of 9
Sensitivity enhancement of a high-resolution
negative-tone nonchemically amplified metal organic
photoresist for extreme ultraviolet lithography
Scott M. Lewis ,
a,b,c,
*
Hayden R. Alty,
a
Michaela Vockenhuber,
d
Guy A. DeRose ,
b
Antonio Fernandez-Mato,
a
Dimitrios Kazazis ,
d
Paul L. Winpenny ,
c
Richard Grindell ,
c
Grigore A. Timco,
a
Axel Scherer,
b
Yasin Ekinci,
d
and Richard E. P. Winpenny
a,c
a
University of Manchester, School of Chemistry and Photon Science Institute,
Manchester, United Kingdom
b
Kavli Nanoscience Institute, California Institute of Technology,
Pasadena, California, United States
c
Sci-Tron Ltd., Walsall, United Kingdom
d
Paul Scherrer Institute, Advanced Lithography and Metrology Group,
Villigen, Switzerland
Abstract.
A new class of negative-tone resist materials has been developed for electron beam
and extreme ultraviolet lithography. The resist is based on heterometallic rings. From initial
electron beam lithography studies, the resist performance demonstrated a resolution of 40-nm
pitch but at the expense of a low sensitivity. To improve the sensitivity, we incorporated
HgCl
2
and
HgI
2
into the resist molecular design. This dramatically improved the resist sensitivity while
maintaining high resolution. This improvement was demonstrated using electron beam and
extreme ultraviolet lithography.
© 2022 Society of Photo-Optical Instrumentation Engineers (SPIE)
[DOI:
10.1117/1.JMM.21.4.041404
]
Keywords:
metal organic EUV photoresist; metal organic electron beam resist; extreme ultra-
violet lithography; electron beam lithography.
Paper 22007SS received Feb. 9, 2022; accepted for publication May 4, 2022; published online
Jun. 2, 2022.
1 Introduction
The ability to produce patterns at the nanoscale using lithography underpins modern society.
The electronic devices we take for granted contain integrated circuits (IC) and the key compo-
nent of those ICs are field-effect transistors (FETs). They have reduced in size by a factor of
two every 2 years for over 50 years, following Moore
s law. The roadmap for the electronics
industry now assumes that this constant reduction of size will continue
at least until the mid-
2020s.
1
However, there is a significant challenge now as the feature size drops to 20 nm and
below. At the end of 2019, extreme ultraviolet lithography (EUVL) was adopted to manufacture
FinFETs (i.e., multigate FETs where the gates wrap around fin-like channels) that are an integral
part of today
s ICs at the 7-nm node and beyond. As future technology nodes will take full
advantage of the EUVL technique, it will become increasingly challenging to transfer the
patterns into the silicon with existing organic resist materials. This is because as the resolution
of the pattern increases, the aggressive conditions of the etch plasma will need to be increased
substantially, leading to a decrease in etch efficiency.
2
,
3
Therefore, inorganic resist materials,
which provide superior etch selectivity, must be explored.
In a previous study, we reported a negative-tone resist that is based on a metal
organic
compound.
2
This resist produced line and space patterns of 20-nm half-pitch (HP) resolution,
while demonstrating extraordinary silicon dry etch performance of
>
100
1
selectivity.
*Address all correspondence to Scott M. Lewis,
scott.lewis@manchester.ac.uk
;
slewis2@caltech.edu
1932-5150/2022/$28.00 © 2022 SPIE
J. Micro/Nanopattern. Mater. Metrol.
041404-1
Oct
Dec 2022
Vol. 21(4)
Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Micro/Nanopatterning,-Materials,-and-Metrology on 07 Jun 2022
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
Unfortunately, this was produced at the expense of resist sensitivity, where the exposure dose
was
61
;
000
μ
C
cm
2
for electron beam lithography (EBL). Although the sensitivity is very low,
it shows great promise from a pattern transfer point of view. Therefore, this resist could provide
an excellent framework to easing the challenges of dry etching in future technology nodes.
From this, we have used our three-dimensional (3D) Monte Carlo simulator to improve the
resist sensitivity without losing the pattern resolution. This gave a critical understanding of the
exposure mechanism, which allowed the exposure efficiency of the resist to be increased.
4
,
5
This
has led us to hypothesize that by incorporating both the diallylammonium at the center of the
molecule [see Fig.
1(a)
] and the addition of heavy metal salts on the outside of the molecule [see
Figs.
1(b)
and
1(c)
] would lead to further gains in sensitivity. It is well known that secondary
electrons (SEs) are essential to lowering the exposure dose for EBL as these electrons are respon-
sible for exposing the resist.
6
The results from our model show that the SEs that are produced
from the alkene group from the diallylammonium collide primarily with the Cr atoms (due to it
having a larger scattering cross section than the surrounding atoms) in the resist and cause a
cascade of SEs that are emitted from the Cr atoms. These electrons are produced at a reduced
electron energy than that of the incident electron beam and have a greater probability of inter-
action with the atoms (C, H, O) that make up the pivalate molecules in the outer region of the
resist. The pivalates on the outside of the molecule will be cleaved by the SEs, and this renders
the resist insoluble in the developing solvent and has the effect of reducing the exposure dose
while maintaining resolution.
One of the key parameters when designing a new resist is its density. Increasing the density
reduces the effective mean free path of the electrons. Therefore, the electron will experience
more collisions due to the reduced distance in between the atoms. This increases the probability
of emitting SEs from the atoms with a high density into the immediate exposure area of the resist
film. Clearly, as more SEs are generated, it means an increase in resist sensitivity, but this comes
at the expense of resolution. To overcome this, the resolution can be controlled by the molecular
weight of the resist molecule. The larger the molecular weight, the fewer the positions at which
the electrons can have a scatter interaction with the resist. It is important to recognize that these
resists do not behave like polymers such as poly meth methyl acrylate (PMMA) or ZEP520A.
Our resists have a very large molecular weight (
>
2000 g
mol
, while a PMMA monomer is
102 g
mol
as the electrons do not see the entire PMMA chain) with a low density, hence the
large hole in the middle. So there is more free space than resist matter for the electrons to scatter
from. The addition of the dopant (mercury-based compounds) increases the density at that local-
ized position. This means that the resist molecule is partially rendered insoluble in the devel-
oping solvent as the electrons expose this area first as less electrons are required to produce the
pattern. Hence, the SEs are confined to the immediate exposure area of the pattern. Thus, pro-
ducing high-resolution patterns. From an ecological point of view, incorporating mercury dichlo-
ride and mercury diiodide products to the resist chemistry sounds like something that would not
be good for the environment. Using these products causes the resist insoluble in all polar solvents
and is only soluble nonpolar solvents, such as tert butyl methyl ether, diethyl ether, and hexanes.
This makes it easy to handle and the resist can be easily be tracked and recovered from the
Fig. 1
The structures of compounds in the crystal: (a) resist
1
, (b) resist
2
, and (c) resist
3
.Cr
green, Ni green with a blue band, F yellow, H atoms omitted for clarity.
Lewis et al.: Sensitivity enhancement of a high-resolution negative-tone nonchemically amplified metal...
J. Micro/Nanopattern. Mater. Metrol.
041404-2
Oct
Dec 2022
Vol. 21(4)
Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Micro/Nanopatterning,-Materials,-and-Metrology on 07 Jun 2022
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
spinner for future use. Also, it should be noted that when bound to the
C
7
NiF
8
ring, the quantity
of mercury dichloride and mercury diiodide is very low, where there is a presence of 11% and
21% of the molecule, respectively.
Preliminary EBL studies led us to study a resist using electrons and EUVas the exposure method
that is based on a metal-organic compound
½
NH
2
ð
CH
2
CH
CH
2
Þ
2
Cr
7
NiF
8
ð
O
2
C
t
Bu
Þ
16

1
where seven chromium(III) centers (in green in Fig.
1
) and nickel(II) (in green with a blue band
in Fig.
1
)formanoctagon.
7
The exterior of the compound consists entirely of
tert-
butyl groups, and
this gives the compound a high solubility in solvents suitable for preparing films on silicon sub-
strates. The compound has a density (
ρ
¼
1.212 g cm
3
) with a large molecular weight (2192 Da).
The resist forms a closely packed film when it is spun on the silicon substrate. This is a high-
resolution but low-sensitivity resist.
To improve the sensitivity, we functionalized the resist
1
with an
iso
-nicotinate (
O
2
CC
5
H
4
N
)
group that can then be bound to mercury dichloride (
HgCl
2
)togive
½
NH
2
ð
CH
2
CH
CH
2
Þ
2

½
Cr
7
NiF
8
ð
O
2
C
t
Bu
Þ
15
ð
O
2
CC
5
H
4
N
HgCl
2
Þ
[resist
2
in Fig.
1(b)
]. The compound has a density
(
ρ
¼
1.7 g cm
3
) with a large molecular weight (2484 Da). We hypothesized that the mercury
dichloride would increase the resist sensitivity because it would generate low-energy SEs upon
absorbing the EUV radiation. These SEs will create a chain reaction of cascading electrons that
will expose the resist in the immediate exposure area.
To further increase the resist sensitivity, we hypothesized that substituting the mercury
dichloride in resist
2
for a mercury diiodide would give the resist a stronger atomic absorption
cross-section to EUV radiation due to the presence of iodine within the molecule.
8
This
was achieved by functionalizing
1
with an
iso
-nicotinate (
O
2
CC
5
H
4
N
) group that can then
be bound to mercury diiodide (
HgI
2
)togive
½
NH
2
ð
CH
2
CH
CH
2
Þ
2
Cr
7
NiF
8
ð
O
2
C
t
Bu
Þ
15
ð
O
2
CC
5
H
4
N
HgI
2
Þ
3
[Fig.
1(c)
]. The compound has a density (
ρ
¼
2.1 g cm
3
) with a large
molecular weight (2667 Da).
We used our Excalibur Monte Carlo simulator to understand the lithographic sensitivity
performance of these resists materials. Figure
2
shows 3D scattering trajectory plots, first they
demonstrate that all of the resists confine the primary electrons to the immediate write area
which suggested that high-resolution nanostructures could be obtained for each resist. Second,
they showed that the introduction of the diallylammonium and
HgCl
2
and
HgI
2
molecule to
resists
2
and
3
generates more SE than resist
1
in the resist. It was calculated that resist
1
pro-
duced 11,114 SEs while resists
2
and
3
produced 24,451 and 40,010, respectively. Therefore, it is
expected that the dose required to render the resist insoluble should be
2.2
and 3.6 times lower
than resist
1
. This is important, as these electrons are responsible for exposing the resist, and they
subsequently increase the overall sensitivity while contributing to the proximity effect.
EBL and EUVL exposures were performed on silicon wafers of
20
×
20 mm
2
. Each resist
(15 mg) was dissolved in tert butyl methyl ether (2 g). Each solution was filtered using a
0.2-
μ
m
polytetrafluoroethylene syringe filter and was spin-coated with a spin rate of 6000 rpm for 30 s,
followed by a 100°C soft bake for 2 min. The resulting thickness was measured to be 30 nm.
2
The EBL experiments were performed using a Raith EPBG5200. The exposure clearing dose
of each resist was determined using a one-dimensional matrix of single-pixel-wide lines with HP
22.5 and 15 nm HP and were
5
μ
m
long. The current and step sizes used were 100 pA and 5 nm
for the 100 keV exposures, respectively. The patterns were exposed in sets of 10 lines with one
pass of the beam per line, and the line dose of each set ranged from 2000 to
30
;
000
μ
C
cm
2
with
incremental steps of
100
μ
C
cm
2
.
The EUVL experiments were performed using an EUV interference lithography (EUV-IL)
technique at the XIL-II beamline at the Swiss Light Source of the Paul Scherer Institute.
9
The provided spatially coherent beam was tuned at an energy of 92 eV (13.5 nm wavelength)
with 4% bandwidth and the measured photon flux was
34 mW
cm
2
. Each resist material was
exposed through a photomask that consisted of patterns of lines and spaces with HP of 22
and 16 nm. To determine the optimum exposure dose, the dose was modified by varying the
exposure time that radiation was incident on the resist. The exposure dose ranged from 100 to
2000 mJ
cm
2
. Following both lithography methods, each resist was developed in hexane for
10 s to dissolve away the unexposed resist, then blow-dried with nitrogen.
To predict the potential behavior of each resist with EUV radiation. We have used the x-ray
interacts with matter CRXO database
10
to calculate the absorption coefficient of each resist.
Lewis et al.: Sensitivity enhancement of a high-resolution negative-tone nonchemically amplified metal...
J. Micro/Nanopattern. Mater. Metrol.
041404-3
Oct
Dec 2022
Vol. 21(4)
Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Micro/Nanopatterning,-Materials,-and-Metrology on 07 Jun 2022
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
It was calculated that resists
1
,
2
, and
3
had an absorption coefficient of
4.93
×
10
7
cm
1
,
6.18
×
10
7
cm
1
, and
8.03
×
10
7
cm
1
, respectively. From this, it was determined that the sen-
sitivity of resist
3
should be at least 1.6 times more sensitive than resist
1
and resist
2
should be
at least 1.25 times more sensitive to the EUV radiation than resist
1
. It should be pointed out that
this only shows how much each resist absorbs the EUV radiation and does not account for the
internal mechanism of producing SE
s cascade which also exposes the resist. Therefore, it would
be expected that the sensitivity difference between each resist should be larger.
The scanning electron microscope (SEM) images seen in Fig.
3
show 22.5 and 15 nm HP
lines in all three resists, patterned using EBL. The patterns are all well-resolved and demonstrate
good line uniformity.
At first glance, it appears that all of the nanopatterns in the SEM images are the same for all
resist materials as expected. The difference is the exposure dose that is required to partially
render each resist insoluble in the developer, which is hexane. The exposure dose that is required
to expose each resist is significantly reduced when functionalizing resist
1
with
HgI
2
to give
3
.
It can be seen that the pattern resolution has not changed because the resist design rules (large
density and large molecular weight to obtain high-resolution nanostructures) that govern the
pattern were met.
It is important to understand the mechanism to how the features are resolved. To do this we
must first determine how far the SEs travel with the resist. We can do this by determining how
they are deposited within the resist and more importantly, what are their associated energies are
Fig. 2
Point spread function of the internal electron scattering interactions inside 30-nm films of:
(a) resist
1
, (b) resist
2
, and (c) resist
3
. The acceleration voltage is 100 KeV. The black lines
represent the primary electrons from the incident beam while the SEs above 500 eV are repre-
sented by the red lines. The SEs that have the associated energies below 500 eV, which were
generated by first-, second-, and third-order collisions, are indicated purple, cyan, and green. The
blue lines are backscattered electrons; 1 million electrons are inserted into a single spot.
Lewis et al.: Sensitivity enhancement of a high-resolution negative-tone nonchemically amplified metal...
J. Micro/Nanopattern. Mater. Metrol.
041404-4
Oct
Dec 2022
Vol. 21(4)
Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Micro/Nanopatterning,-Materials,-and-Metrology on 07 Jun 2022
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
Fig. 3
(a) Scanning electron micrograph of EBL exposures with 22.5-nm HP lines fabricated in
resist
1
; (b) Monte Carlo simulation of SEs energy deposited in resist
1
using 100 KeV acceleration
voltage with 109,180 incident electrons per spot; (c) scanning electron micrograph of EBL expo-
sures with 22.5-nm HP lines fabricated in resist
2
; (d) Monte Carlo simulation of SEs energy
deposited in resist
2
using 100 KeV acceleration voltage with 50,547 incident electrons per spot;
(e) scanning electron micrograph of EBL exposures with 22.5-nm HP lines fabricated in resist
3
;
(f) Monte Carlo simulation of SEs energy deposited in resist
3
using 100 KeV acceleration voltage
with 315,549 incident electrons per spot; (g) scanning electron micrograph of EBL exposures with
15 nm HP lines fabricated in resist
1
; (h) Monte Carlo simulation of SEs energy deposited in resist
1
using 100 KeV acceleration voltage with 124,950 incident electrons per spot; (i) scanning elec-
tron micrograph of EBL exposures with 15-nm HP lines fabricated in resist
2
; (j) Monte Carlo
simulation of SEs energy deposited in resist
2
using 100 KeV acceleration voltage with 73,070
incident electrons per spot; (k) scanning electron micrograph of EBL exposures with 15 nm HP
lines fabricated in resist
3
; (l) Monte Carlo simulation of SEs energy deposited in resist
3
using
100 KeV acceleration voltage with 54,326 incident electrons per spot.
Lewis et al.: Sensitivity enhancement of a high-resolution negative-tone nonchemically amplified metal...
J. Micro/Nanopattern. Mater. Metrol.
041404-5
Oct
Dec 2022
Vol. 21(4)
Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Micro/Nanopatterning,-Materials,-and-Metrology on 07 Jun 2022
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
as they travel through the resist. From this, we can ascertain what the probability of an interaction
with the resist would be. The lower their energy, the more damage in the immediate exposure
area will be. To determine this, we have tracked the SEs generated while retaining their asso-
ciated energy and we created energy deposition plots.
It can be seen from the simulations that patterns with features of 22.5 and 15 nm HP were be
resolved. At first glance, it appears that all the simulations appear to be the same, and that is the
point. The difference is the number of electrons that reflected the dose [shown in Figs.
3(a)
,
3(c)
,
3(e)
,
3(g)
,
3(i)
and
3(k)
] was required to fully render each resist insoluble in the developer.
The simulations show that as the width of line gets larger, the energy toward the edge of the
feature reduces significantly. It is clear that the energy of the electrons in the middle of the feature
starts as 100 KeV (which is the incident energy of the PE) and over a distance of 7.5 nm (for the
case of 15 nm HP features, Figs.
3(h)
,
3(j)
, and
3(l)
] the energy of the SEs essentially is <
10 eV
.
The distance that the majority of SEs travel is significantly less than 10 nm. It shows that the
stopping power of both
HgCl
2
and
HgI
2
dopants in resists
2
and
3
confines the SE to immediate
exposure area.
The SEM images in Fig.
4
show nanopatterns obtained with EUVL with lines and spaces
with 22 and 16 nm HP in all three resists. The patterns are fully resolved in all resists and dem-
onstrate good line uniformity. Like the EBL experiments, all of the patterns appear to be the same
and follow the same reduction of exposure doses that are required to render each of the resists
insoluble to the hexane solvent. It was determined that the doses required to expose resists
1
,
2
,
and
3
were 534, 228, and
152 mJ
cm
2
, respectively, to produce a pattern with an HP of 22 nm,
while to produce a pattern with an HP of 16 nm, the required doses for resists
1
,
2
, and
3
were
625, 322, and
244 mJ
cm
2
, respectively.
The bridging seen in Fig.
4(d)
is due to that the exposure latitude of the resist and when
performing the experiment the exposure dose was on the high end of the dose scale.
Figure
5(a)
shows the exposure doses required to produce the patterns shown in Figs.
3
and
4
.
It is evident that the presence of
HgCl
2
in
2
and
HgI
2
in
3
decreases the exposure dose, dem-
onstrating that we have chemical control of the exposure dose. Thus, for the EBL experiment,
Fig. 4
(a) Scanning electron micrograph of patterns of 22 nm HP lines fabricated in resist
1
; (b) in
resist
2
; and (c) in resist
3
. (d) Scanning electron micrograph of developed patterns of 16 nm HP
lines fabricated in resist
1
, (e) resist
2
, and (f) resist
3
.
Lewis et al.: Sensitivity enhancement of a high-resolution negative-tone nonchemically amplified metal...
J. Micro/Nanopattern. Mater. Metrol.
041404-6
Oct
Dec 2022
Vol. 21(4)
Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Micro/Nanopatterning,-Materials,-and-Metrology on 07 Jun 2022
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
it can be seen from Fig.
5(b)
that the dose required for
3
decreases by a factor of 3.4 and 2.3 times
when compared with
1
for the 22.5 and 15 nm HP, respectively, and this matches the Monte
Carlo simulations of Fig.
2
for an HP of 22.5 nm. The Monte Carlo simulation was demonstrated
using a single point, and the proximity effect was not accounted for, thus, the dose factor for the
higher pitch is lower. In the EUVL experiment, the dose required for
3
is reduced by a factor of
3.5 and 2.5 times when compared with
1
for the 22 and 16 nm HP, respectively. This sensitivity
increase is greater than what was predicted by the absorption coefficient calculations. This shows
that the generation of SEs contributes to increasing the sensitivity.
The immediate observation is that the addition of
HgI
2
in resist
3
increases the sensitivity of
the resist by 2.3 or 2.5 times for EBL and EUVL at an HP of 15 and 16 nm respectively, com-
pared with
1
. For EBL, resist
3
was determined to be 1.3 times more sensitive than resist
2
for
both pitches shown. The design of the resist was based upon density and differed by a factor of
1.2. This means that the probability of the incident electrons colliding with the
HgI
2
to produce
SEs is higher than that of
HgCl
2
. Whereas the density difference between resist
1
and
3
is 1.7;
however, the sensitivity difference between resist
1
and
3
is 3.4 and 2.3 for 45 and 30 pitches,
respectively. It must be noted that the molecular weight of resist
1
is 1.2 times smaller than resist
3
and this accounts for the reduction of the sensitivity.
For EUVL, the exposing mechanism is based upon the metals, and in
3
iodine, absorbing the
radiation and emitting SEs into the immediate exposure area. Resist
3
has mercury diiodide
bound to the outside of the molecule and the atomic absorption cross-section associated with
iodine is one of the strongest of all of the elements on the periodic table.
This gives rise to producing SEs which will collide with the Hg atom in the resist and cause a
cascade of SEs that are emitted from the Hg atoms. These electrons will cleave the pivalates on
the outside of the molecule which are close to the
HgI
2
site. This renders the resist partially
insoluble in the developing solvent and has the effect of reducing the exposure dose while main-
taining resolution. Clearly, resists
1
and
2
do not have any iodine present. The mercury atomic
Fig. 5
(a) Dose-to-clear values for all resist materials using EUVL and EBL. (b) Relative of the
sensitivity of the resists for EUVL and EBL.
Lewis et al.: Sensitivity enhancement of a high-resolution negative-tone nonchemically amplified metal...
J. Micro/Nanopattern. Mater. Metrol.
041404-7
Oct
Dec 2022
Vol. 21(4)
Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Micro/Nanopatterning,-Materials,-and-Metrology on 07 Jun 2022
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
absorption cross-section is 2.1 times lower than that of the iodine.
8
Therefore, it is expected that
the resist sensitivity of
2
when compared with
3
should be lower by a factor of 2.
It was determined that the resist sensitivity difference between them is 1.5 and 1.3 for 22 and
16 nm HP, respectively. Resist
1
does not have any Hg or I in the resist and the Cr atomic
absorption cross-section is 2.6 and 5.4 times lower than the atomic absorption cross-section
of mercury and iodine atoms, respectively.
8
It was determined that the resist sensitivity is lower
by a factor of 3.5 and 2.5 with respect to 22 and 16 nm HP. The discrepancy may be based on the
feature size of the pattern because the probability of the landing of an EUV photon in the expo-
sure area is lower with increasing resolution; and therefore, it will need more photons to interact
with the resist molecule. Also, other atoms of the resist molecule may dominate and absorb
the photons, which lowers the effect of the iodine. It is clear that the sensitivity discrepancies
observed warrant further investigation.
From the results shown here, it is clear that the exposure sensitivity does not seem to be
dependent upon the exposing medium used as the difference in sensitivity demonstrated with
resist
3
when compared with resist
1
is the same for EBL and EUVL.
2 Conclusions
Three metal-organic negative tone resists have been investigated by EBL and EUVL. We have
demonstrated that the exposure sensitivity of the materials can be increased by 2.3 and 2.5 times
for the EBL and EUVL studies by chemical design, i.e., by introducing components such as
mercury diiodide in
3
while maintaining a high resolution of 15 and 16 nm HP, respectively.
The good correlation between EBL and EUVL is perhaps surprising, given that the energy of the
incident electrons and photons is different. For EUVL the radiation has energy of 92 eVand with
EBL we are exposing at 100 keV. The reason for the correlation is that in each case the energy is
sufficient to excite SEs from the resist material; EBL is in fact more energetic than it needs
be to excite SEs from any element. For EUV, the controlling factor is the EUV cross-section
of the elements present but there is still enough energy to excite SEs (the energy of these elec-
trons is 20 eV and below), which control the exposure sensitivity. This contrast with longer
wavelength photolithography where the exciting radiation is breaking covalent bonds through
excitations.
Acknowledgments
We acknowledge the EPSRC (UK) for funding (Grant No. EP/R023158/1) and Innovate UK for
funding (project 72472). The University of Manchester also supported this work. The authors
gratefully acknowledge the critical support and infrastructure provided for this work by the Kavli
Nanoscience Institute at Caltech. REPW thanks the European Research Council for an Advanced
Grant (ERC-2017-ADG-786734) and the EPSRC (UK) for an Established Career Fellowship
(EP/R011079/1). This project has received funding from the EU-H2020 research and innovation
program under Grant Agreement No. 654360 having benefitted from the access provided by
PSI in Villigen within the framework of the Nanoscience Foundries and Fine Analysis Europe
Transnational Access Activity. The manuscript was written through the contributions of all
authors. All authors have given approval to the final version of the manuscript. These authors
contributed equally; Scott M. Lewis designed the research and wrote the manuscript, Hayden R.
Alty designed and programmed the Excalibur Monte Carlo Simulator software, Michaela
Vockenhuber ran the XIL beamline at the PSI and performed the EUV exposures on the photo-
resists, Dimitrios Kazazis performed the EUV exposures and acquired the SEM images of the
photoresist. Paul. L. Winpenny designed and programmed the pattern analysis software, Guy A.
Derose provided the EPBG5200 EBL tool and fabricated the patterns, Axel Scherer provided the
EPBG5200 EBL tool and lab facilities necessary to verify the results, Yasin Ekinci provided the
XIL beam EUVexposure tool and lab facilities necessary to verify the results. Grigore A. Timco
and Richard Grindell made the compounds studied. Richard E. P. Winpenny provided the chem-
istry facilities and lab infrastructure. EPSRC (UK) Grant EP/R023158/1; Innovate UK project
72472; European Research Council ERC-2017-ADG-786734; EPSRC (UK) EP/R011079/1.
Lewis et al.: Sensitivity enhancement of a high-resolution negative-tone nonchemically amplified metal...
J. Micro/Nanopattern. Mater. Metrol.
041404-8
Oct
Dec 2022
Vol. 21(4)
Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Micro/Nanopatterning,-Materials,-and-Metrology on 07 Jun 2022
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
References
1. More Moore,
IEEE international roadmap for devices and systems,
2018,
www.irds.ieee
.org
(accessed 03/02/2020).
2. S. Lewis et al.,
Design and implementation of the next generation electron beam resists
for the production of EUVL photomasks,
Proc. SPIE
10810
, 108100N (2018).
3. S. M. Lewis et al.,
Plasma-etched pattern transfer of sub-10 nm structures using a metal-
organic resist and helium ion beam lithography,
Nano Lett.
19
, 6043
6048 (2019).
4. S. M. Lewis and G. A. DeRose,
Frontiers of Nanoscience, Materials and Processes for
Next Generation Lithography: SML Electron Beam Resist: Ultra-High Aspect Ratio
Nanolithography
, Vol. 11, pp. 421
446, Elsevier (2016).
5. S. M. Lewis et al.,
Use of supramolecular assemblies as lithographic resists,
Angew.
Chem.
129
(24), 6853
6856 (2017).
6. B. Wu and A. R. Neureuther,
Energy deposition and transfer in electron-beam lithography,
J. Vac. Sci. Technol. B: Microelectron. Nanometer Struct. Process. Meas. Phenom.
19
,
2508 (2001).
7. G. F. S. Whitehead et al.,
A ring of rings and other multicompoment assemblies of cages,
Angew. Chem. Int. Ed.
52
, 9932
9935 (2013).
8. R. Fallica et al.,
Absorption coefficient of metal containing photoresists in the extreme
ultraviolet,
J. Micro/Nanolithogr. MEMS MOEMS
17
(2), 023505 (2018).
9. C. Popescu et al.,
Sensitivity enhancement of the high resolution xMT multi-trigger resist
for EUV lithography,
Proc. SPIE
10143
, 101430V (2017).
10. B. L. Henke, E. M. Gullikson, and J. C. Davis,
X-ray interactions: photoabsorption, scat-
tering, transmission, and reflection at E = 50
30000 eV, Z = 1
92,
At. Data Nucl. Data
Tables
54
(2), 181
342 (1993).
Scott M. Lewis
is a research fellow at the University of Manchester. He received his BEng, MSc
degrees, and a PhD in electronic engineering from Cardiff University in 2002, 2004, and 2009,
respectively. His research focuses on developing new Monte Carlo models to design new resist
materials and in nanofabrication by photo and e-beam lithography. He holds the prestigious
position of visiting associate researcher at the California Institute of Technology.
Biographies of the other authors are not available.
Lewis et al.: Sensitivity enhancement of a high-resolution negative-tone nonchemically amplified metal...
J. Micro/Nanopattern. Mater. Metrol.
041404-9
Oct
Dec 2022
Vol. 21(4)
Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Micro/Nanopatterning,-Materials,-and-Metrology on 07 Jun 2022
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use