Ultramicrotome · Lobeworks/17

An ultramicrotome is a cutting machine that slices a block of resin-embedded tissue into sections a few tens of nanometres thick, thin enough for electrons to pass through, so that a transmission electron microscope can photograph them one after another.

Ultramicrotome. An ultramicrotome is a cutting machine that slices a block of resin-embedded tissue into sections a few tens of nanometres thick, thin enough for electrons to pass through, so that a transmission electron microscope can photograph them one after another.

The block, a speck of brain set in hard epoxy, is held on an arm that swings it down past the edge of a knife, usually a diamond ground to an edge a few nanometres wide. Between strokes the arm advances by the thickness of the next section, a step made by letting a metal rod lengthen as it warms or by a piezoelectric crystal that grows when a voltage is applied. Behind the edge the knife carries a small trough of water, and each section slides off onto it and floats, still joined to the previous one, so the cut tissue leaves as a ribbon; the colour the sections reflect, grey, silver or gold, tells their thickness.

The sections are picked up on grids. A grid is a metal disc about 3 mm across with a slot covered by a plastic film tens of nanometres thick; the fly brain of 2018 needed 7,062 sections of about 40 nm, three to a grid.

Collection became automatic. GridTape replaced the grids with a long tape of slotted film and collects more than 4,000 sections a day, over ten times the rate by hand.

One lost section is a hole in the volume. Every neuron that crossed it has to be traced across the gap, which is why throughput and reliability matter as much as thinness.

The thickness sets the depth resolution. A 40 nm section averages the tissue over 40 nm, so its voxels are bricks ten times taller than the 4 nm pixels are wide; FIB-SEM avoids this by milling instead of cutting.

A ribbon of brain on water.

Seen from above, a working ultramicrotome looks like a conveyor belt for soap films: grey strips of plastic, each one a two-thousandth of a hair thick, sliding off a diamond onto a drop of water in a single line.

Questions: When is milling a brain with ions better than cutting it with a diamond? When the fine processes run in every direction and the volume is small enough to wait for. A diamond knife cuts sections about 40 nm thick, so a neurite running nearly parallel to the cut is smeared across one thick slice; an ion beam mills a few nanometres at a time, giving voxels of 8 nm on every side. The price is speed, the loss of the sample as it is imaged, and a depth limit of about 100 µm, which is why whole fly brains are first split into 20 µm slabs with a heated knife and milled on several machines at once. Why must a section be so thin for an electron microscope? Because the electrons of a transmission microscope have to pass through it, and in thick material they scatter again and again until the image is a blur. Sections of a few tens of nanometres let most electrons through, scattered once by the heavy metal on membranes and straight elsewhere, which is the contrast the image is made of. The thinness also sets the depth resolution: the fly brain of 2018 was cut into 7,062 sections of about 40 nm, so each image averages 40 nm of tissue.