Best Student Poster, EIPBN-2008, has been awarded to Thomas Reisingerfor his contribution, "Neutral Helium Microscopy."

Best Student Paper, EIPBN-2008, has been awarded to Chih- Hao Chang for his contribution, "Spatial-frequency Multiplication with Multilayer Interference Lithography."

Neutral Helium Microscopy

Thomas Reisinger,[1] Stefan Rehbein,[2] Gunter Schmah,[3] and Bodil Holst[1]
1.University of Bergen, Department of Physics and Technology, Allegaten 55, N-5007 Bergen
2.BESSY m.b.H., Albert-Einstein-Strasse 15, 12489 Berlin, Germany
3.University of Gottingen, Institute for X-ray Physics,
Friedrich-Hund-Platz 1, 37077 Gottingen, Germany
(Dated: April 29, 2008)

In this article we present new images using a beam of neutral low energy helium as
an imaging probe. Helium microscopy is still in its infancy. The very rst image was
published earlier this year.[1] The new microscopy technique potentially offers signifcant
advantages. Firstly, the atoms are neutral, which means that insulating surfaces can be
investigated without any coating. Furthermore, the energy of the beam (a few tens of meV
for a wavelength of about 1 A) is very low, a factor 1000 less than electrons for a similar
wavelength. This means that fragile samples can be investigated without any damage.
Helium atoms are particularly good at imaging light adsorbates i.e. hydrogen which could
be very important for fuel cell development.

The images were recorded in transmission mode using a zoneplate to focus the beam
(Fig.2). The focus width was 2 um. The images presented in this article (Fig.1(b)) show a
carbon holey foil (commercially available from the company Quantifoil). The paper nishes
with a discussion of a next generation helium microscope where nanometer resolution is
envisaged.

1 (a) M. Koch et al., J. Mic. 229, p.1 (2008) (cover story); (b) see also Nature Research Highlights:
Nature 451, p.226-227 (2008)

FIG. 1: The sample shown in the SEM micrograph (a) is a carbon holey foil on top of a copper
mesh (Quantifoil R, R2/1). The holes have a diameter of about 2 um. The two small images (b)
are transmission helium microscopy scans of the same sample. The upper image is 15 um x 15 um
and took 14.5 hours to record. The bottom one is 30 um x 30 um and took 9.7 hours to record.
The beam focus was less than 2:0 um and about 2:3 um, respectively.

FIG. 2: The supersonic expansion helium beam was focused using a freestanding Fresnel zoneplate
made of nickel on a silicon membrane using electron beam lithography. The scanning electron
microscopy image shown in this gure is a close-up of the innermost zones of the zoneplate. The
broad horizontal lines are support bars and the central circular disc, which can be seen partially to
the right of the image, is used to block the zeroth order for an increased contrast. The zone plate
has 2700 zones with an outermost zonewidth of about 50 nm.

Spatial-frequency Multiplication with Multilayer Interference Lithography

Chih-Hao Chang,* Y. Zhao, R. K Heilmann, and M. L. Schattenburg Space Nanotechnology Laboratory, Massachusetts Institute of Technology, Cambridge, MA 02139

We report progress in achieving large-area spatial-frequency multiplication using
multilayer interference lithography (IL). In this fabrication scheme, multiple grating layers with
a grating period p are patterned with different phase offsets and etched onto a single layer,
effectively dividing the period [1] by an integer number. The linewidth in each grating layer is
controlled with nm-accuracy by high-pressure plasma etching and an image-reversal process.
Using this process, gratings with deep sub-diffraction-limited periods can be achieved with IL
using a 351.1 nm laser.

There are several key challenges to make this process feasible, among which the most
important are linewidth control and overlay accuracy. Linewidth control is important to ensure
that the grating linewidths are consistent in every layer. This condition is essential in
suppressing the fundamental spatial-frequency. In addition, for higher factor frequency
multiplication, the duty-cycle of each layer needs to be as high as possible. High duty-cycle and
linewidth repeatability are achieved through isotropic plasma etching and a novel image-reversal
process, which is shown in Figure 1. Using this process we are able to achieve a duty-cycle of
~0.9 with nm linewidth repeatability using a 200 nm-period grating. Using such high duty-cycle
gratings spatial-frequency multiplication with factors of 4-5 can be theoretically achieved.
High overlay accuracy is achieved by measuring the phase of a reference grating using
homodyne interferometry before patterning each grating layer. By keeping the reference grating
distortion-free in the fabrication process, the overlay accuracy is limited by the repeatability of
our metrology system, the Nanoruler [2]. In previous efforts [3] we have demonstrated overlay
accuracy of better than 3 nm over a 40 x 40 mm2 area with 574 nm period grating layers. Figure
2(a) demonstrates that we now have achieved similar overlay accuracy of 200 nm-period resist
and silicon nitride gratings.

The preliminary result of this process is illustrated in Figure 2(b), where the low dutycycle
resist grating shown in Figure 1(a) has been image-reversed, resulting in a 100 nm-period
grating. The linewidths of the etched grating is on the order of 20-25 nm, and allow the addition
of two more layers to achieve spatial-frequency quadrupling. Note that one set of etched lines
(2nd layer) are straighter than the other (1st layer) — this is because the two layers were etched
with different RIE processes. This will be discussed in more detail.
In this paper we present a multilayer interference lithography process that can achieve
large-area spatial-frequency multiplication. Using this process the optical diffraction-limited
problem can be solved with high precision metrology and well-controlled fabrication techniques.
An image-reversal process that was used to achieve high duty-cycle and nm linewidth control
will be highlighted. These results that we present are important in moving towards our goal of
four-fold spatial-frequency multiplication.

[1] S. R. J. Brueck, Proceedings of the IEEE 93, 10 (2005).
[2] P. T. Konkola et al., J. Vac. Sci. Technol. B 21, 3097 (2003).
[3] Y. Zhao et al., J. Vac. Sci. Technol. B 25, 2439 (2007).

* chichang@mit.edu

Figure 1. Image-reversal process for achieving high duty-cycle grating. Starting with (a) a
low duty-cycle polymer grating, (b) a silicon-containing polymer is spun over the pattern.
Several RIE steps are used to (c) expose the polymer, and (d) transfer the pattern into nitride.

Figure 2. Top-view SEM of (a) 200 nm-period resist grating (narrow light lines) patterned on a
200 nm-period silicon nitride grating (narrow dark lines) with a .-phase offset. (b) The resist
grating is image-reversed and etched, resulting in a 100 nm-period grating in nitride.