Obviously, one of the most important aspects of sample preparation is the choice of which EM grid, which kind of EM grid, and what kind of a surface, is used to support the sample. Now, there are lots and lots of different kinds of grids. They're all three millimeters across, and very thin obviously. But they have different patterns of the support structure. One class of grids are characterized by a mesh, and these have grid bars across the grid, both horizontally and vertically, and the mesh refers to how many grid bars there are per inch. So, for instance, a 200 mesh grid has spacings like this. A 300 mesh is more bars per inch, so the bars are closer together. And you can even get 400 mesh grid bars. Another kind of grid is called a slot grid. And here there's just a very large slot in the middle that's left completely open. Another kind of grid is called a finder grid. And the idea of a finder grid is that there's a number of symbols like this asymmetric object in the middle, and this large arrow here, and then the numbers that are actually imprinted and the letters. A finder grid helps you identify a particular grid square that, of interest and maybe return to it later or label it because all of these landmarks that are placed around it allow you to identify a particular square. These are particularly helpful for clem experiments for instance. Now the grids are commonly made of copper, but you can also buy grids made of nickel, or gold. The advantage of gold is that it's not at all toxic to cells. And so if you want to grow cells on a grid for days in a row, it's gold is a good choice because it's an inert metal. And sometimes people use grids made of molybdenum for reasons that we'll describe later. Now for some samples, it is possible to simply apply the sample to the naked grid itself because a sample will span all the way across the grid bars. However, it's much more common to apply some kind of a coating on top of those grid bars. One common coding is formvar which is a form of plastic. Sometimes on top of the formvar is deposited a thin layer of carbon. Another popular surface is holey carbon. And so here's a picture of holey carbon. Just meaning that there is a carbon sub-straight here at this reticulate network of carbon with lots of holes in between that can provide nice regions to image. Another popular grid surface is called quantifoil. These grids are made by a process that allows both the size of the holes and their pattern to be very carefully controlled. And so you can order a grid surface with any size hole you like in whatever pattern you like. And finally, one might take a grid with holes in it, either the holey carbon film or perhaps a quantifoil grid and cover it with another thin layer of continuous carbon. In that case a cross-section through the grid might look like this, where these are the heavy thick grid bars of copper. And on top of that is a layer of carbon with holes in it, and so in cross-section it might look like this. And then on top of that might be deposited yet another layer of very thin carbon, perhaps just five nanometers or so. Another thin layer of carbon on top of that. Followed by the sample of interest with say individual proteins in suspension in that fluid. Now the way these very thin layers are produced is a process called evaporation. For instance carbon evaporating. The way it's done is imagine a chamber here, and theres a platform on that chamber on which a grid is placed. And then, above that is a thin rod of let's say carbon, a sharpened carbon rod. And the chamber is evacuated. So there's a vacuum inside and then a current is passed through this rod causing it to heat up. And as the rod heats up tiny little clusters of carbon atoms will come raining down away from that tip. I suppose they rain in all directions, but the ones of interest are the ones that rain down and cover our grid with a very thin layer. And depending on how long you apply the current, the, the layers that's deposited gets thicker and thicker. Now it has been observed that as soon as a layer of carbon is deposited through this carbon evaporation, it's quite hydrophilic. And so if you apply a drop of fluid that fluid will just immediately wick all the way to the edges of the grid and behave well for sample preparation. However, if those grids are, stored for weeks in a row. The surface of the grid often changes somewhat, and they become hydrophobic. And so, if you applied a drop of sample on the grid later you would see it bowl up, because the grid itself was so hydrophobic, and this can make getting good grids very challenging. And so, it's common practice before grids are used to do what's called glow-discharging to restore their hydrophilicity. Glow-discharging is done in a device, very much like a carbon evaporator, in fact, some are dual purpose devices, but imagine in a glow discharger. Again, there's a platform upon which your grid rest. And the grid is covered with a layer of carbon. But this time, instead of a rod being evaporated, there's a plate here and a plate here and a big voltage difference between those two plates. And again a vacuum is pulled in this chamber. Because the vacuum isn't actually perfect you'll have still a number of water molecules from the atmosphere that was in the chamber as the situation was set up. And there's always ionizing events that happen. You know cosmic rays can come into this chamber and cut one of these bonds, resulting in a hydroxyl ion and a proton being liberated. And because of the voltage gradient between these two plates, the proton might be accelerated towards the upper plate, while the hydroxyl ion is accelerated down towards the lower plate. And so you have a situation where the carbons surface is hit with a barrage of charged ions. And as these charged ions hit the surface, it's thought that they produce alcohols and carboxylic acid groups as well as aldehydes, ketones, and esters. And perhaps all kinds of different species that are now hydrophilic. And so, after glow discharging it's seen that the grade, the grids behave much better, much like after they were recently made. For certain samples, it's been seen to be advantageous to, after the vacuum is pulled, to leak in through a valve here attached to some kind of vial. Some other substance, for instance, polyamines have been used. And so when you have a, a, s, a low concentration of polyamines here, that are ionized and accelerated towards the grid. That can give the grid a subtly different surface chemistry. Finally, another variant of this device is called a plasma cleaner. And plasma cleaning has the same overall goals. Now unfortunately, if you have a grid or the grid bars are made of say, copper, so, here, for instance, is a grid bar, here's a gird bar, here's a grid bar. Grid bars are made of copper as is the entire outside of the grid as well. But you have a thin layer of carbon over those grid bars. And so here is the thin layer of carbon between the grid bars. And you plunge freeze such a grid. Unfortunately, carbon and copper have very different coefficients of thermal contraction. And so as they rapidly cool down, the carbon contracts less than the copper. And the result has been called cryo-crinkling. And it's seen here, where the size of each grid square is now made quite a bit smaller than the carbon surface that originally covered it. And so the carbon ends up crinkling and rippling. And this can be a severe disadvantage for say, tooty crystallography as we'll see later. And this is why malibdomin grids have been used before because the coefficients of thermal expansion contraction for malibdomin and carbon are much closer to the same. Most recently, Laurie Passmore has developed methods to create the entire grid out of gold, both the bars and also the surface layer with the holes in it, so that it has exactly the same coefficient of shrinking. It has been found that this reduces beam due specimen movement.
