Talk:Molecular diffusion

Definition
Diffusion is not limited to molecules, from a general (mathematical formulation) point-of-view a lot of processes are diffusive. What diffusion is or not is, is best to discuss by means of maths because that is the most accurate and objective way to write down things with. Things like "diffusion is the movement of molecules from high to low chemical potential/concentration, sometimes against the concentration gradient!" I would like to see in it's own section, i.e. not stated as general fact in the first few lines of this article.145.94.65.212 (talk) 07:08, 1 October 2008 (UTC)


 * Diffusion "in general" and the mathematical formulation already have their own articles. Other "diffusive processes" also have their respective articles, e.g., Momentum diffusion. I feel this article should contain itself to a physical phenomenon of molecular diffusion of mass, that does exist, and try to explain it in an accessible language.


 * Alternatively, I think a new article should be considered, like Molecular diffusion of mass. Then this article (Diffusion) changed to represent what it is actually about because right now it does not seem clear at all. Stan J. Klimas (talk) 14:22, 13 October 2008 (UTC)


 * I would like to add that Fourier's brilliant mathematics did not lead him to understand diffusion first, though Fourier was clearly one of the first to describe heat transfer from a mathematical perspective. But he was describing conduction.  It was all Fick, who saw the relevance of the mathematics describing heat and momentum transfer in describing mass transfer.  The thermal conductivity in Fourier's equations of transport can simply be replaced with a diffusion coefficient to describe mass transfer, I know this is a small thing, but I think Adolf Fick was a genius in his work, and I think the discovery of diffusion should be placed on him soley.  Sure on a molecular level, the stretching of bonds is relevant for heat transfer, but I see this as thermodynamics...not mass transfer.Biotechscientist (talk) 03:04, 19 November 2010 (UTC)

Mathematics
It is to specialistic for the general reader looking up diffusion.145.116.1.3 (talk) 07:45, 5 June 2008 (UTC)

Exactly, there seems to be a problem with the management of this page, I tried to put in a general definition in the introduction (see revision of early september) and it disappears; in its place childish statements like "Einstein was clever".

Also there is a problem of definition. I come from a physical chemistry research background where diffusion is only used in the sense of the movement of molecules due to their thermal energy within chemical systems. The transfer of heat may follow the same maths but is not diffusion (to me). I am willing to have a physicist clarify this, but there still needs to be a separate page. The example of the spoon getting hot is confusing and I believe wrong. Manikato (talk) 03:22, 30 September 2008 (UTC)

I would like to keep the page on diffusion general, to fit all the different (natural) processes that show the same behaviour. To me (physicist), diffusion is not restricted to particles/molecules alone, since there can be diffusion of heat, or diffusion of wave function in quantum mechanics (for example). 145.94.65.212 (talk) 19:23, 1 October 2008 (UTC)

General discussion
Diffusion is the movement from a high to low concentration along the concentration gradient. In ternary solutions particles can be tranported from low to high concentrations. Joao

This can happen in binary alloys, too Olof
 * Agreed. Molecular diffusion occurs in solids as well as liquids and gases; even pure, unalloyed metals exhibit molecular diffusion. 99.133.144.210 (talk) 01:56, 29 July 2011 (UTC)

Diffusion as noted in the last paragraph generally implies a dilution or mixing of the medium. Hence I'm not totally happy by a contrast with radiation as this is itself a diffusion mechanism. Indeed the reference to heat through a window is indeed diffusion by radiation. Rjstott

Heat through a window was a bad example. Maybe heat through a window at nighttime was what I was thinking. Perhaps heat through an oven door is a better example. However, I think its correct to say radiation and diffusion are completely different transport phenomena: in heat transport by radiation, photons get emitted by thermal activation and carry the energy until the photons are absorbed somewhere else. heat transport by diffusion is governed by the thermal conductivity of the material and the temperature gradient. Olof

Way back in school, there were three heat transport mechanisms, conduction, convection and radiation. I believe that heat can be diffused by any of these three. Light is an electromagnetic radiation and diffusion is often associated with light too. Perhaps an example based on light diffusion might work better? Rjstott

You are right about the three mechanisms: Radiation is when photons carry the energy from one place to another. Heat disappears from a material when the photon gets created, and reappears at the place where the photon gets absorbed. The key part there is that the material needs to be hot enough so that it can thermally activate a photon creation, as when a metal is 'red hot'. Of course, at lower temperatures, the photons are lower energy and there is some radiative heat transport. Convection is when, for example, the hot air next to the oven rises to the ceiling, and the air that used to be near the ceiling moves to.... and so on. It is only applicable to fluids. Conduction is when heat diffuses through a material, and it is the only one of the three mechanisms which is strictly a diffusion. By this I mean that the mathematical equation which is used to model the heat flow is the diffusion equation. Olof

I was also thinking of adding a section on computer graphics diffusion: the process whereby an image is blurred by an image editing program also follows the same math. Olof

I think the 'in biology' section is completely wrong. Amino acids cannot diffuse across a cell membrane, they must be transported across by specific sodium-dependent transport proteins. Moreover, the citation is not very good - many introductory textbooks are notoriously wrong. —Preceding unsigned comment added by 128.61.94.19 (talk) 02:24, 24 October 2008 (UTC)

Disambiguation
I hate it when one page is used for two topics. Something needs to be done! Either create a page for Diffusion (physics) and Diffusion (anthropology) or kill the anthropology on the diffusion page.

Osmosis= diffusion of water not general diffusion surely. Equally my GCSE biology teacher could have been lying to me...


 * I agree the osmosis reference should be removed, and I did so. Danrs 18:45, 15 August 2007 (UTC)

Types of diffusion in separate pages
I just made a major change to the layout of the page. Since there were alot different types (and sub-types) of diffusion listed inthe article, I grouped them all in one section and moved the relevant information to separate pages on the particular types of diffusion. Hope that made the page a bit clearer than before. Cheers Karol 18:01, 15 September 2005 (UTC)

Diffusion and Turbulence
I think diffusion is not connected to turbulence, the example of smoke dissipating into the air is a bad example since it involves convection. I have removed the parts accordingly. Lirnup 20:28, 28 September 2005 (UTC)
 * Good work catching that. Also, diffusion isn't quite the same thing as Brownian motion, no? Karol 06:06, 29 September 2005 (UTC)
 * As I understand it, in the case of diffusion the particles move one by one by brownian motion, but I'm not really an expert on that. Lirnup 06:21, 29 September 2005 (UTC)
 * Sure, in the case of the diffusion of particles, but that is only one type of diffusion since not only particles diffuse. There is also the subtle difference that BM is a description of the stochastic motion of single particles in itself, whereas diffusion is more a dsecription of a process, in which there is a net transport and flux of the diffusing quantity (related to some kind of gradient). Karol 14:37, 29 September 2005 (UTC)
 * Yes, the brownian motion thing came in only to make a kind of juxtaposition. We better take this out there Lirnup 10:28, 4 October 2005 (UTC)

Good point on the diffusion not being connected to turbulence, some sort of convection would really be necessary as I would consider the velocity of diffusion occurring within some characteristic length to not be able to obtain Reynolds numbers normally associated with turbulence. I consider the dissipation of smoke a laminar flowBiotechscientist (talk) 02:37, 19 November 2010 (UTC)
 * To some extent there is a false dichotomy between convection and diffusion, insofar as "shear induced dispersion" is a coherent concept used to quantify effective diffusivities, see e.g. Taylor dispersion. The key is to establish the scale (in time and space) of the motion of interest- turbulent mixing is an effective diffusion on scales much larger than the scale of the turbulent eddies; for the typically encountered situations the fractal nature of turbulence means that no scale can be found on which it only adds to an effective diffusion, but some media have upper cutoff lengthscales for the turbulent motion. That smoke example was bad and should stay deleted, but convection and diffusion are not always mutually exclusive, and frequently what is called convection on a small scale leads to diffusion on a large scale. 99.133.144.210 (talk) 02:13, 29 July 2011 (UTC)

A redlink story
At 11h04 today (31 october 2005), I added a link to collective diffusion in this article. At 11h06, or 2 minutes later, this link was deleted by user:Karol Langner, on the basis that I should first create the collective diffusion article (I created this article "only" at 11h17). So Karol, a question: are you going to delete all the red links that have existed for more than 2 minutes in this encyclopedia, or should I conclude that you are specifically after my edits? ThorinMuglindir 12:32, 31 October 2005 (UTC)

OK sorry Karol I was a bit upset when I dscovered you had canceled my edit. Now, admit the real reason why you canceled it is because it conflicted with yours, and having to cope with those 2 text boxes is a mess. So that you thought that since your edit as big and mine was just a link, it wouldn't be much of a deal for me to add it back... I prefer that to you being specifically after my edits. ThorinMuglindir 14:27, 31 October 2005 (UTC)

BTW in hopes that you'd visit it quickly, I added a red link to the following page:
 * Talk:Degrees of freedom (physics and chemistry) ThorinMuglindir 14:27, 31 October 2005 (UTC)
 * I wasn't even aware that it was your edit, and deleted the red link, because it didn't add anything to that list. And I didn't look at the time, so I also didn't get that the page would be created soon. In any case, I'm not stalking you. Sorry. My bad. Karol 15:01, 31 October 2005 (UTC)


 * OK no problem. Your edits you do appear sound to me btw. We'd better get across as clearly as possible that diffusion is a very general phenomenon.ThorinMuglindir 16:21, 31 October 2005 (UTC)

Diffusion (mathematics)
I find this page only gives the physical view. There is also the mathematical term of a diffusion (process). Of course these concepts are related, but it's not the same. LARS 18:58, 19 November 2005 (UTC)


 * The mathematical term--and the other uses of diffusion--are in analogy to the mathematical models for the physical process, no? You can discuss the diffusion equation or a random walk without reference to the physical process, but then there are already articles on those topics.  I added a statement about the analogies in the intro. Danrs 18:53, 15 August 2007 (UTC)

Giant animation
Somebody might want to remove the giant animated GIF from this page. Just sayin'.


 * What's your screen resolution? You might want to increase it... &mdash; 0918 BRIAN • 2006-04-3 07:37


 * It's distracting and irritating, and the size contributes to this. My screen resolution is higher than average, yet I still find the image too big; I have a higher resolution because I want a greater density of information.  It really doesn't take a 12x12 square of pixels to represent a particle.  The image should be reduced to half size, and possibly slowed down (most of the animation shows random motion - the reader currently needs to watch the loop several times to catch the three seconds or so of diffusion). --Will 152.78.191.84 12:16, 14 June 2006 (UTC)

Why does diffusion happen faster in warmer systems?
Isn't it because warmer systems have more energy and higher energy molecules move fast? too bad i don't have a reference for that Tsinoyboi 16:16, 21 September 2006 (UTC)

Examples of diffusion
The example of the sugar cube in WATER is wrong, unless we could guarantee that water is completely still. Otherwise there will be some transport due to convection, and it is also likely that the "sugar water" neat the sugar cube will sink to the bottom of the glass due to its higher density and the gravitational force. —The preceding unsigned comment was added by 201.231.152.118 (talk • contribs).


 * I have removed the examples listed as not being examples of diffusion, because the statement implies that they do not really contribute to the article. Perhaps they could be reworded so they can be included in the article. For instance, I would still think diffusion plays an important role (in addition to convection). Perheps this could be reworded to reflect the role of diffusion?--GregRM 21:57, 19 July 2006 (UTC)

Could it be convection and diffusion? i'm finding in my chemistry textbook that convection in an isolated system such as a closed flask are examples. Maybe the difference is that it be closed and a room has too much outside energy such as people or light. Tsinoyboi 17:14, 21 September 2006 (UTC)

Removed section
The following are NOT examples of diffusion:
 * A gas distributes itself over a room by diffusion
 * A sugar cube in a glass of water that is not stirred will dissolve slowly and the sugar molecules will distribute over the water by diffusion
 * Ink in a beaker of water is an example of diffusion. In the end, the ink particles spread evenly throughout the mass of water

According to my textbook, Chemistry & Chemical Reactivity, 6th edition, Thomson Brooks/Cole ISBN 0-495-01013-8, cover seen here http://www.newtexts.com/newtexts/book.cfm?book_id=3038, those count as diffusion that ultimately they become scattered. It says so on page It's supposed to be closed containers so they can be closed systems, it's just that those systems are open to other sources of energy to affect the rate system. Rate doesn't mean it's not diffusion. Warmer molecules will just mix faster. Being open systems, heat may be introduced like windows or people in a room, or escape like water evaporating from a beaker. The listed items are still molecules diffusion.

However, in a quiescent medium, gas could distribute evenly across across a room, and the dissolving of sugar will distribute sugar within the water, and the ink particles distributing is diffusion again. I think the word NOT should have been simply removed. Wiki editors, I am new to this, but removing content that is worthwhile to the general public, and relevant, I think you should correct the inaccuracies, and not just remove content. I know about my little section of the world, but so do other people, even if it is the other way around.Biotechscientist (talk) 02:47, 19 November 2010 (UTC)

Removed examples
Theses have nothing to do with diffusion, they are examples of active transport, require energy and specialized transporters/engulfing machinery:
 * Active transport, pumping material across a cell membrane
 * Pinocytosis, "cell drinking" intake of small droplets of lipids
 * Phagocytosis, carrier proteins are used to transport glucose

Besides Pinocytosis is an intake of small droplets of liquid not lipids, phagocytosis has nothing to do with transporting glucose, it's engulfing of solid particles by cell membranes, read their articles for more info.Enemyunknown (talk) 13:24, 12 December 2008 (UTC)

Diffusion and Thermodynamics
Diffusion and Thermodynamics are definitely related but can you say that diffusion occurs as a result of the second law of thermodynamics? isn't it more that The second law of thermodynamics works because of diffusion of heat? Tsinoyboi 16:15, 21 September 2006 (UTC)

Also, i found on Thermodynamic equilibrium, "diffusive equilibrium" is used but there's no page for it. Tsinoyboi 16:34, 21 September 2006 (UTC)

hmm, i guess in Chemical potential, it states that entropy remains unchanged at equilibrium, but then again how is chemical potential related to diffusion? I still stand by diffusion as a result of the second law is incorrect. Tsinoyboi 16:34, 21 September 2006 (UTC)

I changed the third paragraph of nature of diffusion. I even cited my source --Tsinoyboi 07:28, 29 September 2006 (UTC)

Chemical potential gradients are what cause diffusion. This is why diffusion of, for example, matter, does not have to occur 'down a concentration gradient' - it will go in any direction which reduces the chemical potential. If this weren't true, we couldn't get spinodal decomposition. Mike 17:07, 17 May 2007 (UTC)

I re-added a sentence about the chemical potential gradient's importance to diffusion, and in particular up-hill diffusion. I think it is of great importance to be aware that diffusion in reality is NOT an effect of concentration gradients but of the interactions between individual atoms/molecules, and how these prefer to stick together or to spread out, to minimize the system energy.(Johnsarelli (talk) 12:12, 4 August 2008 (UTC))

Example Animation is Convection not Diffusion
The example animation on the page shows a clear example of convection and bulk flow; the higher-density dye sinks in the glass of liquid (with turbulent fluid flow); barely any of the visible mixing is due to diffusion. After all, diffusion constants for small molecules tend to be a few hundred microns squared per second, meaning that it would take millions of seconds (weeks) for diffusion to equilibrate the dye concentration in the glass.

Does anyone want to keep that graphic as an example of "mixing that isn't diffusion"? Otherwise, even though it looks cool, I'll just delete it. Ichoran 04:29, 15 October 2006 (UTC)

You are right. http://en.wikipedia.org/wiki/Image:Diffusion.gif must be deleted from the diffusion or edited. -V.I. 17 November 2006

WHY?????? goddammit, quit deleting my edit...Will987654321 06:08, 1 December 2006 (UTC)will987654321

Because it is NOT diffusion -V.I.

I agree that the illustration is NOT an example of diffusion in the Fickian sense that this article is referring to. It is an example of convection. The time it would take to see Fickian diffusion at that scale is BIG. I think that the animation is very misleading and should be removed. When I get some time I will make a simulation of Fickian diffusion. -ZBS


 * I agree with the point that convection is not diffusion and doesn't belong here. A quibble about the use of "Fickian": usually it's used to contrast with other (non-Fickian) forms of diffusion, for example in complex systems, or where the flux depends on the magnitude of the concentration (in addition to the concentration gradient).  It's a little odd to use Fickian to contrast with convection, which is not diffusion at all.  Also, see Fick's law section below.  Danrs 19:04, 15 August 2007 (UTC)

Fick's law
The introduction implied that Fick's law and Fourier's law were alternate names of the diffusion equation. Fick's law is used to derive the diffusion equation, but is not identical to it. Also, I think the important aspect of Fick's law is not that it labels the diffusing substance as mass (as opposed heat), which was implied, but that the net transport is proportional to the concentration gradient. There are other possibilities for purely diffusive systems! I tried to clarify this in the introduction. Danrs 20:04, 15 August 2007 (UTC)

Spontaneous Process
See my comment in the Spontaneous Process talk page.

A section for the shmucks
Is it possible for us to add a section into the article where this process is explained simply? I'm staring at the screen and cringing when I read this - maybe I just have unnaturally low intelligence. The Last Melon 01:53, 26 June 2007 (UTC)

I've tried to make the introductory paragraph more accessible to non-experts. Ichoran 19:03, 17 July 2007 (UTC)


 * Unfortunately, the introductory paragraph was edited to say that diffusion *is* the movement of particles, rather than that this is an *example*. I'm restoring the more general (i.e. more accurate) text.  If there is a compelling reason to do otherwise, please mention the reasoning here!  128.177.16.8 19:10, 3 October 2007 (UTC)

Concentration gradient
"A concentration gradient is the difference between the high concentration and the low concentration. It also determines how fast diffusion occurs."

Isn't it more like the concentration gradient is a measure of how fast the diffusion is? // habj 21:17, 3 August 2007 (UTC)


 * No. The diffusion coefficient is the measure of how fast the diffusion is. The concentration gradient is the result of the initial condition, which diffusion tends to erase.


 * Both sentences from the quoted text have problems. The gradient is the difference between the high and low concentration _divided by the distance_.  Also, it's the flux through a surface that is affected by the concentration gradient, not diffusion.  Simple diffusion occurs independent of concentration and can be measured even in a uniform concentration field.  For instance, the motions of particles in a well-mixed solution can be measured (e.g. using a microscope), and assuming pure diffusion their mean squared displacement will increase linearly in time (a dynamical hallmark of diffusion), with the constant of proportionality giving the diffusion coefficient, as described in the "Diffusion displacement" section.  Danrs 22:16, 13 August 2007 (UTC)

Thermal diffusion ≠ Heat flow
Thermal diffusion ≠ Heat flow. Currently Thermal diffusion redirects here, and then this article sends it off to "Heat flow", which redirects to Heat transfer). This is not the best understanding of Thermal diffusion as in the OED or most places. — DIV (128.250.204.118 06:03, 21 September 2007 (UTC))

Diffusion experiments
The diffusion experiment with ammonia and cotton wool seems a little suspect to me. Won't the rapid evaporation of ammonia increase the local pressure and thus cause slight bulk flow within the tube? Wouldn't stabbing gelatin with a needle with ink be a better experiment? The key to doing a good experiment that demonstrates diffusion is to choose a system where non-diffusive flow is greatly retarded by viscosity or somesuch. Ichoran 19:55, 3 October 2007 (UTC)

This etymology is complete nonsense.
The term diffusion is derived from the Latin verb husionere which means "to break" but can also mean "leave way"

There is no such word "husionere" in Latin. In fact, I can only assume this is vandalism, since it doesn't bear even a passing resemblance to reality.

http://www.m-w.com/

diffuse
 * Etymology: Middle English, from Latin diffusus, past participle of diffundere to spread out, from dis- + fundere to pour

dis-
 * Etymology: Middle English dis-, des-, from Anglo-French & Latin; Anglo-French des-, dis-, from Latin dis-, literally, apart;

http://archives.nd.edu/latgramm.htm

diffusilis -e [capable of spreading, elastic]

fusus (1) -a -um partic. from fundo; q.v.

fundo (1) fundere fudi fusum: of liquids [to pour, pour out]; of metals, [to melt, cast]. Transf., [to pour out, shower, give abundantly]; [to squander]; 'se fundere', [to rush, stream]; of sounds, [to utter]; with emphasis on distribution, [to spread, extend, scatter]; milit., [to rout, defeat, scatter, put to flight]. Hence partic. fusus -a -um, [spread out, extended]; 'crines', [flowing free]; of speech, [diffuse]; adv. fuse, [widely, copiously].

Proving once again that an encyclopedia predicated on popular consensus... sucks.

--75.63.48.18 (talk) 08:02, 8 January 2008 (UTC)
 * A quick google search shows that husionere has now successfully diffused around on the net. These usages may now be used as a citation to refute your proposed etymology. 178.38.117.134 (talk) 13:13, 13 April 2014 (UTC)

Proposed move to Molecular diffusion
I proposed to move this article to Molecular diffusion.

Reason: to better define the scope of the article. "Diffusion" is a large and complex topic.


 * For "diffusion in general", there is already an article Diffusion (disambiguation).
 * For specific types of diffusion, there are already multiple English Wikipedia articles, as listed on the above disambiguation page. If any specific diffusion-related topic is missing, a new article can be added.

The obviously central yet missing article is that on the molecular diffusion of mass.

Under the current title, the basic definitions are disputed. The article does not appear to develop into a useful one, despite an overall large number of edits.

Cheers. Stan J. Klimas (talk) 21:32, 22 October 2008 (UTC)
 * Complete Tohd8BohaithuGh1 (talk) 19:22, 23 October 2008 (UTC)

Example diffusion speeds
The article gives a few examples for typical diffusion speeds such as "In gas: 100 mm per minute". Diffusion is inherently something that is described by the square of length per time unit and these examples (which suggest that 100 mm per 60 seconds equals about 1.5 mm per second) are misleading. Han-Kwang (t) 09:14, 2 February 2009 (UTC)
 * Yes, this is much clearer. Yet, perhaps the examples should be even more clarified or removed altogether because they can confuse. Not sure. Stan J. Klimas (talk) 23:03, 2 February 2009 (UTC)

See also section
The See also section is now rather long with 13 entries. Would it be better to shorten this and rely on the Diffusion Category to aid further navigation? --Art Carlson (talk) 11:35, 9 March 2009 (UTC)

Suggested merge of diffusion equilibrium into this article
I just suggested that the page Diffusion equilibrium be merged into this one. There is only a small amount of material on that page, and this article already has a section on non-equlibrium systems, so it seems like it would fit here. Shanata (talk) 21:29, 20 October 2009 (UTC)

Diffusion page figure: Carrier diffusion in instrinsic semiconductors
The animation on carrier diffusion in an intrinsic semiconductor bar is inaccurate. The greater hole concentration relative to electron concentration is attributed to the higher diffusion coefficient of electrons in most semiconductors. While this is true for most values of electron and hole diffusion coefficients, under this configuration the photogenerated electrons and holes actually diffuse at the same rate, i.e. they should be in equal concentration for each position in the bar. The reason for this is the formation of an electric field between the electrons and holes as the electrons attempt to diffuse more rapidly. This electric field drives the electrons back towards the holes and vice versa as any concentration mismatch develops. Because of the induced drift, the carriers actually diffuse at the same rate, otherwise known as ambipolar diffusion. This phenomenon is not a good example to show since it fundamentally involves both diffusion and drift. I would suggest changing to an example not involving the diffusion of charged particles due to this complication, so that it is more accurate.

Diffusion vs convection
[moved here from private talk page:]

In October of 2010 I edited the diffusion page with the example of spraying cologne. Agreed, maybe I should have been more specific about the transport phenomena occuring, but the edit you have made of this example is clearly incorrect. When the cologne leaves the nozzle of the sprayer convective flow will be dominating; however as the cologne "cloud" increases its volume it will eventually reduce its velocity to zero at which point it will be at atmospheric pressure. Please note, this phase of transport is primarily driven by pressure and is therefore the convective regime of this process; however there will still be a concentration gradient between the cologne cloud and the room air(so long as the room is sufficiently large enough to accomodate the cologne cloud at atmospheric pressure and temperature). Transport will then continue via brownian motion, ie diffusion, until the cologne is eventually distributed throughout the room. This process will occur in the presence of no other gradients (pressure, temperture, electromagentic) because of the random motion of the molecules. Or it could be looked at as, it will always occur because the system with a cologne cloud and air is more ordered than a system where the cologne molecules evenly distributed throughout the room. This is to say the process is entropically driven. Since entropy always increases the process will occur spontaneously at any temperature above 0 kelvin. Another way of looking at it, is that the pressure driven flow is an enthalpic process where diffusion is an entropic process.

Please make the appropriate changes to the page. The ink droplet example is also incorrect for similiar reasons. I am not fixing the page because I did not break it.

If you did not make the described changes I apologize. — Preceding unsigned comment added by Anwilson6 (talk • contribs) 17:52, 14 April 2011 (UTC)


 * Eau-de-Cologne in air and ink-in-water are popular examples for mass transport because they are vivid real-life cases known to everybody. In real life, however, these tracers spread in space because of convection, period. All you say is correct in some hypothetical limit of air or water being perfectly at rest, with walls being at perfectly homogeneous and constant temperature - but in this case, diffusion takes extremely long, and you have lost any connection to the real-life situation that originally motivated the choice of these examples. -- Marie Poise (talk) 09:17, 18 April 2011 (UTC)


 * First, please show me the SIMPLE equation that you use to show that the flux of cologne can ONLY be by convection. I want your general equation and assumptions.  Convection is a bulk property.  It is either forced or natural (as with density gradients).  If I just opened the bottle of cologne up, there would be ZERO velocity due to convection.  I have not applied a force and natural convective flow is absent.  This is the same for the drop dye aside from the original movement into the water.  Additionally, you have described diffusion as EXTREMELY SLOW/LONG TIME.  Are you kidding me?  Diffusion of a gas in a gas is THE FASTEST of any molecule in a medium.


 * I would use the Nernst-Planck equation. Is there an electric field being applied?  No.  Is there a velocity of the bulk cologne?  Sure.  Is there a concentration gradient?  Yes.  Explain to me then how both aren't occurring simultaneously?  Regardless of the first persons entropic/enthalpic point, which is valid, you have given ZERO points to validate your claims in this post and the diffusion wikipage.  I BEG YOU, assuming you are educated or in the process of a obtaining a degree, to bring your Diffusion Wikipedia page post to a reputable professor of chemical engineering or mechanical engineering and have them explain to you the fundamental mass flux process.  — Preceding unsigned comment added by JC44567 (talk • contribs) 02:48, 22 April 2011 (UTC)

Major Rewrite Proposal
Diffusion vs convection [moved here from private talk page:] In October of 2010 I edited the diffusion page with the example of spraying cologne. Agreed, maybe I should have been more specific about the transport phenomena occuring, but the edit you have made of this example is clearly incorrect. When the cologne leaves the nozzle of the sprayer convective flow will be dominating; however as the cologne "cloud" increases its volume it will eventually reduce its velocity to zero at which point it will be at atmospheric pressure. Please note, this phase of transport is primarily driven by pressure and is therefore the convective regime of this process; however there will still be a concentration gradient between the cologne cloud and the room air(so long as the room is sufficiently large enough to accomodate the cologne cloud at atmospheric pressure and temperature). Transport will then continue via brownian motion, ie diffusion, until the cologne is eventually distributed throughout the room. This process will occur in the presence of no other gradients (pressure, temperture, electromagentic) because of the random motion of the molecules. Or it could be looked at as, it will always occur because the system with a cologne cloud and air is more ordered than a system where the cologne molecules evenly distributed throughout the room. This is to say the process is entropically driven. Since entropy always increases the process will occur spontaneously at any temperature above 0 kelvin. Another way of looking at it, is that the pressure driven flow is an enthalpic process where diffusion is an entropic process. Please make the appropriate changes to the page. The ink droplet example is also incorrect for similiar reasons. I am not fixing the page because I did not break it. If you did not make the described changes I apologize. — Preceding unsigned comment added by Anwilson6 (talk • contribs) 17:52, 14 April 2011 (UTC) Eau-de-Cologne in air and ink-in-water are popular examples for mass transport because they are vivid real-life cases known to everybody. In real life, however, these tracers spread in space because of convection, period. All you say is correct in some hypothetical limit of air or water being perfectly at rest, with walls being at perfectly homogeneous and constant temperature - but in this case, diffusion takes extremely long, and you have lost any connection to the real-life situation that originally motivated the choice of these examples. -- Marie Poise (talk) 09:17, 18 April 2011 (UTC) First, please show me the SIMPLE equation that you use to show that the flux of cologne can ONLY be by convection. I want your general equation and assumptions. Convection is a bulk property. It is either forced or natural (as with density gradients). If I just opened the bottle of cologne up, there would be ZERO velocity due to convection. I have not applied a force and natural convective flow is absent. This is the same for the drop dye aside from the original movement into the water. Additionally, you have described diffusion as EXTREMELY SLOW/LONG TIME. Are you kidding me? Diffusion of a gas in a gas is THE FASTEST of any molecule in a medium. I would use the Nernst-Planck equation. Is there an electric field being applied? No. Is there a velocity of the bulk cologne? Sure. Is there a concentration gradient? Yes. Explain to me then how both aren't occurring simultaneously? Regardless of the first persons entropic/enthalpic point, which is valid, you have given ZERO points to validate your claims in this post and the diffusion wikipage. I BEG YOU, assuming you are educated or in the process of a obtaining a degree, to bring your Diffusion Wikipedia page post to a reputable professor of chemical engineering or mechanical engineering and have them explain to you the fundamental mass flux process. — Preceding unsigned comment added by JC44567 (talk • contribs) 02:48, 22 April 2011 (UTC

This page needs major work. I wouldn't mind doing it, but I'd like to know if I have the community's approval first, since I'm new to this.

First, and my main gripe, it conflates the relaxation of concentration gradients with the motion of individual particles, which makes it difficult to choose a point of view from which to explain things. I'd like to separate out the continuum concepts of concentration field relaxation from the idea of random molecular motion. That way the thermodynamics and continuum mechanics ideas, like convective mixing, shear-induced dispersion, chemical potential gradients, etc, could be focused on in one article while activated hopping, collisions, and so forth can be focused on in the other. The commonly observed phenomenon, chemical or concentration diffusion, should be detailed separately from its underlying mechanism, particle diffusion, to avoid information overload for people only interested in one or the other. Each one will retain references to the other, but in less confusing mix than right now.

Second, there are numerous misleading statements, starting from the first sentence where it is implied that solids don't undergo diffusion. I'll try to clean these up, using standard engineering and physical chemistry texts like Bird Stewart and Lightfoot and Atkins as terminology references.

Third, there's too much left out. I'll fill in a bit of the history, some additional phenomenology comparing liquid, gas, and solid phase diffusion, some discussion of mathematical methods for dealing with it, e.g. grid-based and meshfree advection-diffusion PDE solvers, stochastic schemes for individual particle trajectories like Brownian and Stokesian dynamics, and so on. I'd also like to flesh out the equilibrium vs. non-equilibrium, linear vs. nonlinear, and frequency-dependent vs frequency independent cases, but I'll wait on that till linear response theory is a respectable page of its own.

That's what I'd like to do. If there are objections, please let me know. Also, I'm not a professor, just a grad student, and I'm not going to want to proofread everything, so it's a sure bet that I'll confuse some terms. Help me out if you see any mistakes. — Preceding unsigned comment added by ChE Fundamentalist (talk • contribs) 05:43, 29 July 2011 (UTC)


 * So you're making a second article? What were you planning to call it? Would this article ("molecular diffusion") be the continuum or the microscopic one? Also, for people coming in with no background knowledge, which page would you want them to view first? (i.e., one of the two pages should presumably have a broader and simpler introduction and overview encompassing both the micro and macro aspects, which are after all related).
 * I'm not an expert and I hope you can educate me...but I always assumed that the word "diffusion" referred only to the process described by random molecular motion that leads to Fick's laws. Therefore, shear-induced dispersion and convective mixing are not within the topic of diffusion (except maybe in the category "small-scale processes that lead to Fick's equations on a larger scale", or the more general topic "mass transport phenomena"). If that's true, I'm not convinced there's enough continuum stuff for a dedicated article (on top of the Fick's laws article). For example, if you want to talk about PDE-solving methods, it seems to me that you should make a new article numerical solution of the convection-diffusion equation, because the topic is self-contained and specialized. (Or add to convection-diffusion equation.) :-) --Steve (talk) 13:49, 29 July 2011 (UTC)


 * I was planning to call the new one "chemical diffusion" and to focus it on continuum processes, while modifying this one to focus more on the microscopic. Do you have any other suggestions for how to handle it? "Molecular diffusion" bothers me in particular, but "Particle diffusion" isn't much better... ideally I could make it general enough to include diffusion of microstructure defects in inhomogenous materials, too, since things like hydronium ions don't really diffuse as molecules but as charge defects, but I can't think of a name that works for it. Any ideas?
 * I would want chemical diffusion to be viewed first in most cases, since it will be more related to the lay interpretation of diffusion as mixing. The introductions will both refer to one another since they are so closely related. Does that sound good?
 * I think "The process described by random molecular motion that leads to Fick's laws" is a great one-sentence summary of what's usually taught about diffusion in school, after you've learned it and if you understand how to read PDEs. However, I also think it's a pretty narrow description and it's not intuitive for people less familiar with the equations. First, it leaves out the role of chemical potential gradients, and second Fick's law is just a basic linear constitutive equation which isn't really fundamental except in the sense linear approximations have been pretty fundamental to the mathematization of physics in general. I'd rather emphasize the non-mathematical, general aspects of diffusion, like the fact that it's a spontaneous and irreversible relaxation of a higher-energy state to a lower-energy state with a rate that depends on the chemistry of the underlying material. From that thermodynamic perspective, there's a lot that can be said just on the continuum level.
 * For the shear-induced dispersion part, I was planning to have a "Role of diffusion in general mixing processes" which describes how convection leads to sharpened chemical gradients, which accelerates diffusion (among other things). I'd agree that that should go in mass transport except that the debates in the talk page here about what is convection and what is diffusion convinced me it would be worth addressing directly. Does that still sound unreasonable?
 * I agree that a detailed discussion of the numerical methods is inappropriate for this page, and if the convection-diffusion equation weren't such a special case, I'd stick everything in that article or the one you suggested making. I am planning to have a "Mathematical Descriptions of Chemical Diffusion" section which will link to articles on all the various PDEs, integral equations, and so forth that come up when you're trying to model it. That's where I was planning to mention the numerical methods, but I think you're probably right that they don't fit there. I'll leave them for the mathematical model pages, like "convection-diffusion equation." Thanks for the suggestion! ChE Fundamentalist (talk) 19:36, 31 July 2011 (UTC)

Welcome, ChE Fundamentalist! There is definitely lot of work to be done. If you are new to Wikipedia, I just suggest you start by small improvements here and there before you attack the major restructuring the diffusion articles definitely deserve. If you need help, feel free to contact me on my talk page. Again: welcome! -- Marie Poise (talk)


 * Thank you for the welcome, Marie. I'm drafting my changes in a separate document, and I'll be sure to get a bit of practice by editing some of the pages I plan to reference before I actually implement any major changes here. ChE Fundamentalist (talk)

I've gotten a bit of experience wikifying, cleaning up typos, and writing big sections on articles, so now I'm planning to return to this. I've come up with a layout, and I'd like to know what anyone watching this thinks.

First, I'm going to go ahead with the separation into Molecular diffusion and Chemical diffusion. Chemical diffusion should be the first one that viewers come across if they're just interested in the mixing process, but both should have disambiguation notices just below the titles. Is there a tool I could use to sort links from other pages to these and make sure each link goes to the right one?

In the Chemical diffusion article, my planned layout is the following:


 * 1) Key Concepts
 * 2) It is a dissipative relaxation process.
 * 3) Chemical potential gradients drive it.
 * 4) In most applications, it can be described as a local, linear, continuum process (said less technically).
 * 5) Characteristics of Chemical Diffusion
 * 6) Advection vs diffusion vs convection
 * 7) Fickian vs. anomalous diffusion
 * 8) Nonlinear vs. linear diffusion (collective vs independent motion)
 * 9) Tracer vs. chemical diffusion (both do belong in this article)
 * 10) Diffusion in Chemical Engineering and Chemistry
 * 11) Definition of diffusion constants and their calculation
 * 12) Fick's law, Maxwell-Stefan, etc
 * 13) Chapman-Enskog, Arrhenius dependence, etc.
 * 14) Application
 * 15) Purely diffusive mass transport processes i.e. membrane separation, gaseous diffusion enrichment of uranium
 * 16) Mass transport processes combining advection and diffusion, i.e. chromatography, mixing, most reactors
 * 17) Coupling of mass transport with other transport phenomena, esp heat transport
 * 18) Modeling
 * 19) Generalized correlations for mass transport, dim'less nos and engineering examples.
 * 20) Continuum equations, i.e. Smoluchowski and linear advection-diffusion

And then the Molecular diffusion page will have the following layout:


 * 1) Key Concepts
 * 2) Random processes in the statistical sense rather than nondeterministic sense
 * 3) Self-similarity and the scaling law  ~ t
 * 4) The idea of long-time, or asymptotic, behavior
 * 5) Phenomenology of Diffusion
 * 6) Mechanisms of Diffusion
 * 7) Elementary deterministic, predictable steps
 * 8) Quantum hopping
 * 9) Ballistic motion
 * 10) Overdamped motion
 * 11) Topological remodeling
 * 12) Randomization
 * 13) Internal nonlinearity (i.e. collisions)
 * 14) Random external forcing (i.e. Brownian motion)
 * 15) Statistical Mechanics of Diffusion
 * 16) Mathematical definitions
 * 17) Definition of the long-time diffusion coefficient
 * 18) Definition of the time-dependent diffusion kernel
 * 19) General Theory
 * 20) Fluctuation-dissipation theorem
 * 21) Onsager regression hypothesis
 * 22) Large Deviation Function
 * 23) Computational models
 * 24) Molecular dynamics
 * 25) Brownian and Stokesian dynamics (overdamped dynamics in general)
 * 26) Kinetic Monte Carlo and Lattice Boltzmann

I'm not trying to make this go straight from start-class to good article status; I know there are things I've left out that should be covered in these articles. If there are any that you think I've left out that simply must be in any remotely acceptable version of these articles, please wait for me to finish this and then add them yourself or type them up on this talk page so I can copy and paste them in. On the other hand, if you have any objections to the general structure that I'm proposing, please let me know immediately so I can consider changing it. ChE Fundamentalist (talk) 18:49, 21 August 2011 (UTC)


 * I am not convinced that the article should be split. It depends too much on your perspective whether the division in chemical vs molecular makes sense or not. In the end, people who do not grasp the distinction will mess thinks up by duplicating contents &c. From my editing experience, I would recommend to have one comprehensive article, covering chemical and physical, macroscopic and microscopic aspects of diffusion, but leaving out all technical details, which should go into specialised articles. At the beginning of each section, a link would point to one of them. This is also the best way to avoid duplication of contents between overview and specialist articles. -- Marie Poise (talk) 15:49, 9 January 2012 (UTC)

Molecular diffusion in solids
There is also molecular diffusion in solids. This is the origin of aging processes in many solid materials. I would therefore suggest to remove the "(liquid or gas)" in the first sentence "Molecular diffusion, often simply called diffusion, is the thermal motion of all (liquid or gas) particles at temperatures above absolute zero." — Preceding unsigned comment added by 128.176.210.208 (talk) 17:14, 10 September 2019 (UTC)

Introductory sentence
Hey authors, I only had a quick view on this article and I find that already the introductory sentence "Molecular diffusion, often simply called diffusion, is the thermal motion of all (liquid or gas) particles at temperatures above absolute zero." misses some precision. To my knowledge and according to the definition in many basic textbooks in physics, diffusion is the mixing process resulting from random motion, not the random motion itself. The random motion of the fluid particles is rather termed thermal motion. It seems I cannot edit the introductory section. Please help! Thanks (Pelicanelson (talk) 10:53, 10 September 2020 (UTC)).