors. (Software comes with the spectrometer to do this calculation.)

My only problems so far has been getting enough light through darker stones. I seem to be able to test medium dark tones, but the length of the vector from the origin is not very long. Still by varying the number of traces that is average for a data point and the integration time, I have gotten reasonable stable hue angles and chroma.

I am done playing around with the set up and now feel ready to do some significant testing once I am sure that I am calibrating the new set up correctly. You will certainly hear if some of the gems I proclaimed by eye to be the same prove to be the same and others, when I analyze them with the spectrometer.


Page 1 of 3 All times are UTC - 4 hours
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group