FridayAFM - Sticky, Sirupy, and Gloopy

Héctor here, your AFM expert at Nanosurf calling out for people to share their Friday afternoon experiments. Today I explore how viscosity affects the way AFM probes behave.

I get a lot of suggestions about new experiments, this is good because it means I don't need to think too hard to come up with new things to explore. At the same time, it is bad because usually it leads me down a rabbit hole exploring new topics, generating new questions... and suddenly it is friday and I need to have something ready. This experiment today is one of those rabbit holes where I got more questions than answers and where, if it wasn't for the need to post something, I will continue exploring forever.

Somebody said to me:

"some time ago we did some studies on viscosity, why don't you experiment with different alcohol concentrations and show how it affects the probe movement?".

You see, the ratio of water/alcohol affects (among other things), the viscosity of the liquid (see Ref 0), and an oscillating AFM probe drags air (or liquid) when it moves, so changing the viscosity will affect the probe movement.

It seems easy enough right? Get some alcohol, see how the probe moves, add some water, see how probe moves, and keep adding water until the liquid becomes mostly water. 

The trap is mostly on "see how the probe moves", there is some part of a trap on the water alcohol mix, but I'll talk about that later.  

A probe oscillating in vacuum has its movement defined by the driving force, the material properties and the geometry. The movement in liquid also depends on the liquid viscosity because when the probe moves in liquid it drags the liquid with it. This converts the probe into a viscosity sensor, the question is, which peak should we use for tracking?

In air, independently if you are using photothermal excitation or piezoelectric, there are nice, isolated, resonant peaks (see below). Some of these peaks are flexural, some torsional, and some might be lateral or combinations. See the post on Torsional resonance imaging for more details and how photothermal allows easily accesing the torsional modes.

Again, which peak or peaks should we use? Moreover, what happens when in-liquid? (In water at room temp to be more precise).

In liquid, there are two stories, one for piezoelectric excitation, one for photothermal. The piezoelectric one is the forest of peaks, the many resonances appearing because the piezoshaker is moving the cantilever, the chip, the water around both, and all couples together. The photothermal one is the nice one, because suddenly the forest of peaks disapears, there are only few well defined peaks, and we are only moving the cantilever and the liquid around it and nothing else. Read more about the benefits of photothermal excitation in liquid in our application note: CleanDrive: Photothermal Excitation of the Cantilever.

Sudenly the photothermal has simplified things a lot. First of all, there only a few peaks to study, and when we look at the dlf and lat channels (to see if these peaks correspond to flexural or torsional or a combination of both), we see that there is a peak that is a mix of flexural and torsional which happens to be very sensitive to viscosity.

So, we have the peak. Time to get some alcohol.

My first idea was to use ice to dilute the alcohol while I keep doing freq sweeps and monitoring the evolution of the peak. It has the advantage that I can keep the pobe submerged and everything is at zero degrees. At intervals I will check the alcohol concentration using a refractometer, which is a instrument based on light refraction. We call this plan A.

So, let me explain why plan A sucks.

The system is not in thermal or concentration equilibrium. This means that the temperature close to the probe is not zero and it depends massively on the water flow. Also, there are places with higher alcohol concentration than others. All this combined results in a plot with the data scattered around as you can see below. also, I don't think it was one of the main effects here, but 

Plan B: use water at room temperature. I added water at intervals, stir, measure alcohol concentration and performed a freq sweep.

As you can see, plan B works and we can see a linear relationship between alcohol concentration and frequency shift. Tiny one in percentage, but measurable.

Can we do better? When looking for bibliography studying this topic I found that sugar also changes water's viscosity. You see, usually there isn't too much diference because in daily life it is strange to add a lot of sugar to water... but it becomes really sticky, and at normal conditions you can add a 200% of sugar into water in terms of mass. So, I repeated the experiment this time adding sugar.

As you can see, the result is more dramatic. almost 50% frequency shift when the water is almost saturated with sugar. Note that I used what I learnt from alcohol to make this experiment better, which is stirring and waitting for the liquid to stabilize.

Let's recap. Viscosity affects AFM probe oscillation. We can easily see this using photothermal excitation and looking at the right peak. With alcohol the change is small but measurable, with sugar the change is massive. There are several potential applications for this, for example, to monitor reactions, or to tune the probe movement and make easier imaging certain objects (see Ref 2).

I hope you find this useful, entertaining, and try it yourselves. Please let me know if you use some of this, and as usual, if you have suggestions or requests, don't hesitate to contact me.