Wikipedia:Reference desk/Archives/Science/2012 July 27

Science desk
< July 26	<< Jun | July | Aug >>	July 28 >

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

July 27

Synthesis of ZnS

I'm trying to make some zinc sulfide for an experiment that I'm doing. I'm using the following reaction:
$ZnSO_{4}(aq)+Na_{2}S(aq)\rightarrow ZnS(s)+Na_{2}SO_{4}(aq)$
At first I didn't bother weighing out stoichiometric amounts since I'm not interested in the yield or anything, so I just mixed the two solutions together, filtered out the precipitate and washed it a few times to remove any remaining reactants and the sodium sulfate and then dried it in a dessicator at room temperature. The result looked like brown sugar, and not at all like the white solid described in the wikipedia article. So, I tried weighing out stoiciometric amounts and washing very thoroughly with hot deionised water, but again, I'm getting dark coloured crystals that don't seem to be very pure. Also, when mixing the two solutions in the first place, I can smell hydrogen sulfide. That indicates to me that I'm not going to have the stoichiometric amount of sodium sulfide present anymore, but the excess zinc sulphate is so water soluble I figure it should easily wash out. Any ideas what's going on here? 203.27.72.5 (talk) 02:32, 27 July 2012 (UTC)

Your sodium sulfide could easily lose some H2S and become more alkaline, so you may have precipitated some zinc hydroxide. If you had sodium hydroxide you can get sodium zincate. in Qualitative inorganic analysis ammonium chloride, ammonium hydroxide, and hydrogen sulfide gas gives a white precipitate with a zinc solution. If you have lead it will be brown or black, and if it is cadmium you will get yellow precipitate. Graeme Bartlett (talk) 12:03, 27 July 2012 (UTC)

I've actually bubbled H₂S through a solution of ZnSO₄ before to make a white precipitate of ZnS (at least I figure that' what it must have been), but the particles were so fine I couldn't filter them with the filter papers I have here; so what does the ammonium chloride and ammonium hydroxide do in that analysis? I've done a similar thing with lead to make PbS which was black as you said and the particles were easy to filter out. For that I used a ferric chloride solution to get the lead to dissolve.

If the impurities are zinc hydroxide and sodium zincate then they're probably not going to interfere with the rest of my experiment anyway. I'm developing a method to distinguish between sulfate and sulfide in ore. The idea is that if I add conc HCl to my sample of ore and heat it to dryness, then any sulfates will remain in the residue but any sulfides will be lost as H₂S. I'm just trying to confirm with pure ZnS that essentially all of the sulfur will be lost as H₂S, and then I want to do the same thing with ZnSO₄ and PbSO₄ to show that sulfate remains. 203.27.72.5 (talk) 23:47, 27 July 2012 (UTC)

You should buy it pure instead. Is there any reason you can't or don't want to? The reaction you've chosen is fraught with practical difficulties and side reactions. 207.224.43.139 (talk) 21:23, 28 July 2012 (UTC)

I'm at remote mine site, so shipping bought products takes weeks and I want to do my test work a fast as possible. 203.27.72.5 (talk) 22:17, 28 July 2012 (UTC)

Don't ask Wikipedia how to do your job. Whoop whoop pull up ^{Bitching Betty | Averted crashes} 22:27, 28 July 2012 (UTC)

I'm not sure of the wisdom, but it's allowed if you're not a physician or an attorney.

Everyone has the right to be stupid, but where one's stupidity might kill or seriously injure others, as in mining, stupidity is less acceptable. Whoop whoop pull up ^{Bitching Betty | Averted crashes} 23:48, 28 July 2012 (UTC)

Consulting a encyclopaedia or library reference desk when doing my job in entirely appropriate. And I'm doing a lab scale experiment in a 500mL beaker, not designing a 400 cube reaction vessel. I'm also qualified to interpret the answers according to a couple of universities. 203.27.72.5 (talk) 23:52, 28 July 2012 (UTC)

Zinc sulfide#Laboratory preparation suggests combustion. Will that work? If not, see if you can get [1] at WP:RX, as long as you use it to expand that section of that article. 207.224.43.139 (talk) 22:36, 28 July 2012 (UTC)

Finely powdered zinc and sulfur do "burn" releasing plenty of heat when they react. I have seen it used as a rocket fuel. Graeme Bartlett (talk) 23:40, 28 July 2012 (UTC)

I don't have any pure elemental sulfur, ethylenediamine or triethanolamine, so I can't do any of that. I looks like I've managed to get reasonably pure stuff using my above method by controling the pH. The assay is 31.9% S and 67.5% Zn which is close enough for me. 203.27.72.5 (talk) 23:52, 28 July 2012 (UTC)

Resolved

Sun

When did people find out that the Sun was a star? 71.146.10.213 (talk) 05:34, 27 July 2012 (UTC)

There's some good content at Star#Observation_history. Basically, the idea that stars were just far away suns is as old as ancient Greece, but was not a popular until the 18th century. It would not be until the 19th century that spectroscopic data was available to really study stars in any detail beyond their motions. Someguy1221 (talk) 05:42, 27 July 2012 (UTC)

Thank you. 71.146.10.213 (talk) 20:12, 27 July 2012 (UTC)

Follow up. Can our scientific knowledge be understood any other way than it is? 2

What if the equations of physics were formulated from the perspective of a single quark? Would they be recognizable?GeeBIGS (talk) 05:50, 27 July 2012 (UTC)

Physicist George Gamow wrote a series of books that elucidate the recognizable and unrecognizable effects of physics, using unlikely hypothetical scenarios where relativistic and quantum-mechanical effects would have observable and macroscopic effects. If I can nitpick the books - the consequences of changing the speed of light or the Planck constant were not followed through to their logical conclusions; in the story-books, these effects were isolated only to certain observed phenomena, as opposed to affecting everything - but altogether, the stories help you stretch your mind a little bit, as far as thinking about fundamental interactions. Nimur (talk) 06:03, 27 July 2012 (UTC)

You can look right now at the various formulations of existing theories, and they can be made to look completely different. Take for example Hamiltonian mechanics, Lagrangian mechanics, and classical mechanics. Or for that matter, the differential and integral forms of Maxwell's equations (there are also Hamiltonian Maxwell's equations, but I can't find an article on them). In the case of the first three and the second two, you have the exact same laws written from different perspectives. Without the necessary mathematical training, they look like completely different physics. But what you will find, however, is that while each form makes different problems easier to solve, they make identical predictions. And so while you could reformulate the Standard model of physics from "the perspective of a single quark" (not entirely sure what that means), if you do it correctly it will make the exact same predictions as the normal formulation, even if the equations are unrecognizable. Someguy1221 (talk) 06:21, 27 July 2012 (UTC)

Also to continue the previous thread.... Just substitute a very useful and recent discovery with a very useful yet old discovery and see what would and would not have been possible.GeeBIGS (talk) 06:30, 27 July 2012 (UTC)

You can change the perspective of practically every equation and end up with the same results. By itself that isn't very philosophically interesting — there are lots of cases where two different approaches to physics give exactly the same outcome, because ultimately they are both different ways of describing the same thing. (The Nobel Prize in Physics for 1965 was to three people who came up with three different ways of describing the same thing, which all turned out to be physically equivalent. Maxwell's Equations, as written up by Heaviside, look quite different than when they were written originally by Maxwell, though they give the same physical results. There are many such examples in the history of physics.) An obvious, straightforward approach would be the differences in describing orbits using traditional Cartesian coordinates or by using orbital elements; two such descriptions of the Moon's orbit would look superficially quite different but, when you know what everything means, clearly describe the same thing. The only interesting bit to this, in my view, is that it highlights that the math itself is hardly canonical — at best, it is just a form of formal description for physical reality, but there are many ways in which even the math can be differently expressed. --Mr.98 (talk) 13:48, 27 July 2012 (UTC)

Spider identification needed

I'm in the UK, just found a spider which I've never seen before. It's about 13mm leg span, body about 6mm long. Colour is white-very pale green with a red patch on its back. Unfortunately my camera isn't up to macro work otherwise I'd have taken a picture. Any ideas as to which spider this is please? Mjroots (talk) 07:13, 27 July 2012 (UTC)

Was it in a web? 203.27.72.5 (talk) 07:43, 27 July 2012 (UTC)

Based on your description, I would suggest that it could be either a Tibellus or a a recently moulted Diaea dorsata. 203.27.72.5 (talk) 07:51, 27 July 2012 (UTC)

Definitely not a tibellus, they aren't native to UK. Diaea dorsata is a distinct possibility though. Mjroots (talk) 09:51, 27 July 2012 (UTC)

According to [2] and [3] they are in the UK. I don't know if they're native or introduced though. 1.124.170.168 (talk) 10:14, 27 July 2012 (UTC)

Other possibilities:

• Flower spider (Misumena vatia) - females can change colors from bright yellow to pale green, and they have red stripes at the sides (sometimes absent). Note that it's a crab spider like Diaea dorsata.
• Candystripe spider (Enoplognatha ovata) - females about the same size, translucent green or white, usually with two red stripes on the abdomen
• Green huntsman spider (Micrommata virescens) - males are green with red/white abdomens
• Cucumber spiders (Araniella cucurbitina, Araniella opisthographa, etc.) - both very similar, body usually pale green, dorsal surface of abdomen has pale stripes making it look like a cucumber (sometimes absent), identifiable by red spot on the spinnerets

If it's not the above. Did it resemble any spider you do know, e.g. orbweavers, wolf spiders, jumping spiders, crab spiders, lynx spiders? -- OBSIDIAN†SOUL 17:39, 27 July 2012 (UTC)

Latent heat

I read the statement "Latent heat of vapourisation of water has higher value than latent heat of fusion of ice". Why is it so? Sunny Singh (DAV) (talk) 15:01, 27 July 2012 (UTC)

This link should explain it for you. Tombo7791 (talk) 15:23, 27 July 2012 (UTC)

It seems intuitive that boiling water, which requires breaking surface tension, would take more heat input than just melting it. μηδείς (talk) 22:24, 27 July 2012 (UTC)

Is the volume of solid ice at 0°C greater than liquid water at 0°C? If so that's negative work being done which will further decrease the energy requirement. Obviously, water at 100°C has a much smaller volume than steam at 100°C, that accounts for a large proportion of the latent heat of vapourisation. 203.27.72.5 (talk) 23:22, 27 July 2012 (UTC)

We should be able to do better than the linked article provided by Tombo. It has 2 answers: the first does not actually answer the question - it essentially just says the vaporisation heat is greater because it's greater. The 2nd part seems to answer it, but not very clearly. I think the author is implying that to go from solid to liquid, inter-molecular bonds are reduced, but to get to a gas, all bonds are subtantially eliminated.

It should be noted that while latent heat of fusion is in practical terms independent of pressure, latent heat of vaporisation varies with pressure, being maximum (= 42.8 MJ/kmol, about 6.5 times the latent heat of fusion) at the triple point pressure (0.617 kPa), and zero at the critical point (22064 MPa). This reduction to zero latent heat at the critical point would be reasonable to expect, if, as pressure rises, the vaporisation/boiling temperature increases - and at the critcal point temperature (647 K), heat absorbed by thermal capacity was equal to the latent heat of vaporisation at the critical point. Actually, its a good bit less, 36.2 MJ/kmol. However, fundamental to the OP's question is why the latent heat of vaporisation at the triple point is the comparitively large positive value that it is. So why is it? Surface tension seems to be a key, as surface tension decreases as temperature increases (does it smoothly become zero at the critical point temperature?). Volume occupied seems not involved, as specific volume at the critical point is most certainly NOT infinite. Ratbone58.169.250.134 (talk) 05:04, 28 July 2012 (UTC)

I'm not sure what you mean about the specific volume not being infinite at the critical point. I'm not great at thermodynamics, but maybe you could explain it in terms I understand. It seems to me that when you take 0.018L of water and turn it into 22.4L of steam you've done some work to push back the atmosphere. When you take 0.018L of ice and turn it into ~0.018L of water, the work done is either very low, zero, or slightly negative. So doesn't that account for some of the difference between the heat of fusion and the heat of vapourisation? 203.27.72.5 (talk) 06:05, 28 July 2012 (UTC)

As every expert, teacher, and self-proclaimed whizz has said as least once when caught out with telling a complete nonsense, I was just testing you.... Of course, specific volume is not the key, but the change in specific volumes upon vaporisation is a key. At the tripple point, the change in specific volume is maximum, smoothly decreasing to zero change in specific volume at the critical point. It's the change in specific volume that means work done on the surroundings, as W203.27.72.5 has said. But, as he indicated, that only accounts for part of the change in heat of vaporisation from the triple point to the critical point, as a glance at a temperature vs internal energy loop chart will show. Ratbone120.145.22.128 (talk) 10:45, 28 July 2012 (UTC)

Even after this discussion I don't get a simple answer of my question. Will you please make it more clear? Sunny Singh (DAV) (talk) 13:18, 30 July 2012 (UTC)

Partly because for vapourisation you have to overcome the surface tension of water for the molecules to escape as a gas whereas for melting you don't, and; partly because for vapourisation you have to do significant work pushing back the atmosphere in order to increase the volume of water to that of steam whereas for melting the volume is about the same. 203.27.72.5 (talk) 20:27, 30 July 2012 (UTC)

Calculate properties of chemical compounds given the properties of chemical elements

Given some chemical elements, why do you have to perform a real life experiment to discover the properties of a derived chemical compound? Couldn't you just calculate (almost) everything in a kind of chemical CAD on steroids? Where does the unpredictability comes from? OsmanRF34 (talk) 16:02, 27 July 2012 (UTC)

It depends on the elements, number of atoms, and their arrangement in the molecule (i.e., "what specific derived chemical compound?") and the properties (and accuracy and precision of them, i.e., "what do you want to know about it?") of interest. Some are easily doable, some are easy to approximate, some are possibly doable but have many known exceptions/mismatches vs reality especially depending on computational power, some are fairly hopeless to get anything much better than a wild guess or mental approximation based on other knowns. DMacks (talk) 16:09, 27 July 2012 (UTC)

But, is our knowledge about properties of elements accurate and definitive? Or is it something like decimal places of Pi, which can be always expanded? OsmanRF34 (talk) 17:38, 27 July 2012 (UTC)

Again, it depends on the elements and properties of interest. DMacks (talk) 18:38, 27 July 2012 (UTC)

I suspect that even if we did have a formula or a simulation to predict the melting point of a novel compound, the requisite computing power to carry out the simulation would be greater than all such power available on the Earth. Someguy1221 (talk) 19:08, 27 July 2012 (UTC)

See protein folding and protein structure prediction. μηδείς (talk) 21:48, 27 July 2012 (UTC)

In principle it all follows from quantum mechanics. The bound state of atoms forming a particular molecule is a particular eigenvector of the the Hamiltonian, the eigenvalue is the energy of the bound state. The properties referring to the dynamics of the molecule follow from the enitre spectrum of eigenvectors and eigenvalues. In practice when doing such first principle calculations, elaborate approximation techniques have to be used, such as the Density functional theory. Count Iblis (talk) 22:09, 27 July 2012 (UTC)

And to follow on from Count Iblis, since you can't exactly solve the quantum mechanical Schroedinger equation for a two electron system, you usually can't make accurate predictions of the properties of compunds. Just doing the experiment isn't usually that hard anyway. You may be interested in computational chemistry. 203.27.72.5 (talk) 23:28, 27 July 2012 (UTC)

Indeed. This video illustrates how small perturbations in outer electron bands can have profound structural implications, e.g. on volume and density. It's absurdly difficult to try to derive melting and boiling points or even density from simulations for this reason. However, with Monte Carlo simulations of the Schroedinger equations it is certainly possible with sufficient computing power, which often isn't too much to be practical for reasonable amounts of accuracy. 207.224.43.139 (talk) 21:36, 28 July 2012 (UTC)

Rectifier

• If I install two rectifiers in parallel configuration, would that double the possible draw of amps? Since rectifiers with lower amps rating are much cheaper I'd rather install two small ones if possible. Thanks,TMCk (talk) 17:23, 27 July 2012 (UTC)

• Add a question: this is the rectifier I intent to use. If I solder the connections, do I have to be overly careful so that the heat doesn't destroy it? It has male terminals (will look up the kind and post shortly Done and according to WP they're called blade connectors) but I need to save as much space as possible. That's why I'd like to solder the connections. Again, thanks for any helpful response.TMCk (talk) 00:14, 28 July 2012 (UTC)

• Add yet another question: Rectifier is rated for 50 V but I'll only draw about 20 V from it. Is there a problem I should know about?TMCk (talk) 00:18, 28 July 2012 (UTC)

Two bridge rectifiers in parallel can carry twice the current of one, as long as they're not so close that they interfere with each other's cooling. Running at a lower voltage won't be a problem. --Carnildo (talk) 00:44, 28 July 2012 (UTC)

It is quite foreseeable that when two rectifiers are connected in parallel, one might have a smaller voltage drop/lower resistance and "hog the load," such that it overheated and burned out, leaving the surviving rectifier to carry all the load, whereupon it also burned out. If you're very lucky, they would divide the load equally and all would be sunshine and roses. Edison (talk) 02:52, 28 July 2012 (UTC)

Ah, thanks, to both of you. @Edison: since one could carry the constant current and the other would be mostly to make sure they both together can handle the starting current I guess I have a good chance this configuration with the proper cooling of course would handle what I have in mind with little to no danger for breakdown. Is this a fair conclusion?TMCk (talk) 03:45, 28 July 2012 (UTC)

You have a better chance of survival. You can purchase matched rectifier diodes, which should have similar characteristics and be less prone to the problem. Graeme Bartlett (talk) 08:16, 28 July 2012 (UTC)

I bought an identical pair so they should be very close in their actual characteristics. When I'll get on with the project I'll give it a try :) TMCk (talk) 16:24, 28 July 2012 (UTC)

This can depend on the characteristics of the other elements of the circuit. Answering correctly is a job for electronic circuit simulation, e.g. with SPICE. 207.224.43.139 (talk) 21:40, 28 July 2012 (UTC)

That would be above my paygrade ;) The other elements of the circuit are simply an altered transformer and some cordless tools. The transformer will get a cooling fan and maybe the former battery pack where I'll place the rectifiers will get one two. I like it save and fancy :) TMCk (talk) 01:46, 29 July 2012 (UTC)
I know something about electric but when it comes to electronics I'm pretty much lost with my very limited knowledge about it. I know how a transistor works but don't ask me how to use it in a circuit.TMCk (talk) 01:52, 29 July 2012 (UTC)

There are some fantastic beginner simulators these days, with white-box block libraries for tens of thousands of commercial components including common power tool motors. Try https://www.circuitlab.com/ and http://ngspice.sourceforge.net/ before spending money on something like http://www.simetrix.co.uk/ (which has a free demo that can probably handle your circuit anyway, so maybe start with that.) 207.224.43.139 (talk) 05:53, 29 July 2012 (UTC)

As I said, that goes beyond my needs and expertise and needs but thanks, I'll keep it in mind in the future for more delicate projects.TMCk (talk) 23:34, 29 July 2012 (UTC)

Camel urination and defecation

Camels have far less water in both, so how does this work ? Specifically:

1) How do they avoid kidney stones ?

2) How can they defecate with dry stool ? In humans this causes constipation and impaction. StuRat (talk) 18:00, 27 July 2012 (UTC)

Without knowing anything specific about camels per se, it should be noted that the digestive and excretory systems of mammals shows a wide enough variation in other animals that the answer is that camels don't work like humans. Cows and other ruminants have an organ called the Rumen, which doesn't exist in humans, that produces the cud they chew. There is no reasonable analogue for that process or that organ in humans. Rabbits don't chew cud, but they do have two types of droppings, one of which the re-eat, again something humans don't often do (at least, unless there are two girls with one cup). Camels can defecate with dry stool because they have evolved a digestive canal that allows them to defecate with dry stool. Likewise, their kidneys likely work differently from humans. Humans do their own weird things. For one thing, we sweat; something many mammals don't do. The simple answer is that camels are not humans, so don't look for them to work exactly as you do. It must work for them, because they exist and thrive. If it didn't work well enough, they wouldn't and that'd be the end of that. --Jayron32 03:57, 28 July 2012 (UTC)

Feces is (are?) excreted covered in mucus. Mucus is low in water content--think of slug trails and the fact that they don't dessicate because of them. They probably have large turds, because the larger the less relative surface area to lubricate. μηδείς (talk) 05:24, 28 July 2012 (UTC)

Jayron, do you have a citation that many mammals don't sweat? 203.27.72.5 (talk) 05:49, 28 July 2012 (UTC)

See perspiration. Many mammals sweat to some extent but very few use sweating for thermoregulation. Gandalf61 (talk) 08:14, 28 July 2012 (UTC)

Thanks, I had already read that though. I'm interested in the many mammals that don't sweat at all, regardless of whether it's for thermoregulation or some other purpose. 203.27.72.5 (talk) 09:10, 28 July 2012 (UTC)

It's likely that water is absorbed through the camel bladder, concentrating the urine and recycling water. Similarly, many ruminants produce hard pellet feces, from which water can be absorbed by the intestines before defecation more readily than from smooth primate feces. 207.224.43.139 (talk) 21:43, 28 July 2012 (UTC)

Thanks. Can you verify that a camel's bladder can absorb water, unlike a human bladder ? Also, wouldn't this just lead to bladder stones rather than kidney stones ? And how are hard pellets moved through the digestive tract ? StuRat (talk) 21:46, 28 July 2012 (UTC)

[4] says that urine production is reduced in a dehydrated camel, but with no indication of whether that's a function of the kidneys or the bladder. "Camel kidneys have long loops of Henle and can manage salts. Camel can take sea water without any side effects and can excrete sea water with a salt concentration almost double that of sea water," suggests that it might be more of an adapted kidney function. Pellets don't constipate like dehydrated smooth feces with ordinary intestinal contractions. Think peanuts versus peanut butter. 207.224.43.139 (talk) 22:05, 28 July 2012 (UTC)

Camels have the reputation of being fractious beasts, prone to kicking, biting, and spitting, for reasons only understood by the individual camel. Imagine the veternary assistant assigned to don a rubber glove and manually "deimpact" the colon of a constipated camel. The scenario makes many marginal jobs seem idyllic. Edison (talk) 04:51, 31 July 2012 (UTC)

Densest substance at standard conditions?

It's known that osmium is the densest known (naturally occurring) element, but what is the densest known substance at standard conditions? A discussion at Talk:Osmium#Densest_element_or_densest_substance.3F did not bring definite results. This (reliable?) source claims that an Ir-Os alloy is slightly denser than pure osmium, but is it actually the densest substance known? Is there an upper limit for how dense a substance could be at standard conditions? --Roentgenium111 (talk) 22:59, 27 July 2012 (UTC)

It is possible that strangelets can exist under standard conditions (although there is no evidence either way, as far as I know). If they can, then they would be extremely dense - orders of magnitude denser that anything made of atoms. --Tango (talk) 23:25, 27 July 2012 (UTC)

Hassium is probably far denser than osmium. 203.27.72.5 (talk) 00:18, 28 July 2012 (UTC)

Hs could have a density of 41 g/cm³, almost twice that of Os. Double sharp (talk) 13:42, 29 July 2012 (UTC)

Thanks for the answers, but I was thinking of a "normal" substance, i.e. made up of atoms of the naturally occurring chemical elements (Z<99). Any ideas there? User: Stonemason89 suggested doting osmium with hydrogen atoms, this might also be denser than pure osmium (see below). --Roentgenium111 (talk) 15:09, 29 July 2012 (UTC)

To have a higher density than osmium, a material could have either a higher average atomic mass or denser packing or a smaller average atomic radius, or some combination of the three. The atomic mass is dependent on the nuclear characteristics, and the packing and atomic radius is dependent on the electronic characteristics. Obviously, plently of materials have higher atomic masses than osmium, but their packing is so much less efficient and/or their atomic radii are so much larger that they're less dense over all. Osmium, and the other elements in the platinum family, happen to occur in a region of the periodic table where the atomic radius is at a minimum, the crystal structure is very efficiently packed, and they're in a period far enough down that they have high atomic masses, but not so far that they're unstable (as in hassium). Any compounds made with elements from outside the platinum group would have to significantly increase the average atomic mass without offsetting that increase by changing the packing struture or average atomic radius too much. I don't think any of the heavier elements are suitable, because they all have much longer radii, and may not want to sit in the same packing structure as osmium. Just having atoms of a different size in the same lattice causes interstitial defects which increases the void between the packed atoms thereby decreasing your packing efficiency. Maybe you can slightly increase the density by making an alloy of different platinum group metals (as indicated by your source), but I don't really see what mechanism is at work there. 203.27.72.5 (talk) 01:08, 28 July 2012 (UTC)

Thanks for that great overview on the problem. According to Atomic_radius#Empirically_measured_atomic_radii, at least fluorine and deuterium seem to have substantially larger "atomic density" (:= atomic mass/volume of the atom) than osmium, but of course they're gases at STP. (But OsF_x could be denser than Os, if it can be packed as closely; but osmium hexafluoride is in fact much less dense.) All the elements with higher atomic mass given there have lower atomic densities.

Do you know which exact crystal structure osmium has? Our article says only "hexagonal", but is it the hexagonal close-packed or something less closely packed? In the first case I would also doubt that IrOs can be denser, but in the second case the alloy might form a denser crystal structure than that of pure osmium. And I would expect that interstitials could in principle also increase the density of a material, if you fill the "empty spaces" left between the closely packed osmium atoms with atoms small enough to completely fit into such a hole (hydrogen or helium should be small enough for this). --Roentgenium111 (talk) 15:57, 29 July 2012 (UTC)

Osmium hexafluoride has a much lower average atomic mass than pure osmium (~43Da vs 190.2Da), and it's crystal structure is orthorhombic so the packing is much less efficient. That accounts for the very low density relative to osmium metal. The structure of osmium is very close to a perfect close packed hexagonal configuration. I calculated the density of osmium from the atomic mass (190.2Da) and radius (135pm) assuming a hcp structure and actually got 21.39g/cm³ which is less than the empirically measured value of 22.59g/cm³. I found that very small changes to the atomic radius changed the result wildly, with 132.6pm giving the accepted value for the density. The data page for atomic radii gives 130pm for the empirical value and 180pm for the calculated, but 135pm for the metallic element. In any case, with such great variation I think it's fair to say that 132.6pm is a fair fudge and the structure is hcp. Both iridium and osmium already have the most efficient packing possible (or very very close to it), so I can't see how Os-Ir alloy could have a denser crystal structure. I think the answer most likely has to do with electronic interactions in certain Os-Ir alloys reducing the effective atomic radius and thereby leading to a higher density. The behaviour of electrons in high atomic mass elements is extemely complicated due to the relativistic speeds of the electrons. You're quite right about the very small interstitial particles like Hydrogen. You could perhaps boost the density a bit by saturating the metal with hydrogen. 203.27.72.5 (talk) 02:03, 30 July 2012 (UTC)