Towards Reconciling Issues of Color Grading Blue Flourescent Diamonds
The Gemological Institute of America, (GIA), recently published in their quarterly journal, "Gems & Gemology",[1] the results of a human experiment in the observation of diamonds with various amounts of blue fluorescence. The article was entitled "A Contribution to Understanding the Effect of Blue Fluorescence on the Appearance of Diamonds". An introductory editorial by GIA's president William Boyajian indicates that the study "challenges the perception held by many in the trade that UV fluorescence generally has a negative effect on the overall appearance of a diamond". He makes it clear that this is a false perception, which the GIA hopes this study will help to dispel. He says this study "should bring into question the trade's lower "bid" prices for moderate to highly fluorescent diamonds in the better colors".
Historically, the GIA has been the principal source of information in the area of diamond color grading for gemologists, appraisers and the diamond trade, so this article will generate much discussion. Because of the GIA's stature and credibility with the trade and the buying public, this article will be quoted by many to improve the relative salability of diamonds with the property of medium or strong blue fluorescence indicated on their grading report.
To help disseminate the message of this study to the trade, an abbreviated version of this article was published in the April 1998 issue of the "Rapaport Diamond Report". In the same issue Martin Rapaport responded with a thoughtful article, entitled "Blue White" in response to the GIA study's conclusions. In it he explained the trades historical handling of blue fluorescence in diamonds from a pricing and marketing perspective. He points out why the trade discounts fluorescence in the higher gem quality diamonds, and how it came about that fluorescence comments, indicated on a diamond grading report, became undesirable. He concludes: "Obviously from the market perspective there appears to be a reasonable basis for price discrimination against fluorescence". He acknowledges "If fluorescent diamonds are graded accurately under ideal laboratory lighting conditions blue fluorescent stones have an advantage in that their color improves in normal daylight". Finally, he admonishes that "changing buyers perception about the negative impact fluorescence has on higher color diamonds . will have to backed up by solid results. In other words the labs are going to have to be very serious about not over-grading the color of fluorescent stones even though these stones tend to appear whiter than they are".
As a gemologist and member of the diamond trade, the author has been involved with this issue since the early 80's and would like to make a contribution by addressing the following questions, which both articles should raise in the minds of their readers:
Are the labs over-grading blue fluorescent diamonds? If so, why? How can the labs insure they are not over-grading blue fluorescent diamonds and convince the trade of this? What can be learned from this new study beyond Boyajian's and Rapaport's conclusions? These questions are examined in the historical context of what gemologists have learned and taught the diamond trade regarding fluorescence and diamond color grading over the last 30 or so years.
The majority of natural gem quality diamonds contain nitrogen as an impurity. Depending upon the amount of nitrogen and how it is distributed throughout the diamond, this impurity causes the slight tints of yellow to which the GIA assigns the letter color grades from D to Z. This type of diamond is categorized as Type Ia. For simplicity and because it covers most of the gem color white diamonds, the GIA and this discussion address mainly this Ia diamond type. However the principles discussed are extensible to the other categories. This same nitrogen impurity in the type Ia diamond, depending upon the percentage and distribution of nitrogen, may cause fluorescence of varying intensities of blue or bluish-white.
Fluorescence is the emission of visible light from a substance being stimulated by invisible, ultra violet, (UV), energy such as that contained in sunlight and indirect sunlight. Gemologists look for this property of fluorescence in diamonds and other gems by employing ultra violet lamps. These UV sources use either a high or low-pressure mercury vapor tube with a filter that removes visible light emitted by the tube. This leaves energy or radiation in the short and long wave UV wavelength range of 200 nm to 400 nm. The two highest energy wavelengths put out by these tubes are 366 nm, which is in the long wave UV range, and 254 nm, which is in the short wave UV range. These wavelengths are selectively filtered to produce long and short wave UV energy for testing fluorescence. Another popularly used source of long wave UV is a low-pressure mercury vapor tube that excites a special coating which then emits long wave UV. When a diamond with blue fluorescence is illuminated with invisible long wave UV energy from one of these lamps in an otherwise dark environment the diamond glows blue. Short wave UV energy also stimulates blue fluorescence, but to a much lesser extent, leaving long wave UV the dominant source that produces this effect.
The GIA Article makes the point that a group of two nitrogen atoms, the A aggregate, tends to quench fluorescence while a group of three nitrogen atoms, called the N3 center, produces blue fluorescence. Since a single diamond may contain varying amounts of these nitrogen defects, the result may be anything from no fluorescence to extremely strong blue fluorescence. The GIA says this confounds "the trade notion that nonfluorescent diamonds are more "pure" than those that fluoresce, since there are nitrogen-related centers that extinguish fluorescence as well as those that cause blue fluorescence." They are making the important point that blue fluorescence or lack thereof is a property of diamond, not an indicator of "purity", since a nonfluorescent diamond may often have a greater percentage of nitrogen impurity than one that is blue fluorescent.
The light yellow tints in a Type Ia diamond combine with the various amounts of blue fluorescence to give the diamond its perceived color. Gemologists and the trade define and refer to the diamond's "true color" as that color that a diamond exhibits when viewed under illumination devoid of any UV energy that would excite fluorescence. Because all the light sources regularly used to view and grade diamonds contain some amount of long wave UV energy the "true color" of a strongly blue fluorescent diamond is never seen. What is seen is the perceived net color. This net color includes whatever amount of blue fluorescence is stimulated by the light source's UV energy. To explore the perplexity of perceived color verses "true color", we need to discuss in detail how gemologists have historically measured and graded color in a diamond.
Of primary importance in diamond color grading is the illumination used to view the diamond during grading. Under what illumination should the diamond be viewed in order to properly grade diamond color?
There are three general types of illumination, which have been used in diamond color grading. The first is the various forms of natural daylight. The second is fluorescent lighting, and the third is the various incandescent, filament type, light bulbs from high intensity tungsten to quartz halogen. Each of these forms of illumination has their advantages and disadvantages for use in color grading. Historically, the best form of natural daylight for color grading was found to be northern daylight. Eric Bruton's classic book from the 70's, [2] defines northern daylight to be "daylight from a North-facing window in the Northern Hemisphere and from a South facing one in the Southern". Grading was typically done during the daylight hours when the sun was high enough in the sky to avoid the reddish tint in early morning and late afternoon light. The very slightly bluish-white northern daylight made an ideal illuminant to detect the slight tints of yellow typical in gem quality diamonds.
To further the understanding of the properties of this light, which made it desirable for color grading diamonds, it is useful to introduce the unifying concept of color temperature. The measure of temperature described here is degrees Kelvin (K). Degrees Kelvin is just degrees centigrade (C) plus 273, because absolute zero, to which Kelvin temperature is referenced, is -273C.
When a substance such as the tungsten filament in an incandescent light bulb is heated to sufficiently high temperatures it radiates energy over a whole spectrum of different wavelengths. The spectrum extends from the very long infra red waves, which produce the sensation of heat, thru the 750 nm to 400 nm waves which produce the sensation of color from red to orange, yellow, green, cyan, blue, violet and then on into the invisible ultra violet short wavelengths between 400 nm and 200 nm. The perceived color of this spectrum is a function of the balance or distribution of energy over the visible wavelengths. When the filament temperature reaches 1500K to 2500K the visible wavelengths combine to give a reddish tint to the illumination. As the filament is heated to 2800K to 3200K the yellowish- white light of incandescent bulbs is produced. The higher the temperature is raised as in quartz halogen bulbs, the less yellow and more neutral-white the emitted light becomes. Incandescent lights reach a practical temperature limit at about 3400K, but at a color temperature of 5500K to 6500K the balance of energy in the visible wavelengths tilts to the very slightly bluish-white of northern daylight.
Compare color grading under a yellowish-white, incandescent illumination of 3200K to grading under the 6500K, bluish-white, northern daylight. You will find, as did the early color graders, that slight tints of yellow are more easily observed in the complimentary very slightly bluish-white northern daylight than in the yellowish-white light of incandescent sources. Thus, the color temperature close to northern daylight of 5000K to 6500K became the standard for color grading diamonds.
However, natural, northern daylight has another property that complicates the diamond color-grading picture. It is rich in both long and short wave UV energy. This invisible UV part of the spectrum of daylight excites the blue fluorescence in diamonds having this property. Early on, it was recognized that the perceived color grade of typical gem quality diamonds was a combination of the faint yellowish tint in the diamond and any blue fluorescence caused by UV from the illuminating source. Since blue and yellow light are complimentary, they tend to cancel. A highly blue fluorescing, yellowish diamond will show its yellowish nature when viewed under illumination with a minimum of UV such as incandescent light, but will appear increasingly colorless in lighting with increasing amounts of long wave UV. Finally, in the UV rich illumination of daylight it may appear colorless or even slightly bluish-white. The plate glass windows, thru which the northern daylight enters, such as the floor to ceiling windows of the pictured, Antwerp diamond bourse, filter out most of the short wave UV energy. However, most all the long wave UV that is the main cause of fluorescence remains present.
This presented a real dilemma to early diamond graders including Shipley, the GIA's founder. On the one hand the color temperature of northern daylight was ideal for color grading diamonds with slight tints of yellow, but at the same time daylight's long wave UV energy was causing blue fluorescent diamonds to look whiter than they did when compared under incandescent illumination having a minimum of UV energy. Since the color of a fluorescent diamond changes as a function of the illumination under which it is observed, what should be the standard illumination to grade this color? Early on the GIA and others decided to separate the whitening effect of blue fluorescing diamonds from the basic yellowish tint by designing a standard light source with a close to northern daylight 0K, but unlike nocolor balance of 5000K to 650rthern daylight having a minimum of UV.
The GIA Diamond Course from 1969,[3] stated "A large percentage of diamonds fluoresce, usually blue; and the fluorescence, if sufficiently intense will alter the color of such a stone when observed under a light source emitting ultraviolet rays. Since this occurs under daylight examination the most desirable conditions are to be encountered under a balanced artificial light with a minimum of ultra violet content." This is perhaps the most logical approach to the dilemma presented by fluorescence. By changing the standard illumination for color grading from northern daylight with a lot of UV to incandescent illumination with a minimum of UV, the color advantage that blue fluorescing diamonds possessed in daylight was minimized in the grading.
The GIA produced their first diamond color grading instrument, called the DiamondLite, using an incandescent filament type of light source and a "daylight filter" which produced "the equivalent of north light without the UV radiation."[4] At the same time the GIA Diamond Course stated that "a reasonably good substitute for the DiamondLite can be made by adapting a simple desk lamp fixture containing cool white fluorescent tubes."[5] However, they caution "the disadvantage of this kind of illumination is that fluorescent tubes emit a significant percentage of ultra violet radiation. Although this does not affect the grading of nonfluorescent stones, it causes fluorescent diamonds to be graded higher then is actually warranted due to the neutralizing, or masking, effect of the fluorescent color on the true body color."[6] Gemologists graduating from the GIA in the 1969 time frame came away with the principle that diamonds should be graded for color under a daylight balanced artificial illumination absent the UV radiation energy contained in natural northern daylight and also contained in fluorescent lighting. They learned that grading in daylight or fluorescent light with the attendant UV radiation will result in over grading a blue fluorescing diamond. This has become the conventional wisdom among gemologists since that time. Eric Bruton's book, "Diamonds", published in the 70's, indicated that gemologists worldwide shared these views on illumination for diamond color grading. In a section of his book under conditions for color grading,[7] he said a "very important consideration is that any fluorescence in the stone must be suppressed".."It is therefore important to grade stones in white light that is relatively free of ultra-violet".
What has happened since that time? By the 70's we find gemologists and the diamond trade worldwide are universally using some form of fluorescent light to grade color in diamonds. The later versions of the GIA DiamondLite have substituted Verilux brand, fluorescent tubes for the daylight corrected incandescent light source in the early model. This is the very same source of grading illumination that was said to result in the over grading of fluorescent diamonds. Does that mean that by their original color grading standards, gemologists are over grading blue fluorescent diamonds? If so, would that not make a good additional argument for discounting the "bid" price of blue fluorescent diamonds?
First of all there is indeed evidence to suggest that going by the original, UV free, diamond grading standards, the trade and anyone grading with fluorescent lighting is over grading diamonds with strong or very strong blue fluorescence and possibly even medium blue. This varies little, except in degree, from the over grading that many felt was occurring when natural northern daylight was being used for color grading. By using fluorescent lighting as the standard illumination for diamond color grading, gemologists have in effect redefined the standard amount of UV to be that contained in fluorescent lighting. Perhaps this is an acceptable alternative to grading in the absence of UV, but it presents its own set of problems. Most importantly, fluorescent illumination containing significant UV energy certainly does not reveal the "true color" as defined by the early gemologists and as understood by the trade. In addition, the amount of UV exciting the fluorescence in the diamond being graded varies with the tube's manufacturer, with the tube's age and with the distance the diamond is held from the tube during grading.
The new GIA article indicates that a digital radiometer was used to measure the UV content of each of the light sources that were used in the study. They found no appreciable differences in long wave UV content from one source of fluorescent lighting to the next including the Verilux tubes used in their standard DiamondLite. They also found "indirect daylight through our windows has about as much UV radiation as the fluorescent light sources". If fluorescent lighting and daylight have a similar amount of UV radiation as the GIA measured, it would be expected that blue fluorescing diamonds would be perceived to be relatively whiter in daylight and under the DiamondLite than they would be if no UV radiation were present.
The four, color sets of diamonds used in the GIA observation studies were carefully chosen from more than 1000 diamonds, so they would be similar to one another in all respects except their fluorescence. Of course, their colors were graded in the Verilux illumination of the DiamondLite, illumination that is the standard for grading diamond color at the GIA Gem Trade Labs. Excellent quality, color photos were published with the article of each of the four, six stone sets whose colors were E, G, I and K. These photos appear to provide an indication that at least some of these diamonds have been graded higher than they would have been under the old color corrected incandescent illumination of the first DiamondLites. Because the photos were taken in incandescent illumination, relatively free of UV energy, the blue fluorescing diamonds should loose most of their color advantage. Under incandescent illumination, comparing within a set of identical color grade should reveal the approximate, relative "true color" of these diamonds. For example, the I color set's diamonds have measured fluorescence, from left to right, of: 1. Medium, 2. Very Strong, 3. Faint, 4. Strong, 5. None, and 6. Strong. If you study carefully those six stones in the face up photo, stones 2, 4 and 6 appear to have more color than the other three in spite of having been graded identically as I colors under the Verilux tube of the DiamondLite. It is more than a coincidence that these are the three with the strongest blue fluorescence, indicating that they have been overgraded relative to what they would have been had the illumination used for grading been the low UV, incandescent illumination used in the photo. The other sets are less consistent in showing this effect, but the highest fluorescing member of each of the E, G, and K color sets appears to be the most tinted of its respective color set. It is important to note that these differences in color appearance are not apparent in photos taken of the table-down view of these same diamonds, the view under which they were graded. If these same photographs were retaken using natural daylight with more UV than the Verilux tubes have, what would be the expected outcome? Using daylight film to take the photographs, you would find that the most yellow tinted diamonds, which were the most fluorescent members in each set, would now be the least tinted and might even appear bluish-white.
All this brings up another interesting problem for gemologists and anyone who uses colorimeters or spectrophotometers to assist them in diamond color grading. The GIA and many other organizations and manufacturers sell these instruments to assist the human eye in color grading diamonds. Colorimeters use incandescent light sources to make their measurements of color. Because of the differing UV content in incandescent vs. fluorescent lighting the colorimeter's measurement of color in blue fluorescing diamonds will differ from the measurements the human eye makes under fluorescent lighting. This is a further dilemma for gemologists to ponder.
So we have come full circle. Thirty or more years later, we are again facing the question of whether the illumination under which diamonds are graded should contain long wave UV, and if so how much? We can stand pat with the GIA's and the trade's use of fluorescent light sources with their attendant somewhat variable amounts of UV as the de facto standard, and live with the fact that a blue fluorescent diamond will grade whiter due to the UV energy contained in this illumination. After all a blue fluorescent diamond does appear a whiter color in all but perhaps incandescent lighting. The result will likely be a continuation of the distrust of color grading done on blue fluorescent diamonds as discussed in Rapaport's article.
There is an alternative course of action that is being taken recently by a number of gemologists and gem labs. They are adhering to the early gemologists' teaching that the body color of the diamond should be graded in an illumination that is absent UV energy. They have found a simple, inexpensive method to filter out the UV energy from all light sources be they natural, fluorescent or incandescent. GE manufactures a window glass substitute called Lexan, which is a plastic that filters out virtually all long and short wave UV. By filtering the light thru a sheet of Lexan or other UV filter, all the three types of light sources used for color grading would be put on the same footing, namely no UV. This eliminates the possibility of the so-called over-grading of blue fluorescing diamonds due to UV. Of course it is even stricter a standard than the original DiamondLite since the tungsten filament light source it used still contained the minor amount of UV present in all incandescent lights. This solution to the problem should go a long way toward correcting the trade perception that labs are over-grading blue fluorescent diamonds. There would be much better correlation, when it comes to blue fluorescing diamonds, between colorimeter and spectrophotometer grading and human graders using fluorescent lights and natural lighting environments.
Whichever choice a gemologist or a member of the diamond trade makes, whether to continue using unfiltered fluorescent lighting or eliminate the UV energy by filtering, he should be aware of the impact of his decision on the color grading of blue fluorescent diamonds. Keep in mind that there are differing views about the degree to which the UV content in standard fluorescent lighting is affecting the grading of blue fluorescent diamonds compared to their nonfluorescent counterparts. While the GIA's study concluded that "in general the results revealed that strongly blue fluorescent diamonds were perceived to have better color appearance when viewed table up" they found "no discernable trend table-down". Since this is the position used for color grading, perhaps, except for strong blue fluorescence, the grading is close to the "true color." A large number of gemologists and the trade appear to believe differently, and perhaps by filtering out the UV, these concerns can be allayed.
A final issue that both articles address is the perception held by most that blue fluorescence is sometimes associated with an attendant milky or hazy appearance. Rapaport admonishes "the labs must clearly indicate on their grading reports instances where milky fluorescence detracts from the quality of the diamond". At what point along the scale of fluorescent strength from faint to very strong blue can this milky appearance be noticed? Here again there is a difference in opinions. However, everyone acknowledges that this effect, which is a loss in transparency, has a significant negative impact on value and salability when it is present. This was a quality factor that the GIA study asked its participants to evaluate in order to make a contribution to answering this question. They did not include in the study "diamonds with extremely strong blue fluorescence and a distinctive oily or hazy appearance". In the range of fluorescent strengths, including very strong, the study found that "most observers did not detect any differences in transparency among diamonds in a given color set. Of those who did see a difference under fluorescent lighting, it was only apparent in the table-down position. These results challenge the notion that strongly fluorescent diamonds typically have a hazy appearance". The GIA believes that the study shows, up to and including strong blue fluorescent diamonds, there is no noticeable decrease in transparency in the normal face up viewing position. The study's findings indicate that this may be one negative perception of blue fluorescent diamonds, harbored by the trade that may not be warranted except in the case of extremely strong blue fluorescence.
Concluding Remarks
The author encourages you to read or reread the GIA's article and Rapaport's response in the light of this discussion. Two possible simplified conclusions may be reached as to the correct direction the gemological community and the trade should take. They both have merit.
The first direction is to pursue measuring the "true body color" and minimize the fact that most diamond viewing environments enhance the whiteness of blue fluorescing diamonds because they contain significant amounts of UV energy. The grading labs and the trade could implement, using Lexan or other UV filters, a color grading environment absent UV in order to grade the "true body color" as taught by the early GIA Diamond Course. If all color-grading measurements employ a good UV filter, there should be better consistency in color grading among grading labs and between colorimeter and human grading. This will eliminate the charge that blue fluorescing diamonds are being over-graded. If the "true color" is graded, a medium or strong blue fluorescing diamond of I color, for example, will have a perceived color in most viewing environments much whiter than a corresponding I color with no fluorescence. As long as there is no loss in transparency this should give fluorescent diamonds a real selling advantage, something they used to enjoy.
The second direction is to minimize the importance of "true body color" and embrace the fact that grading with a DiamondLite or other UV rich fluorescent light is yielding a color grade closer to the perceived color seen in the diamond in the environment it is most often viewed (natural daylight and office or home florescent lighting). Since everyone is currently using some form of fluorescent lighting as their color grading illumination why not stay with the status quo? Without consciously making any decision and by continuing to use unfiltered fluorescent light for grading, the trade has in effect already abandoned trying to measure the diamond's elusive "true body color". Because of the inertia of human nature, this is the most likely path the trade may continue following by default. And of course this will continue to leave the labs open to the charge that blue fluorescing diamonds are being over-graded. On the other hand, anyone grading in natural daylight, such as that pouring thru the floor to ceiling windows of the pictured, Antwerp diamond bourse, is also open to the same criticism of over-grading blue fluorescent diamonds.
Epilogue
The September 1998 issue of "Jewelers' Circular Keystone" has an article by Senior Editor Gary Roskin addressing "What GIA's Fluorescence Study Ignored" wherein he states "To fully understand color grading, you need to appreciate the often overlooked matter of the light under which grading is done." He asks "Does strong blue fluorescence, whether perceived by a consumer or not, enhance the color grade of a diamond?" He then answers his own question in the affirmative saying that he has compared diamonds with strong and very strong blue fluorescence in UV free light compared to "normal", "traditional" light and "the difference can be quite dramatic, possibly by two or three color grades". As a gemologist with much trade experience he appears to be heading in the first direction acknowledging the difference between the GIA color grade using the unfiltered DiamondLite verses the traditional "true body color". However, he points out thru the words of John King, GIA Gem Trade Lab's director of special projects, why GIA doesn't use UV-free lights. King says "It's not relevant because it doesn't really exist anywhere". Roskin goes on to say "Remember, all diamonds from a century ago until now have been graded in light that has a small percentage of UV. Says James Shigley, GIA's director of research, "We have this traditional way to look at diamonds, and that is the benchmark."" From Shigley's and King's remarks as well as those of Boyajian, it is clear that the GIA is headed in the second of the two directions.
Footnotes:
