How Thick Should a Section Be? Paraffin, Frozen and Semi-Thin

The same tissue cut at 2 micrometres and at 10 micrometres looks like two different things under the microscope. The thin one is crisp, with clear nucleoli. The thick one is layered and hazy, as if viewed through fog. Section thickness is the most basic and most often overlooked parameter in pathology and research.

1. Thickness decides how much you can see through

Microscopy has an inescapable physical limit: depth of field. A 40X, NA 0.65 objective has a depth of field of roughly plus or minus 1 micrometre, and a 100X oil lens only about plus or minus 0.2 micrometres.

Any section appreciably thicker than that brings information from outside the focal plane into the image as background haze. The thicker the section, the denser the haze and the lower the contrast.

Choosing a thickness is therefore a trade-off between seeing one clean optical plane and seeing the wider architecture. Subcellular detail wants thin; tissue relationships can tolerate thicker.

2. Four methods, four thickness ranges

Paraffin sections are the workhorse. Routine H&E pathology uses 3 to 4 micrometres, thin enough that nuclei do not overlap and staining stays even. The workflow is fixation, dehydration, clearing, infiltration, embedding and sectioning, taking two to three days. Morphology is best preserved, blocks archive for decades, and serial sections allow multiple stains. The limitation is that solvents remove lipids and destroy all enzyme activity.

Frozen sections run 5 to 10 micrometres. Tissue is snap-frozen without dehydration and cut in a cryostat. The great advantage is speed: results in twenty to thirty minutes, which is why intraoperative frozen diagnosis relies on it. It also preserves lipids and some enzyme activity, making it mandatory for fat staining and enzyme histochemistry. The cost is ice crystal damage to cells, morphology inferior to paraffin, and less uniform thickness.

Semi-thin sections run 0.5 to 2 micrometres. Resin embedded and cut with glass or diamond knives, they serve mainly to locate regions before electron microscopy and for optical work needing very high resolution. Organelles are clearly resolved, but the workflow is long and expensive and unsuited to high throughput.

Vibratome sections run 50 to 300 micrometres. Tissue is neither fixed nor embedded; a vibrating blade cuts it under buffer, leaving cells alive. This suits electrophysiology, brain slice culture and floating immunohistochemistry where antigenicity must be preserved.

3. Choosing by purpose

Routine H&E diagnosis: paraffin, 3 to 4 micrometres, the default.

Intraoperative benign versus malignant: frozen, 5 micrometres, slightly thicker to keep the section intact within limited time.

Immunohistochemistry: paraffin at 3 to 4 micrometres, or vibratome sections at 40 to 50 micrometres for floating staining. The latter reveals three-dimensional fibre tracts.

Fluorescence and confocal: 5 to 10 micrometres, or a Z stack reconstructed in three dimensions. Too thick and out-of-focus fluorescence contaminates the signal.

Electron microscopy: 60 to 70 nanometres, beyond optical microscopy entirely, requiring an ultramicrotome and diamond knife.

Organoids and live tissue: vibratome sections of 100 to 300 micrometres with light-sheet imaging for long-term observation.

4. Quality is not only about thickness

Even with the right setting, sections can fail for other reasons.

Incomplete or excessive freezing. Samples for frozen sections must be snap-frozen in liquid nitrogen or isopentane. Slow freezing forms large intracellular ice crystals that rupture structure, while over-freezing makes tissue brittle.

Knife angle and sharpness. Clearance angle is usually 3 to 8 degrees; too much compresses tissue, too little fails to cut cleanly. A dull blade produces folds and chatter marks.

Cutting speed. Too fast causes compression and static attraction; too slow lets tissue rebound and thickness varies.

Temperature. Warm rooms soften paraffin blocks so sections curl. A wrong cryostat temperature means either the block will not cut or it shatters.

5. Judging whether the section is right

Scan at low power first: is the tissue complete, free of folds and knife marks? Then move to 40X and check three things: nuclei in a single layer without overlap, nuclear membranes and nucleoli clearly defined, and no background haze.

Nuclei clumped into a mass usually means too thick. Fragmented tissue with pale staining usually means too thin or poor fixation.

6. The short version

Section thickness is not a fixed number but a parameter tied to your purpose. To study architecture, err on the thick side rather than fragment the tissue. To study cellular detail, err thin rather than stack. Get that logic straight and most sectioning problems resolve themselves.