Tuesday, December 29, 2009 Tungsten Plugged Gold? A month ago a friend sent me a gold investment article (at bottom) regarding the risk of buying gold ETF's backed by fake gold bars and a conspiracy theory involving the U.S. government, so I did a little investigating. I concluded that, while it is possible to fake the density of gold with gold-plating on tungsten bars, it wasn't possible to fake gold bars for even moderately sophisticated testing using other criteria. Since this dovetails with my last two posts on gold, I thought it might be of general interest so I'm posting here. I replied as follows:
----- begin reply ----- I see from google (searching "gold tungsten plugged bars fraud") that this story also was on kitco, but do you really think the U.S. Government was involved (even indirectly) in making plugged gold bars to suppress the price of gold? You didn't express an opinion.
Did they really do it? That's the $1123/oz question. I couldn't find anything on snopes. It would be a fraud the likes of which I've never heard, and the consequences would be, well, earth shaking. It seems rather unlikely, but I've gotten so suspicious of the government that I could almost believe it.
I looked up the density of gold and tungsten to verify the claim of equal density. It's hard to get an exact definition for a given temperature and isotopic composition, and web values are vague, but for best information I could quickly find, gold comes is 19.32 gm/cc at 293Kelvin and Tungsten comes is 19.3 gm/cc @ 293Kelvin (http://www.chemicalelements.com/elements/w.html).
The half-lives of gold isotopes are relatively short -- seconds to days, so it's not possible to have any appreciable mass fraction in any isotope but Au-197. Tungsten, however has stable isotopes of W-180, W-182, 183, 184, and W-186. The given values of tungsten density do not, as I say, indicate the mass fraction for each isotope, but W-184 is the most common isotope (31% of mined tungsten). That means, a bar of pure W-186 could be 1.1% more dense (or 19.5gm/cc) than a bar of W-184. So it's entirely feasible to mix 10% W-186 with 90% W-184 to get exactly the same density of tungsten as gold.
W-184 constitutes about 31% of known tungsten, and W-186 is 29% (http://en.wikipedia.org/wiki/Isotopes_of_tungsten). Even if we assume this is the fraction going into the density of 19.3gm/cc, it is still possible to skew the mass fraction of W-186 upward to give the same density of gold.
That said, there are tests other than density to detect a tungsten bar plated with gold. (eg, http://reactor-core.org/~djw/myblog/archives/2008/10/24/T12_14_56/) Nuclear resonance imaging would NOT help (gold plating shields the measurements), but magnetic fields reportedly have unique eddy-current signatures. I didn't investigate further here. It's possible. Eddy currents excited by an oscillating magnetic field will cause a material to heat up in proportion to the resistivity of the material. Some cook top ranges use this effect, and might be employed to analyze a gold bar. You could probably put a gold bar in a caloric bomb surrounded by a 60Hz field from a transformer, and measure the temperature rise over some minutes. A tungsten plugged bar should give higher temperature for the same heating time.
DC electrical resistance measurements could detect a plugged bar, if you contacted it carefully at both ends. (You'd probably needs a hydraulic press.) The resistivity of gold is 2.271 micro-ohm-cm at 300K. The resistivity of tungsten is 5.65 micro-ohm-cm. A bar plugged 50% with tungsten should have a resistivity about 3.971 micro-ohm-cm, about 74% higher. But you have to measure the resistance of an entire gold bar, which will be VERY low.
A 400 oz gold bar has dimensions of 20cm x 8cm x 4.5cm. The resistance end-to-end is 1.26167 micro-ohms. Applying a voltage of 10 microvolts across it would generate a current of 7.926 amps. A W-plugged gold bar would have a current of about 4.56amps. This would be a very noisy measurement, but if you filter the signal, it could be accurately measured with a cheap multimeter. The effect of contact resistance at both ends of the bar, however are significant. What it will do is reduce both currents and narrow the difference in dc current. That's why I say it's important to get a very good electrical contact which is very low resistance (less than the bar itself, preferably) and consistent from bar to bar measurements. But not impossible.
However, a dead sure test would be to measure the heat capacity of a gold bar, the amount of heat needed to raise the temperature of a material by 1 degree kelvin. Gold has a heat capacity of 0.129 joules / gm-K, and tungsten has a heat capacity of 0.134 J/gm-K. That's a difference of 3.88% which should be easily measurable -- unless someone was REALLY clever and used two or more elements to plug the gold to match both heat capacity and density. Not impossible in principle, but it would require research to ascertain if practical. But it would be difficult if not impossible to match heat capacity, density AND electrical resistivity, because gold is the third lowest resistance material, behind only copper and silver. But Cu and Ag are so much less dense than gold, I'm not sure you could compensate the higher resistivity of tungsten without lowering the density of the gold bar too much.
I could also imagine a sound test -- strike a gold bar with an exact force and duration with a computer controlled metal hammer, and measure the sound generated, then run that through a spectrum analyzer. A quick check shows tungsten has a velocity of sound of 5174 m/s and gold has a velocity of 1740m/s. That's a huge difference. You could hear that difference. For sure by dropping a tungsten plugged gold coin on the floor.
Another test is thermal conductivity. Gold is 320W/m-K and Tungsten is 174W/m-K. That would be just as hard to conceal, though the measurement would be awkward.
What I conclude is that plugged gold bars are easy to detect (and one reason gold is such an ideal monetary standard), but you do have to use the right techniques. I wonder what methods are used by the mints? I could see them getting defrauded by using only a density test, and once they take possession, even if they discover the fraud, they probably wouldn't want to reveal it.
It makes me wonder how silver bars / coins could be faked?
You know, in this day and age, there could be a market for low-cost gold testing machines, but of course, the thieves could fake those, too.
---- end reply to friend ----
The poster mentioned sound resonance. Coin acceptors on vending machines test electronic resonance as do gold detectors. He didn't explore that.
