ArticlePDF Available

Abstract and Figures

Field emission devices are promising candidates to replace silicon FinFETs as next-generation nanoelectronic components. For these devices to be adopted, nanoscale field emitters with nanoscale gaps between them need to be fabricated, requiring the transfer of e.g. sub-10 nm patterns with sub-20 nm pitch into substrates like silicon and tungsten. New resist materials must therefore be developed that exhibit the properties of sub-10 nm resolution and high dry etch resistance. A negative tone, metal–organic resist is presented here. It can be patterned to produce sub-10 nm features when exposed with helium ion beam lithography at line doses on the order of 10s of pC/cm. The resist was used to create 5 nm wide, continuous, discrete lines spaced on a 16 nm pitch in silicon, and 6 nm wide lines on 18 nm pitch in tungsten, with line edge roughness of 3 nm. After the lithographic exposure, the resist demonstrates high resistance to silicon and tungsten dry etch conditions (SF6 and C4F8 plasma), allowing the pattern to be transferred into the underlying substrates. The resist’s etch selectivity for silicon and tungsten was measured to be 6.2:1 and 5.6:1, respectively; this allowed 3-4 nm thick resist films to yield structures that were 21 and 19 nm tall, respectively, while both maintained sub-10 nm width on sub-20 nm pitch.
Content may be subject to copyright.
Plasma-Etched Pattern Transfer of Sub-10 nm Structures Using a
MetalOrganic Resist and Helium Ion Beam Lithography
Scott M. Lewis,*
Matthew S. Hunt,
Guy A. DeRose,
Hayden R. Alty,
Jarvis Li,
Alex Wertheim,
Lucia De Rose,
Grigore A. Timco,
Axel Scherer,
Stephen G. Yeates,
and Richard E. P. Winpenny*
School of Chemistry and Photon Science Institute, The University of Manchester, Oxford Road, Manchester M13 9PL, United
Department of Applied Physics and Materials Science and the Kavli Nanoscience Institute, California Institute of Technology, 1200
East California Boulevard, MC 107-81, Pasadena, California 91125, United States
ABSTRACT: Field-emission devices are promising candidates
to replace silicon neld-eect transistors as next-generation
nanoelectronic components. For these devices to be adopted,
nanoscale eld emitters with nanoscale gaps between them need
to be fabricated, requiring the transfer of, for example, sub-10
nm patterns with a sub-20 nm pitch to substrates like silicon and
tungsten. New resist materials must therefore be developed that
exhibit the properties of sub-10 nm resolution and high dry etch
resistance. A negative tone, metalorganic resist is presented
here. It can be patterned to produce sub-10 nm features when
exposed to helium ion beam lithography at line doses on the
order of tens of picocoulombs per centimeter. The resist was
used to create 5 nm wide, continuous, discrete lines spaced on a
16 nm pitch in silicon and 6 nm wide lines on an 18 nm pitch in tungsten, with line edge roughness of 3 nm. After the
lithographic exposure, the resist demonstrates high resistance to silicon and tungsten dry etch conditions (SF6and C4F8
plasma), allowing the pattern to be transferred to the underlying substrates. The resists etch selectivity for silicon and tungsten
was measured to be 6.2:1 and 5.6:1, respectively; this allowed 3 to 4 nm thick resist lms to yield structures that were 21 and 19
nm tall, respectively, while both maintained a sub-10 nm width on a sub-20 nm pitch.
KEYWORDS: Metalorganic resist, ion beam resist, helium ion beam lithography, high-resolution pattern, high dry etch resistance
The ability of integrated circuit technology to follow
Moores law has depended on the continuous reduction
in the size of eld-eect transistors (FETs), rst in the planar
metaloxidesemiconductor eld-eect transistor (MOS-
FET) architecture and now more recently in the 3D n
eld-eect transistor (FinFET) architecture. This has been
accomplished by reducing the FETs channel length, width,
and gate oxide thickness and by changing the gate dielectric
material according to Dennards scaling rules.
these scaling rules have begun to break down because as the
gate length is reduced to dimensions of 32 nm or smaller, the
supply voltages need to be scaled down as well, but doing so
does not provide enough voltage to turn on the pn junction.
Furthermore, the power density in the newest microprocessors
has become so large that powering all transistors simulta-
neously would rapidly exceed the thermal power budget for the
chip, resulting in diminished performance, decreased lifetime
and, eventually, permanent device failure. Overheating can be
addressed by powering 50% of transistors on a single chip on a
single clock cycle,
but this presents a signicant technical
design challenge. Considering these problems together, it has
been predicted by the International Technology Roadmap for
Semiconductors (ITRS) that it will no longer be economically
feasible to decrease FET device dimensions past the 7nm
thus imbuing a sense of uncertainty on the future
direction of the semiconductor industry.
Field-emission devices are promising candidates to succeed
silicon FinFETs because they can operate in high-power-
density regimes where chip temperatures can reach 300 °C.
Solid-state transistors fail in this regime because the pn
junctions functionality is lost when electrons in the p-doped
regions are thermally excited to the same conduction electron
concentration as that in the n-doped regions.
eld-emission devices remain operational because as the
temperature is elevated, the current remains exponentially
dependent on the eld until the temperature is sucient to
initiate thermionic emission, which usually occurs hundreds of
degrees above 300 °C.
These devices are also attractive
because they are capable of operating at frequencies of
hundreds of gigahertz; this has been achieved by fabricating
150 nm vacuum gaps using optical lithography and resist
Other researchers recently demonstrated that when
Received: May 9, 2019
Revised: July 20, 2019
Published: August 19, 2019
Cite This: Nano Lett. 2019, 19, 60436048
© 2019 American Chemical Society 6043 DOI: 10.1021/acs.nanolett.9b01911
Nano Lett. 2019, 19, 60436048
Downloaded via CALIFORNIA INST OF TECHNOLOGY on September 2, 2020 at 18:07:04 (UTC).
See for options on how to legitimately share published articles.
sub-50 nm emittercollector gaps were fabricated, electric
elds high enough for eld emission could be achieved at <10
V; the devices were complementary metaloxidesemi-
conductor (CMOS)-compatible, functional at atmospheric
pressure, and able to be independently gated on a single
integrated chip.
Turn-on voltages can be further reduced by
fabricating sharper emitters with smaller emittercollector
gaps, incentivizing the creation of new fabrication techniques
that yield tightly spaced, sub-10 nm structures.
Whereas electron beam lithography (EBL) oers high-
resolution patterning to create sub-10 nm structures in the
it is dicult to pattern these with high density, for
example, with sub-10 nm wide lines spaced <20 nm apart. This
is because secondary, Auger, and backscattered electrons (SEs,
AEs, and BSEs) scatter in between the nanostructures during
patterning, which exposes the resist in that space, resulting in
bridges between lines following the development process.
This proximity eectlimits the resolution of the pattern that
can be produced. To alleviate this issue, a technique that uses a
focused helium ion beam instead of an electron beam has been
explored over the past decade. Previous helium ion beam
lithography (HIBL) studies have demonstrated a reduced
proximity eect
owing to less backscattering, the smaller
interaction volume with the substrate, and the subnanometer
beam diameter,
resulting in sub-10 nm resolution.
This is
accompanied by orders of magnitude higher resist sensitivity
than can be achieved with EBL
due to a higher SE yield per
incident helium ion compared with each incident electron.
Once the pattern has been dened in the resist by
lithography, it must be transferred to the underlying material,
which is often done using inductively coupled plasma reactive-
ion etching (ICPRIE). The most common metal used to
produce eld-emission devices is tungsten, which exhibits a low
work function and has a high thermal conductivity, preventing
the device from being destroyed via Joule heating.
ferring the desired nanoscale pattern (e.g., sub-10 nm
structures with sub-10 nm gaps in between) to tungsten is a
challenge because the probability of landing ions in ever
smaller gaps becomes ever lower. This leads to a decrease in
etch eciency, which inherently decreases the etch rate and
selectivity. To increase the etching eciency, the ICP forward
power must be increased, but this also increases the etch rate
of the resist. The thickness of the resist would then need to be
increased to achieve the desired etch depth, which would
require a higher dose, which, in turn, would reduce the
resolution of the pattern. To avoid this problem, one may use a
hard mask to withstand the aggressive nature of the plasma
but doing this introduces more processing steps and
leads to higher production costs. Another route is to enhance
the etch selectivity of the resist by introducing into the
molecular chemistry a metal species that eectively oxidizes
upon lithographic exposure to become the hard mask. This has
previously been demonstrated by our group using supra-
molecular Ni- and Cr-containing assemblies while maintaining
a sub-10 nm patterning capability,
albeit at a relatively low
pattern density compared with what is needed for modern
In this Letter, a metalorganic, negative tone resist
candidate, Cr8F8(O2CtBu)16 (Figure 1), rst introduced by
our group in ref 15 and henceforth denoted as
Cr8F8(pivalate)16, is presented. It is formed by the binding of
eight chromium atoms (in green in Figure 1) in a ring-like
structure, with an exterior composed entirely of tert-butyl
groups (pivalates).
The pivalates provide a high solubility in
nonpolar solvents, which allows the resist molecule to be
dissolved in hexane and spun onto substrates (e.g., Si and W).
The molecule achieves high-resolution patterning because it is
simultaneously low density (ρ= 1.212 g cm3), meaning that it
does not have many lateral scattering centers for the
lithography beam to interact with as it travels through, and
has a high molecular weight (2192 g mol1), meaning that the
number of resist molecules that are required to produce a thin
lm is signicantly reduced, leading to a high-resolution
pattern. Upon exposure, SEs and AEs break carbon bonds in
the resist, liberating some C and O atoms while permitting
other O and Cr atoms to react to form a chromiumoxide
hard mask that is particularly resistant to the ICPRIE
chemistry used to etch both silicon and tungsten.
Prior to the spin-on application of the Cr8F8(pivalate)16
resist, atomic force microscopy (AFM) was used to evaluate
the surface morphology of silicon and tungsten substrates
(Figure 2a,b). The root-mean-square (RMS) roughness was
measured to be 0.29 nm for silicon. For tungsten, which was
sputter-deposited onto silicon as a 100 nm thick lm on top of
a 5 nm sputter-deposited titanium adhesion layer, the RMS
roughness was 0.42 nm. The tungsten was 45% rougher than
silicon; topographical contrast revealed that the lm was
composed of nanograins that individually were 5 nm wide
and as long as 50 nm. For all sputter processes, wafers were
rst cleaned inside the chamber with argon plasma, and targets
were presputtered for 60 s to remove surface oxides.
The exact nature of the resist lm in this stage is uncertain.
Previous studies of similar compounds sublimed onto gold
show that an ordered monolayer forms,
but subsequent
layers are not ordered because there are only weak van der
Waals interactions between the molecules of metal rings. The
lms formed here, shown here by AFM (Figure 2c,d) to be
3.5 nm thick (approximately two layers), are therefore
amorphous. The resist is monodispersed and in some ways
resembles the molecules studied by Ober and coworkers
form molecular glasses rather than conventional polymeric
resists. We have not observed a glass-transition temperature
because Cr8F8(pivalate)16 sublimes before such a transition is
observed. This low sublimation temperature is again due to the
very weak intermolecular forces within the resist lms.
Samples were created by dicing wafers into 20 mm ×20 mm
pieces. Both substrate types, bare silicon and tungsten-coated
silicon, provided a smooth enough surface upon which sub-10
Figure 1. Structure of the Cr8F8(pivalate)16 molecule in a ball-and-
stick representation. Chromium atoms are green and uorine atoms
are yellow. Hydrogen atoms are omitted for clarity.
Nano Letters Letter
DOI: 10.1021/acs.nanolett.9b01911
Nano Lett. 2019, 19, 60436048
nm features could be clearly resolved. This is evident in Figure
3, which shows plan-view helium ion microscope (HIM)
images of Cr8F8(pivalate)16 resist nanostructures following
HIBL and the subsequent development in hexane. The
Cr8F8(pivalate)16 resist (30 mg) was dissolved in hexane (3
g); then, the solution was ltered using a 0.2 μm
polytetrauoroethylene syringe lter before being spun onto
substrates with a spin rate of 6000 rpm for 40 s, followed by a
100 °C soft bake for 2 min to evaporate the solvent. The
spacing of adjacent lines (i.e., the pitch) was set to be 1622
nm using a Raith ELPHY MultiBeam pattern generator, which
controlled the helium focused ion beam (35 keV, 0.50 pA) on
a Zeiss ORION NanoFab apparatus. The exposure clearing
dose of the resist on each substrate was determined using a 1D
matrix of single-pixel-wide lines that were 5 μm long. The
width of the line was therefore the width of the ion beam,
which is estimated at 0.5 nm;
the beam step size was 1 nm.
At any pitch, patterns were exposed in sets of 20 lines with one
pass of the beam per line, and the line dose of each set ranged
from 10 to 100 pC/cm with incremental steps of 1 pC/cm.
Following lithography, the resist was developed in hexane for
10 s to dissolve away the unexposed resist, then blown dry with
It can be seen in Figure 3 that discrete, continuous lines
were successfully patterned at all pitches on silicon, with no
bridging between any adjacent lines. On tungsten, patterning
was likewise successful at 18, 20, and 22 nm pitches; at a 16
nm pitch (Figure 3h), the line uniformity was poor and
bridging had occurred, a hallmark of being just beyond the
lithographic resolving limit. The line width, on average, was
measured to be 5.5 nm (standard deviation, σ= 0.9 nm) on
silicon at a 16 nm pitch and 5.6 nm (σ= 0.9 nm) on tungsten
at an 18 nm pitch. The line edge roughness (LER), dened as
3σ, was 3 nm for both sets of Si and W lines. Tungsten
performed slightly worse than silicon in both the minimum
achievable pitch and the minimum line width because tungsten
has a signicantly larger atomic number (Z= 74 for W, Z=14
for Si) and therefore leads to a larger number of SEs and AEs
generated by the primary ion beam; this eect is triggered by
both incident electrons in EBL
and incident He ions in
The ejected SEs can be calculated using the Joy
to have a scattering angle of 80°relative to the
incident beam vector,
which leads to the exposure of the
resist material adjacent to the beams entry point. A similar
mechanism is at play with low -energy ion recoil events
initiated by incident ions, which scatter SEs at the same high
angle in addition to physically displacing atoms.
The more
SEs and AEs that are generated, the wider the exposure radius
is that surrounds the beam entry point, leading to wider lines
and, when the pitch is too small, bridging between them.
Whereas this proximity eect diminishes the smallest
achievable line width and pitch, the generation of more SEs
and AEs also has the benet of decreasing the necessary
exposure dose, which was as much as 1.9 times lower at an 18
nm pitch for tungsten (11 pC/cm) compared with silicon (21
pC/cm). The necessary exposure dose also decreased on
tungsten as a function of decreasing pitch, whereas it did not
for silicon due to the intensity of the proximity eect when
lines are written ever closer to each other on a high-Zmaterial.
On the basis of these results, the outlook for patterning sub-10
nm wide lines on tungsten is that the achievable pitch may be
slightly higher compared with silicon (18 versus 17 nm), in
exchange for nearly half of the exposure dose. It must also be
noted that these HIBL doses are an order of magnitude below
the threshold dose at which He implantation has been shown
to induce dislocation damage in Si.
Figure 2. AFM images of substrates prior to spin-on application of the
resist: (a) silicon and (b) 100 nm tungsten lm (on silicon). (c)
Roughness and (d) thickness of the patterned resist are also
demonstrated by AFM.
Nano Letters Letter
DOI: 10.1021/acs.nanolett.9b01911
Nano Lett. 2019, 19, 60436048
In Figure 4, the resist is shown at its smallest successfully
etched pitch for silicon (17 nm) and tungsten (18 nm), both in
plan view (Figure 4a,d), and when tilted to 87°to better show
its thickness (Figure 4b,e). The same spin settings were used
to apply the resist to both silicon and tungsten. The resist
thickness was measured at the front edge of the tilted lines (the
higher tungsten roughness perhaps accounts for the thinner
measurement) and conrmed by AFM (Figure 2c,d). It should
be noted that the resist had been changed to a chromium
oxide material by the time the lines were imaged as a result of
the lithographic exposure. It has been observed that this resist
can shrink under the exposure of an electron or helium ion
beam; as bonds are broken and carbon and oxygen are
volatilized, the resist lm volume consolidates slightly into the
oxide material. The initial resist thickness, which was not
measured, is therefore necessarily larger than depicted here.
Regardless, the tilted view images in Figure 4 show more
clearly than the plan-view images that the resist structures are
resolvable against the roughness of the substrates beneath
them. The ability to spin the resist into a sub-5 nm thick lm
also helps to reduce the smallest feature size and dose; a
thinner resist yields fewer lateral scattering sites for the
traversing beam and also means that fewer ions are needed to
generate enough SEs and AEs to change the small volume of
resist material into the chromiumoxide material.
It is important to note that when characterizing these
nanostructures, each HIM image was captured via a single scan
of a 600 nm ×600 nm area, meaning that sputtering of the
Figure 3. Plan-view HIM images of lines spaced with pitches of 22, 20, 18, and 16 nm on silicon substrate (ad, respectively) and on a 100 nm
thick tungsten lm that was sputter-deposited onto a silicon substrate (eh, respectively). Average width (w), standard deviation (σ) and line edge
roughness (LER) (3σ) to the nearest 0.1 nm were determined using GenISys ProSEM software.
Figure 4. HIM images of lines spaced with pitches of 17 and 18 nm on silicon substrate (ac) and on a 100 nm thick tungsten lm that was
sputter-deposited onto a silicon substrate (df), respectively. In the top row of images (a,d), developed resist structures are shown in plan view
prior to an ICPRIE etch. In the middle row (b,e), developed resist structures are shown when tilted to 87°prior to the etch. In the bottom row
(c,f), n-like structures are shown following the etch. Measurements to the nearest 0.1 nm were made using GenISys ProSEM software. The LER of
etched Si lines was determined via the plan-view image (not shown); the LER of etched W lines was not determined because the triangular shape of
the cross-section does not lend itself to the LER calculation.
Nano Letters Letter
DOI: 10.1021/acs.nanolett.9b01911
Nano Lett. 2019, 19, 60436048
nanostructure material by the low-current He ion beam (1.0
pA, 30 keV) was negligible; tests were done to show that even
multiple scans of the same area at these settings did not alter
the size of structures. This ensured that HIM imaging could be
used as a nondestructive technique while oering higher
resolution (and higher depth of eld, important for imaging
tilted structures) than, for example, a scanning electron
microscope operated with an immersion lens. Furthermore,
measurements made on AFM and HIM micrographs were
calibrated against images taken of a NIST traceable standard
(50 nm wide gold lines spaced on a 100 nm pitch); the same
microscope settings were used on both the experimental
samples and the standard to ensure measurement accuracy.
After characterizing the resist on substrate, samples for both
silicon and tungsten were subjected to the same 30 s ICPRIE
process (forward power: 20 W RIE, 1200 W ICP) using SF6
and C4F8gases owing at 22 and 35 sccm, respectively. Figure
4c shows that the average width of the resultant silicon ns was
6.4 nm, with an average height of 21 nm; the resist has been
completely etched away. The eective etch rates, based on a
3.4 nm resist thickness (Figure 2d), were therefore determined
to be 0.11 and 0.70 nm/s for the resist and silicon, respectively.
This indicates that silicon etches 6.2 times faster than the
resist when subjected to these etch conditions (i.e., the
selectivity is 6.2:1). For tungsten (Figure 4f), the structures
etched with less of a straight-sidewall n shape and more of an
angled-sidewall triangular shape, suitable for a sharp-tipped
eld emitter. The average width at the top and bottom of the
triangle was 5 and 16 nm, respectively, with an average height
of 19 nm. The resist was completely etched away from the top
of these W structures in 30 s, resulting in etch rates of 0.11 and
0.63 nm/s for the resist and tungsten, respectively (i.e., the
selectivity is 5.6:1). It is impossible to compare these results
directly with other common resists because the etch
selectivities of those resists have not been reported for sub-
20 nm pitches. At a 100 nm pitch, the etch selectivities of
common resists on silicon are 2.0:1 for poly(methyl
methacrylate) (PMMA), 2.9:1 for ZEP520A, and 4.2:1 for
hydrogen silsesquioxane (HSQ).
Etch selectivity is expected
to decrease at a smaller pitch due to the decreasing probability
of landing ions between the features, so the 6.2:1 selectivity for
Si reported here at a 17 nm pitch is especially notable, by
comparison. The improvement in etch performance demon-
strated here by Cr8F8(pivalate)16 on silicon has also been
previously demonstrated for related metalorganic resists,
where selectivities greater than 100:1 could be achieved at
larger pitches.
It is also notable that whereas lines narrower
than those shown in Figures 3 and 4have been patterned by
other groups in the resist (e.g., 4 nm lines on a 8 nm pitch by a
combination of HIBL and nanoimprint lithography
), the
etched structures reported here are both the narrowest and the
tallest to be transferred to substrate on a sub-20 nm pitch. The
next smallest transferred patterns found in the literature are on
a 22 nm pitch via thermal scanning probe lithography.
In comparison with a previous study of Cr8F8(pivalate)16
with EBL (100 keV, 300 pA),
HIBL required a dose three
orders of magnitude smaller to achieve its smallest pitched
lines. (EBL achieved 40 nm pitch lines at a 30 500 pC/cm line
dose compared with a 16 nm pitch at 22 pC/cm here.) It must
be mentioned that an orders-of-magnitude smaller dose with
HIBL is accompanied by an orders-of-magnitude smaller
current as well (e.g., 0.5 pA He ions versus 300 pA electrons).
Whereas at rst glance that might indicate that HIBL writing
speeds are approximately equivalent to EBL writing speeds, it
is important to also consider how the resist thickness impacts
doses. In the EBL study, the resist was 10 times thicker (30
nm) than in this HIBL study. If the resist thickness were to
increase here, then we might expect the HIBL dose to actually
decrease because we would be taking advantage of a cascade of
scattering events that cannot similarly take place when the
thickness is conned to something as small as 3 nm. (That
decrease in dose with increasing thickness, it must be noted,
would come at the expense of resolution.) Additionally, it is
well known that the clearing dose increases as a function of the
decreasing pitch, as was the case in the comparative EBL study,
where the smallest pitch was 40 nm. If we were able to
compare the 17 nm pitch EBL lines with the 17 nm pitch
HIBL lines shown here, then we would expect the HIBL dose
to be even more favorable than the three orders of magnitude
dierence noted above.
For mass manufacturing, the high exposure doses inherent to
EBL, which translate into long writing times, have always
outweighed the allure of EBLs small-probe, high-resolution
capability. Much work has been put into developing EBL tools
that split one primary beam into many beamlets to decrease
writing times by exposing many patterns in parallel.
Here we
see a demonstration of HIBL yielding both better resolution
and an orders-of-magnitude smaller dose than EBL. Whereas
single-beam HIBL, with the same pixel-by-pixel exposure
mechanism as EBL, may still not oer the lithographic speed
desired by the industry, perhaps this study indicates that if any
beam is to be split and operated in parallel, then it is a beam of
helium ions and not electrons.
In conclusion, it has been demonstrated that the molecule
Cr8F8(pivalate)16, when used as a resist, is capable of
producing sub-10 nm structures in silicon and tungsten,
spaced on a sub-20 nm pitch, following pattern transfer with an
ICPRIE. This result is due to several interrelated factors
associated with the resist material and the method of
lithography, HIBL. First, the ability to spin the resist into
sub-5 nm thick lms reduces the lateral scattering as the beam
travels through the resist, resulting in high resolution. Second,
the materials high molecular weight and low density limit the
number of scattering sites that the beam encounters, which
also improves resolution. Third, the nature of helium ion beam
interactions yields more SEs and AEs per incident beam
species than is achievable by the more traditional EBL; the
HIBL dose can therefore be orders of magnitude lower, which
allows for a low current to be selected, which results in a
subnanometer probe diameter that further improves the
patterning resolution. Finally, because exposing the resist
changes it from a metalorganic compound to a chromium
oxide material, the material exhibits extremely high etch
selectivity to both silicon and tungsten in the presence of an
SF6/C4F8etch, allowing for the transfer of 6 nm wide lines into
the substrates, even when the etch eciency is reduced by
tightly spacing the lines on a sub-20 nm pitch. It is therefore
possible to fabricate sub-10 nm wide, 19 nm tall silicon and
tungsten structures in a single lithography-and-etch step,
opening new possibilities for future nanoelectronics. The role
of HIBL in the future should also not be discounted.
Corresponding Authors
Nano Letters Letter
DOI: 10.1021/acs.nanolett.9b01911
Nano Lett. 2019, 19, 60436048
Scott M. Lewis: 0000-0002-4183-1906
Richard E. P. Winpenny: 0000-0002-7101-3963
The authors declare no competing nancial interest.
We acknowledge the EPSRC (U.K.) for funding (grant EP/
R023158/1). The University of Manchester also supported
this work. We gratefully acknowledge the critical support and
infrastructure provided for this work by the Kavli Nanoscience
Institute at Caltech.
(1) Dennard, R. H.; Gaensslen, F. H.; Rideout, V. L.; Bassous, E.;
LeBlanc, A. R. Design of Ion-Implanted MOSFETs with Very Small
Physical Dimensions. IEEE J. Solid-State Circuits 1974,9(5), 256
(2) Shafique, M.; Garg, S.; Henkel, J.; Marculescu, D. The EDA
Challenges in the Dark Silicon Era: Temperature, Reliability, and
Variability Perspectives. Proceedings of the The 51st Annual Design
Automation Conference on Design Automation Conference - DAC 14
(3) International Technology Roadmap for Semiconductors 2.0; ITRS,
06/0_2015-ITRS-2.0-Executive-Report-1.pdf (accessed on Aug 19,
(4) Jones, W. M.; Lukin, D.; Scherer, A. Practical Nanoscale Field
Emission Devices for Integrated Circuits. Appl. Phys. Lett. 2017,110
(26), 263101.
(5) Neudeck, P. G.; Okojie, R. S.; Chen, L.-Y. High-Temperature
Electronics - a Role for Wide Bandgap Semiconductors? Proc. IEEE
2002,90 (6), 10651076.
(6) Han, J.-W.; Sub Oh, J.; Meyyappan, M. Vacuum Nano-
electronics: Back to the Future?Gate Insulated Nanoscale Vacuum
Channel Transistor. Appl. Phys. Lett. 2012,100 (21), 213505.
(7) Maile, B. E.; Henschel, W.; Kurz, H.; Rienks, B.; Polman, R.;
Kaars, P. Sub-10 nm Linewidth and Overlay Performance Achieved
with a Fine-Tuned EBPG-5000 TFE Electron Beam Lithography
System. Jpn. J. Appl. Phys. 2000,39 (12), 6836.
(8) Yang, J. K. W.; Cord, B.; Duan, H.; Berggren, K. K.; Klingfus, J.;
Nam, S.-W.; Kim, K.-B.; Rooks, M. J. Understanding of Hydrogen
Silsesquioxane Electron Resist for Sub-5-Nm-Half-Pitch Lithography.
J. Vac. Sci. Technol. B Microelectron. Nanometer Struct. 2009,27 (6),
(9) Winston, D.; Cord, B. M.; Ming, B.; Bell, D. C.; DiNatale, W. F.;
Stern, L. A.; Vladar, A. E.; Postek, M. T.; Mondol, M. K.; Yang, J. K.
W.; et al. Scanning-Helium-Ion-Beam Lithography with Hydrogen
Silsesquioxane Resist. J. Vac. Sci. Technol. B Microelectron. Nanometer
Struct. 2009,27 (6), 2702.
(10) Hill, R.; Faridur Rahman, F. H. M. Advances in Helium Ion
Microscopy. Nucl. Instrum. Methods Phys. Res., Sect. A 2011,645 (1),
(11) Sidorkin, V.; van Veldhoven, E.; van der Drift, E.; Alkemade, P.;
Salemink, H.; Maas, D. Sub-10-nm Nanolithography with a Scanning
Helium Beam. J. Vac. Sci. Technol. B Microelectron. Nanometer Struct.
2009,27 (4), L18.
(12) Shi, X.; Prewett, P.; Huq, E.; Bagnall, D. M.; Robinson, A. P.
G.; Boden, S. A. Helium Ion Beam Lithography on Fullerene
Molecular Resists for Sub-10 nm Patterning. Microelectron. Eng. 2016,
(13) Notte, J.; Hill, R.; McVey, S.; Farkas, L.; Percival, R.; Ward, B.
An Introduction to Helium Ion Microscopy. Microsc. Microanal. 2006,
12 (S02), 126127.
(14) Lewis, S. M.; Fernandez, A.; DeRose, G. A.; Hunt, M. S.;
Whitehead, G. F. S.; Lagzda, A.; Alty, H. R.; Ferrando-Soria, J.; Varey,
S.; Kostopoulos, A. K.; et al. Use of Supramolecular Assemblies as
Lithographic Resists. Angew. Chem., Int. Ed. 2017,56 (24), 6749
(15) Lewis, S.; DeRose, G.; Hunt, M.; Scherer, A.; Yeates, S.;
Winpenny, R. E. P.; Alty, H. R.; Werthiem, A.; Li, J.; Fowler, T.; et al.
Design and Implementation of the next Generation Electron Beam
Resists for the Production of EUVL Photomasks. Photomask
Technology 2018, 24.
(16) Lin, Y.; Joy, D. C. A New Examination of Secondary Electron
Yield Data. Surf. Interface Anal. 2005,37 (11), 895900.
(17) Ghirri, A.; Corradini, V.; Bellini, V.; Biagi, R.; del Pennino, U.;
De Renzi, V.; Cezar, J.; Muryn, C. A.; Timco, G. A.; Winpenny, R. E.
P.; Affronte, M. ACS Nano 2011,5, 70907099.
(18) Dai, J.; Chang, S. W.; Hamad, A.; Yang, D.; Felix, N.; Ober, C.
K. Chem. Mater. 2006,18, 34043411.
(19) Joy, D. C. Monte Carlo Modeling for Electron Microscopy and
Microanalysis; Oxford University Press, 1995.
(20) Lewis, S. M.; DeRose, G. A. SML Electron Beam Resist. Front.
Nanosci. 2016,11, 421446.
(21) Ziegler, J. F. Handbook of Stopping Cross-Sections for Energetic
Ions in All Elements; Pergamon Press: New York, 1980.
(22) Livengood, R.; Tan, S.; Greenzweig, Y.; Notte, J.; McVey, S.
Subsurface Damage from Helium Ions as a Function of Dose, Beam
Energy, and Dose Rate. J. Vac. Sci. Technol. B Microelectron. Nanometer
Struct. 2009,27 (6), 3244.
(23) Goodyear, A.; Boettcher, M.; Stolberg, I.; Cooke, M. Direct
Comparison of the Performance of Commonly Used E-Beam Resists
during Nano-Scale Plasma Etching of Si, SiO2, and Cr. SPIE Advanced
Lithography 2015, 94280V.
(24) Li, W.-D.; Wu, W.; Stanley Williams, R. Combined Helium Ion
Beam and Nanoimprint Lithography Attains 4 nm Half-Pitch Dense
Patterns. J. Vac. Sci. Technol., B: Nanotechnol. Microelectron.: Mater.,
Process., Meas., Phenom. 2012,30 (6), No. 06F304.
(25) Ryu Cho, Y. K.; Rawlings, C. D.; Wolf, H.; Spieser, M.; Bisig,
S.; Reidt, S.; Sousa, M.; Khanal, S. R.; Jacobs, T. D. B.; Knoll, A. W.
Sub-10 Nanometer Feature Size in Silicon Using Thermal Scanning
Probe Lithography. ACS Nano 2017,11 (12), 1189011897.
(26) Chang, T. H. P.; Mankos, M.; Lee, K. Y.; Muray, L. P. Multiple
Electron-Beam Lithography. Microelectron. Eng. 2001,5758, 117
Nano Letters Letter
DOI: 10.1021/acs.nanolett.9b01911
Nano Lett. 2019, 19, 60436048
... Moreover, free-standing 3D structures can serve as current collectors [56] or heat sinks [57], while doping Cu or W phases into a primary metal matrix can enhance thermal properties [47,48]. Several additive and subtractive manufacturing methods, such as photolithography, plasma etching, wet etching, dry reactive etching, ion beam lithography, machining, physical vapor deposition, chemical vapor deposition, powder metallurgy, laser melting, and combinations of these methods have been used to fabricate these two transition metals or dope these materials into another material [38,48,[58][59][60][61][62][63][64][65]. ...
Full-text available
Copper (Cu) and tungsten (W) possess exceptional electrical and thermal conductivity properties, making them suitable candidates for applications such as interconnects and thermal conductivity enhancements. Solution-based additive manufacturing (SBAM) offers unique advantages, including patterning capabilities, cost-effectiveness, and scalability among the various methods for manufacturing Cu and W-based films and structures. In particular, SBAM material jetting techniques , such as inkjet printing (IJP), direct ink writing (DIW), and aerosol jet printing (AJP), present a promising approach for design freedom, low material wastes, and versatility as either stand-alone printers or integrated with powder bed-based metal additive manufacturing (MAM). Thus, this review summarizes recent advancements in solution-processed Cu and W, focusing on IJP, DIW, and AJP techniques. The discussion encompasses general aspects, current status, challenges, and recent research highlights. Furthermore, this paper addresses integrating material jetting techniques with powder bed-based MAM to fabricate functional alloys and multi-material structures. Finally, the factors influencing large-scale fabrication and potential prospects in this area are explored.
... 11,12 Moreover, unlike thermionic emission, field emission does not require heating, thus reducing power consumption and eliminating the need for thermal management. In addition, nanoscale vacuum gaps can be manufactured using either a top-down approach with current nanofabrication techniques and high-resolution electron lithography 13 or via a bottom-up approach using nanostructured materials such as nanocrystals, 2D materials, nanowires, and nanotubes. [14][15][16][17][18] This allows for operation at low voltages, 19 which is advantageous in terms of energy efficiency, Joule heating reduction, and minimizing the effects of ion sputtering that can result in device destruction. ...
Nanoscale field emission devices are promising candidates to design high-frequency electronics due to the lack of scattering in the vacuum channel that enables ballistic transport. In-plane devices are relatively easy to fabricate with current fabrication techniques and offer sub-fF capacitance. In this work, the characteristics of lateral gold multi-tip field emission arrays are studied. Vacuum gaps between the electrodes of 30 nm are fabricated, which allow < 10 V operation. The effect of number of emitting tips on measured current is investigated. By taking advantage of the strong non-linearity in the emission characteristic, frequency mixing in the MHz range is also demonstrated.
... Likewise, sub-10 nm resolution was achieved on an alumina-based resist and the pattern was later transferred to Si, with an aspect ratio of 10, by means of a low-bias reactive ion etching process 1034 . Recent work has made use of a metalorganic negativetone resist, Cr 8 F 8 (pivalate) 16 , which is capable of sub-10 nm resolution and requires low He + irradiation dose around 20 pC/cm (three orders of magnitude lower compared to a similar process by EBL), being used on Si and W substrates and with good behaviour against a subsequent dry etching process 1035 . Interestingly, hybrid organic-inorganic resists based on Ni perform well in HIBL, requiring 22 µC/cm 2 He + irradiation dose and producing sub-10 nm resolution with low line and width roughness 1036 . ...
Full-text available
The focused ion beam (FIB) is a powerful tool for the fabrication, modification and characterization of materials down to the nanoscale. Starting with the gallium FIB, which was originally intended for photomask repair in the semiconductor industry, there are now many different types of FIB that are commercially available. These instruments use a range of ion species and are applied broadly in materials science, physics, chemistry, biology, medicine, and even archaeology. The goal of this roadmap is to provide an overview of FIB instrumentation, theory, techniques and applications. By viewing FIB developments through the lens of the various research communities, we aim to identify future pathways for ion source and instrumentation development as well as emerging applications, and the scope for improved understanding of the complex interplay of ion-solid interactions. We intend to provide a guide for all scientists in the field that identifies common research interests and will support future fruitful interactions connecting tool development, experiment and theory. While a comprehensive overview of the field is sought, it is not possible to cover all research related to FIB technologies in detail. We give examples of specific projects within the broader context, referencing original works and previous review articles throughout.
... This, in turn, calls for a joint multidisciplinary effort to increase not only spin coherence but also the coupling of spins to superconducting resonators. Increasing the values estimated under section III by pushing the circuit miniaturization is a challenge, although the application of ultrahigh resolution nanolithography methods might still provide room for some improvement [67]. The alternative is to look for a different coupling regime between spins and photons. ...
Full-text available
The implementation of a universal quantum processor still poses fundamental issues related to error mitigation and correction, which demand to investigate also platforms and computing schemes alternative to the main stream. A possibility is offered by employing multi-level logical units (qudits), naturally provided by molecular spins. Here we present the blueprint of a Molecular Spin Quantum Processor consisting of single Molecular Nanomagnets, acting as qudits, placed within superconducting resonators adapted to the size and interactions of these molecules to achieve a strong single spin to photon coupling. We show how to implement a universal set of gates in such a platform and to readout the final qudit state. Single-qudit unitaries (potentially embedding multiple qubits) are implemented by fast classical drives, while a novel scheme is introduced to obtain two-qubit gates via resonant photon exchange. The latter is compared to the dispersive approach, finding in general a significant improvement. The performance of the platform is assessed by realistic numerical simulations of gate sequences, such as Deutsch-Josza and quantum simulation algorithms. The very good results demonstrate the feasibility of the molecular route towards a universal quantum processor.
... Alternatively, block copolymer lithography can be used to fabricate wafer scale arrays of relatively monodisperse nanoseeds at sub-50 nm pitches. [42][43][44] These simulations provide ample motivation toward the exploration of these techniques for the fabrication of nanomeshes with improved charge transport properties. Additionally, in order to isolate the properties of the meshes from those of the Ge, we opted to transfer the nanomeshes onto SiO 2 /Si substrates for our electrical measurements. ...
Full-text available
The synthesis of functional graphene nanostructures on Ge(001) provides an attractive route toward integrating graphene-based electronic devices onto complementary metal oxide semiconductor-compatible platforms. In this study, we leverage the phenomenon of the anisotropic growth of graphene nanoribbons from rationally placed graphene nanoseeds and their rotational self-alignment during chemical vapor deposition to synthesize mesoscale graphene nanomeshes over areas spanning several hundred square micrometers. Lithographically patterned nanoseeds are defined on a Ge(001) surface at pitches ranging from 50 to 100 nm, which serve as starting sites for subsequent nanoribbon growth. Rotational self-alignment of the nanoseeds followed by anisotropic growth kinetics causes the resulting nanoribbons to be oriented along each of the equivalent, orthogonal Ge⟨110⟩ directions with equal probability. As the nanoribbons grow, they fuse, creating a continuous nanomesh. In contrast to nanomesh synthesis via top-down approaches, this technique yields nanomeshes with atomically faceted edges and covalently bonded junctions, which are important for maximizing charge transport properties. Additionally, we simulate the electrical characteristics of nanomeshes synthesized from different initial nanoseed-sizes, size-polydispersities, pitches, and device channel lengths to identify a parameter-space for acceptable on/off ratios and on-conductance in semiconductor electronics. The simulations show that decreasing seed diameter and pitch are critical to increasing nanomesh on/off ratio and on-conductance, respectively. With further refinements in lithography, nanomeshes obtained via seeded synthesis and anisotropic growth are likely to have superior electronic properties with tremendous potential in a multitude of applications, such as radio frequency communications, sensing, thin-film electronics, and plasmonics.
... With the improvement of nanofabrication technology [1,2] and the demand for highperformance nanophotonic devices, the footprint of these devices is greatly reduced for high integration density. Nanophotonic devices are widely used in imaging [3], optical computing [4], medical diagnosis [5,6], etc. ...
Full-text available
The inverse design method based on a generative adversarial network (GAN) combined with a simulation neural network (sim-NN) and the self-attention mechanism is proposed in order to improve the efficiency of GAN for designing nanophotonic devices. The sim-NN can guide the model to produce more accurate device designs via the spectrum comparison, whereas the self-attention mechanism can help to extract detailed features of the spectrum by exploring their global interconnections. The nanopatterned power splitter with a 2 μm × 2 μm interference region is designed as an example to obtain the average high transmission (>94%) and low back-reflection (<0.5%) over the broad wavelength range of 1200~1650 nm. As compared to other models, this method can produce larger proportions of high figure-of-merit devices with various desired power-splitting ratios.
Full-text available
Understanding the (dis)assembly mechanisms of large metallosupramolecules is critical in their design, stability and application. The inherent complexity of these structures leads to many potential pathways for combining (or separating) the constituent building blocks, which makes this task difficult. Here we use collision-induced dissociation mass spectrometry to study the disassembly of heterometallic complexes. Collisional activation leads to the formation of a series of previously unknown smaller ring products and we characterize their geometry using ion mobility. The disassembly of both { Cr x Cu 2 } hourglass structures ( x = 10, 12) and of a { Cr 12 Gd 4 } cluster shows the formation of rare closed, heptametallic species { Cr 6 Cu }, { Cr 5 Cu 2 } and { Cr 5 Gd 2 } as dominant products, as well as other closed ions such as { Cr 5 Cu }, { Cr 10 Cu }, { Cr 12 Cu }, { Cr 10 }, { Cr 12 } and { Cr 6 Gd 2 }. The collision cross-section of cyclic products and precursors has a linear correlation with ion mass—a relationship that does not hold for acyclic systems. As these rings are non-trivial to synthesize individually in solution, we propose the presented workflow to identify and characterize feasible molecules for bulk phase synthesis.
The implementation of a universal quantum processor still poses fundamental issues related to error mitigation and correction, which demand investigation of also platforms and computing schemes alternative to the main stream. A possibility is offered by employing multilevel logical units (qudits), naturally provided by molecular spins. Here we present the blueprint of a molecular spin quantum processor consisting of single molecular nanomagnets, acting as qudits, placed within superconducting resonators adapted to the size and interactions of these molecules to achieve a strong single spin-to-photon coupling. We show how to implement a universal set of gates in such a platform and to readout the final qudit state. Single-qudit unitaries (potentially embedding multiple qubits) are implemented by fast classical drives, while an alternative scheme is introduced to obtain two-qubit gates via resonant photon exchange. The latter is compared to the dispersive approach, finding in general a significant improvement. The performance of the platform is assessed by realistic numerical simulations of gate sequences, such as Deutsch-Josza and quantum simulation algorithms. The very good results demonstrate the feasibility of the molecular route towards a universal quantum processor.
Full-text available
We report the synthesis and structural characterization of a series of heterometallic rings templated via alkylammonium or imidazolium cations. The template and preference of each metal's coordination geometry can control the structure of heterometallic compounds, leading to octa-, nona-, deca-, dodeca-, and tetradeca-metallic rings. The compounds were characterized by single-crystal X-ray diffraction, elemental analysis, magnetometry, and EPR measurements. Magnetic measurements show that the exchange coupling between metal centres is antiferromagnetic. EPR spectroscopy shows that the spectra of {Cr7Zn} and {Cr9Zn} have S = 3/2 ground states, while the spectra of {Cr12Zn2} and {Cr8Zn} are consistent with S = 1 and 2 excited states. The EPR spectra of {(ImidH)-Cr6Zn2}, {(1-MeImH)-Cr8Zn2}, and {(1,2-diMeImH)-Cr8Zn2} include a combination of linkage isomers. The results on these related compounds allow us to examine the transferability of magnetic parameters between compounds.
Conference Paper
Full-text available
Electron beam writing remains one of the reference pattern generation techniques, and plasma etching continues to underpin pattern transfer. We report a systematic study of the plasma etch resistance of several e-beam resists, both negative and positive as well as classical and Chemically Amplified Resists: HSQ[1,2] (Dow Corning), PMMA[3] (Allresist GmbH), AR-P6200 (Allresist GmbH), ZEP520 (Zeon Corporation), CAN028 (TOK), CAP164 (TOK), and an additional pCAR (non-disclosed provider). Their behaviour under plasma exposure to various nano-scale plasma etch chemistries was examined (SF6/C4F8 ICP silicon etch, CHF3/Ar RIE SiO2 etch, Cl2/O2 RIE and ICP chrome etch, and HBr ICP silicon etch). Samples of each resist type were etched simultaneously to provide a direct comparison of their etch resistance. Resist thicknesses (and hence resist erosion rates) were measured by spectroscopic ellipsometer in order to provide the highest accuracy for the resist comparison. Etch selectivities (substrate:mask etch rate ratio) are given, with recommendations for the optimum resist choice for each type of etch chemistry. Silicon etch profiles are also presented, along with the exposure and etch conditions to obtain the most vertical nano-scale pattern transfer. We identify one resist that gave an unusually high selectivity for chlorinated and brominated etches which could enable pattern transfer below 10nm without an additional hard mask. In this case the resist itself acts as a hard mask. We also highlight the differing effects of fluorine and bromine-based Silicon etch chemistries on resist profile evolution and hence etch fidelity.
Conference Paper
A new class of resist materials has been developed that is based on a family of heterometallic rings. The work is founded on a Monte Carlo simulation that utilizes a secondary and Auger electron generation model to design resist materials for high resolution electron beam lithography. The resist reduces the scattering of incident electrons to obtain line structures that have a width of 15 nm on a 40 nm pitch. This comes at the expense of lowering the sensitivity of the resist, which results in the need for large exposure doses. Low sensitivity can be dramatically improved by incorporating appropriate functional alkene groups around the metal-organic core, for example by replacing the pivalate component with a methacrylate molecule. This increases the resist sensitivity by a factor of 22.6 and demonstrates strong agreement between the Monte Carlo simulation and the experimental results. After the exposure and development processes, what remains of the resist material is a metal-oxide that is extremely resistant to silicon dry etch conditions; the etch selectivity has been measured to be 61:1.
High resolution lithography often involves thin resist layers which pose a challenge for pattern characterization. Direct evidence that the pattern was well defined and can be used for device fabrication is provided if a successful pattern-transfer is demonstrated. In the case of thermal scanning probe lithography (t-SPL), highest resolutions are achieved for shallow patterns. In this work we study the transfer reliability and the achievable resolution as a function of applied temperature and force. Pattern-transfer was reliable if a pattern depth of more than 3 nm was reached and the walls between the patterned lines were slightly elevated. Using this geometry as benchmark we studied the formation of 10 - 20 nm half-pitch dense lines as a function of the applied force and temperature. We found that the best pattern geometry is obtained at a heater temperature of ~ 600°C, which is below or close to the transition from mechanical indentation to thermal evaporation. At this temperature, there still is considerable plastic deformation of the resist, which leads to a reduction of the pattern depth at tight pitch and therefore physically limits the achievable resolution. By optimizing patterning conditions, we achieved 11 nm half-pitch dense lines in the HM8006 transfer layer and 14 nm half-pitch dense lines and L-lines in silicon. For the 14 nm half-pitch lines in silicon we measured a line-edge roughness (LER) of 2.6 nm (3σ) and a feature size of the patterned walls of 7 nm.
Nanoscale field emission devices promise many advantages over traditional solid-state devices including fast switching speeds, extreme operating temperatures, and radiation hardness. Despite this, practical circuits have long been hampered by the extreme requirements of nanoscale field emitters. Devices have required vacuum packaging, or extremely sharp emission points that are difficult to reproduce, or cannot be integrated on a single wafer with independent gating. We demonstrate CMOS compatible, integratable two- and three-terminal devices operating at near atmospheric pressures with high single tip currents at low voltages that can be used as building blocks for future circuits.
A new resist material for electron beam lithography has been created that is based on a supramolecular assembly. Initial studies revealed that with this supramolecular approach, high-resolution structures can be written that show unprecedented selectivity when exposed to etching conditions involving plasmas.
A new resist material for electron beam lithography has been created that is based on a supramolecular assembly. Initial studies revealed that with this supramolecular approach, high-resolution structures can be written that show unprecedented selectivity when exposed to etching conditions involving plasmas.
In this chapter, the focus was on investigating a suitable technology to fabricate the next generation optical photomasks with a one-step process. This was achieved by a fabricating a novel nanocomposite electron beam resist that incorporated azo dyes into polymethylmethacrylate (PMMA). The azo dyes were introduced to attenuate the ultraviolet radiation propagating through the electron beam resist as the PMMA was found to be transparent at the wavelength of the incident radiation, thus, a metallic blocking layer is not required. When the nanocomposite resist was patterned by the electron beam, the pattern was transferred to a photoresist via contact printing using the conventional photolithography technique. To push the resolution limits, the SML resist films were spun to 600-nm and 300-nm thicknesses. The resulting pattern had linewidths of 55 and 33 nm, respectively. Therefore, an aspect ratio of 10:1 and 9:1 has been demonstrated in resist.
Helium ion beam lithography (HIBL) is an emerging technique that uses a sub-nanometre focused beam of helium ions generated in the helium ion microscope to expose resist. It benefits from high resolution, high sensitivity and a low proximity effect. Here we present an investigation into HIBL on a novel, negative tone fullerene-derivative molecular resist. Analysis of large area exposures reveals a sensitivity of ~ 40 μC/cm2 with a 30 keV helium beam which is almost three orders of magnitude higher than the sensitivity of this resist to a 30 keV electron beam. Sparse line features with line widths of 7.3 nm are achieved on the ~ 10 nm thick resist. The fabrication of 8.5 half-pitched lines with good feature separation and 6 nm half-pitched lines with inferior but still resolvable separation is also shown in this study. Thus, sub-10 nm patterning with small proximity effect is demonstrated using HIBL using standard processing conditions, establishing its potential as an alternative to EBL for rapid prototyping of beyond CMOS devices.
Conference Paper
Technology scaling has resulted in smaller and faster transistors in successive technology generations. However, transistor power consumption no longer scales commensurately with integration density and, consequently, it is projected that in future technology nodes it will only be possible to simultaneously power on a fraction of cores on a multi-core chip in order to stay within the power budget. The part of the chip that is powered off is referred to as dark silicon and brings new challenges as well as opportunities for the design community, particularly in the context of the interaction of dark silicon with thermal, reliability and variability concerns. In this perspectives paper we describe these new challenges and opportunities, and provide preliminary experimental evidence in their support.
The authors demonstrated a promising technique that yielded single-digit nanometer features for nanotechnology research and possible future electronic circuit fabrication by combining high resolution helium ion beam patterning and nanoimprint lithography. They fabricated a series of line patterns with single-digit nanometer half-pitches by exposing a layer of hydrogen silsesquioxane (HSQ) resist with a scanning focused helium ion beam. The smallest half-pitch of clearly resolved line patterns was 4 nm. Using the HSQ patterns as a nanoimprint template, nanoscale patterns down to 4 nm half-pitch were transferred into nanoimprint resist through a UV-curable nanoimprint process.