I have been fascinated for a few years now by observations of vanishing droplets by Hajian and Hardt. Their droplets demonstrate Kamikaze behaviour, they move into a region of the solution where they are soluble, and so dissolve. Hajian and Hardt have good, and likely correct, ideas for what is going on, but I have wanted to find a simple model that shows this behaviour, so I can play around with it and understand what is going on. I have not yet succeeded but have now made some progress. As you can see if you run the above YouTube movie, the model droplet does move and dissolve.
Hajian and Hardt generate droplets using the ouzo effect (also shown and explained in first 20 s of this YouTube video). Drinks like ouzo (also pastis, sambuca, …) are strong (around 40% ethanol (alcohol)) and have an aniseed flavour, which comes from the molecule anethole. Anethole is highly soluble in ethanol, but very poorly soluble in water. This is presumably why you can’t get aniseed-flavoured soft drinks.
It also means that when you dilute ouzo by adding water, the anethole comes out of solution in droplets. This makes the drink cloudy, and this is the ouzo effect. The droplets are around one micrometre in size, and they scatter light, which is what makes the drink cloudy.
So this is how you create these droplets, how do you destroy them? The secret is to create a gradient in the ethanol concentration, then if the droplets move up this gradient, they move to regions with high-enough ethanol concentration to dissolve them. In Hajian and Hardt’s work it looks like their droplets may be spontaneously moving up the gradient in ethanol concentration, due to what is called the Marangoni effect.
I have not managed to see that in a simple model. But a bunch of people, including recently Katzmeier and Simmel, have used flow from a tiny pipette immersed in a solution, to generate concentration gradients. This made me wonder if you just use flow to push droplets up concentration gradients, and so dissolve them.
And as you can see from the video up top, you can.
The video is of Lattice Boltzmann simulation, using a method from Shan and Chen to generate the droplet. I first generate a droplet. Then at the start of the video, I impose flow to the right. This carries the droplet to right. Although you can’t see it directly in the video, there is an imposed gradient of effective solvent quality, such that the droplet is stable on the left but not on the right. So the flow carries the droplet to regions where it dissolves.
Presumably in experiment this could be done by flowing an anethole solution at a low ethanol concentration, and so with anethole droplets, into a solution with much higher ethanol concentration. Then the flow would push the droplets up an ethanol-concentration gradient and so dissolve them.
richardsear.me 