6

As I known all the matter consists of atoms. What is the simplest possible experiment (with as little scientific equipment as possible) that can prove that there are atoms?

HDE 226868
  • 10,821
bessarabov
  • 161
  • 2
    It's not actually true that all matter consists of atoms. It's believed that perhaps 75% of the mass of the universe is "dark matter" whose nature is unknown but is certainly not atoms. Neutron stars are made of neutronium, which does not consist of atoms (unless you're going to claim that a neutron star is a single colossal atom). And so on. All the matter we're familiar with on Earth is made of atoms, but that's by no means all matter. – Mike Scott Aug 03 '15 at 19:49
  • 1
    @NeuroFuzzy - I disagree - this question is asking about the simplest experiment for a demonstration, whereas the linked question is about experiments in geneeral e.g. mass spectrometer etc., but maybe I am wrong – tom Aug 03 '15 at 21:02
  • After some thought, how would you define an atom? Would an experiment have to prove its structure? – HDE 226868 Aug 03 '15 at 23:06

5 Answers5

13

Brownian motion would be considered pretty good evidence of atoms. Originally, Brown observed pollen grains moving in water in ways that could not be explained otherwise. 78 years later, Einstein came up with a theoretical model of Brownian motion1 that was later borne out by experiments by Jean Perrin (see this). A translation of Einstein's paper is available here.2

You could reproduce it easily with other small particles suspended in a different liquid (though water might be the best). You would need to make sure that the liquid is not disturbed by outside forces, and you would of course want to try other substances to show that it is a general phenomenon.

Note: It looks like this was explained better here, which is absolutely worth reading.

1 Marian Smoluchowski also did work on Brownian motion.

2 A note on Brownian motion: DanielSank pointed out in a comment that the Langevin equation, one of the most important parts of the mathematical model behind Brownian motion, does not necessarily imply the existence of atoms or any particles. However, Einstein's use of molecular-kinetic theory does imply the existence of atoms.

Community
  • 1
HDE 226868
  • 10,821
  • 1
    So Brownian motion of stock prices implies there is a smallest unit of money? – arivero Aug 03 '15 at 22:28
  • @arivero It's more that the same mathematical framework behind Brownian motion can be applied to the stock market, because of the theory of random walks. – HDE 226868 Aug 03 '15 at 22:32
  • Hmm I answer myself: of course Brownian motion of stock prices implies the existence of individuals, exactly in the same sense that Brownian motion of pollen coordinates implies the existence of atoms. It is actually the same word, (a=in, toms==dividuals). – arivero Aug 03 '15 at 22:42
  • 2
    @arivero Brownian motion doesn't prove anything about the discretized nature of the system. If I write $\ddot{x}(t) + \gamma \dot{x}(t) = \zeta(t)$ (i.e. Langevin equation) I recover the Brownian motion process, yet there's absolutely no sign of anything discrete. It's only in connecting $\gamma$ and $\zeta$ to the kinetic theory that gives evidence of atoms. I hope HDE will note this in the answer. – DanielSank Aug 03 '15 at 22:45
  • @DanielSank note my other comment, individuals=atoms: there I agree that we do not want to prove the discretized nature of configuration space -polen positions $x(t)$ or stock prices $S(t)$ -but the existence of discrete actions, or discrete actors... here is where the kinetic theory takes command indeed – arivero Aug 03 '15 at 22:56
  • 1
    @arivero ok so aren't all these answers about Brownian motion kind of bogus? I understand that historically Brownian motion was a huge step toward accepting the atomic theory, but it seems that this was mostly because of all the kinetic theory behind it and not he Brownian motion itself. That said, I don't think calling out "Brownian motion" is a very good answer to this SE question. – DanielSank Aug 03 '15 at 23:02
  • @DanielSank I noted your previous comment. I do think that the observations themselves were quite important in showing that there must be small particulate constituents in matter. – HDE 226868 Aug 03 '15 at 23:03
  • If we are into kinetics probably we could try other statistical mechanics results. The log variation of pressure with height, for instance. We should ask elsewhere is Brownian motion requires discrete actors or if it just happens that the main examples have this property. – arivero Aug 04 '15 at 00:32
  • 2
    Even putting aside the fact that, as @DanielSank says, Brownian motion proves nothing about discreteness, isn't Brownian motion at best evidence of molecules, not atoms? – WillO Aug 04 '15 at 01:00
  • @WillO I can't imagine how the distinction between molecules and atoms could come into it. The issue, as I see it, is that Brownian motion alone needs only a stochastic force and friction to happen. The interesting thing about Einstein and Smoluchowski's work is that it connected Brownian motion to kinetic theory of little particles. – DanielSank Aug 04 '15 at 02:31
  • 1
    @DanielSank: my point exactly. One could certainly believe in the little particles of E/S and still not believe in atoms. – WillO Aug 04 '15 at 03:48
  • @DanielSank on other hand, chemistry and stochiometry can distinguish between atoms and molecules. – arivero Aug 08 '15 at 23:56
7

To show kids atoms in action I use Brownian Motion similar to what Robert Brown did with his pollen grains. You need a microscope with at least X400 magnification, some whole milk, a glass slide and cover slip. Place a drop of milk on the slide and a drop of water. Cover the milk-water with a cover slip and view under the microscope. For a whole class it's good to have a camera and projector on the microscope.

What you will see are small fat globules bouncing around. The motion is random walk caused by random bombardment of the water molecules against the globules. The dynamics of this motion were predicted by Einstein in 1905 as a function of viscosity, particle size and temperature. Given those known quantities, the dynamic model and the mean size of the random walk gives an estimate of the size of the atoms (or molecules in this case).

Einstein's work was a turning point in the belief that atoms actually exist. Before that time scientists were divided.

docscience
  • 11,663
  • 2
  • 32
  • 69
  • 1
    As I said in my comment to the OP: This seems to be at best evidence of molecules, not of atoms. – WillO Aug 04 '15 at 01:00
  • I agree with @ WillO because students tend to be very literal and might assume from this milk demonstration that the globules are atoms or even molecules, even if it's explained otherwise. – suse Dec 19 '23 at 04:57
6

Chemistry. The things we call "atoms" are chemical objects, different of the atoms of the ancients (closer to "elementary particles"). The existence of chemical atoms implies integer proportions in chemical reactions.

I'd vote for electrolysis of water. The volume of $H_2$ and $O_2$ increases in integer proportion 2:1.

EDIT: from discussion in the comments, it seems that a single proportion is not enough to settle the question, in this sense stoichiometry is not the "simplest" possible, as a whole corpus of proportions is needed to show that some basic multiple is going on. EDIT2: note also that weighting the products of electrolysis you could have a first hint of nucleon mass. Again, lot of measures needed to be sure, isotopes and all that.

arivero
  • 1,943
  • 1
    If you want a "pure physics" experiment you need to go to crystallography and scattering. Not so cheap. – arivero Aug 03 '15 at 19:18
  • 7
    Good experiment! However, I don't think chemistry or electrolysis proves that matter comes in discrete bits. – DanielSank Aug 03 '15 at 19:49
  • Stochiometry was historically one of two pillars of argument (along with the Brownian motion calculation/observation agreement) that together brought people around. Crystalography came decades after the atomic hypothesis was accepted as fact. – dmckee --- ex-moderator kitten Aug 03 '15 at 21:47
  • @DanielSank why? – arivero Aug 03 '15 at 22:25
  • It's suggestive, but I don't see how electrolysis rules out a theory wherein infinitesimal quantities of oxygen and hydrogen "fluid" combine to make water. Kinetic theory seems like the real key here. – DanielSank Aug 03 '15 at 22:32
  • @DanielSank Well, becouse such combination theory does not explain the integer proportions. In the same spirit, would you say that quantisation of atomic mass did not imply the existence of protons and neutrons? – arivero Aug 03 '15 at 22:40
  • Oh no that's different. Quantized mass is a pretty clear signal that the parts are discritized. Discrete ratios don't seem the same to discrete masses. – DanielSank Aug 03 '15 at 22:41
  • Note that the mass quantization of atoms also follows from quotients. I think that it could be possible to weight the gas out from electrolysis and obtain the mass proportions. – arivero Aug 04 '15 at 00:25
  • 1
    Arivero, what you propose here is essentially to find the law of definite proportion. You also want the laws of multiple and reciprocal proportions to complete the stochiometric picture. This was done by the mid 1800s, but it wasn't considered completely convincing at the time for exactly the reasons the @DanielSank points out. It's finished when combined with evidence of discrete little mass lumps existing below the microscopic scale. – dmckee --- ex-moderator kitten Aug 04 '15 at 01:47
  • @dmckee what I am surprised is that Brownian is considered a more important leg that stochiometry. As WIIIO points out, it can no go beyond molecules. I wonder if the question fares differently in chemistry.stackexchange.com. – arivero Aug 08 '15 at 23:57
3

Oil drop on liquid is about the simplest I think, - take a bowl of still water - dust it with some fine powder - put a small drop of oil of top - The oil spreads out over the surface and generally forms a circle the size of which you can see from the dusting. The thickness of the drop on the liquid can be calculated if you can estimate the volume of your initial oil drop. The thickness should come out at about 1 Anstrom - roughly 1 atom thick.

I saw this demonstrated in School (US High School) as a demonstration of atomic nature of matter.

tom
  • 6,948
  • This gives you data about the size of the molecules assuming that they are there. It isn't clear evidence that matter comes in that form in the first place. – dmckee --- ex-moderator kitten Aug 03 '15 at 21:41
  • @dmckee - I agree, I suggest this experiment mostly because it is so simple and easy and yet correctly gives an estimate about atomic dimensions - in effect the experiemnt shows that matter is not continuous, but at the smallest scale atomic, or perhaps rather molecular. – tom Aug 04 '15 at 19:57
0

About the question ,(case 1)Is it that You actually believe that matter is made of atoms in theory and just want to cross check with a simple practical experiment? or (case 2) You actually doubt the existence of atoms itself and want to check it out with an experimental proof.

(case 1) Take a flask/container with an attached barometer and fill it with a mole(2 grams) of hydrogen gas . Let the temperature maintained be T. Let the pressure reading by the barometer be P. Now at the same temperature T remove the hydrogen gas completely and fill it with a mole of nitrogen gas(28 grams) and you find the pressure reading P. Hence the number of molecules in 2 grams of hydrogen gas is same as that in 28 grams of nitrogen.

Because according the ideal gas law PV=nRT where P is the Pressure , V is the Volume ,T is the Temperature , R is the gas constant and n is the number of molecules, since P,V and T were same n is also same. Theoretically the mass of hydrogen molecule is 2 atomic mass unit(amu) or 2*1.6605*(10^-24) grams. Then how many number of molecules constitute 2 grams of hydrogen gas each with a mass of 2 amu. Just by doing math we get 6.023*(10^23)(avagadro number) molecules count. Now same number of molecules of nitrogen are present corresponding to 28 grams which would give the molecular mass of 28 amu for nitrogen molecule which is theoretically correct.

(case 2) In case you doubt the existence of atom itself then you could consider Rutherford’s gold foil experiment where a thin gold foil is bombarded with a beam of alpha radiation. With out going to the deeper details of the experiment the scattering of some alpha rays in different directions and some being able to pass right through, gives an abstract view that matter consists of smaller constituents. Of course the experiment was actually conducted to know the structure of atom and that existence of nuclei was an outcome of the experiment but the scattering of alpha rays just pushes us to the thought that the matter is made of some smaller constituents. The Brownian motion also gives a notion of atoms and molecules.

Yash
  • 3