The testers a typical jeweler or diamond merchant might use only identify between diamond, cubic zirconia or moissanite. They do not identify origin of a diamond.
The testers that can identify a diamond's origin are quite expensive and usually only found at major gemological labs.
All synthetic white diamonds are "Type IIa" (no or trace amounts of nitrogen) while only 1-2% of mined diamonds are IIa. 97% of mined diamonds are "Type Ia" (clusters of nitrogen). It is possible that is what they were referring to about purity, but the amount of nitrogen is not a definitive indicator of origin.
Major gemological labs can certainly identify a diamond's origin (mined or grown). There are many characteristics that can identify origin, and color or clarity is rarely used as the conclusive indicator. Here is a semi-recent study of synthetic diamonds from my company:
https://www.gia.edu/gems-gemology/spring-2014-ulrika-hpht-sy...
You are right though, gem-quality synthetic diamond technology didn't come around for several decades after this mine started producing.
Correct. The Soviet technology for growing diamonds is HPHT (high pressure high temperature). There are a handful of known producers and research facilities in the former Soviet Union that can grow diamonds, but all using HPHT. These were all industrial grade until the late 90s-early 00s, and colorless diamonds did not consistently grow with this technology for several more years.
The other technology you mention is CVD (chemical vapor deposition), which was not developed in Russia, and was not commercially viable for gem-quality diamonds until the last decade either.
> These were all industrial grade until the late 90s-early 00s
I believe the point being made is that the only ones publically admitted to being lab made were the industrial grade. The lab made gem grade were made to look like mined stones. The belief is the Soviets kept this a closely guarded secret because it was very profitable for them.
Even today there are only a couple places in the former Soviet Union that can grow gem-quality synthetic white diamonds. Their production numbers are far less than the output of a single mine, and the results are still rather random. Prior to 2004 or so, the only synthetic diamonds that could be produced commercially were colored and primarily industrial-grade.
Well, I say "gem-quality" because the article did, but it also points out that the diamonds were a greenish color, and incredibly consistent in size and quality.
When DeBeers finally got a 20-minute tour of the Siberian mine, far less earth had been excavated than they expected for the number of diamonds that the USSR sold.
There are many additional circumstantial details in the article.
Diamond growers try their best to keep out impurities and imperfections. Elemental impurities do add color (nitrogen=yellow, boron=blue), same as mined diamonds, and faults and inclusions do naturally happen in the process. However, at an atomic level, they are different than inclusions and imperfections in mined diamonds.
As a producer, I don't see any incentive to make them imperfect on purpose, yet still distinguishable from mined diamonds. All larger diamonds intended for gemstones come with independent grading reports, which will still identify it as grown regardless of presence or lack of impurities and imperfections.
While not technically correct, some of the advanced detection equipment can be thought of like looking at growth rings on a tree. You (probably) can't replicate the growth rings of a 200 year old tree in a 5 year old tree, yet both are perfectly valid sources of wood (sustainability, etc. aside).
I have other comments in this thread about cost, but it basically comes down to "they cost a lot more to grow than people think", and the costs after the rough diamond is grown are basically the same as mined diamonds (cutting, grading, etc.). Diamond is the hardest substance on earth, and it costs far more to polish one of those, regardless of origin, than it does to polish an imitation or other gemstone.
If you're buying raw carbon in the form of coal, it costs about $40 per ton.
The rest is manufacturing details. I know there's a lot of manufacturing details, but the cost of the inputs suggests that diamonds in bulk could be much much cheaper.
If a large enough production facility (no R&D) had nearly every growth cycle yield large enough, colorless, flawless diamonds (or close to it), the retail prices could come down more than they currently are. However, that does not reflect the current reality and all producers today are basically in R&D mode with mixed success rates and mixed finished quality.
If someone showed up tomorrow with a check for several million dollars, I believe we could "disrupt" the current diamond pricing within 3-5 years. However, as I mentioned in another post in this thread, raw cost of production will still keep those "disrupted" non-R&D diamonds much more expensive than moissanite, CZs, sapphires and other gemstones.
Diamond is a form of carbon, and its growth is more-or-less determined by nature. The cost of production over time will be more similar to an industry like steel, than assembled goods like TVs or computers. There are improvements to be made, but they are more linear over time, than exponential breakthroughs.
For raw carbon, we use a highly refined and purified form of graphite. If you used the $40/ton coal, you would not be successful in growing jewelry-quality diamonds, and even if you did, the other costs wouldn't really change your ultimate price much.
Another analogy is gold. You can buy a 1 Troy-ounce bar of 24 karat gold for a bit over $1000. However, a solitaire engagement ring with much less actual gold content can cost more than $1000. People ask why the jewelry is more expensive, but don't include the cost for refining and alloys, design, 3D printing or molds, casting and investments, polishing, setting, shipping, ring boxes, and all the other overhead associated with turning that gold bar into a sold, ready-to-wear ring.
Do you only grow gemstones or do also for industrial applications? And in case of the latter, how hard is it to make monoisotopic diamonds? I've read they have about the best thermal conductivity of any solid.
Gemstone and high-value industrial. Industrial diamonds fit into a couple different levels. Low-grade diamonds are made spontaneously by the ton, mostly in China. They are generally small (<2mm) with poor color and clarity and are used in cutting tools, drill bits, etc. Mid-grade industrial diamonds are larger (3~6mm) seeded-crystals with good clarity, but color rarely matters, and are used in cutting tools and heat sinks. High-value industrial diamonds are usually higher quality than jewelry-grade diamonds (IF-VVS clarity, D-F color, low stress, etc.) and are used in applications like high pressure anvils and optics for lasers. A large portion of our current production ends up in the last category, as well as in jewelry.
We are currently focused on high-quality single-crystal diamonds. There are many potential research paths with diamond, but slow and limited research capacity and market potential. Element Six (a De Beers subsidiary) currently has the most potential for developing non-standard diamond applications.
For HPHT, there are three primary machine designs, but they all create intense heat and pressure, and dissolved graphite slowly builds up on a diamond seed (<1mm).
1. BARS press. It is a Russian design from the 1980s capable of up to 2-3 carat polished sizes, and was one of the first methods to commercially grow jewelry-grade diamonds. It is much less efficient than modern presses though.
2. Cubic press. This is a much larger 3-axis press used primarily in China to spontaneously grow diamond grit and powder. Some of these have been converted and upgraded to run longer cycles needed for large single crystals. These can grow multiple diamonds at a time or fewer larger diamonds, and have been used to grow the largest diamonds currently available (5-10 carat), however success rates and control within the larger growth cell are still low. We are in the process of developing our own modern cubic press (rather than a refurbished grit press).
3. Single axis. This our own in-house design. It has similar growth capabilities as a BARS press, but is much more efficient with greater control and can grow multiple diamonds simultaneously. For qualified parties, we can sell these presses as well as license the IP and diamond growth "recipes".
CVD reactors are basically a vacuum chamber with plasma over a growth surface. That surface holds diamond plates, which are just thin slices of diamond, and usually come from larger CVD or HPHT single-crystal diamonds. Methane provides a carbon source, which is broken up into its elemental components by the plasma. The carbon "rains" down onto the diamond seed plates and basically grows straight up, so the finished dimensions are limited by the starting length and width of the seed plate. Some CVD reactors use microwaves to assist while others do not. There are a few companies that sell complete CVD reactor systems, but no one I am aware of that offers IP or "recipes", so those would have to be developed on your own.
I guess a follow up question would be, if one wanted their own diamond making machine where would they get it. Obviously your company is an option "for qualified parties"
Yes, we can sell brand-new presses along with the IP and recipes, or can consider a JV, production contract or some similar arrangement.
Otherwise, for HPHT, you can find used BARS presses, mostly in Russia. They are certainly capable of growing diamonds, but are like comparing emissions, performance and fuel efficiency of a 1980s carbureted vehicle to a 2016 fuel-injected vehicle. Used cubic presses can be found in China, however those presses were built for grit and powder, which run short cycles (<1 hour), and would need to be converted and upgraded to sustain strict parameters for multi-day and multi-week cycles.
Companies sell new CVD reactors, as they have more applications than just diamonds, but they do not come with any diamond knowhow.
With all of those, you will have to develop your own recipes and growth cells or methods. Unless you have plenty of time and advanced degrees in physics, chemistry and/or material science, it will probably be an exercise in futility.
The original Florida-based Gemesis is a good example of this. They were VC funded and bought BARS presses, that at-the-time, could only grow orange diamonds. Gemesis made dozens if not hundreds of these presses however, when they ran out of money years later, they could still only grow orange diamonds. A Singapore company bought those presses (mostly to grow CVD seeds and HPHT-treat CVD rough) and to my knowledge, still can't reliably grow blue or colorless diamonds with those BARS presses.
For a code analogy, they basically "forked" the orange recipes when they bought the machines from Russia, and haven't been able to merge any upstream advancements since. BARS have basically been deprecated, so that stack is no longer developed.
They are smaller than people think. Not counting the years of R&D, one average growth cycle for one white diamond actually costs more than what mining companies typically report for their cost-per-rough-carat dug from the ground. Mines use giant diesel earth moving and processing equipment, while we use scientists, electricity, and advanced machinery and alloys, but the variable cost-per-carat for rough diamonds are somewhat comparable.
From there, the cost for polishing, grading reports, jewelry, logistics, marketing, etc. are all basically the same whether the diamond was grown or mined. White grown diamonds are generally 10-30% less expensive at retail than mined diamonds, however most of that is just due to lean, vertically integrated companies. We grow, cut, distribute and retail our diamonds as well as jewelry, so have a bigger slice of the rough-to-retail pie. That extra margin isn't huge, but is enough to sustain the business as well as progress the research, development and production. Mined diamonds pass through many more hands between the mine and the consumer, and each step adds some markup. A mined diamond entity that only sells rough, only polishes, or only buys and sells polished mined diamonds at wholesale would see smaller margins, but that can be made up for in greater volume of mined diamonds traded.
It is also worth pointing out that the diamond success rate is sustainable, but not every growth cycle is profitable. While we can monitor the external HPHT equipment, we cannot see the actual diamond growth until after the cycle is complete. For example, a cycle may run for two weeks, but we find out afterward that it stopped growing on day three, or there were some bad inclusions on day six, and what could have been polished into a one carat gemstone (and cost the same to grow) may not yield anything, or may only yield a much smaller gemstone. The total production cost has to be averaged over the total successful sales. This includes growing diamonds that sit in inventory for a long time due to lower clarity, less desirable colors, etc.
Do the lab grown diamonds come out of the process as the cubish shape as rough diamonds? Do you still send them off to be cut/polished by 3rd parties? How do you sell them?
They all need polished into their final shapes (round brilliant, princess, pear, anvil, etc.), or can be sliced into plates, cubes or cylinders with lasers.
HPHT (high pressure, high temperature) grows from a tiny diamond seed (<1mm) and the growth sort of "snowballs". Blue and white/colorless are a hexacubic type shape while yellow are a truncated octahedron shape.
CVD (chemical vapor deposition) grows basically straight up from a diamond plate, so are generally cuboids. Spontaneous polycrystalline diamond can grow on the sides, but is hard, black and unusable for the same applications as single-crystal.
CVD (poly is cut off): http://d.neadiamonds.com/images/rough-cvd-brown.jpg
When CVD grows brown or gray, it is usually due to atomic-level defects in the diamond, which can be healed through post-growth treatment, turning it a near colorless or light yellow color. This CVD photo was HPHT treated to a light yellow, then further treated to become a nice fancy pink. It is also possible to have CVD grow near colorless, without requiring treatment.
We sell diamonds for jewelry on our retail website: https://d.neadiamonds.com
They are also available for high-value industrial and wholesale jewelry applications.
Yes. Cutting blades are in the middle of the industrial spectrum. They are generally 2-5mm in dimension and need to be grown relatively clean, but color does not matter, so they are commonly yellow or brown (those colors grow faster than white or blue).
Are other lab grown gemstones made with roughly the same HPHT process? I ended up getting a diamond from Gemesis for my wife years ago, this is a very fascinating topic for us.
The core science of diamond growth inside an HPHT cell is essentially the same (a diamond seed, heat, pressure and graphite). There are a lot of variations in the composition of the cell though, which can control color (brown, orange, yellow, blue, colorless, near colorless), as well as overall quality, quantity and size of the diamond.
Outside of the growth cell, the machine and environment also influence growth. Temperature and pressure ranges, tolerances and gradients of the press all factor in to the final quality of the diamond. Ambient room temperature and humidity as well as conditions during the growth cell preparation play a role too.
For consumers though, when considering two 1.0ct G color, VS2 clarity round polished diamonds (for example), it doesn't really matter what process it came from, who grew it or what size or shape the rough diamond was.
Why are lab grown diamonds still priced in the 1000+ range. Does it actually cost that much to manufacture.
Making up research cost?
Just high profit margin?
TL;DR They have very similar cost structures as mined diamonds (rough [mine vs. grow], polishing, grading, etc.), plus the R&D and failure/success rates.
They are 100% real diamond, just grown instead of mined. A typical jeweler cannot conclusively identify a diamond's origin, however gemological grading labs with more advanced equipment can. Companies are trying to develop inexpensive testing devices, like exist for cubic zirconia and moissanite, but so far they are expensive and limited to those major grading labs.
There are a few ways to identify the grown origin:
-Most grown diamonds sold for jewelry have independent grading reports from gemological grading labs (GIA, IGI, EGL, GCAL, etc.) identifying them as grown. Part of this grading process is to laser inscribe the diamond with wording like "Laboratory Grown". This inscription can be read on the diamond with 10-20x magnification.
-Inclusions can be different in grown diamonds, however are graded on the same clarity scale as mined diamonds (VVS, VS, SI, etc.). Metallic inclusions are extremely rare in mined diamonds, but common in HPHT-grown, since they grow in a molten metal solution. CVD inclusions can be graphite or have planar characteristics.
-Grown white/colorless diamonds are all "Type IIa", which means few to no impurities. Less than 2% of mined diamonds are type IIa. ~98% of mined diamonds are type Ia and actually contain more nitrogen than grown fancy yellow diamonds (type Ib). Equipment that can check these impurity levels are good initial screening tests (2% false positive for IIa mined diamonds). For completeness, type IIb contain trace amounts of boron, and make the diamond blue and electrically conductive. These IIb diamonds can eventually be used as semiconductors.
-Disclosure. We are proud of our grown diamonds, and the other producers are too. Most of our customers buy them because they are made by scientists and technicians in high tech labs. It took longer to grow jewelry quality diamonds than it did to put humans on the moon. Mined diamonds do support economies in remote and third-world regions of the world, but can come with their own environmental and social issues.
The core HPHT technology was originally developed by Soviet research institutions. After the fall of the Soviet Union, this information basically became public domain. It took decades of incremental improvement to get the diamonds up to jewelry-grade qualities. Prior to jewelry-grade though, the production could be used in industrial applications (cutting blades, optics, etc.).
Traffic lights. While not directly personal or "Internet of Things", traffic lights could really use some AI. It would be fantastic if they could learn routine traffic patterns (rush hour, weekends), detect flow (green light, but no one is there), even receive traffic data from Google/Apple/etc., then automatically adjust timing accordingly. Even mesh networking with nearby intersections.
I can't count the number of times I have been waiting at a red left turn light, with a green straight light and no other cars around. Or, backed up at a red light with no cross traffic, yet cross traffic has a long green.
It is probably a complex problem to solve and suspect the biggest barriers are bureaucracy and control. Is there anyone on here that works with traffic lights? It seems like they are setup once with a predefined timing and are rarely ever changed.
This already happens in some cities. There are traffic lights in my city, for example, which adjust for trains, presence of cross-traffic, etc. The cross-traffic one is neat because a light will, quite literally, stay red forever and only change when a car pulls up to actually use the green light.
From this[0] article:
> Today, cities use computer-controlled traffic lights that adjust their timing based on traffic levels, the time of day and even the number of trucks on the road. In Los Angeles, for example, city officials use traffic management to control their 4,400 traffic lights, reducing travel time by 12 percent.
My hometown had one of those "cross-traffic" lights a few blocks from my house at an intersection that I had to cross to reach one of the major bike paths.
Naturally, it didn't register bikes. So if you wanted the light to ever change, you had to cross over to the left side of the street, go up on the sidewalk, and push the pedestrian button (no crosswalk on the right side of the intersection). Not much of a fan.
I'm not saying they're always bad, but they're not always good either. Maybe newer ones have improved on detection, but my understanding was that the sensor was induction based and required a big chunk of metal in order to spot you.
I've seen the induction loops "presence" sensors too. One near where I used to live would never change to green unless the sensor was triggered, but wouldn't register until your car was beyond the crosswalk, nearly in the intersection. That means if someone in front of you stopped where they were supposed to, the light would never change.
I suppose there is bias though, in that people don't notice as much when traffic lights are operating efficiently.
I looked around some before I posted that. This article seems respectable enough and at least supports the idea that bikes are intended to be able to trigger some of them:
It looks like those are specifically designed to be triggered by bikes. I've been at intersections with a 50-100cc motorcycle that would not trigger the light.
In my area I find that the cross-traffic ones are poorly implemented. I'll drive up, come to a complete stop, the light will change for me and will already be back to yellow before I clear the intersection. This forces me to always stop when there was no traffic on the preferred route. It also means that if there is a car more than a few car lengths ahead of me it will change for them and force me to wait a full cycle even when there isn't traffic on the preferred route.
This shouldn't be a hard problem to fix. Sensors just need to be installed farther from the light to detect approaching vehicles earlier.
I don't think they're always looking for efficiency of traffic flow. I live on the south side of Chicago. I swear the lights along major streets are timed to make it inconvenient for traffic flow ON PURPOSE. My unscientific theory is that this is done to prevent speeding, rack up violations for idiots that like to race, and make it easier for cops to pursue in their cars. Case in point: my street has a stretch of at least 3 blocks without a light; hence, we get idiots (cars and cycles) that like to gun it in a 30 mph zone.
When I was a kid, we went through some small town when on vacation (don't remember which one). It didn't have speed limits - just signs that said that the traffic lights were synchronized for 25 MPH. After a light or two, you believed them, and so you drove 25 MPH.
Many lights and transportation networks already do (NYC is one great example). It's not just lights - there are all sorts of inputs into that system.
It does take some resources, so you typically don't see that sort of intelligent routing outside of the major cities (at least in the US), but the tech is definitely there (not that it couldn't be improved).
Yes, I suspect major cities already have complex and optimized traffic routing, along with full-time traffic engineers. The mid and small cities are what I experience most, but I suppose these also have less traffic, so less need to be fully optimized.
I find most lights these days have sensors to detect when cars are present and some basic logic to alleviate many of the problems you mention. No system is perfect, of course.
If looking only at jewelry-quality diamonds, there are millions of carats of diamonds mined per year, while there are generally thousands of carats of diamonds grown per year. In that regard, lab-grown white diamonds are much more rare than mined diamonds. It will take billions of dollars in capital to have a diamond growing facility with enough capacity to output more jewelry-quality diamonds than a single large diamond mine (though I suppose that is less than one WhatsApp, so is within the realm of possibility).
The Wired article about synthetic diamonds from 2003 was full of hype and misconceptions. Most all synthetic diamonds grown today are not flawless, and that is not by design. However, in the last couple years lab-grown white diamonds have become much more available in normal jewelry-quality ranges:
http://d.neadiamonds.com/lab-created-diamonds/White-Diamonds
^ Disclosure: I'm an owner of D.NEA and have been selling jewelry-quality synthetic diamonds for many years.
Thanks for the additional input, EF. The disparity is much larger than I initially thought. In time, as methods & costs are refined, perhaps the ratio will shrink and reverse.
I also believe that is the article I was referring to... IIRC the 'white' conundrum was the big impediment to mass production... much like the recent breakthrough in LED tech's progression to white light.
As a producer of synthetic diamonds (http://d.neadiamonds.com), I can say the production costs for jewelry-quality diamonds are not as low as people seem to think. It is one thing producing brown/yellow diamond powder/grit for cutting tools, but is orders of magnitude more difficult growing a large single-crystal diamond colorless and clean enough to set into jewelry.
The capital equipment for HPHT and CVD are both still quite expensive. It is possible to find some used BARS presses for reasonable prices, but you will be hard-pressed to make a large colorless diamond with one of those machines, even if you know the right "recipe" to use. Gemesis has many of these BARS presses and they have only been able to produce orange yellows and to treat CVD material with them.
CVD does grow more crystals per machine cycle, but also has much higher labor, power and support costs than the latest generation of HPHT machines. CVD diamonds also typically grow as a brownish or grayish color and have to be HPHT-treated (different process than HPHT-growing, but can be done in the same machines) at additional cost to whiten them, healing defects in the crystal lattice.
The cost to grow a rough white diamond is generally comparable with the cost to mine one from the ground. From there, the cutting, grading, logistics and jewelry all cost essentially the same.
Lab-grown diamonds are a raw good, more similar to steel, than they are an assembled good, like a TV or laptop. There will certainly be more improvements along the way, but diamond synthesis only occurs under certain conditions defined by nature. Changing the crystalline structure of carbon is a bit more involved than heating up some filament for a 3D printer.
Thanks for your informative comment. I know less about HPHT diamond synthesis, and may not understand the cost structure of that technology as well as I would like.
I agree re the current high capital cost of CVD diamond synthesis equipment, but I'm pretty certain it need not remain so. For example, in microwave plasma assisted diamond CVD, significant slices of the cost pie are in the microwave power source and the deposition chamber. The former tends not to take advantage of 2.45 GHz consumer sources and is, I think, overpriced in $/Watt compared to what it could be with some additional electronic design work. The latter suffers because diamond microwave CVD chambers tend to be one-offs. Building them in hundreds or thousands would allow lower cost manufacturing technologies to be used.
Power costs are an issue, but there are ways of extending the lifetime of atomic hydrogen, which is a key cost determinant of CVD diamond. Labor costs will be reducible to the extent that CVD processes can be automated, which I regard as largely a matter of getting reproducible processes in hand. When you have a predictable process, you can automate it.
I concur it will be awhile before my Replicator 2 can spit out a diamond filament. But I think the current manufacturing cost of diamonds is far higher than what it might be. The missing link is somebody willing to fund the volume manufacturing process development.
The testers that can identify a diamond's origin are quite expensive and usually only found at major gemological labs.
All synthetic white diamonds are "Type IIa" (no or trace amounts of nitrogen) while only 1-2% of mined diamonds are IIa. 97% of mined diamonds are "Type Ia" (clusters of nitrogen). It is possible that is what they were referring to about purity, but the amount of nitrogen is not a definitive indicator of origin.