UV gem testing. Time to change!

[Garry H (Cut Nut)]

As a diamond dude with an interest and love of blue fluorescent diamonds, I know that cheap LED UV lights create a Strong UV effect in diamonds that GIA grades as Faint.

This also applies to many colored gems.
One of my real scientist friends who is also a gemologist wrote this email to me today when we were discussing the diamond issue.
Paraphrased from Grant Pearson FGAA:

"....for instance the red chromic fluorescence of ruby, red spinel & even of alexandrite etc. The erroneous use of the supposed classic 365nm LWUV wavelength mercury vapor tube has long thought to be the effective exciting wavelength.
From my own investigations the effective actual active excitation even for exciting ruby red fluorescence from an ordinary LWUV tube is a longer-wavelength minor emission because there is some about 395nm or so in the output. The effect is not from the long-supposed 365nm which is next to useless actually."

Back to diamonds - when you use one of these more expensive (still cheap) 405nm visible violet laser pointers almost all diamonds fluoresce - so much for GIA claims that about a third fluoresce!

 

[V_sh]

As a diamond dude with an interest and love of blue fluorescent diamonds, I know that cheap LED UV lights create a Strong UV effect in diamonds that GIA grades as Faint.

This also applies to many colored gems.
One of my real scientist friends who is also a gemologist wrote this email to me today when we were discussing the diamond issue.
Paraphrased from Grant Pearson FGAA:

"....for instance the red chromic fluorescence of ruby, red spinel & even of alexandrite etc. The erroneous use of the supposed classic 365nm LWUV wavelength mercury vapor tube has long thought to be the effective exciting wavelength.
From my own investigations the effective actual active excitation even for exciting ruby red fluorescence from an ordinary LWUV tube is a longer-wavelength minor emission because there is some about 395nm or so in the output. The effect is not from the long-supposed 365nm which is next to useless actually."

Back to diamonds - when you use one of these more expensive (still cheap) 405nm visible violet laser pointers almost all diamonds fluoresce - so much for GIA claims that about a third fluoresce!

Thanks for sharing Garry,

By the third kind of fluorescence you mean LWF, SWF, and something for high-energy visible light (405nm)?

That seems cool, but what would be the advantage, especially for the Diamond industry?

 

[Garry H (Cut Nut)]

Thanks for sharing Garry,

By the third kind of fluorescence you mean LWF, SWF, and something for high-energy visible light (405nm)?

That seems cool, but what would be the advantage, especially for the Diamond industry?

Not sure what you mean re 3 frequencies?
For gem ID 365 and 254 are the supposed standards. But the science is sloppy.
It was based on what was possible with mercury vapour tubes and various filters.
But for gem ID now we have cheap narrow band LEDs the science can be much more precise.

Re diamonds, the whole flour thing and GIA science until Yun Luos paper 2021 was very sloppy.
The discounts are a total mistake.
My book will explain it all.
Pre publish print promo is being printed now.

 

[Avondale]

I've tried reading this several times over now and I still don't understand.... what?

395nm activates the stronger fluorescence in chromium compared to 365nm? This does not coincide with my personal experience with two corundum gems with chromium content that I own, which fluoresce distinctly under my 365 torch but remain inert under the so called cheap UV light. Granted, I cannot verify the sources of the cheap UV, I only know they're marketed as such, whereas my UV torch is from a world-wide recognised brand.

 

[V_sh]

Not sure what you mean re 3 frequencies?
For gem ID 365 and 254 are the supposed standards. But the science is sloppy.
It was based on what was possible with mercury vapour tubes and various filters.
But for gem ID now we have cheap narrow band LEDs the science can be much more precise.

Re diamonds, the whole flour thing and GIA science until Yun Luos paper 2021 was very sloppy.
The discounts are a total mistake.
My book will explain it all.
Pre publish print promo is being printed now.

So happy to hear about the progress of your book, that will be exiciting to read it.

I do agree that many old methods that seemed so scientific are now easy to know that they aren't.

I know a very old guy who do diamond setting, once I supprised when he said that he prefers diamonds with blue flour. and also I saw a threat by @0-0-0 that proof something is just not right in the market.

Can't wait to read your book to know your complete opinion.

 

[oncrutchesrightnow]

Re diamonds, the whole flour thing and GIA science until Yun Luos paper 2021 was very sloppy.
The discounts are a total mistake.
My book will explain it all.
Pre publish print promo is being printed now.

Don’t say the discounts are a total mistake! Say more discount! More discount!

0.60 rose cut pear

I am so close to finishing my collection of natural diamonds in the traditional shapes. I was only missing two — heart and pear. Heart is not going to be a problem; I think they’re cute, it’s just a matter of finding the right one. But, no offense to anyone, I hate pears. Well, I was helping...

www.pricescope.com

 

[V_sh]

Don’t say the discounts are a total mistake! Say more discount! More discount!

0.60 rose cut pear

I am so close to finishing my collection of natural diamonds in the traditional shapes. I was only missing two — heart and pear. Heart is not going to be a problem; I think they’re cute, it’s just a matter of finding the right one. But, no offense to anyone, I hate pears. Well, I was helping...

www.pricescope.com

I don't think about discount any more, this new LGD method got into my mind.

Diamond will be as cheap as Moissanite

What's your opinion Garry?

 

[Autumn in New England]

Interesting! Thank you for sharing!!

 

[LilAlex]

You have to use one of these with a band-pass filter (there is virtually no visible light emission -- which can be dangerous, of course):

A great little UV light!

I have always used a cheap little UV penlight that I got from Esslinger whenever I want to check for fluorescence in a colored stone. But I started reading a bit more and learned that the cheap ones are mostly ~ 395nm with tons of spillover into visible (violet) light. That's what mine was --...

www.pricescope.com

 

[Garry H (Cut Nut)]

I've tried reading this several times over now and I still don't understand.... what?

395nm activates the stronger fluorescence in chromium compared to 365nm? This does not coincide with my personal experience with two corundum gems with chromium content that I own, which fluoresce distinctly under my 365 torch but remain inert under the so called cheap UV light. Granted, I cannot verify the sources of the cheap UV, I only know they're marketed as such, whereas my UV torch is from a world-wide recognised brand.

most of the 365nm torches are nothing near 365 Avondale.
Have you access to a spectroscope analysis to see the actual spectra?
I just spent 2 hours with Grant and none of the light sources we looked at were anything like what we were told or they were marketed as.

 

[Garry H (Cut Nut)]

Don’t say the discounts are a total mistake! Say more discount! More discount!

0.60 rose cut pear

I am so close to finishing my collection of natural diamonds in the traditional shapes. I was only missing two — heart and pear. Heart is not going to be a problem; I think they’re cute, it’s just a matter of finding the right one. But, no offense to anyone, I hate pears. Well, I was helping...

www.pricescope.com

Yes - shooting myself in the foot too - but looking forward to trade up day when I get loads of fluoro stones sold over decades back for credit payed at the time and upgrades!

 

[Garry H (Cut Nut)]

I don't think about discount any more, this new LGD method got into my mind.

Diamond will be as cheap as Moissanite

What's your opinion Garry?

I bought perfectly cut and polished 6.5mm Moissanite for $10 each at the HK show last September V.
I do not know how they grow it - but growing diamond will always cost more as will polishing it.
I think $50 to $100 per carat for under 2ct is about the limit and maybe $35 to $80 per carat for +2ct.

 

[Garry H (Cut Nut)]

You have to use one of these with a band-pass filter (there is virtually no visible light emission -- which can be dangerous, of course):

A great little UV light!

I have always used a cheap little UV penlight that I got from Esslinger whenever I want to check for fluorescence in a colored stone. But I started reading a bit more and learned that the cheap ones are mostly ~ 395nm with tons of spillover into visible (violet) light. That's what mine was --...

www.pricescope.com

That is what is used with the mercury vapor tubes - but the spectra are still all over the place Lil!
Shorter wave LEDs tend to have fairly narrow spectra. A 252 laser LED we checked today was from 245 to about 260nm.

 

[LilAlex]

That is what is used with the mercury vapor tubes - but the spectra are still all over the place Lil!
Shorter wave LEDs tend to have fairly narrow spectra. A 252 laser LED we checked today was from 245 to about 260nm.

This is very narrow -- 245 - 260nm. I have used fluorescent excitation professionally for many years (and for much less forgiving and more critical applications than "high/medium/low" in a few gems).

The spectrum that the tube produces is not the point -- it's the bandpass filters and cutoffs that determine the excitation wavelength reaching the gem (or cell). And, yes, these filters can capture the lower-intensity shoulder of a broader emission peak from the lamp but still do their job properly. At least in all of my applications.

Plus, the concept of blue-fluorescing diamonds predates any of this new technology that will now be your gold standard?

And did you mean 252nm or 352nm? The light I referenced was 365 with a long wavelength cut-off that I know is working because there is no visible light coming from this otherwise blazingly bright UV light.

And how confident are you that your "checking" equipment is better than the UV-generating equipment?

If your point is that super-short UV is not covered by the light I referenced, you'll get no argument from me; that is not the excitation wavelength I am looking for (although it's pretty good for nucleic acid). I am not even a diamond guy.

I don't understand what your science friend wrote.

If you are proposing to re-define this phenomenon of diamond UV fluorescence using new tools, you will face an uphill battle, imo -- creating a new gold standard that is not a refinement but is instead completely unrelated to common practice and parlance. That's a bit like my coming up with a special light for grading sapphire blue saturation and then being shocked that many of the "supposedly-vivid" sapphires that labs report and that my eye perceives are all in fact not vivid by my new lamp. It may be very precise, but it has no relation to what has been in common use and seems unlikely to replace it, imo.

I sense a commercial opportunity more than a quantum leap (no pun intended) in methodology.

ETA: If I were going to design a new method, and considering that the utility of gauging fluor is in predicting appearance degradation under sunlight, I would want an excitation source that mirrored the solar spectrum over some defined slice like 345 - 360nm, etc. Using a single wavelength could give you great precision but may bear little relation to the real-world phenomenon.

 

[V_sh]

I bought perfectly cut and polished 6.5mm Moissanite for $10 each at the HK show last September V.
I do not know how they grow it - but growing diamond will always cost more as will polishing it.
I think $50 to $100 per carat for under 2ct is about the limit and maybe $35 to $80 per carat for +2ct.

Just kidding about the Moissanite, but take a look at my new threat in LGD forum.

I've sent you the article on ResearchGate also

It seems that this new method gonna drop the price for a few tens of percents

 

[Garry H (Cut Nut)]

This is very narrow -- 245 - 260nm. I have used fluorescent excitation professionally for many years (and for much less forgiving and more critical applications than "high/medium/low" in a few gems).

The spectrum that the tube produces is not the point -- it's the bandpass filters and cutoffs that determine the excitation wavelength reaching the gem (or cell). And, yes, these filters can capture the lower-intensity shoulder of a broader emission peak from the lamp but still do their job properly. At least in all of my applications.

Plus, the concept of blue-fluorescing diamonds predates any of this new technology that will now be your gold standard?

And did you mean 252nm or 352nm? The light I referenced was 365 with a long wavelength cut-off that I know is working because there is no visible light coming from this otherwise blazingly bright UV light.

And how confident are you that your "checking" equipment is better than the UV-generating equipment?

If your point is that super-short UV is not covered by the light I referenced, you'll get no argument from me; that is not the excitation wavelength I am looking for (although it's pretty good for nucleic acid). I am not even a diamond guy.

I don't understand what your science friend wrote.

If you are proposing to re-define this phenomenon of diamond UV fluorescence using new tools, you will face an uphill battle, imo -- creating a new gold standard that is not a refinement but is instead completely unrelated to common practice and parlance. That's a bit like my coming up with a special light for grading sapphire blue saturation and then being shocked that many of the "supposedly-vivid" sapphires that labs report and that my eye perceives are all in fact not vivid by my new lamp. It may be very precise, but it has no relation to what has been in common use and seems unlikely to replace it, imo.

I sense a commercial opportunity more than a quantum leap (no pun intended) in methodology.

I did mean the shorter wave 252nm - it was an example of how much more effective LED laser sources are.
Grant uses Ocean Optics spectrphotometer - as did GIA in this study which shows how inefective the filters are.

Fluorescence Produced by Optical Defects in Diamond: Measurement, Characterization, and Challenges | Gems & Gemology

How variations in UV lamp output can affect observed fluorescence color and intensity.

www.gia.edu

See figurue 4. top left graph

 

[Garry H (Cut Nut)]

Just kidding about the Moissanite, but take a look at my new threat in LGD forum.

I've sent you the article on ResearchGate also

It seems that this new method gonna drop the price for a few tens of percents

Will see it when I believe it - it might work for coatings V

 

[LilAlex]

Grant uses Ocean Optics spectrphotometer - as did GIA in this study which shows how inefective the filters are.

But this is what GIA uses as the emission source vs. the UV LED-powered handlight that I wrote about. They look pretty similar (and pretty good) to my eye -- if anything, the LED emission spectrum is narrower ("peakier"). This is Fig 4 that you referred me to:

 

[oncrutchesrightnow]

ETA: If I were going to design a new method, and considering that the utility of gauging fluor is in predicting appearance degradation under sunlight, I would want an excitation source that mirrored the solar spectrum over some defined slice like 345 - 360nm, etc. Using a single wavelength could give you great precision but may bear little relation to the real-world phenomenon.

Results that actually mean something to the wearer? That’s crazy talk!

 

[Garry H (Cut Nut)]

But this is what GIA uses as the emission source vs. the UV LED-powered handlight that I wrote about. They look pretty similar (and pretty good) to my eye -- if anything, the LED emission spectrum is narrower ("peakier"). This is Fig 4 that you referred me to:

The spectrum is probaly as wide as this one (centered on 373) from a commonly available cheap Chinese LED said to be 365nm. Most of the excitation in a diamond, and according to Grant, in the case of rubies too - from the right hand side of the 380 - 385.
A better result can come from a $2 light around 400ish.
It is not my feild - I am a diamond dude - but thought this might be useful for gem hounds

 

[Avondale]

most of the 365nm torches are nothing near 365 Avondale.
Have you access to a spectroscope analysis to see the actual spectra?
I just spent 2 hours with Grant and none of the light sources we looked at were anything like what we were told or they were marketed as.

No access to fancy toys, so can't do a spectroscope analysis, but my torch is a Convoy which is the same handheld that GIA gemologists use, I've read the brand name on their website several times before.

 

[Garry H (Cut Nut)]

No access to fancy toys, so can't do a spectroscope analysis, but my torch is a Convoy which is the same handheld that GIA gemologists use, I've read the brand name on their website several times before.

No confidence in GIA

 

[Avondale]

No confidence in GIA

Well, if we've reached the point where people can't be sure of GIA... good luck to everyone who has any dealings with diamonds, I guess.

 

[Garry H (Cut Nut)]

Well, if we've reached the point where people can't be sure of GIA... good luck to everyone who has any dealings with diamonds, I guess.

I agree. It's a shambles across multiple fronts

 

[Starstruck8]

most of the 365nm torches are nothing near 365 Avondale.
Have you access to a spectroscope analysis to see the actual spectra?
I just spent 2 hours with Grant and none of the light sources we looked at were anything like what we were told or they were marketed as.

The spectrum is probaly as wide as this one (centered on 373) from a commonly available cheap Chinese LED said to be 365nm. Most of the excitation in a diamond, and according to Grant, in the case of rubies too - from the right hand side of the 380 - 385.
A better result can come from a $2 light around 400ish.
It is not my feild - I am a diamond dude - but thought this might be useful for gem hounds

I wondered about this. It doesn't seem to be so easy to find actual spectra on the net. Of course, I don't have a spectrophotometer. But I checked my UV flashlight (which is sold as 365nm) using a cheapo pocket spectroscope and a camera:


Top to bottom: flashlight; flashlight + fluorescent (to confirm alignment); fluorescent (for calibration); fluorescent with extra exposure to show the faint line at 404nm; scale.

It seems that the flashlight has a peak at about 376nm and a fair spread above that. (The brightness in the visible region is to be expected, because the camera is, by design, much more sensitive to visible light than to UV).

Of course, this is very rough and ready. Is it showing the emission spectrum of the flashlight or is it showing the sensitivity spectrum of my camera (Nikon D7000) in the near-visible UV? It's just luck that my camera doesn't filter this out completely. And the positions won't be perfectly linear with wavelength. But broadly, this is consistent with Garry's example.

I'm surprised. It seems that, as Garry suggested, my '365nm' flashlight isn't 365nm. If anyone has the gear to check their flashlight properly, or can find credible spectra on the net, I'd be interested to hear.

 

[Garry H (Cut Nut)]

I wondered about this. It doesn't seem to be so easy to find actual spectra on the net. Of course, I don't have a spectrophotometer. But I checked my UV flashlight (which is sold as 365nm) using a cheapo pocket spectroscope and a camera:


Top to bottom: flashlight; flashlight + fluorescent (to confirm alignment); fluorescent (for calibration); fluorescent with extra exposure to show the faint line at 404nm; scale.

It seems that the flashlight has a peak at about 376nm and a fair spread above that. (The brightness in the visible region is to be expected, because the camera is, by design, much more sensitive to visible light than to UV).

Of course, this is very rough and ready. Is it showing the emission spectrum of the flashlight or is it showing the sensitivity spectrum of my camera (Nikon D7000) in the near-visible UV? It's just luck that my camera doesn't filter this out completely. And the positions won't be perfectly linear with wavelength. But broadly, this is consistent with Garry's example.

I'm surprised. It seems that, as Garry suggested, my '365nm' flashlight isn't 365nm. If anyone has the gear to check their flashlight properly, or can find credible spectra on the net, I'd be interested to hear.

Not sure why that spectrum is showing red in an invisible violet range?
But if whatever your 'pocket spectroscope' is showing indicates a faint range from 376 up into the visible range then you have a better excitation than from the shorter wave lengths. In addition it seems your flashlight goes up into the green range???? That is weird.

 

[Starstruck8]

Not sure why that spectrum is showing red in an invisible violet range?

Well, as I said, is this showing the emission spectrum of the flashlight, or is it showing the response spectrum of the camera?

I haven't been able to find UV response spectra for my camera. That the camera is sensitive to UV at all shows that the filtering is less than ideal. Here are visible response curves grabbed from the net:

You can see that the red response extends further into the blue and green than might be expected. It's at least plausible that this could extend into the UV range. This could explain the red at the lower end.

In addition it seems your flashlight goes up into the green range???? That is weird.

The camera is (as it should be!) much more sensitive to visible light than to UV. So with enough exposure to show the UV, it will show even small amounts of visible light. The flashlight isn't filtered. I can easily see the light when looking at non-reactive things. Here's a photo I took a while ago, lit by the flashlight only:


As I said, I'd be interested to see curves from people with more suitable gear, or who are better that I am at searching the net. It is interesting, as you found, and as my efforts at least suggest, that many flashlights sold as 365nm aren't even close to sharp 365nm.

But if whatever your 'pocket spectroscope' is showing indicates a faint range from 376 up into the visible range then you have a better excitation than from the shorter wave lengths.

True, from what I've been able to find on the net. Maybe the main benefit of using LWUV to show fluorescence is that it doesn't distract from the fluorescence, not that it's especially good at activating it?