# How does osmosis work?

I have (very) basic knowledge in chemistry.

I am curious as to why osmosis works. I read a couple of introductory chemistry books and also searched online but most explanations I found only explained the phenomena (i.e. solvent molecules move through semipermeable membrane from low solute concentration area to high concentration area), but not the reason behind it at the molecular level (i.e. what makes more solvent molecules move towards the high solute concentration area than in the reverse direction).

Wikipedia seems to go a step further:

... there is an interaction between the solute and water that counteracts the pressure that otherwise free solute molecules would exert.

And a bit later:

The virial theorem demonstrates that attraction between the molecules (water and solute) reduces the pressure, and thus the pressure exerted by water molecules on each other in solution is less than in pure water, allowing pure water to "force" the solution until the pressure reaches equilibrium.

I think I understand what it means, but what I don't understand is this: osmotic pressure is said to be a colligative property, i.e. only affected by the concentration of the solute and not by its type. If the reason for osmotic pressure is the attraction between water and solute molecules - wouldn't we expect different osmotic pressures for different kinds of solutes?

• This should probably be asked at Physics, since osmosis is a physical process. – Jan Dec 4 '16 at 15:37
• @obe try to think about entropy changes, third law of thermodynamics and remember the fact that in osmosis solutes do not cross membrane. – JM97 Dec 4 '16 at 15:44
• Hmm, imho it would probably be OK here, although if it doesn't get answers, please feel free to flag for migration to physics. – orthocresol Dec 4 '16 at 15:44
• @JM97 honestly I don't know much about thermodynamics. I did read about it a little. I read that a system always moves towards a higher entropy state. The explanation didn't go into the low-level mechanics but the way I imagined it was that since free particles move randomly within a system, they are bound to spread out evenly (on average), thus increase the system's entropy. If this physical description is correct (is it..?) - then it makes sense to me that solvent molecules should still go through the permeable membrane at the same rate in both directions... What am I missing here? – obe Dec 4 '16 at 16:01
• @JM97 yes, but how come the interaction doesn't depend on the type of the solute? Wouldn't the interaction be different depending on the polarity / size / geometry / etc... of the solute? – obe Dec 4 '16 at 16:24

According to this document (https://arxiv.org/pdf/physics/0305011v1.pdf)

"Water molecules can pass through the membranes in either direction, and they do. But because the concentration of water molecules is greater in the pure water than in the solution, there is a net flow from the pure water into the solution." Hydraulic equilibrium is achieved when the lower concentration of water in the high pressure side is balanced by the higher pressure and therefore higher energy pV on that side

This is the diffusion model of osmosis which is easy to understand at your level.

• Thanks! This paper looks very relevant. I'll try to read it and come back... – obe Dec 4 '16 at 20:30
• I don't think that your answer answers my question - it doesn't describe the process at the particle level but rather uses derived concepts, however the article you linked to is the most relevant response I received for this question. However, while the article acknowledges that osmosis is not properly explained in a qualitative manner in chemistry texts, I am not sure it itself does it properly. It seems to go a bit further in the qualitative description but then jumps to equations instead of following through. Or maybe I just lack sufficient background to thoroughly understand... – obe Dec 7 '16 at 12:01

To first order the solute doesn't matter, though it probably has a small influence. The driving force for osmosis is entropic. The osmotic pressure then is not due to the forces between the solute and solvent, but mostly from the fact that you can mix the two together.

Indeed, osmosis works even when there's less interaction between the solvent and solute, but at some point, the solute is no longer soluble in solvent, so there is no mixing.

• Can you explain the forces that are exerted on solvent molecules that cause more of them to move or stay to or at one side of the membrane? – obe Dec 4 '16 at 16:13
• I think you're looking at this wrong. Imagine a container that only had gas molecules in one half it. What force causes the gas to diffuse into the other half? There is no force, but the molecules are moving, and there is a greater likelihood that they will occupy the whole box than just half of it. Well, osmosis is a kind of diffusion driven by the entropy of mixing. We can describe the fact that the process is favorable by defining a chemical potential, but there is no force per se that is operating. – Zhe Dec 4 '16 at 16:50
• I actually mentioned my understanding of the phenomena of gas filling up the whole box in a comment above. It is my understanding that free particles are in constant movement in random directions, and as such it makes sense to me that they will fill out a container evenly. I admit I have some open questions about the constant movement of particles as well but let's take it as an axiom for now. I still don't see how it explains osmosis. With a membrane that is not a barrier for water, according to the analogy, and ignoring interactions with the solute, I'd expect water to spread out evenly – obe Dec 4 '16 at 17:02
• See, that's the issue. It does spread out evenly, but you're using the wrong definition of evenly. A simplistic way to think about it is that when you add the solute, "evenly" requires that you need more water to make it look like the other side that doesn't have solute. – Zhe Dec 4 '16 at 17:13
• @Zhe is correct in writing that it is the increase in entropy that drives the process. Entropy here means the increase in probability (number of possible arrangements) which arises from distributing particles throughout a bigger rather than smaller volume. As the solute becomes diluted its entropy increases. – porphyrin Dec 5 '16 at 18:01

I think you can find what you are looking for by googling "entropy" http://www.mpcfaculty.net/mark_bishop/solubility_entropy.htm

To summarize the article : "In general, particles of matter tend to become more dispersed (spread out)."

My guess is that it has something to do with the leonard jones force. Here is a video I made where the "text" is like a positive force and all the particle are a negative force. So the particles are slightly attracted to the text but when they get too close to the text, they get launched away (because I made the repulsive force to be huge).

- But in reality every single particle would have a positive and a negative so they all are attracted to each other but push away from each other once they get close enough. https://vimeo.com/182775330

• I read the article. Interesting, thanks :) I agree that entropy seems to be key here. However, I still have difficulty with this. I think my main issue is natural tendency toward greater dispersal. I understand why it's the case with gas, at the atomic level, and why it's the case when mixing water and ethanol. I imagine the particles and their random movements and I can understand why they disperse. However applying this to osmosis is a leap I don't understand because when I imagine solvent molecules moving freely, they end up with the same amount on both sides of the membrane... – obe Dec 4 '16 at 20:12
• On one side, the atoms are repelling each other more because they are concentrated. But after they have the same amount on both sides, the pressure is equal....I imagine. – eromod Dec 4 '16 at 20:19
• @obe It's in the wikipedia quote you have in your question - when you dissolve something in a solvent, it changes both the something and the solvent. The water no longer behaves entirely like water, which in this scenario means that the side with solvent "feels" like it has less water than it really does. Read up on how solutions (or e.g. acids) work - it's not just a mechanical mix of two chemicals. Think about a salt solution - pure water is an isolant, pure salt is an isolant, but dissolve some salt in pure water, and you get a great conductor - both are "torn apart" in a solution. – Luaan Dec 4 '16 at 22:50
• @Luaan I realize that but I'm having trouble with the "feel" concept... if I believed water could "feel" things, I'd post on homeopathy.stackexchange.com :) jokes aside, having learnt a little chemistry, I now understand the molecular forces that underlie various macro phenomena. For example, I can explain why water is liquid in room temperature while oxygen is gas, using the intrinsic properties of the nucleus and its sub-particles. I'd like to be able to do the same with the osmosis phenomena... – obe Dec 4 '16 at 23:04
• @Luaan as far as I know the explanation for saturated solution is that the reaction works in both ways (i.e. solute dissolved and precipitates at the same time). The solution is said to be saturated when the rate of the reaction in both directions is the same. The exact point when it happens depends on several factors such as polarity, geometry, temperature, etc... It can be explained in these terms without using words like "feels"... can you explain how this manifests as thermodynamic pressure across the membrane? – obe Dec 4 '16 at 23:41