Why do metals smell?

Every fusor and fusion system seems to need a vacuum. This area is for detailed discussion of vacuum systems, materials, gauging, etc. related to fusor or fusion research.
JohnCuthbert
Posts: 339
Joined: Mon Aug 19, 2002 8:30 pm
Real name:

Re: Why do metals smell?

Post by JohnCuthbert » Sun May 23, 2010 11:59 am

Many, if not most, finely divided metals are spontaneously flammable in air. Clusters would be even more reactive.
Cleaving a bunch of atoms from the bulk would take more energy than taking a single atom. The vapour pressure is so low that we know the number of atoms coming off a metal is practically zero. The number of clusters would be even less.

Some metals like selenium, arsenic and osmium have strong smelling oxides that are fairly readily formed in air and would explain the odour associated with them.
There's a definite smell associated with working aluminium. It is due to things like arsine and phosphine formed by reduction of the metals present as impurities in the Al in the presence of atmospheric water.
The odour of iron and copper has been explained as due to decomposition products of skin oils formed in the presence of air and the finely divided metal.

User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 7:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: Why do metals smell?

Post by Doug Coulter » Sun May 23, 2010 7:56 pm

While Frank is correct that a lot of things are (mostly fortunately) odorless to us, that's far from the whole story, and in fact, smell can detect levels bordering on a single molecule. That is fact, well proved, in many sorts of animals (and non animals) that have some sensitivity of which humans are just one of the samples, and we're not the best at this by far in the living kingdom.

In the '70's, when I was an engineer for a beltway bandit (ENSCO), we worked on a project for the forerunner of DHS (FBI and the DOD security crew) that used rats to detect high explosives. Yes, really. At the time, the only thing that could match a rat's nose for sensitivity was a combination of GC and mass spectrometry, and it took best of breed of both together to rival a rat's nose. Neither portable or quick enough for the requirement.

You could train them to get excited when they smelled TNT etc by the usual methods every experimental psych student learns, but there turned out to be a glitch. They were so sensitive that you could only use the same cage, or even lab just once!

We were using EEGs to detect their excitement when they sniffed a TNT sample, and at first it worked pretty well. The Air Force (Rome labs) then contracted out for bulk rat training since the experiment/prototype had been so successful. Well, the next trainers weren't quite as careful, and couldn't get rats trained correctly, as they contaminated their own lab with ~ fraction of a milligram of TNT and from then on, the rats they tried to train *always* smelled TNT and so learned to cue off other things in the experimental procedure that were irrelevant in the field where the rat was in a dark suitcase with a little vent fan and a Z80 processor and analog amps. This was proved when we went back to the original, more careful lab for another try (they were using techniques similar to serious bio-hazard or chem warfare as their norm) and it all started working again.

Needless to say, something so hard to do and fairly impractical didn't make it into mainstream use, but it's a true story just the same. Rats (and some other more high maintenance animals) are so sensitive *and* selective (for some things) that it becomes the problem rather than the solution.

Feynman had some good words on rats everyone should look at that kind of show the problems.
I can provide text of the lecture to any interested.

Humans aren't as good, and we don't train ourselves (mostly) as well as a being who may depend on the skill for survival moment by moment more than we do, but that doesn't make the fact that you (and a lot of people) *can* smell metals and even tell which sort bunk -- the proof lies right here and in a lot of other places.

The fact that it tends to happen most when the metal has just been made really clean indicates that the well known fact that super-clean surfaces are highly reactive takes part in this. Remember, it only takes a few molecules, as in a single-digit number, if it is something you are sensitive to. So all the usual arguments about low vapor pressure, non-reactivity have to be viewed in that light. Could be that clean metal reacts with some rare air component (say a sulfur or chlorine atom) and then does have a higher vapor pressure, for example. Some people can smell the difference between clean and dirty copper, there must be a reason, because IT EXISTS. Calling it bunk does not make it go away.

I can detect fusion even if only a tiny fraction of the D's in there get involved, too -- with the right gear, and that gear is mostly less sensitive than the nose is in terms of input particles per successful detection. See many of Carl's posts for detector sensitivities that work out to big negative exponents that still work fine for us. What's so different here? Nothing at all. We are simply debating the mechanism, not the existence of the effect.
Why guess when you can know? Measure!

JohnCuthbert
Posts: 339
Joined: Mon Aug 19, 2002 8:30 pm
Real name:

Re: Why do metals smell?

Post by JohnCuthbert » Tue May 25, 2010 6:23 am

Can you cite evidence for the "single digit" molecules for odour detection in humans?

Also, since it's documented that, for example, dogs have at least 10 fold better sense of smell, does this mean that they can detect a few tenths of a molecule?

User avatar
Chris Bradley
Posts: 2931
Joined: Fri May 02, 2008 11:05 am
Real name:

Re: Why do metals smell?

Post by Chris Bradley » Tue May 25, 2010 9:19 am

Maybe there's a little hearsay in this thread but, hey, when did dogs ever smelt lead tin and copper? Maybe it's an evolutionary thing - if you could sniff a prehistoric smelting site and head towards it, then you were destined to evolve into modern man, whereas the neanderthals who did not smell it died out. Maybe it's just that the human nose has evolved to smell metals, as they have been so important to us? Just a thought!

User avatar
Frank Sanns
Site Admin
Posts: 1621
Joined: Fri Jun 14, 2002 6:26 pm
Real name: Frank Sanns
Location: Pittsburgh, PA USA

Re: Why do metals smell?

Post by Frank Sanns » Tue May 25, 2010 2:28 pm

John,

The way I see it, a single atom could produce an reaction at an olifactory center if the atom finds its way to the exact proper location. If it falls near but not near enough then no productive reaction so statistic are in full play here.

Even if the atom or cluster of atoms do find a single smell detecting nerve, I see no reason why we would detect the "smell". Nerves experience noise just like any detectors that we may use. The body learns to ignore the background nerve firing so to get a meaningful stimulus, many neurons need to be involved which means many atoms or molecules or clusters.

I do not have a reference but I think the reason dogs can smell better is simply that they have a much bigger olifactory surface area (bigger detector) than humans. Bigger detector usually means more counts for a given concentration and better signal to noise ratio. I see no reason why physics should change for a biological detector over a non biologic.


Frank Sanns

JohnCuthbert
Posts: 339
Joined: Mon Aug 19, 2002 8:30 pm
Real name:

Re: Why do metals smell?

Post by JohnCuthbert » Tue May 25, 2010 5:53 pm

It's perfectly possible that a single molecule or atom of a material landing on a receptor will cause a signal. However that's not the real question.
The question is how much stuff does it take before a person perceives the presence of a smell? (in fact, we are asking for a more difficult job- the identification of a particular smell iron vs. copper vs. Aluminium...).

A bit of Googling indicates that the lowest odour thresholds are about 0.01 ng/litre and these compounds have molecular masses of about 200.
That means we can detect ( given a good healthy sniff of a litre or so) about
30,000,000,000 molecules assuming I have kept track of the zeroes correctly
If anyone really thinks that's "single digit" they should be in advertising.

It's also possible that there are molecules out there that are smellier, but we would know about them.
There's no doubt that metals have a smell associated with them.
That smell is not the metal per se.
Incidentally, TNT has a vapour pressure of about 10 ppb so it's fairly well into the detectable range of odours- it just doesn't smell very strong to us.
Rats, like many animals, have a better sense of smell than ours.

Mankind has not been smelting metals for long enough to affect evolution so the lack of rats that smelt metals is a red herring.

User avatar
Chris Bradley
Posts: 2931
Joined: Fri May 02, 2008 11:05 am
Real name:

Re: Why do metals smell?

Post by Chris Bradley » Tue May 25, 2010 7:50 pm

John Cuthbert wrote:
> Mankind has not been smelting metals for long enough to affect evolution so the lack of rats that smelt metals is a red herring.
I don't dispute the possible irrelevance of my observation, but evolution can take place not necessarily through genetic changes but also through selection of existing traits. If some select group of a coastal species senses an earthquake is coming and makes for high ground, whereas the rest do not and drown in a tsunami, then that species will have selected for the complex attribute of 'sensing earthquakes and going for high ground' within a single day.

There is no greater evolutionary event than "not dying"!

A similar thing has been found with lactose intolerance, so I seem to recall. Before domesticated cow milk, lactose intolerance was widespread, but mostly died out, as a general genetic trait, within just a couple of generations for communities where cows were domesticated - all the babies fed on the stuff died! (which was probably nothing too unexpected anyway for folks in those days, just a slightly higher rate than an already high rate of infant mortality, I guess)

I was just speculating that, just perhaps, the size of one's organ doesn't necessarily dictate its efficiency at its particular task! Maybe the human nose is particularly sensitive to metal smells and needs be no bigger? A sensitivity in mass/volume is clearly just an approximation and will differ for different substances. My posit may be wrong, but I do not feel it is *so obviously* incorrect without some substantive evaluation.

I guess a simple initial test question would be - Steven S.; having worked with gold and platinum, d'you think you could smell which one was which in a blacked out room? I presume there would be no chemical intermediaries with noble metals?

JohnCuthbert
Posts: 339
Joined: Mon Aug 19, 2002 8:30 pm
Real name:

Re: Why do metals smell?

Post by JohnCuthbert » Tue May 25, 2010 9:13 pm

When did an individual or an entire community last die out due to the inability to smell iron?

Babies are not lactose intolerant (where the definition of baby is someone young enough to subsist largely or solely on milk) no matter what their genetic make-up unless they have a very unfortunate lethal mutation.
Lactose intolerance only troubles those old enough to have been weaned.

User avatar
Chris Bradley
Posts: 2931
Joined: Fri May 02, 2008 11:05 am
Real name:

Re: Why do metals smell?

Post by Chris Bradley » Tue May 25, 2010 9:45 pm

John Cuthbert wrote:
> When did an individual or an entire community last die out due to the inability to smell iron?
I could invent scenarios, but they would be inventions. I'm not trying to convince anyone here, it was just a comment!

Seems like there are a load of guesses coming up as to what people actually smell when they perceive a 'metal' smell, so if others can make guesses then I feel at liberty to make some aswell!

Wikipedia says for iron, the smell is aldehydes released on contact with skin secretions. Well, that may be so but it doesn't smell like any aldehydes I know. If these metal smells are released chemicals, then it should be easy enough to prove by getting hold of, or reproducing a proposed chemical, and seeing if it smells like the metal.



> Babies are not lactose intolerant (where the definition of baby is someone young enough to subsist largely or solely on milk) no matter what their genetic make-up unless they have a very unfortunate lethal mutation.
Yeah!!... not any more...that's my point! I'll look out for a citation on this research. It was one of these studies looking at distributions of genes in the population, and by whatever means they have developed to analyse changes in gene patterns, they identified the lactose intolerance died out essentially instantaneously around the time of cattle domestication. The cause would seem pretty obvious!!

JohnCuthbert
Posts: 339
Joined: Mon Aug 19, 2002 8:30 pm
Real name:

Re: Why do metals smell?

Post by JohnCuthbert » Wed May 26, 2010 6:22 am

I don't mind guesses, but silly ideas like people being able to smell single molecules are just a waste of bandwidth.
Also, no baby mammal is lactose intolerant. All milk has lactose in so any baby born with this condition would die.

Post Reply