The phenomenon of osmosis is familiar to most readers from their junior high school science class. A dialysis bag containing sugar is placed in beaker containing pure water. It is explained that the bag is semi-permeable: water molecules can pass through microscopic pores in the membrane, but sugar molecules are too large to pass. The student watches as water gradually flows into the bag, inflating it until it seems ready to burst. The mystery of osmosis is that water continues to flow in, even after the pressure within the bag exceeds that outside.
This seemingly simple phenomenon is vitally important for plants and animals. Osmosis plays a role in the blood circulation, keeping just the right balance between the water content of the blood and the surrounding tissues. Osmosis drives fluid flow in the kidneys, preventing waste products from accumulating to dangerous levels. Osmosis is also the driving force behind plant cell expansion, playing a role in flower and fruit growth.
It's likely that you learned the following explanation for the flow of water: the sugar molecules inside the bag displace some of the water molecules, so the number of water molecules per unit volume is lower inside that out. Thus, water molecules undergo diffusion, through the membrane pores, from the region of higher water concentration to the region of lower concentration. In short, osmosis is a special case of diffusion.
Easy as pie, right?
Except that it's wrong.
When a salt dissolves in water, it tends to disrupt the hydrogen bonds that keep water molecules evenly spaced. Most salts tend to increase the spacing, but there are a few - sodium fluoride, for example - that decrease the spacing. The number of water molecules per unit volume actually goes up. In such cases, the diffusion theory would predict an osmotic water flow in the wrong direction!
The correct atomic theory of osmosis emerged gradually during the first half of the twentieth century. But it wasn't until 1951 that an English language textbook presented a clear and fully realized alternative to the diffusion model (Theoretical Physics, Georg Joos). Shortly after this, Harvard biophysicist Arthur Solomon decided to tackle the issue from the experimental side. He led a research group that spent several years measuring the rate at which water flows in and out of red blood cells during osmosis. They showed that the flow was much faster than the diffusion theory could explain.
Thus, the 1950s saw the end of the diffusion explanation for osmosis, at least as far as physicists were concerned. Modern biophysics textbooks present an explanation of osmosis that does not differ from that first published in 1951. Here's my own description of the force that drives water across the membrane, as it appears in the April 2013 issue of Trends in Plant Science, ("solute" means any sugars, salts, etc., dissolved in the water):
The key interactions take place in the small region of space adjacent to a pore aperture that allows water molecules to pass but repels solute ... Each time a solute molecule enters this region, it is repelled. That is, the aperture gives to the solute molecule a small amount of momentum directed away from the membrane. Due to viscous interactions between solute and water, this momentum is rapidly shared among all nearby molecules, including both solute and water ... Thus, although the pore aperture repels only the solute, the net effect is a force directed away from the membrane acting on the solution as a whole.
Unfortunately, communication on this topic between physics, chemistry and biology has not been good. In the 1960s, most introductory college-level textbooks in chemistry and biology continued to repeat the discredited, diffusion-based view.
In the 1970s the situation grew even worse. A research group led by physiologist Harold Hammel at the University of California in San Diego began promoting a third theory of osmosis, which they called the solvent tension theory. In brief, they suggested that the sugar and the water in a solution could permanently co-exist at two different pressures. Water flowed into the dialysis bag, they suggested, because the water (i.e. the solvent) was only affected by the difference in water pressures across the membrane. A series of papers published in the September 1979 issue of the American Journal of Physiology refuted this theory point-by-point, after which it was regarded as fringe science. (Hammel's conviction was not shaken. When he passed away in 2005, the preferred theory was carved onto his headstone.)
As far as I know, 1979 was the last time that the osmosis of dilute solutions was discussed as an active research topic. Physicists regard the matter as long-settled. However, chemistry and biology textbooks continue to repeat the incorrect, diffusion explanation. And it's not just textbooks. If you try an internet search, you're sure to find plenty of authoritative-sounding discussions that explain osmosis as a special case of diffusion. There are also links to educational movies and video games designed to amplify the point.
In collaboration with my colleague, chemist David Myers, I have recently written two papers [2012, 2013] that try to bridge the gap between physics and biology. We are also working with the author of a popular physiology textbook to help him improve the next edition.
After so many years of confusion, will we finally succeed in fixing the problem? Here's a quote from the biologist Howard J. Stein, who also recognized the problem and tried to address it back in 1966: "The time has passed when authors of textbooks should continue with the older, and deceptively simple, idea of osmosis as a special case of diffusion."
Hope springs eternal.
Image: Internet archive.