11.2.
THEORETICAL ASPECTS
347
optimal visualization of the specimen presupposes that it be thin enough
so that it has good electron transparency, and that the film in which it is
embedded be of approximately the same thickness, or perhaps somewhat
less. The embedding material should be as dense as possible to achieve
maximum contrast, and it should be essentially amorphous, even at very
high resolution. Also, if the embedding material is to fill the tiniest
cracks and crevices in the specimen, it must consist of relatively small
molecules which must wet the specimen. It is also obvious that the
embedding material must not act as too powerful a "stain" for the objects
to be studied or contrast would be lost.
One might think that many compounds containing atoms of high
atomic density would reasonably fulfill these requirements. However, a
good negative stain must also be at least fairly soluble even though the
solutions used may be only from 0.5-2%. It is important that the stain
does not crystallize until the last possible moment during drying. Thus
the staining molecules must remain in solution, perhaps in a supersaturated state, until drying is essentially complete. Thus, Valentine and
Home (1962) point out that thallium carbonate is a poor stain in spite of
its very great density (anhydrous density, 7.1), and although it has a
solubility of 4%. However, apparently it is impossible to avoid its crystallization which effectively prevents its use as a good negative stain.
These authors also emphasize the importance of solubility by comparing
the staining properties of sodium dihydrogen phosphate, which is very
soluble, with the far less soluble disodium hydrogen phosphate. The
former gives a good negative stain (although, of course, of low contrast).
The latter gives an unsatisfactory granular background.
A substance to be useful as a negative stain must also withstand electron
bombardment in the microscope. Thus it must not volatilize or "run."
The special virtue of a good negative stain lies in the fact that its
anhydrous density can be three or more times greater than the weight
density of biological material embedded in it. Valentine and Home
(1962) indicate theoretical reasons for believing that under such circumstances particles of approximately 33 A would be visible. Identical calculations make it clear that even the most intense positive staining cannot be
expected to reveal small objects as well, for it is hard to imagine that
positive staining could ever do much more than double the weight density
of a biological substance. Under these hypothetical circumstances an
object of 50 A would have the same contrast differential as the 33 A
particle negatively stained. In actual practice this seems to be a very
THEORETICAL ASPECTS
347
optimal visualization of the specimen presupposes that it be thin enough
so that it has good electron transparency, and that the film in which it is
embedded be of approximately the same thickness, or perhaps somewhat
less. The embedding material should be as dense as possible to achieve
maximum contrast, and it should be essentially amorphous, even at very
high resolution. Also, if the embedding material is to fill the tiniest
cracks and crevices in the specimen, it must consist of relatively small
molecules which must wet the specimen. It is also obvious that the
embedding material must not act as too powerful a "stain" for the objects
to be studied or contrast would be lost.
One might think that many compounds containing atoms of high
atomic density would reasonably fulfill these requirements. However, a
good negative stain must also be at least fairly soluble even though the
solutions used may be only from 0.5-2%. It is important that the stain
does not crystallize until the last possible moment during drying. Thus
the staining molecules must remain in solution, perhaps in a supersaturated state, until drying is essentially complete. Thus, Valentine and
Home (1962) point out that thallium carbonate is a poor stain in spite of
its very great density (anhydrous density, 7.1), and although it has a
solubility of 4%. However, apparently it is impossible to avoid its crystallization which effectively prevents its use as a good negative stain.
These authors also emphasize the importance of solubility by comparing
the staining properties of sodium dihydrogen phosphate, which is very
soluble, with the far less soluble disodium hydrogen phosphate. The
former gives a good negative stain (although, of course, of low contrast).
The latter gives an unsatisfactory granular background.
A substance to be useful as a negative stain must also withstand electron
bombardment in the microscope. Thus it must not volatilize or "run."
The special virtue of a good negative stain lies in the fact that its
anhydrous density can be three or more times greater than the weight
density of biological material embedded in it. Valentine and Home
(1962) indicate theoretical reasons for believing that under such circumstances particles of approximately 33 A would be visible. Identical calculations make it clear that even the most intense positive staining cannot be
expected to reveal small objects as well, for it is hard to imagine that
positive staining could ever do much more than double the weight density
of a biological substance. Under these hypothetical circumstances an
object of 50 A would have the same contrast differential as the 33 A
particle negatively stained. In actual practice this seems to be a very