vented into the liquid nitrogen dewar to prevent frost formation
that may potentially block gas flow and to prevent depositing liquid
nitrogen in the lab environment.
Within the chamber, one of the lines of cooled nitrogen gas is
passed through the lower part of the stage. The specimen is located
on the upper part of the stage and is cooled by thermal conduction
with the lower stage. The second line is passed through the cryoshield, a separate anti-contamination fixture that has a large surface
area to adsorb residual water molecules in the chamber. The stage
and shield nitrogen flow rates are regulated separately from the
input nitrogen line by the dedicated flow controller. Temperatures
are monitored and recorded by computer software through thermocouples embedded in the stage and shield.
1.2 Considerations
for Sample Milling
Angle and Orientation
The target milling angle depends on the specimen. Typically, a large
milling angle relative to the grid surface will result in shorter
lamellae (i.e., measured from the front edge to the back edge)
due to reduced cross-sectional area (Fig. 2d). High milling angles
may be useful to create lamellae in thicker specimens or in specimens where it is difficult to identify a cell at low angles. Conversely,
low milling angles result in longer lamellae, but may be more prone
to obstruction from surrounding material or may make it difficult
to identify a cell due to the nearly parallel viewing angle. Practically,
it is best to mill lamellae at as low an angle (i.e., as parallel to the
substrate as possible) for two reasons:
1. The lamella angle limits the tilt range of tomography in the
TEM, contributing to resolution anisotropy [32].
2. Milling at low angles results in long lamellae, maximizing the
usable area for tilt-series acquisition.
The lower limit to milling angle is set by sample geometry–at
low angles, the edge of the autogrid or the grid bars will block the
beam from reaching the grid square surface. This lower limit is
about 11
stage tilt, corresponding to a beam incident angle of
about 4
with respect to the grid surface. The upper limit to milling
angle is set by the TEM stage. Due to the pre-tilt of the lamella, in
one tilt direction on the TEM stage, the apparent sample thickness
will be greater than compared to the other direction,
corresponding to a greater relative tilt between the lamella and
the beam. The apparent thickness of the sample increases proportionally to 1/cosine of the stage angle and at high TEM stage tilt
angles, the sample will become too thick to image. If the initial
lamella angle is too high, it will unnecessarily limit the range of
TEM stage angles for tilt series acquisition. This upper limit for
lamella milling is typically around 22
on the SEM stage,
corresponding to an incident angle of about 15
(see Note 1). In
56
Vinson Lam and Elizabeth Villa
that may potentially block gas flow and to prevent depositing liquid
nitrogen in the lab environment.
Within the chamber, one of the lines of cooled nitrogen gas is
passed through the lower part of the stage. The specimen is located
on the upper part of the stage and is cooled by thermal conduction
with the lower stage. The second line is passed through the cryoshield, a separate anti-contamination fixture that has a large surface
area to adsorb residual water molecules in the chamber. The stage
and shield nitrogen flow rates are regulated separately from the
input nitrogen line by the dedicated flow controller. Temperatures
are monitored and recorded by computer software through thermocouples embedded in the stage and shield.
1.2 Considerations
for Sample Milling
Angle and Orientation
The target milling angle depends on the specimen. Typically, a large
milling angle relative to the grid surface will result in shorter
lamellae (i.e., measured from the front edge to the back edge)
due to reduced cross-sectional area (Fig. 2d). High milling angles
may be useful to create lamellae in thicker specimens or in specimens where it is difficult to identify a cell at low angles. Conversely,
low milling angles result in longer lamellae, but may be more prone
to obstruction from surrounding material or may make it difficult
to identify a cell due to the nearly parallel viewing angle. Practically,
it is best to mill lamellae at as low an angle (i.e., as parallel to the
substrate as possible) for two reasons:
1. The lamella angle limits the tilt range of tomography in the
TEM, contributing to resolution anisotropy [32].
2. Milling at low angles results in long lamellae, maximizing the
usable area for tilt-series acquisition.
The lower limit to milling angle is set by sample geometry–at
low angles, the edge of the autogrid or the grid bars will block the
beam from reaching the grid square surface. This lower limit is
about 11
stage tilt, corresponding to a beam incident angle of
about 4
with respect to the grid surface. The upper limit to milling
angle is set by the TEM stage. Due to the pre-tilt of the lamella, in
one tilt direction on the TEM stage, the apparent sample thickness
will be greater than compared to the other direction,
corresponding to a greater relative tilt between the lamella and
the beam. The apparent thickness of the sample increases proportionally to 1/cosine of the stage angle and at high TEM stage tilt
angles, the sample will become too thick to image. If the initial
lamella angle is too high, it will unnecessarily limit the range of
TEM stage angles for tilt series acquisition. This upper limit for
lamella milling is typically around 22
on the SEM stage,
corresponding to an incident angle of about 15
(see Note 1). In
56
Vinson Lam and Elizabeth Villa