Page 47 - Chip Scale Review_November December_2021-digital

image of 3μm, 3.5μm and 4μm isolated
Figure 5: a) Bossung plot generated from FEM data showing less than 10% deviation over 60μm DOF; b)
Lower resolution image of 3μm, 3.5μm and 4μm isolated and dense area line/space arrays; c) Cross-section
image of 3µm lines in 10µm thick dry film resist on copper substrate; the line critical dimension is 3.181µm, and
the resist height is 9.873µm in the cross-sectional image.
Figure 6: a) Bossung plot generated from FEM data showing less than 10% deviation over 40μm DOF for 5μm
lines and 70μm DOF for 6μm lines. b) Lower resolution image of 4.5μm, 5μm, 6μm and 7.5μm isolated and
dense area line/space arrays. c) Higher resolution cross-sectional images of 5μm and 6μm lines in a 10μm-thick
dry resist on copper substrate.
Chip Scale Review
and dense line/space arrays. A higher
resolution cross-sectional image of 3μm
lines (Figure 5c) shows dimensions for
the middle line: 3.181µm line width and
9.873µm line height (resist thickness).
5µm and 6µm lines. Larger feature
sizes were also investigated. A CCL/
ABF substrate with a 25µm-thick dry film
resist was selected for this demonstration,
resulting in lines with an aspect ratio of
about 1:5. Best dose and best focus were
determined using FEM. Best dose was
used for the resolution demonstration.
The 5μm line CDs showed less than 10%
deviation from -40μm to -80μm, and a
DOF of 40μm. The 6μm line CDs showed
less than 10% deviation from -30μm to
-100μm, and a DOF of 70m. The data from
the FEMs were used to generate Bossung
plots (Figure 6a). The plots show a 40μm
DOF for 5μm lines and a 70μm DOF for
6μm lines. Figure 6b also includes a lower
magnification image of 4.5μm, 5μm, 6μm
and 7.5μm isolated and dense line/space
arrays and higher resolution cross-sectional
images of 5μm and 6μm lines in a 10μm-
thick resist (Figure 6c).
15µm and 20µm square vias. Via
resolution was also investigated (Figure 7).
Best dose and best focus were determined
using a FEM and a CCL/ABF substrate
with 40µm-thick dry film resist; best
dose was selected for this demonstration.
Bossung plots were generated for both via
sizes. The 15μm vias showed less than 10%
deviation from -30μm to 80μm, and a DOF
of 110μm. The 20μm vias showed less than
10% deviation from -40μm to 110μm, and
a DOF of 150μm.
Uniformity. We used a 1.4µm-thick
liquid resist film on a 510mm x 515mm
glass panel and 3µm lines to test uniformity
across the panel. The uniformity data in
Figure 8 show a maximum CD of 3.258μm,
a minimum CD of 2.988μm and an average
CD of 3.099μm. Deviation ranges from
-0.20% to 4.12% for an overall uniformity
of 4.32%. The deviation chart shows no
peaking or trending and indicates a stable
exposure field.
Overlay. Overlay accuracy is essential.
We used a 510mm x 515mm glass panel
with a 1.4µm-thick liquid resist as the
overlay test vehicle. The exposure field was
250µm x 250µm. Four shots covered the
entire panel. The test procedure comprised
the deposition and patterning of an initial
layer, followed by deposition and patterning
of a second, overlying layer. Patterning
45
November • December • 2021 [ChipScaleReview.com] 45