Article Over Grading of Blue Fluorescent Diamonds Revisited

[drk14]

Is that your only possible objection?

Another problem with the graph you are using is that these spectra were measured at near liquid nitrogen temperature (-196°C), so they will not be directly relevant to room-temperature observations. Cooling the diamonds to cryogenic temperatures results in significant quenching of the phonon sidebands in both the absorption spectrum and the emission spectrum. This completely alters the shape and amplitude of the spectra, compared to what would happen at room temperature.

 

[drk14]

drk,
Would you mind telling us a little bit about your scientific background? You are extremely adept at slicing and dicing this stuff. Many of us struggle to fully comprehend some of the highly technical studies.

Bryan,
Thank you for the compliment. I hope that my arguments stand on their own and do not need the appeal to authority, but in case you're asking out of curiosity, I am now a few decades into an academic career as an engineering professor. My training is in physics, mechanical engineering, and biomedicine, and my current research interests are at the interface of materials science and bioengineering. To maintain my online anonymity, I probably can't divulge any further details of my academic biography.

Reading (and writing) scientific articles is just another day at the office lol!

 

[OoohShiny]

Bryan,
Thank you for the compliment. I hope that my arguments stand on their own and do not need the appeal to authority, but in case you're asking out of curiosity, I am now a few decades into an academic career as an engineering professor. My training is in physics, mechanical engineering, and biomedicine, and my current research interests are at the interface of materials science and bioengineering. To maintain my online anonymity, I probably can't divulge any further details of my academic biography.

Reading (and writing) scientific articles is just another day at the office lol!

I did think you came across as a clever chap lol

 

[drk14]

This photo shows a different also cheap LED with and without screening by two 3mm sheets of Lexan polycarbonate (the exposure and aperture are the same in both photos).
The visible violet we see is generally above 400nm and this passes virtually unimpeded through glass and Lexan - this light contains the very powerful 415nm radiation that we can see and creates a strong violet shade in a blue fluorescent diamond.

Lexan totally blocks UVA like the 365nm UV radiation that is (wrongly) used to grade the strength of fluorescence by labs like GIA. Wrong because it was once the only easily made long wave UV, and wrong because it does not cause much of the blue fluorescence we observe in a diamond. It is mainly used by gemologists to ID various gem types and separate synthetic gems from natural etc/.

Garry,
Again, you are using some evidence that does not actually support your point. In your photos, the blue channel is saturated (B=255) in both images of the LED, so the fact that the amount of blue is similar in both pictures does not tell us anything. It is interesting to note that the red and green channels are not saturated (left image has values around R=160, G=240, B=255; right image has values around R=80, G=220, B=255), and that the intensities of those colors suggests that the Lexan reduces transmission of red and green by 50% and 8%, respectively. This seems to conflict with the transmission spectrum you posted for Makrolon Polycarbonate, but this apparent discrepancy is due to the fact that the Bayer filter used to produce color in digital cameras is not perfectly selective -- i.e., the "red" and "green" channels also pick up some wavelengths in other ranges, even down to violet/UV.

The transmission spectrum makes your point more accurately, if your point is that polycarbonate is opaque to UV light (<400 nm), but largely transparent to visible wavelengths.

ETA: I'm not understanding your point about 365nm being the "wrong" wavelength to evaluate blue fluorescence. The maximum excitation of the N3 fluorescence occurs at 395nm, but dropping the excitation frequency to 365nm only reduces the fluorescence intensity by around 50%, so it should still be detectable if it is present in any appreciable amount. Possibly some stones with extremely faint fluorescence may fall under the detection threshold by using the a non-optimal wavelength, but as long as fluorescence is visible in the master set, it will also be visible in the diamond being graded.

 

[OoohShiny]

ETA: I'm not understanding your point about 365nm being the "wrong" wavelength to evaluate blue fluorescence. The maximum excitation of the N3 fluorescence occurs at 395nm, but dropping the excitation frequency to 365nm only reduces the fluorescence intensity by around 50%, so it should still be detectable if it is present in any appreciable amount. Possibly some stones with extremely faint fluorescence may fall under the detection threshold by using the a non-optimal wavelength, but as long as fluorescence is visible in the master set, it will also be visible in the diamond being graded.

I apologise in advance as it's been a while since I've read through this whole thread, so I may be asking duplicate questions or just re-stating things that have already been said!


Is it the case that the argument is around how much excitation of fluorescence should be taking place during assessment for fluor?

i.e. should the assessment of stones be under lamps that maximally excite fluor, so that the full effects of fluor are realised and a (more?) accurate assessment of 'worst case scenario' negligible/faint/medium/strong/very strong fluor undertaken? (so 395nm wavelength)

Rather than using lamps with the 'wrong' wavelength (365nm) that don't bring out as much fluor as is present under the right (maximally excited) conditions?

(And that grading of the body colour of stones should be done under lamps with zero UV emissions, so that body colour is accurately assessed without fluor causing misgrading?)


If the above is the case, does it matter that 365nm is used instead of 395nm, as long as the grading is consistent across all stones assessed, so (accurate) comparisons between stones can be made? As noted, if the only drawback with using 365nm is a few faint fluor stones slipping through the net, is that an issue?

 

[drk14]

If the above is the case, does it matter that 365nm is used instead of 395nm, as long as the grading is consistent across all stones assessed, so (accurate) comparisons between stones can be made? As noted, if the only drawback with using 365nm is a few faint fluor stones slipping through the net, is that an issue?

Yes, that's my point, I don't think this should be an issue (unless there is some more nuanced concern behind the objections raised by @Garry H (Cut Nut) ).

(And that grading of the body colour of stones should be done under lamps with zero UV emissions, so that body colour is accurately assessed without fluor causing misgrading?)

I haven't read every post in all 19+ pages of this thread, but my impression from what I have read is that the above assertion is the focus of most of the contention and argumentation in this discussion. As I understand it, @michaelgem claims overgrading of fluorescent diamonds does occur, @Texas Leaguer concurs, and @Garry H (Cut Nut) does not.

 

[Texas Leaguer]

I apologise in advance as it's been a while since I've read through this whole thread, so I may be asking duplicate questions or just re-stating things that have already been said!


Is it the case that the argument is around how much excitation of fluorescence should be taking place during assessment for fluor?

i.e. should the assessment of stones be under lamps that maximally excite fluor, so that the full effects of fluor are realised and a (more?) accurate assessment of 'worst case scenario' negligible/faint/medium/strong/very strong fluor undertaken? (so 395nm wavelength)

Rather than using lamps with the 'wrong' wavelength (365nm) that don't bring out as much fluor as is present under the right (maximally excited) conditions?

(And that grading of the body colour of stones should be done under lamps with zero UV emissions, so that body colour is accurately assessed without fluor causing misgrading?)


If the above is the case, does it matter that 365nm is used instead of 395nm, as long as the grading is consistent across all stones assessed, so (accurate) comparisons between stones can be made? As noted, if the only drawback with using 365nm is a few faint fluor stones slipping through the net, is that an issue?

Two main issues have been discussed. Essentially where the rubber meets the road for consumers. First, the topic that is the subject of the article - color grading practice. Michaels premise is that while UV ideally would be eliminated in grading in order to prevent against color masking by blue fluoro, as long as the UV is not of sufficient intensity to stimulate fluoro, there is no problem. Intensity drops off dramatially with distance from the source, such that even grading under tubes emitting UV wavelengths, if your grading position vis a vis the tube is a foot or so away, you can remove potential masking. By grading the stones up within a few inches of the tubes, which is common practice, there is a likelihood of overgrading.
The second and related question is whether diamonds with blue fluoro actually look whiter than their color grade in real world viewing environments due to color masking in the real world. This commonly held belief is very questionable for the same reason. With the exception of direct sunlight, most viewing environments contain insufficient UV intensity to stimulate the fluoro effect. Overhead light sources, even those containing ample UV wavelenghs are typically many feet away from the diamond in normal lighting scenarios.
Visible Violet wavelenghts are also shown to stimulate fluoro, but the same rules of physics apply in terms of intensity.

With regard to the use of 365 nm to observe and describe fluoro, as drk14 just pointed out (which I think is an important point), lab reporting of fluorescence is done against master stones. So it's relativistic. As long as there is enough fluoro to stimulate the faint masters, the comparisons can be made accurately.

As a side note, this is probably a good argument for the AGSL method of reporting: their designation 'negligible' includes both inert and faintly fluorescent diamonds. Essentially, the philosophy is that faint fluoro has negligible impacts and is thus a non issue. And this broader approach mitigates some of the variations between different devices on the market used to stimulate the fluoro effect.

 

[drk14]

I am open to proof that I am wrong, but everything I have read and heard so far has not come close.

I will give it a shot...

A good place to start is one of your posts on page 1 of the thread:

Look at this chart - I chopped off an experimental error - You can see there is excitation taking effect right down to very slightly greenish blue.

(emphasis added)

What you call an "experimental error" actually holds a very important information. This peak is a indication of the intensity of the excitation light, because some (but not all) of the excitation light gets reflected back to the sensor by the diamond facets. Even though only a fraction of the excitation light gets reflected, the intensity of the reflected light is so large (relative to the intensity of the emitted fluorescence) that it swamps (saturates) the sensor, which is why the peak appears to be flat in the range 435-445 nm. The real peak value is probably at least an order of magnitude larger, and then this represents only the fraction of the illuminating light that gets reflected back to the sensor. The bottom line is that the intensity of the excitation light is many orders of magnitude larger than the intensity of the emitted fluorescence, i.e., fluorescence intensity is typically a minuscule fraction of the available excitation intensity.

In general, the visible brightness of emitted fluorescence (If) can be expressed as a product of the excitation brightness (I0) and coefficients that describe the efficiency of the conversion:

If = 2.3 x CE x QE x A x I0

In the above equation, A is the absorbance (negative of the logarithm of transmittance) of the excitation light by the diamond, QE is the quantum efficiency (probability of converting an absorbed photon into a fluorescence photon), and CE is the collection efficiency (probability of seeing any given fluorescence photon).

For the N3 center, QE=0.29 (Rand & DeShazer, 1985), which means that only one in about three absorbed photons will result in the emission of a fluorescence photon.

The absorbance A is the product of the concentration of N3 centers within the diamond (on the order of 100 ppm, per Thomaz & Davies, 1978 ), their extinction coefficient, and the pathlength of the excitation light within the diamond (which is very short in well-cut diamonds facing up). I have not (yet) found data on the extinction coefficient of N3 centers. Nonetheless, if one assumes that >90% of incident light is transmitted, then one can deduce that the absorbance is on the order A ~ 0.01.

Because the flux of emitted fluorescence photons radiates in all directions, not just towards the eye, the collection efficiency (CE) must be considered. For a pupil area on the order of ~10 mm^2 (both eyes), and a viewing distance of around 12 in (300 mm), the probability of a given fluorescence photon (which has been emitted in a random direction) actually hitting the eye is around 10 in a million. In the eye, only 2% of cones detect blue light, so one can crudely estimate an upper bound for the collection efficiency to be approximately CE = 0.0000002.

The end result is that for N3 center fluorescence, the fluorescence brightness is about a billionth of the excitation brightness:

If = (0.000000001) x I0

Now, because in a well-cut diamond viewed face-up, a large fraction of the excitation light will be reflected towards the viewer (just like it is reflected into the sensor of the spectrometer in the graph above), the fluorescence signal will be swamped by the reflected excitation light. If the excitation light is visible (and of a similar color as the emitted fluorescence), then the eye will not be able to discern the tiny increase in brightness that is contributed by the fluorescence.

We can see the fluorescence when it is excited by a bright UV source, which is invisible to the eye (and therefore does not swamp the visible fluorescence).

 

[Texas Leaguer]

I will give it a shot...
.........calculations removed from reply by TL to save space
Now, because in a well-cut diamond viewed face-up, a large fraction of the excitation light will be reflected towards the viewer (just like it is reflected into the sensor of the spectrometer in the graph above), the fluorescence signal will be swamped by the reflected excitation light. If the excitation light is visible (and of a similar color as the emitted fluorescence), then the eye will not be able to discern the tiny increase in brightness that is contributed by the fluorescence.

We can see the fluorescence when it is excited by a bright UV source, which is invisible to the eye (and therefore does not swamp the visible fluorescence).

I think this is a very important contribution to the discussion. It helps to explain why the visual impact of fluorescence on apparent color in normal light environments is negligible.

Drk, can you explain why visible whitening is obviously taking place in close proximity to fluoro tubes during grading, when those tubes are also emitting a wide range of visible wavelengths? Why is aparent color improving even though the emitted fluoro is being 'swamped' by other wavelengths in this particular lighting scenario?

I assume the answer has something to do with the color cancelling phenomenon (blue/yellow) still taking place behind the scene, beneath all the reflected light. But it would be informative to understand technically what intensities must be present to enable masking.

 

[OoohShiny]

Drk, can you explain why visible whitening is obviously taking place in close proximity to fluoro tubes during grading, when those tubes are also emitting a wide range of visible wavelengths? Why is aparent color improving even though the emitted fluoro is being 'swamped' by other wavelengths in this particular lighting scenario?

I assume the answer has something to do with the color cancelling phenomenon (blue/yellow) still taking place behind the scene, beneath all the reflected light. But it would be informative to understand technically what intensities must be present to enable masking.

I also have this question but my phone is so slow to type that I didn't post it previously lol

I was thinking specifically about photos such as @Dancing Fire's picture of his lavender Octavia in the sun! (haha, the board autocorrects Octavia to have a capital O! )

 

[drk14]

Drk, can you explain why visible whitening is obviously taking place in close proximity to fluoro tubes during grading, when those tubes are also emitting a wide range of visible wavelengths?

Since I have no personal experience with this phenomenon (whitening of fluorescent diamonds during color grading), I cannot provide a definitive answer. Nonetheless, the main point of my post above is that the conversion of illuminating light to detected fluorescence is an extremely inefficient process. Therefore, in practice, detection of fluorescence usually requires that a long-pass (LP) filter be placed between the specimen and the detector. The LP filter blocks short wavelengths (including the excitation wavelengths), but is transparent to long wavelengths (including the emission wavelengths), thus preventing the latter from being swamped by the former. When excitation occurs in the UV, then the eye itself acts as an LP filter, because it cannot detect the UV light.

Thus, one reason that fluorescence can be seen at all in diamonds with N3 centers is that significant excitation is occuring in the UV, while the reflected UV is blocked by the LP properties of the eye (thus preventing swamping). Therefore, the presence of detectable fluorescence during color grading is in part explainable by the fact that a significant fraction of the excitation is coming from the UV emitted by the Verilux lamps (see King et al., 2008 ). As shown by Luo & Breeding (2013), the majority of excitation of the N3 fluorescence occurs from UV frequencies:



Conservatively, only around 20% of the excitation (colored blue in the graph above) is from visible frequencies (probably less, because the width of the ZPL at 415nm has been artificially broadened by the finite bandwidth of the spectrophotometer).

There will be other factors at play, related to the geometry of the observer and diamond relative to the illumination environment. In my back of the envelope calculations above, I assumed that the illuminating light and the observer where both located above the face-up diamond (the same assumption used in ASET and IS analysis). However, in color grading, the diamond is viewed table down through the pavilion. In this geometry, only a fraction of the excitation light is reflected into the eye of the observer (which could significantly increase the value of CE). This is another contributing factor to the mitigation of swamping of the fluorescence signal.

 

[drk14]

I was thinking specifically about photos such as @Dancing Fire's picture of his lavender Octavia in the sun! (haha, the board autocorrects Octavia to have a capital O! )

Essentially the same answer that I gave to @Texas Leaguer , with a few quantitative differences:

The amount of UV excitation provided by direct sunlight is many orders of magnitude larger than the UV component of Verilux grading lamps, so the amount of emitted fluorescence is much, much higher.

Simultaneously, if you want to photograph or view the fluorescence produced by sunlight, you will not angle the diamond so that table glare reflects the sun directly into your eye/camera (if you did this, the fluorescence would indeed be swamped by the excitation light).

Thus, you have a very bright, UV-intense light source (the sun) exciting the N3 electrons but not being reflected into the detector (camera or eye), and you have the much less intense indirect illumination (from the sky or surroundings) bouncing off the diamond into the camera. Relative to the non-blinding brightness of this indirect illumination, the emitted fluorescence can indeed be visible.

In a more conventional lighting environment, the light that bounces from the diamond into your eye will be more or less the same light that is available to excite any fluorescence (i.e., there is no separate source of intense UV light). In such a scenario, I posit that the fluorescence will be too faint to observe.

 

[Texas Leaguer]

Essentially the same answer that I gave to @Texas Leaguer , with a few quantitative differences:

The amount of UV excitation provided by direct sunlight is many orders of magnitude larger than the UV component of Verilux grading lamps, so the amount of emitted fluorescence is much, much higher.

Simultaneously, if you want to photograph or view the fluorescence produced by sunlight, you will not angle the diamond so that table glare reflects the sun directly into your eye/camera (if you did this, the fluorescence would indeed be swamped by the excitation light).

Thus, you have a very bright, UV-intense light source (the sun) exciting the N3 electrons but not being reflected into the detector (camera or eye), and you have the much less intense indirect illumination (from the sky or surroundings) bouncing off the diamond into the camera. Relative to the non-blinding brightness of this indirect illumination, the emitted fluorescence can indeed be visible.

In a more conventional lighting environment, the light that bounces from the diamond into your eye will be more or less the same light that is available to excite any fluorescence (i.e., there is no separate source of intense UV light). In such a scenario, I posit that the fluorescence will be too faint to observe.

Regarding the last sentence(my bold). I think everyone generally agrees with that in terms of not actually seeing the blue ( as can sometimes be seen in photos taken in direct sunlight). The question for me remains, can this unseen fluorescence in conventional lighting environments with far weaker stimulation of N3 centers, still somehow cause visible whitening of the body color of the diamond? If so, it must have something to do with a special power of blue/yellow cancellation that has yet to be fully explained.

Another point I will make about photos of fluorescence in sunlight - the blue of the sky can be reflecting in the diamond such that even inert diamonds can also appear blue.

 

[drk14]

Regarding the last sentence(my bold). I think everyone generally agrees with that in terms of not actually seeing the blue

I should have been more clear -- when I say the fluorescence cannot be "observed", I mean that is too weak to have any observable effects (including any whitening).

I'm not sure if there is also a confusion about how whitening occurs, so I will add my explanation:

A full spectrum of visible light will be perceived as white. If a diamond absorbs some fraction blue/violet light (with an absorption maximum, say, near 450 nm), the eye will perceive a spectral peak in the complementary color, i.e. yellow/orange (~600 nm), because that part of the spectrum now has the largest amplitude compared to the minimum at 450 nm. If we can add back some photons in the blue wavelengths (e.g., by N3 fluorescence emission), then the spectrum will flatten out again, and the yellow tinge will disappear (or at least be lessened).

Against this background, let's consider the possibility of a whitening effect caused by visible light excitation only (which is what has been proposed by @Garry H (Cut Nut) ). To make things simple, let's more or less ignore the collection efficiency (CE), which will additionally reduce the fluorescence efficiency by many orders of magnitude (as shown above). We can contrive an example (e.g., a diamond contact lens resting directly on the pupil, and collimated illumination shining directly into the eye) in which CE approaches unity (actually CE ~ 0.5, since half of the fluorescence photons travel away from the eye).

Let's say that a 5% decrease in the transmission of blue/violet light causes a one-grade decrease in the body color. To whiten this body color by one grade, we would need to add an equivalent number (+5%) of blue/violet photons, emitted by N3 fluorescence. Now, due to the effects of CE (only 1 in 2 photons will travel into the eye) and the quantum efficiency (only 1 in 3 photons absorbed by the N3 center will be converted to a fluorescence photon), the whitening effect can occur only if the diamond's N3 centers absorb 6 times as many photons as the number of fluorescence photons that reach the eye. Therefore, to create the required +5% increase in whitening blue/violet photons, there has to be a 6 x 5% = 30% increase in absorption of photons by the N3 centers.

Perhaps you already see where this is going. If all of the excitation light comes from the visible (400-415 nm), then the diamond will have significantly reduced transmission of violet light, which will cause the eye to perceive the diamond as having a significant tinge of yellow-green (550-565 nm) body color. This will result in a color grade reduction that is worse than the original color grade reduction which we were trying to correct!

The conclusion is (again) that if the excitation light is primarily in the visible part of the spectrum, and if the diamond does not have such a high concentration of N3 centers that it takes on a noticeable yellow-green body color, then the amount of fluorescence produced is not sufficient to cause noticeable whitening of yellow/orange body color.

 

[Texas Leaguer]

Drk,
I feel a bit like I’m enrolled in an upper level physics class without having the prerequisites. I hope I don’t start having those post-college dreams where you realize there is a final exam and you haven’t been attending class all semester!

Seriously though, thank you for taking the time to bring your considerable expertise into this forum, and in particular to this discussion. There are a lot of myths and misinformation surrounding this subject in the trade and in the consumer marketplace. Because fluorescence can have such a dramatic impact on diamond pricing, value, and liquidity, it is very important for the consumer to have accurate understandings around this topic.

 

[drk14]

I feel a bit like I’m enrolled in an upper level physics class without having the prerequisites.

That wasn't my intention! My goal was to explain the phenomena so that they can be understood by non-physicists, but I felt I should be explicit with regards to the physical assumptions and reasoning that my arguments depend on, in case anyone wants to check my work (or challenge my conclusions).

I can attempt to rephrase my most recent post in a more succinct manner:

You can't get something for nothing. To produce whitening fluorescent light, the diamond needs to absorb light (because the energy contained in the emitted light comes from the energy in the absorbed excitation light). If visible light is absorbed, the diamond will appear darker to the eye. Because conversion of excitation light to emitted fluorescent light is inherently inefficient, the diamond must absorb more photons than it emits (3:1 ratio for the N3 center). If we propose that a 1-grade whitening can result from fluorescence produced by visible light excitation only, this requires a diamond in which the absorption of violet light by N3 centers produces a darkening by 3 or more grades. This darkening should result in a noticeable yellow/lime-green body color. So in diamonds that are do not have a visible yellow/green tinge, we cannot get any grade-changing whitening, unless there is a significant amount of UV light present.

This analysis is very conservative, because it neglects the effects of collection efficiency.

Let me know if anything is not clear.

I will make an additional post with some experiments to try at home.

 

[Texas Leaguer]

Drk,
If I understand your interpretations of the science correctly, you would agree that in the vast majority of conventional lighting environments, due to the relatively weak UV and VV signals of artificial lighting, the inefficiency of blue emissions generated by fluorescence, and the predominance of other wavelengths in the reflected light (including yellow), there is no scientific basis for the widely promoted notion that a diamond with blue fluorescence delivers the benefit of body color whitening to the consumer.

It that a fair statement?

 

[drk14]

Drk,
If I understand your interpretations of the science correctly, you would agree that in the vast majority of conventional lighting environments, due to the relatively weak UV and VV signals of artificial lighting, the inefficiency of blue emissions generated by fluorescence, and the predominance of other wavelengths in the reflected light (including yellow), there is no scientific basis for the widely promoted notion that a diamond with blue fluorescence delivers the benefit of body color whitening to the consumer.

It that a fair statement?

Sort of, but you have made several stipulations that are unnecessary (for example, we don't need to assume that VV is weak, or that yellow is more dominant than other wavelengths).

The key is that fluorescence is an extremely inefficient process for converting absorbed light into emitted light. Therefore, some unusual circumstances are required in order to see any visible effects (e.g., whitening) of emitted blue fluorescence. In particular, at least one of the following things must be true:

1. The diamond must be illuminated by a very bright light source that has significant UV component.
2. Some artifice must be introduced to reduce the amount of light that travels from the light source to the eye (e.g., by arranging the light source and observer to minimize the amount of illuminating light that the diamond is able to reflect or transmit towards the observer, and/or by using a long-pass filter that blocks VV light coming off the diamond).
3. The diamond body must have an extremely high concentration of N3 centers, giving it a yellow-green body color.

I think that the third scenario is just hypothetical, as I have never heard of a diamond with such properties. The first scenario is not relevant to "normal" viewing environments (unless in direct sunlight).

Under scenario #2, it is possible to that a bright light source that is mostly devoid of UV (but containing VV) would produce some detectable blue fluorescence (manifesting as blue glow or whitening of yellow/orange body color), but only if one blocks the illuminating light from reaching the eye (thus preventing the fluorescence from being swamped by the brightness of the illuminating light). This would require some artificial setup (e.g., viewing a table-down diamond through the pavilion from the side, with the illumination at a perpendicular angle from the top). In real-life viewing circumstances, however, the illuminating light is deliberately made to reflect off the diamond facets into the eye, so scenario #2 is not applicable (unless one is wearing yellow-colored glasses and illuminating with a light source that is mostly violet).

It really just comes down to the inefficiency. Absent any of the unusual scenarios enumerated above, producing detectable whitening requires blinding amounts of excitation light. If the illumination is not blinding, then I would expect the fluorescence to be so weak as to have no perceptible effects on the appearance of the diamonds.

 

[Texas Leaguer]

So, direct sunlight would be an example of #1. Therefore a consumer might observe some whitening outdoors. Would diffusion by cloud cover be enough to negate the effect?

Observation within a few inches of a grading light with the diamond table down would be an example of scenario #2. (consistent with Michael Cowing's conclusion in the over-grading study)

Assuming scenario #3 does not exist for practical purposes, then there are few, if any, conventional lighting scenarios that would result in whitening.

My experience in observing diamonds over a fairly lengthy career is consistent with what the science would predict. In my early years I operated under the assumption that the conventional wisdom about the effects of fluorescence on real-world observation was valid. As I began to question those assumptions and avail myself to more objective information, I have come to the conclusion that the commonly held belief that blue fluorescent diamonds provide a benefit to consumers in terms of whitening is fundamentally untrue and a product of self-serving promotional efforts within the trade over a very long period of time.

 

[Rockdiamond]

Hi DRK- the intense scientific explanations are beyond me- but it seems that in theory, you're agreeing that whitening in an indoor room during daylight hours, when the sun is shining, but not directly on the stone does not occur.
It does. No theory, but reality.
That's my issue with this entire discussion.
I've been observing fluorescent diamonds in person, for over 40 years. Unlike Bryan, my experience has only strengthened what I learned back at Winston. Back 20 pages or so I explained how in the late 70's Harry Winston charged a premium for colorless SB stones precisely due to the whitening- or blue-ing, in the case of a D-E or F. This wasn't imaginary.
If there's enough light to discern small differences in body color, medium or strong fluorescence may be apparent in the way it affects the color- in some cases whitening. Not even all cases- some stones that appear similar in darkness under a UV light react differently in a room with natural lighting.
This is based on extensive physical observation, not theoretical studies....

 

[Rockdiamond]

I have come to the conclusion that the commonly held belief that blue fluorescent diamonds provide a benefit to consumers in terms of whitening is fundamentally untrue and a product of self-serving promotional efforts within the trade over a very long period of time.

Bryan, all due respect, but this type of statement is precisely why it's a contentious issue.
I don't find a lot of "deception" going on in this area.
There's a cutter that advertises FL stones, but in general, they're scorned by the current market.
We don't scorn them- I happen to love them.
If I see a color improvement and tell a client, and they don't see it, we get a return. We work extremely hard on accurate descriptions because returns hurt a company in many ways.
It's not self serving to describe what one sees and represent it honestly.

 

[Texas Leaguer]

Bryan, all due respect, but this type of statement is precisely why it's a contentious issue.
I don't find a lot of "deception" going on in this area.
There's a cutter that advertises FL stones, but in general, they're scorned by the current market.
We don't scorn them- I happen to love them.
If I see a color improvement and tell a client, and they don't see it, we get a return. We work extremely hard on accurate descriptions because returns hurt a company in many ways.
It's not self serving to describe what one sees and represent it honestly.

All due respect David, you put quotes around the word deception as if you were quoting me. I said "promotion".

That promotion has been going on for decades and continues today, and that is why so many consumers are under the impression that there is a clear benefit from something that is quite dubious. I believe that many sellers who promote fluorescent diamonds also assume as I did once, that the whitening effect they tout is real. Therefore it does not qualify as deception. More like a convenient story that they choose not to question.

I personally think fluorescent minerals and gems are extremely cool. I have been mesmerized by fluoresecence and phosphorescence in nature since I was a kid. From glowing minerals, to lightning bugs, to grunion eggs that leave glowing footprints on the beach as you step on it, to blinking sea creatures on a night dive, to black light posters!

In the majority of cases fluorescence has no impact on visual appearance of a diamond. In many cases it lowers the cost. There are legitimate reasons for consumers to consider them when shopping. The assumption that it will make the diamond look whiter than its color grade is not one of them, yet that is the property most commonly cited by sellers. That is self-serving on the part of the seller and a disservice to the consumer.

 

[Rockdiamond]

My apologies on badly chosen word Bryan.
I think we're both agreeing on some core issues- that being, there's a lot of confusion about fluorescence among many consumers.
We both agree it can be a very cool characteristic.

We diverge on a few crucial points.
I don't feel like there's promotion of fluorescent diamonds in advertising- other than one place we can both think of.
If someone is buying a stone in person, and the seller is touting a J as looking whiter due to MB/SB, the consumer is in the position to see if they can see the difference.
It's not a characteristic used to upsell folks- because we also agree MB/SB stones are, as a rule, lower priced than their inert counterparts.

But of course, the real difference we're having is about the actual experience of looking at MB/SB stones, in a variety of lighting environments and seeing a difference- specifically in any sunlit room.
You don't notice it, I do. I know of countless others who have also noticed this effect.
There can be reasons for our difference on this.
I was trained as a color grader, so many years ago- and I have sensitive eyes when it comes to color.
In my expereince, the ability to detect small differences in shade is not a universal ability, even for sharp-eyed people.
Some people just see color better.
Maybe the light is different in Texas

In the end, I believe we're both defenders of fluorescence- and actually, we're on the same side.

 

[drk14]

Hi DRK- the intense scientific explanations are beyond me- but it seems that in theory, you're agreeing that whitening in an indoor room during daylight hours, when the sun is shining, but not directly on the stone does not occur.
It does. No theory, but reality.

Hi David,
As Thomas Huxley said, the great tragedy of Science is the slaying of a beautiful hypothesis by an ugly fact! The corollary of that statement is that the triumph of Science is the ability of hypotheses to evolve in response to experimental data. I am definitely open to debating my assumptions, analysis, or conclusions. However, to change my opinion, I would need something more definitive than an anecdotal observation. (This doesn't mean that I don't believe you, just that a statement such as the one you've made doesn't give me any constructive path for figuring out how to reconcile your experience with the laws of physics).

An interesting publication that I don't think has been referenced yet in this thread is Moses et al. (1997), Gems & Gemology 33(4):244-259. Unfortunately it is somewhat of a preliminary study and not very well controlled, but it contains some interesting data and observations that provide grist for the mill of both sides of the "whitening" debate. Here are some take-away points from that report:

• Untrained observers (average consumers) could not notice any differences in apparent body color in diamonds of equal color grade but different levels of blue fluorescence.
• Members of the diamond trade, when viewing diamonds of equal color grade through the pavilion table-down, claimed to see differences in apparent body color among diamonds with differing levels of fluorescence, but were evenly split about whether the fluorescent diamonds appeared to have more color or less color than the non-fluorescent diamonds.
• Among members of the diamond trade, when viewing diamonds of equal color grade through the crown table-up, approximately 60% of observers perceived diamonds with very strong fluorescence to have less body color than diamonds with less fluorescence; about 40% of such trained observers (just like the untrained consumers) did not notice any color improvement in the diamonds with very strong fluorescence.
  
• Among those who claimed to see differences in body color, this was more likely to occur in the I-K color range than in E-G color diamonds.
• For diamonds with equal color grade and equal amounts of fluorescence (strong), certain specific stones were more frequently identified as having less apparent color than other stones (of the same color grade and fluorescence level). This suggests that there are other factors that affect body color appearance more strongly than does fluorescence.

I was especially intrigued by the last point, which suggests one possible explanation under which my claims and yours may both be true: it could be that while fluorescence emission itself is not sufficiently strong to cause whitening, strongly fluorescent diamonds may be more likely to possess some other property that does cause some noticeable whitening effect.

 

[Rockdiamond]

As Thomas Huxley said, the great tragedy of Science is the slaying of a beautiful hypothesis by an ugly fact! The corollary of that statement is that the triumph of Science is the ability of hypotheses to evolve in response to experimental data. I am definitely open to debating my assumptions, analysis, or conclusions. However, to change my opinion, I would need something more definitive than an anecdotal observation. (This doesn't mean that I don't believe you, just that a statement such as the one you've made doesn't give me any constructive path for figuring out how to reconcile your experience with the laws of physics).

Point well taken DRK!
And thank you for forwarding the discussion as you have.

• For diamonds with equal color grade and equal amounts of fluorescence (strong), certain specific stones were more frequently identified as having less apparent color than other stones (of the same color grade and fluorescence level). This suggests that there are other factors that affect body color appearance more strongly than does fluorescence.

I was especially intrigued by the last point, which suggests one possible explanation under which my claims and yours may both be true: it could be that while fluorescence emission itself is not sufficiently strong to cause whitening, strongly fluorescent diamonds may be more likely to possess some other property that does cause some noticeable whitening effect.

We also share interest on this point.

Here's some of my anecdotal observations that may or may not contribute.
Fluorescence in Fancy yellows may provide insight- GIA's MB/SB Fancy COlor grading is far more controversial than in colorless.
My experience is that GIA will downgrade the color grade of a yellow diamond with MB/SB.
If I'm looking at a tray of goods ( on a desk, during daylight hours, with indirect southern sunlight)- all Fancy Yellow ( GIA graded) there are times when the strongest ( the ones that looked most yellow face up) ones are MB/SB.
Some of my colleagues feel that since GIA grades Fancy Yellow diamonds face up, maybe they are using some strange lighting methods. This line of thinking among dealers agrees with part of what Michael is saying- about a flaw in GIA's methodology.
Or they're just tougher on MB/SB yellow stones.
There's also a fair percentage of Yellow ( all ranges) of MB/SB stones where the color shift is a definite negative. Either in terms of lessening the yellow color, but still a "shiny diamond"- and some with the dreaded dullness.
In the dull cases, the haze is visible in any lighting bright enough to view the diamond in detail.

The other question goes to what exactly is "normal lighting". Maybe indirect sunlight, and or other sources of bright light are producing effects based on factors/ wavelengths (?) we've not discussed/discovered.
Even using a simple light meter I can see that even areas of our offices that have no direct sunlight whatsoever are exponentially brighter in daytime than nite.

 

[Texas Leaguer]

Drk,
During your research on the topic of fluorescence have you come across any good explanation for why some strongly fluorescent diamonds have an oily or hazy appearance? What would be the mechanism for emissions to degrade transparency? Is this also a case where there are other factors at play either seperate from or in conjunction with fluoro?

I would appreciate your thoughts as I have spoken to a number of experts who could not give me a definitive answer. I formulated a couple of my own hypotheses but they did not really fly with some of the lab folks I know!

This is clearly the most impactful visual aspect of fluorescence and is one of the reasons that fluoro diamonds are discounted to the extent they are. (potential overgrading of color is another).
A better understanding of this phenomenon would help consumers considering diamonds with strong or very strong fluorescence.

 

[drk14]

David,
Your observations about the appearance of yellow diamonds are very intriguing. So far, I have focused my reading on colorless diamonds, but if I have an opportunity, I will try to research a little bit of the literature on fancy yellows. It is hypothetically possible that a yellow diamond contains a much higher concentration of N3 centers than do colorless diamonds (which could make for stronger fluorescence), but whether this is the case or not I can only speculate until I do the research. Maybe some of the other experts on the board know something about this.

The other question goes to what exactly is "normal lighting". Maybe indirect sunlight, and or other sources of bright light are producing effects based on factors/ wavelengths (?) we've not discussed/discovered.

This is an excellent question, and none of the published studies cited in this thread (Moses et al., 1997; King et al., 2008; Cowing, 2010; Luo & Breeding, 2013) actually address this point in a satisfactory manner.

Note that what we're calling indirect sunlight should probably better be described as diffuse radiation or skylight. Due to Rayleigh scattering, the sky is blue, and in fact also radiates a significant amount of UV (the peak irradiance is at ~400 nm):

(source)

How much of this UV makes it through window panes, I'm not sure, and it certainly would vary depending on what type of window you have.

 

[Rockdiamond]

After all these pages, I think we may be making some progress- thank you DRK!!
You too Bryan- your input has been instrumental in moving the discussion forward.
This discussion is infinitely more fun to focus on than.....other stuff.

DRK- here's another weird experiential detail that may or may not shed light ( pun intended)
Diamonds, or observers, react differently to light based on geographical factors.
Here's what I mean: I have gone to Israel, and Belgium to buy diamonds on numerous occasions.
The stones look different there, than they do in NYC.
I've spoken to many dealers who've experienced the same phenomenon.
Say I am right 75% of the time when I grade the color of a stone before sending it to GIA ( which is about accurate over the past 20 years)
When I have purchased, and graded goods in Israel, my average of predicting GIA grades goes down below 50%.

Remember that GIA is specifically using clinical lighting and methodology ( no sunlight)- while real, practical grading requires one to be able to account for more environmental factors that indirect sunlight introduces.

Another interesting experience I had- in the '90's my job was supplying diamonds to stores- in the US and Caribbean. The stores in St Maarten and St Thomas ( both now sadly devastated) could not keep enough 2ct K-L color diamonds in stock, while guys in North Carolina couldn't even sell I color.
Maybe the cruise passengers were drunk....but maybe the differences in natural lighting had an effect.
The island guys LOVED MB/SB stones in the K-L-M range. They just looked whiter- and not in direct sunlight. Most of the stones had windows and a lot of indirect sun.

 

[drk14]

When I have purchased, and graded goods in Israel, my average of predicting GIA grades goes down below 50%.

Very interesting! Would you say that your batting average for these Israeli stones was worse for fluorescent stones than for nonfluorescent (or weakly fluorescent) stones? In the cases where you were wrong, were you more frequenlty overgrading or undergrading compared to GIA? And at the stores in the Carribean, was the demand for K-L stones limited to those with fluorescence, or also nonfluorescent stones?

 

[drk14]

During your research on the topic of fluorescence have you come across any good explanation for why some strongly fluorescent diamonds have an oily or hazy appearance? What would be the mechanism for emissions to degrade transparency? Is this also a case where there are other factors at play either seperate from or in conjunction with fluoro?

Bryan,
I have not come across any explanations for this phenomenon.

Moses et al. (1997) claim that such "overblue"/milky stones are very uncommon (<1%). A very interesting result of their study was that even though they did not include any overblue diamonds in their experiments, they found that 50% of observers (from the diamond trade) claimed to discern a difference in the apparent transparency of diamonds that had differing levels of fluorescence, and that when observing the stones table-down in a grading environment, the very strongly fluorescent diamonds were more likely to be perceived as less transparent than the faintly fluorescent diamonds.

So again, there may be some material property responsible for haziness that is correlated with (but not caused by) fluorescence strength.

If I may speculate, I would guess that elastic scattering of light is an important phenomenon. So called Mie scattering occurs when photons bounce off of objects that are similar in size to the wavelength of the incident light, with minimal loss of energy. A common example is fog or clouds, in which visible light (400-700 nm wavelength) is scattered by water condensation (droplet diameters 100-10,000 nm).

So, it is possible that microscale (100-10,000 nm) crystal defects may be present within the diamond, and cause light to be scattered. Although N3 centers are point defects with dimensions less than 1 nm, it may be possible that strain fields surrounding the point defects, or aggregation of multiple point defects, create microscale structures that cause Mie scattering. If present in sufficient concentration, a haze may become visible within the diamond. Conversely, if other stones contain N3 centers that are not associated with microscale defects, then you can have fluorescence without the haze.

If you'll allow me to be more bold in my speculation, I think that at least hypothetically, incident light may undergo Rayleigh scattering when bouncing off smaller (nanoscale) crystal defects. Rayleigh scattering is another form of inelastic scattering (photon does not lose energy), which occurs when the scattering object is much smaller than the wavelength of the light. This type of scattering is the reason why the sky is blue. Wouldn't it be interesting if some diamonds take on a blue appearance for the same reasons that the sky is blue? In the sky, the light bounces off nitrogen molecules (N2), which are very similar in size to the nitrogen optical centers responsible for diamond fluorescence. Again, if this idea is correct, then whitening is not caused by fluorescence, but by elastic scattering. However, fluorescent diamonds would be more likely to produce Rayleigh scattering, which explains why strong fluorescence appears to be correlated with a whitening effect in some cases.

I'll leave it to someone else to run the calculations for this scenario and let me know that I'm way off base!