Microelectronic Engineering xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Microelectronic Engineering
journal homepage: www.elsevier.com/locate/mee

Review Article

Nanofabrication by electron beam lithography and its applications:

A review
Yifang Chen
Nanolithogrophy and Application Research Group, State Key Lab of ASIC and System, School of Information Science and Engineering, Fudan University, Shanghai 200433, China

a r t i c l e i n f o a b s t r a c t

Article history: This review covers a wide range of nanofabrication techniques developed for nanoelectronic devices,
Received 1 August 2014 nanophotonic metamaterials and other nanostructures, based on electron beam lithography (EBL).
Received in revised form 30 January 2015 Differing from earlier publications, this review particularly focuses on how to apply the property of
Accepted 20 February 2015
EBL resists for constructing multilayer stacks towards pattern transfer. Most frequently used resists
Available online xxxx
and their lithography property are first introduced, followed by categorizing multiple layers of resists
for fulfilling various tasks in nanofabrication. Particularly, T shape gates for high electron mobility tran-
Keywords:
sistors (HEMTs), metallic tunneling junctions (MTJs) in single electron tunneling transistors (SETs), chiral
Nanofabrication
Electron beam lithography
structures and photonic crystals for optical metamaterials, templates for NIL and etching masks for
Nanostructure nanoscale reactive ion etch (RIE) are reviewed. In the description of process development, scientific
Nanoelectronics advances behind these fabricated nanostructures are described at the same time. By this way, this review
Nanophotonics aims to indicate that the development of nanofabrication techniques is essential for the rapid advances of
Metamaterial nanoscience as a whole.
Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction techniques at the present. This review is to summarize the signifi-

The current top-down nanofabrication technique is, in practice,
the pattern transfer at nanometer scale. Despite the large variety of
nano processes, similar to well-developed semiconductor pro-
cesses, nanofabrication at laboratory stage mainly involves nano-
lithography, dry etch, and/or lift-off of unwanted metal films that
are deposited on patterned resists. Since the booming of nano-
lithography in 1980s, numerous nanolithography techniques have
been developed, among which the state-of-the-art electron-beam
lithography has been most widely implemented for patterning
mesoscopic structures or systems with unique advantages of high
resolution in feature size, high reliability in processing, high accu-
racy in positioning/alignment, and high flexibility in pattern repli-
cation. Nowadays, resolution capability as good as sub 10 nm by
EBL has been repeatedly reported [1], which is fine enough to sat-
isfy most of the demands as far as feature size concerned. Dry etch
as opposed to wet etch which is difficult in dealing with nanostruc-
tures, belongs to the subtractive approach in pattern transfer
following the replicated patterns by nanolithography. On the other
hand, the lift off process after metal deposition belongs to the
additive approach. These pattern transferring processes are the
fundamental techniques to form the major nanofabrication
E-mail address: yifangchen@fudan.edu.cn
cant progresses in the development of these processes by EBL and
demonstrate their wide applications.
2. Development of electron beam lithography and its applications
2.1. Processing study of electron beam lithography
Processing study in electron beam lithography (EBL) mainly
covers resist property, resist profile control by EBL of either single
layer or multiple layers, and pattern transfer by either lift-off or
etch. In this part of the review, an overview of various common
resists will be given at first, then three typical cases of processing
study will be described.
2.1.1. EBL properties of most common resists
Up to date, the most common high resolution EBL resists can be
categorized, according to their working principles, into two big
groups, one contains PMMA (e.g. MW350K [2], MW100K [3]),
PMMA/MAA [2], ZEP [4,5] and HSQ [6,7], etc. The other one, termed
as chemically amplified resists (CARs), includes UVIII [8–10],
UV1116 [11], UVN-30 [12–14], SU-8 [15], SAL-601 [16], NEB
[17,18] and Calix[n]arene [19,20], etc. Resists in the first group
can also be distinguished, according to their chemical structures,
into organic (PMMA for example) and inorganic (HSQ) for the

http://dx.doi.org/10.1016/j.mee.2015.02.042
0167-9317/Ó 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
2 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx

Fig. 1. Collections of contrast curves of most frequently used EBL resists by the author. The dash line with purple color is added as a guide for calculating the sensitivity and
the contrast of a particular resist. In this particular figure, the line passes through two points of the PMMA (350 K) at 25% and 75% of the normalized remaining thickness,
respectively. Its intersect point on the dose-axis measures the clearing dose (the reverse of sensitivity value) and its slope is the definition of contrast of the resist [5–22]. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 1
Comparison of lithography property for most frequently used e-beam resists.

Resist Resist tone Sensitivity (lC/cm2) Contrast CD (nm) Developer Chemical structure Refs.
PMMA (350 k) Positive 230 (100 keV) 3–4 65 MIBK:IPA, O-xylene Poly methyl methacrylate [10]
PMMA(100 k) Positive 163 (100 keV) 3–4 65 MIBK:IPA, O-xylene Poly methyl methacrylate [3]
PMMA–MAA Positive 84 (100 keV) 3–4 10 MIBK:IPA, O-xylene Poly(methyl methacrylate-co-methacrylic acid) [10]
UVIII Positive 18 (100 keV) 11.8 30 CD26 TMAH [8,10]
UV1116 Positive 38–90 (100 keV) 10.4 30 CD26 TMAH [11]
ZEP520 Positive 73 (100 keV) 7.8 10 O-xylene a-Hloromethacrylate + a-methylstyrene [4,5]
UVN-30 Negative 5 (100 keV) 6 90 CD26 Unavailable [14]
HSQ Negative 334 (100 keV) 1.1 65 CD26, MIBK, TMAH Hydrogen silsequioxane [7,18]
SU-8 Negative 3–7 (100 keV) 3.5 20–30 EC solvent, cyclopentenone 8 epoxy groups [15,18]
SAL-601 Negative 10–30 (50 keV) 5–6 60 TMAH – [16]
NEB Negative 9–14 (50 keV) 4.7 Sub 10 CD26 – [17,18]
Calix[n]arene Negative 80–150 (50 keV) 1.5–1.9 10–20 MIBK p-Methylcalix[n]arene [19,20]

purpose of resolution enhancement. The Resist sensitivity and con-

trast are the two basic characteristics to describe its EBL property.
Resolution, or minimum line-width in terms of critical dimension
(CD) of a resist is certainly an important character, but it is strongly
related to its environment (substrate, single or multiple layer, line-
edge roughness, etc.), resist thickness and developing dynamics.
Fig. 1 presents the contrast curves of these resists as a whole family
picture for comparisons. The sensitivities and contrasts derived
from these curves are concluded in Table 1. Also summarized in
the table are the minimum critical dimensions (CD) of these resists.
Fig. 2 is the bar chart of the feature size ranges of each family
member. The concluded characteristics above can be used as a
good guide for designing the resist layer stacks to fulfill various
kinds of tasks for pattern transfer. Table 2 concludes the practical
applications of these resists and their stacks.
Fig. 2. The bar chart for the feature size ranges of the most frequently used e-beam
In designing multiple layers of resists, three major kinds of pro- resists. The resolution limit of each resist is from: Refs. [17,20] for SU-8, Ref. [16] for
files are desired as schematically shown in Fig. 3, respectively UVN-30, Refs. [19,20] for NEB, Refs. [9,20] for HSQ, Refs. [2,12] for UVIII, Ref. [13] for
corresponding to three arrangements of sensitivity for the applica- UV1116, Refs. [6,7] for ZEP-520A, Ref. [12] for PMMA–MAA and Ref. [12] for PMMA.
tions in undercut profile for lift-off [21], multistep profile for T
shape gates [22] or Aztec configuration for structural color
[23,24] and vertically periodical gratings for 1D photonic crystals
as sensors [25]. The technical methods will be reviewed in the sub- [28] and nanodevices [30,31], etc. Extensive studies have been
sequent sections. attempted, including processing conditions such as curing tem-
peratures [32], developer composition [33], aging and delay time
before exposure [34]. In addition, developing conditions such as
2.1.2. A hot developing process for dense HSQ patterns prebake temperature and developer concentration have been para-
In the resist family of EBL as described above, one of the fre- metrically studied [35]. However, one of the key issues in the EBL
quently used resists is the inorganic resist, HSQ [26] for its broad of HSQ for dense patterns such as equal lines/spaces is the diffi-
applications such as a low permittivity (low-k) interlayer dielectric culty in removing the residual resist in between lines as shown
(ILD) in ICs technology [27], a resist in nanoimprint lithography in Fig. 4. Plasma etch had to be used with very limited effect but
[28], etch masks [29] in reactive ion etch, nanoscale metamaterials significantly deteriorated the lithography quality.

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx 3

Table 2
Summary of profiles and applications for the most frequently used e-beam resists.

Multilayer resist stacks Profiles Applications

PMMA–MMA/PMMA, PMMA–MMA/Ge/PMMA, ZEP/PMMA, PMMA(LW)/PMMA(HW), UVIII/Ge/ Undercut Lift-off, narrow metallic lines, pyramid gates, binary
PMMA, UVIII/LOR/PMMA, LOR/PMMA, PMMA/HSQ, PMMA/NEB, PMMA/UVN-30 gratings, nano metallic devices, etch masks
PMMA/PMMA–MAA, PMMA/Ge/PMMA–MAA, PMMA/Al/PMMA–MAA, PMMA/ZEP, PMMA/Al/ Steps T shape gates, optical components, surface relief
UVIII, PMMA/LOR/UVIII, ZEP/UVIII structures, structural color
PMMA/PMMA–MAA/PMMA, PMMA/ZEP/PMMA, PMMA/LOR/PMMA, ZEP/UVIII/ZEP Vertical Biosensors, gas sensors, structural color, vertical 1-D
gratings photonic crystals

Fig. 3. Three different trilayer stacks of resists for three profiles. The sensitivities of the resists (Si, i = 1–3) have the relationship of S1 > S2 > S3.

Fig. 4. Schematic diagrams showing residual resist existing in the gaps of dense lines. This happens frequently to negative tone resists such as HSQ and UVN-30 etc. [38]. (a)
The residual resist exists in the gaps. (b) By a hot developing process the gap bottom is clear. Raised temperature in the developing procedure increases the dissolution rate of
the developer. In the same principle, this hot developing process should be applicable to both negative tone and positive tone resists.

Fig. 5. Comparisons of lithography results of HSQ dense lines/spaces with different developing temperatures. (a) and (b) are for 100 nm equal lines/spaces. (c) and (d) are for
30 nm line-width and 100 nm pitch. The residue between HSQ lines is observed for the developing temperature at 30 °C (a and c), which is clearly removed at the 50 °C (b and
d) [36].

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
4 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx
Fig. 6. SEM micrographs of typical undercut profiles in the PMMA/LOR bi-layer structure, showing the opening trench width effect on the undercut length. The trench widths
were measured as: (a) 120 nm; (b) 80 nm; (c) 30 nm, respectively [21].
150
Resist thickness
30 nm
Undercut length (nm)
70 nm
100 120nm
50
0
50 100 150
Opening line width (nm)
Fig. 7. The resist thickness dependence of the undercut length for three LOR
thickness, 30, 70 and 120 nm. The dissolution time is fixed to be 80 s and the CD26
concentration is 60% [21].
The author of this review developed a hot developing process
[36] by raising the developing temperature up to 50 °C, effectively
solving this problem as demonstrated in Fig. 5. The contrast curve
is the tool in this work to figure out the processing window. The
hot developing method has been widely applied since this work
even for other resists [37–40]. Ultrasonic agitation could also be
applied for the removal of residual resist, creating clear trenches
[36] for subsequent processes.
2.1.3. A LOR/PMMA bilayer technique for lift-off process
Lift off process is generally the key step in nanofabrications for
pattern transfer, but is also widely recognized as unstable and
unreliable such that it is not implemented in industrial manufac-
ture. The way to increase its success chance is to either increase
the resist thickness when a single layer is applied, or using a
bilayer resist stack with an undercut profile as shown in Fig. 3a.
However, in practice, it is always a challenge to generate such a
clear and controllable undercut as expected [41]. A LOR/PMMA
bilayer resist stack was proposed in 2003 and systematically inves-
tigated [21]. Undercut length up to 300 nm can be achieved by
selectively dissolving the LOR layer in a diluted alkali solution such
as CD26 that is a developer for UVIII, as presented in Fig. 6. Clear
undercut can also be observed for as narrow as 30 nm trenches
as shown in Fig. 7. This process has enabled the successful lift-off
for narrow lines of metals with 20 nm in width [42]. This technique
has found other applications as well [42–45].
Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
EB exposure for head
EB exposure for foot
UVIII
Al layer
PMMA (a)
GaAs
(b)
200
Ti/Pd/Au
(c)
Fig. 8. The process flow chart and the UVIII/Al/PMMA sandwiched structure for T-
shaped gates on GaAs substrate. (a) Step 1: coating of PMMA, Al film and UVIII, and
one step e-beam exposure, (b) step 2: development of UVIII and PMMA, and
chemical recess, (c) step 3: metalization and lift-off [10].
2.2. EBL for T shape gates and HEMTs
T shape profiles are used in high performance metal–semicon-
ductor-field effect transistors (MESFETs) [46,47] and high electron
mobility transistors (HEMTs) for low gate resistance. This is essen-
tial for high-gain, low-noise, and high-power devices. To this end, T
shape gates, or mushroom gates as appearing in literature as well,
were first proposed back to 1980s, fabricated by angle evaporation
process [47], X-ray lithography [48], optical lithography [49] and
electron beam lithography [50]. Since then, as many as over 50
kinds of nano-processes for T shape gates have been developed
[51–54], indicating the importance of semiconductor microwave
devices. The concentration of EBL study for T shape gates are on
the profile control of T shape, including its foot-width, head-width,
aspect ratio (head-width/foot-width), height of head, effect of
resist sensitivity ratio on T shape profile, etc. Some hybrid pro-
cesses using optical exposure for the head and electron beam expo-
sure for the feet have been reported [9,55,56]. These have the
 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx 5

Fig. 9. SEM micrographs of resist profiles and fabricated gates. (a) and (c) are for T shape profile and T gate, respectively; (b) and (d) are for C shape profile and gate,
respectively. The foot-widths for all the gates are 100 nm [2].

advantage of speed but involve a more complex process and
require accurate pattern alignment. Despite the broad variety of
the processes, gate fabrications are often carried out using a one-
step electron beam lithography process in multilayer resists with
different sensitivities, as schematically shown in Fig. 3b. Bilayers
of polymethylmethacrylate (PMMA) and the copolymer (PMMA–
MAA) have been dominantly used for T shape gates for over two
decades, but the relatively small difference in sensitivity between
these resists gives rise to a small process latitude and poor profile.
By the year of 2000, several novel nanoprocesses using UVIII resist
to replace the top copolymer emerged, having significantly
improved the processing control over T shape profile [2,10].
2.2.1. A PMMA/Al/UVIII process for T shape gates and HEMTs
Traditional method of T shape gate fabrication is to use bilayer
of PMMA/PMMA–MAA. When the top layer of resist, copolymer
was replaced by a Shipley chemically amplified UVIII resist [2,10]
as shown in Fig. 8, a number of advantages appeared. First, the
UVIII resist is 12.8 times more sensitive to electron beam exposure
than PMMA (the contrast curves in Fig. 1), allowing the head parts
of T- and C- shaped profiles to be written without significantly
affecting the footprint, meanwhile facilitating the replication of
resist profiles with large aspect ratios. Second, the time taken to
write the head is approximately five times shorter than when using
a bilayer of PMMA and PMMA–MAA copolymer. Third, the foot-
width change by exposure dose of PMMA/Al/UVIII bilayer is smal-
ler than that of traditional PMMA/PMMA–MAA by a factor of 3,
indicating a better CD control [10]. Finally, the UVIII resist is based
on environmentally friendly solvents [8] and has improved dry
etch resistance compared to PMMA–MAA copolymer and PMMA
resists. This has benefits for dry etch gate recess processes.
However, the only drawback is the necessity of an extra thin layer
of Al to separate the UVIII from the bottom layer of PMMA for
avoiding polymer intermixing. Much better defined T shape pro-
files as well as T-shape gates have been achieved (Fig. 9) with
broader processing windows as shown in Fig. 10.
Owing to the large difference in the sensitivity between the top
layer and bottom layer of the resists, ultra broad head-widths were
achieved as shown in Fig. 11 for gates with both T- and C-shape,
respectively [2]. The aspect ratio as high as 40/1 for the T shape
and 35/1 for the C-shape have been the world record for the high-
est aspect ratio to date.

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
6 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx

(a) 200 9 outstanding performance. The fabricated p-HEMTs had [57] a gm of

600 mS/mm and ft of 200 GHz. Metamorphic HEMTs had a gm of
180 8 1500 mS/mm and ft of 350 GHz, which were the fastest transistors

Headwidth/footwidth
of their kind in the world in 2003 [57].
160
Footwidth (nm)

7
140 2.2.2. A PMMA/SiNx/UVIII process for sub-30 nm T shape gates
6 Starting from PMMA/Al/UVIII process for T shape gates in 1999
120
[2,10], further advance was made to replace the thin Al layer by a
100 5 20 nm LOR, saving a vacuum evaporation process [57,58], foot-
width was further reduced from 100 nm down to as short as
80 4 30 nm, as presented in Fig. 12.
60 To further reduce the foot-width below 30 nm and enhance the
3 mechanical reliability of the ultra short T shape gates, a 60 nm SiNx
600 800 1000 1200 1400 1600 1800 2000 layer was added to the sandwiched layer structure to define the
Dose (μC/cm2) foot-width, sub-30 nm foot-width was achieved thanks to the
9 tapering effect in the reactive ion etch on the thin SiNx film, which
(b) 240 was one of the shortest T shape gates in early 2000s [59].
220 8
200 Head-width/foot-width
Foot-Width (nm)

2.2.3. A two-step EBL process for ultra-short T shape gates

180 7 The investigation into the foot-width by EBL [60] reveals that
160 there is a minimum foot-width related to both the head layer
6 thickness and e-beam tension, owing to the forward scattering in
140
the top layer of the resist [60]. The ultimate limitations on foot-
120 5
width were found to be around 30 nm at 100 kV and 50 nm at
100 50 kV, when the head layer was about 1 lm of UVIII. Therefore,
80 4
to further reduce the foot-width well below 30 nm, some other
60
3 measures need to be taken. Chen et al. [61] proposed a conceptual
600 800 1000 1200 1400 1600 1800 2000 idea of a two-step EBL process to write the foot and the head sepa-
rately through a high precision registration, as schematically illus-
Dose (μC/cm2)
trated in Fig. 13. 20–25 nm foot-width was measured in that work.
Fig. 10. The foot-width (solid dots) and the ratio of headwidth over footwidth The key advantage of this process over the one-step lithography
(open circles) changing as the dose increasing from 600 lC/cm2 to 2000 lC/cm2. [50] is that it is able to replicate much shorter foot-width without
EBL was carried out using a Leica Microsystems Lithography Ltd Electron Beam any limitation to the head thickness, which is desired for a good
Pattern Generator (EBPG5 HR 100). An accelerating voltage of 50 kV and a beam
lift-off and a tall metal head.
spot size of 20 nm were used to write the T-shaped and C-shaped resist profiles. (a)
The nominal foot-width is 40 nm; (b) The nominal foot-width is 80 nm. For the
footwidth of 120 nm in (a), a 1% change in dose results in a 0.7% change in actual 2.2.4. A ZEP/Al/UVIII process for T shape gates
footwidth, whereas a 2.3% change of footwidth is measured for the same footwidth
Another extension of T shape gate technique from PMMA/UVIII
in PMMA/PMMA–MAA bilayer [2].
is the replacement of the PMMA bottom layer by a faster ZEP resist
[4] in 2001. As shown in Fig. 1, ZEP resist possesses advantages
T shape gates with 50 nm foot-width were applied on epitaxial over PMMA on two folds, one is that ZEP is four times faster than
material grown specifically for the fabrication of pseudomorphic PMMA, the other is it has considerably higher contrast, giving rise
and metamorphic HEMTs. DC and rf characterization demonstrated to a significantly improved foot profile as shown in Fig. 14. At the

Fig. 11. SEM micrographs of (a) broad head T-shape gates with head-width of 4 lm and (b) broad head C-shape gates with head-width of 3.5 lm. The foot-width is 100 nm in
both of the gates [2].

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx 7

Fig. 12. SEM micrographs of sub 30 nm T shape gates by a PMMA/SiNx/UVIII trilayer stack. (a) The resist profile as a template; (b) A fabricated sub 30 nm T shape gate. The
60 nm-thick SiNx layer ankles the 30 nm foot, significantly enhances the mechanical strength of the whole gate [59].

e beam
e beam for foots
for heads

UVIII
LOR
PMMA
Substrate Substrate
(a) (b)
Fig. 13. A schematic diagram describing the two-step electron beam lithography for ultra short T gates. (a) In the first step, the head-part was exposed by an electron beam on
the top UVIII layer with a light dose followed by a development in CD26, which also cleared away the LOR layer (described on the left hand side). For comparison, traditional
one-step exposure process was also carried out on the same substrate (the right hand part). (b) In the second step, a fine beam was aligned accurately inside the head trench
and directly exposed the PMMA layer on the bottom. Development in O-xylene formed the foot trenches written by both exposure methods [58].

same time, ZEP/UVIII bi-layer technique is as stable as PMMA/UVIII
for the foot-width at 150 nm [2].
2.3. EBL for quantum tunneling devices
Single electron tunneling transistors (SETs) as quantum tunnel-
ing devices were one of the major nanoelectronics in 1990s, in
which its nanoscale tunneling junctions are the key part requiring
nanofabrication. Based on the structure of tunneling junctions,
there have been mainly three different categories of tunneling
junctions. The first category is a one-dimensional tunneling barrier
with a saddle shaped potential in quantum dots as defined by
metallic confining gates on two dimensional electron gas (2DEG)
in III-V hetero-structures [62]; the second is the metal–insulator–
metal (MIM) tunneling junctions and finally the third one is the
resistive nanowire as a tunneling barrier. The typical example is
the nanostructured silicon on insulator (SOI). Nanofabrication with
EBL for these three kinds of tunneling junctions in SETs will be
reviewed.
2.3.1. EBL for 1D tunneling junctions in 2DEG in III-V heterojunctions
With the development of EBL into nanometer scale by 1980s as
well as the advance of material growth in molecular beam epitaxy
(MBE) for high quality layer structures in III-V semiconductor com-
pounds, nanoelectronic devices entered ballistic regime when elec-
tron mean free path is much longer than tunneling length. The
pioneering research for discovering quantization effect of ballistic
conduction was first conducted by two groups respectively from
University of Cambridge [63] and Delft University of Technology
simultaneously [64], as reported in 1988. Point contact gates, as
shown in Fig. 15, or split gates were first defined by electron beam
lithography, followed by metalization and lift-off process on 2DEG
localized in the triangle well in the heterojunctions of III-V semi-
conductors grown by MBE [65]. One dimensional quantization of
conductance was observed in GaAs/AlGaAs based 2DEG when the
channel is shorter than electron mean free path (ballistic trans-
port). Two pairs of point contact gates are fabricated together with
a pair of central gates, forming a quantum dot with tunable 0D
energy as shown by the SEM micrograph in the inset of Fig. 15b.
Similar behavior was also observed in 1DEG defined by point
contact gates in GaSb/InAs quantum well for the first time in
1996 [66]. The two barriers are the two tunneling junctions for a
single electron tunneling transistor based on 2DEG. Coulomb
blockade oscillations were measured. Technical development by
EBL for such kind of devices is pretty straightforward and numer-
ous work in quantum tunneling can be found in Refs. [67–69].

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
8 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx

(a) (b)

250
(c)
200
50kV
Footwidth (nm)

150

100 kV
100

50

0
300 400 500 600 700 800 900 1000 1100
Dose (µC/cm 2)

Fig. 14. The results of T shape gate by ZEP/UVIII bilayer technique. (a) and (b) The SEM micrographs of the T shape profile in resist (a) and the T shape gate (b). (c) The
measured relation of footwidth against exposure dose at 50 kV (dots) and 100 kV (open circles), respectively [4].

Fig. 15. One dimensional quantization in magnetoresistance (qxx) and conductance (1/qxx) measured in a 1D conducting channel defined by the point contact gates as
schematically shown in the inset of (a). The nanoscale point contact gates are fabricated by electron beam lithography and metallization of Cr/Au on 2DEG in GaSb/InAs/GaSb
quantum well. The space between the two points were measured to be 250 nm. (b) Coulomb blockade oscillations are observed through a quantum dot defined by two pairs
of point contacts with a pair of central gate. For a detailed description, refer to Ref. [63,64].

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx 9

2.3.2. EBL for metallic tunneling junctions in SETs

Quantum tunneling junctions can also be formed by a metal–in-
I sulator–metal sandwich structure, termed as metallic tunneling
gate junctions (MTJs) in literatures [70]. Fig. 16 schematically shows
C1,R1 the typical layout of a single electron tunneling transistor (SET)
C
island
g as well as its equivalent circuit. To observe Coulomb blockade
effect [71], the contact area of the MTJs has to be reduced to nanos-
V cale so that the charging energy of the central island is much
s=t-2d2tg θ C2,R2 greater than both thermal and quantum mechanical fluctuations
V [71]. Therefore, the key technology in this area is to fabricate ultra
g
small tunneling junctions.
source drain
Fulton [72] was the first to use electron beam lithography to
fabricate tunneling junctions by a bilayer of PMMA on the top of
(a) (b)
copolymer, as shown in Fig. 17. Shadow evaporation of Al film
Fig. 16. Schematic diagrams show the typical layout of the SET with double was carried out and the whole process ended by a lift-off process.
junctions to be patterned by EBL (a), and its equivalent circuit (b). The equivalent Fig. 17c presents a SEM micrograph of a typical double junction SET
circuit of single electron tunneling transistor with double metallic junctions. The with MTJs of Al/1-nm oxide/Al fabricated by EBL on SiO2 insulator
purple color defines the area to be exposed and developed in the EBL on copolymer/
[73]. The junction size is around 20 nm in this particular case.
PMMA bilayer stack. t is the distance from the two tips to the edge of the island.
After EBL, a two-step shadow evaporation of 100 nm Al film is carried out to form Coulomb blockade gap can be observed in the I–V curves measured
tunneling junctions between the island and the two tips, sandwiched by a 1-nm at 4 K in a cryostat as shown in Fig. 18a. Clear Coulomb blockade
native oxide of Al. After the metallization of Al, two overlap contacts between the oscillations in the SET output Vout, corresponding to single electron
tips and the edge of island form the metallic tunneling junctions, as symbolized by tunneling through from the source to the drain was shown in
the two capacitor/resistor boxes in (b) [73]. (For interpretation of the references to
Fig. 18b [73].
color in this figure legend, the reader is referred to the web version of this article.)

(a) e-beam (b) Free standing

bridge of PMMA
2nddep of Al 1stdep of Al

PMMA
The exposed The exposed
Copolyme volume volume Al AlOx

Substrate Substrate
substrate

(c)

Fig. 17. Schematic illustration for the formation of metallic tunneling junctions by shadow evaporation in PMMA–MAA/PMMA bilayer resist. (a) The exposure by e-beam
generates two exposed volumes (defined by dashed lines) which are connected to each other, thanks to the proximity effect; (b) after developing in MIBK:IPA (1:3), a free-
standing bridge in the PMMA layer is formed. The shadow evaporation creates overlap contacts of the two deposited Al films, sandwiched by a native AlOx layer as a tunneling
barrier in the middle (green color). (c) SEM micrograph of a fabricated SET with two overlap junctions [73]. (For interpretation of the references to color in this figure legend,
the reader is referred to the web version of this article.)

(a) 1.0 (b) 0.5

B=0T, 0mV T=300mK 60(pA)
B=2T, 0.25mV B=1T
0.4 70(pA)
B=2T, 1mV
0.5 80(pA)
B=2T, 1.25mV
Vsd (mV)
Isd (nA)

B=2T, 2mV
0.3
B=2T, 2.25mV
0.0
0.2

-0.5
0.1

-1.0 0.0
-0.004 -0.002 0.000 0.002 0.004 0 2 4 6 8 10 12 14 16
Vsd (V) Vg (mV)

Fig. 18. Characterizations of the SET device with double metallic tunneling junctions as shown in Fig. 17c. (a) DC source–drain I–V curves under various gate biases, measured
at 4 K. (b) The record of output voltage when sweeping the gate bias, showing Coulomb blockade oscillations [73].

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
10 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx

The SECOP technique can also be applied for the fabrication of

shortly spaced electrodes. As schematically illustrated by the dot-
dash lines, if the second evaporation of Al is at the angle slightly
over the normal, a short gap between two electrodes can be suc-
cessfully achieved [73]. Fig. 20c presents the fabricated electrodes
with a gap of 2–3 nm. Such kind of shortly spaced electrodes are
applied for the study of electrical transport property in single
molecule when it is positioned in between the two electrodes.
Apart from nanofabrication techniques by EBL combined with
shadow evaporation as introduced in this review, numerous alter-
natives for nanoscale tunneling junctions have been developed
[75–80].

Fig. 19. Schematic diagrams of a step-edge cut off process for 10 nm connection
junctions. (a) The layout of the device in PMMA layer after EBL from top view. The
brown color is for PMMA and the unfilled areas are open window for the deposition
of Al. (b) The cross-sectional view of the SECOP method. The detailed description of
the process is given in the text [76]. (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of this article.)
Since the first SET was lithographically fabricated [72], the
majority of the technical development in this area has been to find
the solution for ultra small junctions for high temperature opera-
tion. Fig. 19 schematically illustrates a step-edge cut off process
(SECOP). As shown in Fig. 17b, after the first step deposition of
Al, a 1-nm thick native AlxOy is grown by filling the evaporation
chamber with pure oxygen gas. The second layer of Al is then
deposited vertically down, which is cut off by the edge of the first
layer of Al, a connection junction is thus formed. By SECOP, 10 nm
MTJs were achieved by Chen at TU Delft in 1996 [73], as shown in
Fig. 20, which was one of the smallest tunneling junctions during
1996. Nakamura [74] from NEC also made 10-nm MTJs and
observed the Coulomb blockade effects at 30 K.
2.3.3. EBL for resistive tunneling junctions in silicon on SOI
The third form of tunneling junctions are made from short and
narrow constrictions in a semiconductor such as a thin layer of sili-
con on insulator (SOI) [81–83]. The nano-constriction in silicon
layer behaves like a resistor to block charge tunneling. In general,
for the junctions in SOI, an EBL process is first applied to pattern
nanoscale constrictions, followed by a metallization of metals such
as Cr or Al to form etching mask. The resistive tunneling junctions
are finally formed by a dry etching process in fluorine based
plasma using the patterned metallic layer by EBL as an etching
mask. Fig. 21 presents the fabricated SETs with 10 nm tunneling
junctions [3] on SOI. Silicon SETs exhibit important advantages
over those on 2DEG as described in Section 2.3.1 as well as those
with MTJs in Section 2.3.2. First, the micro-nanoprocess on Si is
already well established thanks to the fast development of Si semi-
conductor manufactures. The junction size well below 10 nm are
feasible [84]. Second, comparing to both III-V compound based
and metallic junction based single charge tunneling transistors,
SETs and their circuits built on SOI are more compatible to the

Fig. 20. SEM micrographs of a SET transistor with 10 nm connection junction between two deposited Al films, fabricated by the SECOP process. (a) The top view; (b) The tilt
view. (c) SEM micrograph of closely spaced electrodes in Al with a gap of 2–3 nm. For detailed explanation, refer to the text [73].

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx 11

Fig. 21. SEM micrographs of a fabricated SET on SOI with 10 nm tunneling junctions in short and narrow Si wire, which is located in between two large Si regions. (a) A top
view of the device. The sharp tip is the gate made by Si on SOI. The dot area with the diameter of 70 nm is the central island. (b) A tilt view of the SET. The junction part is so
thin such that it looks transparent [3].

Fig. 22. SEM micrographs for dielectric chiral structures with perforated profile in UVIII (a) and protrude in HSQ (b). Both of them were achieved by direct write of e-beam
with a VB6HR beam writer supplied by Vistec Lithography Ltd. [13].

Fig. 23. SEM micrographs for metallic chiral structures fabricated by EBL and metallization of 100 nm Au on Si/SiO2 substrate. (a) Perforated chiral structure and (b) Ridge-
like chiral structure [98,103].

Fig. 24. Subwavelength multiple chiral structures. (a) Schematic diagrams of multiple layer chiral structures with a twisted angle with respect to each other (on the top) and
the tilt view without a twist. (b) The SEM micrograph for the fabricated bilayer chiral structures of Al on quartz. The solid and dashed lines highlight two aligned elements
sandwiched by a 50 nm SiNx film. The twist angle is 15°. (c) The measured manipulation of polarization status by the bilayer chiral planes. The solid line is a numerical fit. The
dashed line indicates the true level of the optical activity component in the rotation. The rotary power is enhanced by one order of magnitude, comparing to single layer of
chiral ones [104].

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
12 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx

Fig. 25. SEM micrographs for fish scale patterns. Top row: resist pattern; bottom row: the fish scale pattern with 50 nm Al lines on quartz. Both the line width and thickness
of Al is 50 nm [43].

E-beam 200 nm Al

quartz quartz quartz quartz

quartz

(a) (b) (c) (d)

Fig. 26. The process flow from (a) to (d) for nano holes array in 100-nm thick Al as planar photonic crystals on quartz. (a) E-beam exposure, (b) development, (c) deposition of
100 nm Al and (c) fabricated hole array after lift-off [13].

Fig. 27. SEM micrographs for the fabricated quasi crystals of hole array in 100 nm Al on quartz. (a) The whole photonic quasi crystal. (b and c) The close-up view of crystal.
Because of the illusion, the holes appear as if dot array [107].

existing Si-based IC technique. Finally, it is relatively more cost
effective than MTJs and those in quantum dots on III-V compounds.
In recent years, similar works on SETs by replacing the silicon
on insulator (SiO2) with a single atomic layer of graphene have also
been seen [85]. Even though, the main technical procedure remains
unchanged. SETs based on a single molecule such as carbon nan-
otubes have also been widely investigated [86–92].
2.4. EBL for metamaterials and nanophotonic structures
In the discovery of novel physics for photonic metamaterials
and photonic nanostructures from infrared to visible wave lengths,
nanolithography has played an essential role. Similar to nanoelec-
tronic physics as described earlier, the development of nanofab-
rication techniques has made various kinds of photonic
metamaterials available for pioneering scientific research.
Numerous novel processes have been developed, involving the
state-of-the-art electron-beam lithography (EBL), nanoimprint
lithography (NIL) and hot embossing. This section will focus on
the technical development of EBL on optically functional metama-
terials and nanophotonic structures including dielectric and
metallic structures with both single and multiple layers.
Particularly, they are chiral structures [93], fish-scale patterns as
magnetic mirrors [94], quasi-periodic hole arrays with extraordin-
ary transmission [95], metallic gratings for the generation of tra-
velling surface plasmon waves by free-electron impact [96] and
local surface plasmonic structures, etc.
2.4.1. EBL for chiral structures in optical frequencies
EBL technique is capable of fabricating planar metamaterials
and 2D photonic structures functional in a broad frequency range
from near infrared (0.75–1.4 lm) down to deep ultraviolet wave-
lengths (100 nm) [13,97–99]. This section will cover the develop-
ment of electron beam lithography for planar chiral metamaterials
(both dielectric and metallic) and 2D nanophotonic structures in
single layer, and then describe multilayer patterning.
2.4.1.1. Dielectric chiral structures. The advantages to use dielectric
chiral metamaterials are that they can be formed entirely from
loss-less dielectric materials, and consequently have no free charge
carriers or collective charge excitations (such as excitons or plas-
mons) that can attenuate the transmitted (or reflected) intensity.

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx 13

Fig. 28. Various quasi photonic crystals as nanolens based on super oscillation theory for super resolution focus. All of the planar photonic crystals are fabricated on quartz by
the process as described in Fig. 26. The overall diameters of the crystals are in the range from 50 lm to 250 lm and each hole diameter is in the range from 150 nm to 200 nm
[107].

Fig. 29. Quasicrystal sample and its reciprocal lattice. (a) SEM image of a fragment 2020 lm2 of the quasicrystal array of holes. (b) Far-field diffraction pattern of the
quasicrystal array, obtained with white light illumination. (c) Reciprocal lattice of the quasicrystal array wherein spot diameter is proportional to the magnitude of the
spectral component. The minimum distance between two neighboring holes d = 1.2 lm. Red dashed circles show several ‘‘partial’’ Montgomery. (d) Schematic of the
experiment showing the scanning of electrical field pattern (E2) over the array illuminated with coherent light (Courtesy to [107]). (For interpretation of the references to
color in this figure legend, the reader is referred to the web version of this article.)

If such structures were fabricated on dielectric substrates such as different approaches have been reported. One is to use electron
fused silica, these devices could operate in transmission in the visi- beam direct write to sculpture the polymeric resists which
ble region of the electromagnetic spectrum. In this scenario, two becomes the dielectric chiral structures after developing. The other

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
14 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx

Fig. 30. Surface stiffness modification by focused e-beam radiation on HSQ. (a) Schematic diagram depicts the process of e-beam exposure with/without development
afterwards. The Young’s modulus (D) and RMS roughness (
) were then measured on undeveloped HSQ (b) and developed (c). The stiffness shows strong increase with the e-
beam exposure dose, indicating that the surface stiffness can be controlled by e-beam radiation [118,119].

belongs to a subtractive method [97], that is, to remove unwanted
dielectric material by dry-etch using the EBL patterned masks. In
EBL, the most frequently used e-beam resists are PMMA, PMMA–
MAA (copolymer), UVIII [8], UVN-30 [12], ZEP [5] and HSQ (flow-
able oxide) [6].
Fig. 22 presents dielectric chiral structures by e-beam direct
writer. Fig. 22a shows perforated style of chiral structure by the
direct write of e-beam on positive tone UVIII after developing
[13]. Protruded chiral structures of dielectrics were also replicated
by negative tone HSQ as shown in the Fig. 22b [13]. A hot develop-
ing process was applied in this case [36] for clearing HSQ scum as
discussed in an earlier section.
2.4.1.2. Metallic chiral structures. Metallic chiral structures with
both protruded and indentation profile are fabricated by EBL fol-
lowed by metallization as shown in Fig. 23. Three metals, Au, Al
and PdAu alloy have been used for nanoscale metamaterials and
photonic structures. So far, four types of bilayer stacks have been
reported, which are PMMA–MAA/PMMA bilayer,
PMMA(MW100k)/PMMA(MW350k), LOR/PMMA (350 k) and ZEP/
PMMA. Chirality from the metallic structures have been widely
investigated [28,100–103]. However, it was found that multiple
layers of chiral structures on top of each other should exhibit much
stronger rotatory power than single layer [104]. Fig. 24 demon-
strates the enhanced chirality by a double layer of chiral structures
aligned by a high precision registration technique. Stronger rota-
tory power and negligible circular dichroism in the visible to
near-IR spectral range were measured.
2.4.1.3. Chiral and non-chiral fish scales by a PMMA bilayer
technique. Planar metamaterials possess novel physical properties
beyond the classical limit in bulk materials. One of the meta struc-
tures is fish scale (both chiral and non-chiral), which has been dis-
covered to behave like a magnetic mirror [94] in micrometer/
millimeter wavelength. Using a PMMA (MW100k)/PMMA
(MW350k) bilayer with different molecular weights, dense fish
scale patterns with relatively small undercut were replicated [3].
The contrast curves of these two resists are presented in Fig. 1.
The clearing dose for PMMA (MW100k) and PMMA (MW350k) is
198 lC/cm2 and 240 lC/cm2, respectively. Such similar sensitivi-
ties lead to relatively small undercut which is essential for dense
pattern. As illustrated in Fig. 25, fish scales composed of 50 nm
wide Al lines with both reflective mode (fabricated on the surface
of Si/Al) and transmission mode (fabricated on the quartz surface)
were fabricated [102]. Novel nano photonic physics including mag-
netic mirror behavior [94] and asymmetric optical transmission
was observed [93].
2.4.2. EBL for nano quasi photonic structures as nanolens
It has been clear that quasi-periodical crystals in dielectrics
with various point-symmetries also exhibit complete photonic
band gap (PBG) property as periodical ones do [105]. The novel
physical properties related to the interactions of electromagnetic
waves with metallic hole arrays in quasi periodical order (metallic
quasi crystals) is being discovered both theoretically and experi-
mentally, demanding technical development for the construction
of theoretically designed structures [106]. Again, for two dimen-
sional planar photonic crystals with quasi period, electron beam
lithography is the prime method at laboratory stage for basic scien-
tific research [106–108]. A typical example is the nanofabrication
technique for the replication of a quasi crystal in 100 nm Al on a
slab (quartz wafer in this work) by electron beam lithography
using chemically amplified resist, UVN-30 as well as positive CAR
resist UVIII. In this work, a particular design with 10-fold symme-
try was made in the technical development, which should be
applicable to other designs including both regularly distributed
dot arrays as planar photonic crystals and random dot arrays.
Fig. 26 are the schematic diagrams of the layer structure and the
process flow [13]. Fig. 27 presents the SEM micrographs of the fab-
ricated quasi crystals. Fig. 28 shows various fabricated photonic
crystals as nanolens based on super oscillation theory [108].
Novel photonic behaviors for the light vertically incident
through the q-crystal were observed. Fig. 29 presents the diffrac-
tion pattern through the crystal [108]. Extraordinary optical

Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx
transmission (EOT) through such kind of planar quasi-crystal was
observed [109]. Optical self-imaging of the planar quasi-periodic
arrays of holes (Montgomery Effect) was revealed by SNOM at
600 nm for several distances from the quasi-crystal [107,108].
Furthermore, it was demonstrated that such a quasi-photonic crys-
tal is able to focus light beyond diffraction limit [106].
2.5. EBL for bioscience and medications
Electron beam lithography can also find applications for bio-
science and biomedicine. For example, micro and nanofluidic chan-
nels can be readily fabricated in e-beam resist (polymers) and solid
substrates such as Si and dielectric materials [110,111]. EBL also
offers solutions in patterning a variety of periodical-submicron
structures used to promote cell growth in tissue engineering
[112–114]. One of the exciting experiments is to control stem cell
growth by surface stiffness. It is now clear that the fate of stem
cells is decided by the surface condition [115–117], on which stem
cells grow. Chen et al. was the first to propose the surface modi-
fication by e-beam radiation [118,119], as shown in Fig. 30. E-beam
radiation has the advantages of fine control of surface stiffness,
large flexibility in geometry and 2D arrays with negligible change
of surface roughness. Several e-beam curable polymers including
SU-8, PMMA (tone reversing mode) and HSQ have been tested
[120]. The Young’s modulus measurement by AFM on HSQ shows
a broad change in stiffness on the same surface by e-beam radia-
tion, successfully providing a platform for the study of stem cell
growth control [118,119].
3. Summary
Electron beam lithography based nanofabrication for a wide
range of nanostructures, nanoelectronic devices, nanophotonic
crystals and metamaterials in optical frequencies are reviewed.
The main focus in this review lies in achieving pattern transfer
by multiple layer approaches, making great use of the differences
in sensitivity, contrast and dissolution rate between resists. Most
frequently used EBL resists are compared in association with their
lithography property, to illustrate the lithography principles with
multiple layers of resists for pattern transfer as a whole. Various
combinations of these resists for multiple layer stacks enable peo-
ple to create a wealth of tactics for constructing desired nanostruc-
tures. A great number of practical examples are given in the
illustration of nanofabrication by EBL. It is believed that the princi-
ple and experience of multiple layer lithography, concluded in this
paper, are applicable for other kinds of nanostructures and nanode-
vices as well.
Acknowledgements
The author acknowledges all the important contributions from
his colleagues in the preparation of this review. Notably, most of
the work in T shape gates was carried out together with Dr. D. S.
Macintyre, Dr. Stephen Thoms and many others at the James
Watt Nanofabrication Centre (the former Nanoelectronics research
centre), School of Engineering, University of Glasgow in the UK
under the EPSRC grant (Nanofabrication technology driven by high
speed devices: GA/M93383). My PhD student, Dr. Jiarui Tao, Dr.
Adnan Malik, the visiting postdoc, Dr. Xudi Wang, as well as my
colleagues, Dr. E. Huq and others at Micro Nanotechnology
Centre, Rutherford Appleton Laboratory in the UK had made bril-
liant contributions in dry etch and nanoimprint under TSB project
(Ref. CHBS/004/00090C) and EU FP7 project (MICROCARE, Ref.
PIRSES-GA-2009-247641). The author enjoyed the long time col-
laborations with Prof. Genhua Pan from University of Plymouth
Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
15
under EU FP6 grant (BMR project: NMP4-CT-2003-505282), and
Prof N. I. Zheludev and his team from University of Southampton
under the EBSRC BT project (Nanoscope: EP/F040644/1). It is
impossible to make this review happen without their marvelous
input. The author is also grateful to his current colleagues from
Fudan University, both Prof Ran Liu and Prof Xin-Ping Qu for their
constant support and friendship.
References
[1] W.W. Hu, K. Sarveswaran, M. Lieberman, G.H. Bernstein, J. Vac. Sci. Technol., B
22 (2004) 1711–1716.
[2] Y. Chen, D. Macintyre, S. Thoms, J. Vac. Sci. Technol., B 17 (1999) 2507–2511.
[3] Q. Wang, Y.F. Chen, S.B. Long, J.B. Niu, C.S. Wang, R. Jia, B.Q. Chen, M. Liu, T.C.
Ye, Microelectron. Eng. 84 (2007) 1647–1651.
[4] Y. Chen, D. MacIntyre, S. Thoms, Microelectron. Eng. 57 (2001) 939–943.
[5] D.M. Tanenbaum, C. Lo, M. Isaacson, H. Craighead, M. Rooks, K. Lee, W. Huang,
T. Chang, J. Vac. Sci. Technol., B 14 (1996) 3829–3833.
[6] A. Grigorescu, C. Hagen, Nanotechnology 20 (2009) 292001.
[7] H. Namatsu, Y. Takahashi, K. Yamazaki, T. Yamaguchi, M. Nagase, K. Kurihara,
J. Vac. Sci. Technol., B 16 (1998) 69–76.
[8] D. Macintyre, S. Thoms, Microelectron. Eng. 35 (1997) 213–216.
[9] M. Ahmed, H. Ahmed, J. Vac. Sci. Technol., B 15 (1997) 306–310.
[10] Y. Chen, D. Macintyre, S. Thoms, Electron. Lett. 35 (1999) 338–339.
[11] B.-R. Lu, Y. Chen, E. Huq, X.-P. Qu, R. Liu, J. Nanosci. Nanotechnol. 10 (2010)
7130–7133.
[12] Z.S. Masnyj, P.J. Mangat, E.S. Ainley, K.J. Nordquist, D.J. Resnick,
Microlithography 2000, International Society for Optics and Photonics,
2000. pp. 525–529.
[13] Y.F. Chen, A.S. Schwanecke, N.I. Zheludev, in: G. Dewar, M.W. McCall, M.A.
Noginov, N.I. Zheludev (Eds.), Complex Photonic Media, Spie-Int Soc Optical
Engineering, Bellingham, 2006 [P3200–P3200].
[14] A.J. van Dodewaard, W. Ketelaars, R.F.M. Roes, J.A.J. Kwinten, F. van Delft, A.J.
van Run, A.K. van Langen-Suurling, J. Romijn, Microelectron. Eng. 53 (2000)
461–464.
[15] A. del Campo, C. Greiner, J. Micromech. Microeng. 17 (2007) R81–R95.
[16] E.A. Dobisz, C.R.K. Marrian, J. Vac. Sci. Technol., B 15 (1997) 2327–2331.
[17] L.E. Ocola, C. Biddick, D. Tennant, W. Waskiewicz, A. Novembre, J. Vac. Sci.
Technol., B 16 (1998) 3705–3708.
[18] R. Bonam, P. Verhagen, A. Munder, J. Hartley, J. Vac. Sci. Technol., B 28 (2010)
C6C34–C36C40.
[19] S. Manako, Y. Ochiai, H. Yamamoto, T. Teshima, J. Fujita, E. Nomura, J. Vac. Sci.
Technol., B 18 (2000) 3424–3427.
[20] H. Sailer, A. Ruderisch, W. Henschel, V. Schurig, D.P. Kern, J. Vac. Sci. Technol.,
B 22 (2004) 3485–3488.
[21] Y. Chen, K. Peng, Z. Cui, Microelectron. Eng. 73 (2004) 278–281.
[22] M.A. Thompson, L.M. Jelloian, L.D. Nguyen, U.K. Mishra, J. Vac. Sci. Technol., B
8 (1990) 1339–1342.
[23] J.J. Cowan, JOSA A 7 (1990) 1529–1544.
[24] J.J. Cowan, J. Opt. Soc. Am. A Opt. Image Sci. Vis. 7 (1990) 1529–1544.
[25] B. Cui, T. Veres, Microelectron. Eng. 85 (2008) 810–813.
[26] H. Namatsu, T. Yamaguchi, M. Nagase, K. Yamazaki, K. Kurihara,
Microelectron. Eng. 41 (1998) 331–334.
[27] P.-T. Liu, T. Chang, T. Tsai, Z. Lin, C. Chen, B. Chen, S. Sze, Appl. Phys. Lett. 83
(2003) 4226–4228.
[28] Y. Chen, J. Tao, X. Zhao, Z. Cui, A.S. Schwanecke, N.I. Zheludev, Microelectron.
Eng. 78 (2005) 612–617.
[29] J. Tao, Y. Chen, X. Zhao, A. Malik, Z. Cui, Microelectron. Eng. 78 (2005) 665–
669.
[30] S. Park, J. Liddle, A. Persaud, F. Allen, T. Schenkel, J. Bokor, J. Vac. Sci. Technol. B
Microelectron. Nanometer Struct. 22 (2004) 3115–3118.
[31] W.S. Hwang, K. Tahy, L.O. Nyakiti, V.D. Wheeler, R.L. Myers-Ward, C. Eddy Jr.,
D.K. Gaskill, H.G. Xing, A. Seabaugh, D. Jena, J. Vac. Sci. Technol., B 30 (2012)
03D104.
[32] H.-C. Liou, J. Pretzer, Thin Solid Films 335 (1998) 186–191.
[33] G.M. Schmid, L.E. Carpenter II, J.A. Liddle, J. Vac. Sci. Technol., B 22 (2004)
3497–3502.
[34] F.C. Van Delft, J. Vac. Sci. Technol., B 20 (2002) 2932–2936.
[35] W. Henschel, Y. Georgiev, H. Kurz, J. Vac. Sci. Technol., B 21 (2003) 2018–
2025.
[36] Y. Chen, H. Yang, Z. Cui, Microelectron. Eng. 83 (2006) 1119–1123.
[37] S. Choi, N. Jin, V. Kumar, I. Adesida, M. Shannon, J. Vac. Sci. Technol., B 25
(2007) 2085–2088.
[38] M. Yan, J. Lee, B. Ofuonye, S. Choi, J. Jang, I. Adesida, J. Vac. Sci. Technol., B 28
(2010) C6S23–C26S27.
[39] M.G. Wood, L. Chen, J.R. Burr, R.M. Reano, J. Nanophotonics 8 (2014). 083098–
083098.
[40] M. Rommel, B. Nilsson, P. Jedrasik, V. Bonanni, A. Dmitriev, J. Weis,
Microelectron. Eng. 110 (2013) 123–125.
[41] A. Ketterson, M. Tong, J.W. Seo, K. Nummila, K. Cheng, J. Morikuni, S. Kang, I.
Adesida, J. Vac. Sci. Technol., B 10 (1992) 2936–2940.
[42] Y. Chen, Z. Lu, X. Wang, Z. Cui, G. Pan, Y. Zhou, M. Muñoz, C. Hao, L. Yonghua,
N. Garcia, Microelectron. Eng. 84 (2007) 1499–1502.
16 Y. Chen / Microelectronic Engineering xxx (2015) xxx–xxx
[43] Y. Chen, A.S. Schwanecke, V. Fedotov, V. Khardikov, P. Mladyonov, S.
Prosvirnin, A. Rogacheva, N.I. Zheludev, E. Huq, Microelectron. Eng. 86
(2009) 1081–1084.
[44] X. Xia, H. Yang, Y. Sun, Z. Wang, L. Wang, Z. Cui, C. Gu, Microelectron. Eng. 85
(2008) 1433–1436.
[45] S. Tanner, C. Rogers, J. Vac. Sci. Technol., B 26 (2008) 481–486.
[46] M. Matsumura, K. Tsutsui, Y. Naruke, Electron. Lett. 17 (1981) 429–430.
[47] P. Chao, W. Ku, P. Smith, W. Perkins, Electron Device Letters, IEEE 4 (1983)
122–124.
[48] N. Yoshioka, T. Fujino, H. Morimoto, Y. Watakabe, H. Abe, J. Vac. Sci. Technol.,
B 8 (1990) 1535–1538.
[49] G. Metze, P. Laux, S. Tadayon, Electron. Lett. 28 (1992) 1261–1263.
[50] N. Samoto, Y. Makino, K. Onda, E. Mizuki, T. Itoh, J. Vac. Sci. Technol., B 8
(1990) 1335–1338.
[51] H. Takenaka, H. Watanabe, Y. Todokoro, M. Inoue, Jpn. J. Appl. Phys. 29 (1990)
2879–2883.
[52] K. Nummila, M. Tong, A. Ketterson, I. Adesida, J. Vac. Sci. Technol., B 9 (1991)
2870–2874.
[53] A. Marten, H. Schneider, H. Schweizer, H. Nickel, W. Schlapp, R. Lösch, H.
Dämbkes, P. Marschall, J. Vac. Sci. Technol., B 9 (1991) 2861–2865.
[54] F. Ren, S. Pearton, D. Tennant, D. Resnick, C. Abernathy, R. Kopf, C. Wu, M. Hu,
C. Pao, B. Paine, J. Vac. Sci. Technol., B 10 (1992) 2949–2953.
[55] K. Hosogi, N. Nakano, H. Minami, T. Katoh, K. Nishitani, M. Otsubo, M.
Katsumata, K. Nagahama, Electron. Lett. 27 (1991) 2011–2012.
[56] Y.-L. Lai, E.Y. Chang, C.-Y. Chang, H.-P.D. Yang, K. Nakamura, S. Shy, Jpn. J.
Appl. Phys. 35 (1996) 6440–6446.
[57] Y. Chen, D. Macintyre, X. Cao, E. Boyd, D. Moran, H. McLelland, M. Holland, C.
Stanley, I. Thayne, S. Thoms, J. Vac. Sci. Technol., B 21 (2003) 3012–3016.
[58] Y. Chen, K. Peng, Z. Cui, Microelectron. Eng. 73 (2004) 662–665.
[59] Y. Chen, D. Edgar, X. Li, D. Macintyre, S. Thoms, J. Vac. Sci. Technol., B 18
(2000) 3521–3524.
[60] Y. Chen, D. Macintyre, S. Thoms, Microelectron. Eng. 53 (2000) 349–352.
[61] Y.F. Chen, K.W. Peng, Z. Cui, Microelectron. Eng. 73–4 (2004) 662–665.
[62] B.J. Vanwees, H. Vanhouten, C.W.J. Beenakker, J.G. Williamson, L.P.
Kouwenhoven, D. Vandermarel, C.T. Foxon, Phys. Rev. Lett. 60 (1988) 848–
850.
[63] D. Wharam, T. Thornton, R. Newbury, M. Pepper, H. Ahmed, J. Frost, D. Hasko,
D. Peacock, D. Ritchie, G. Jones, J. Phys. C: Solid State Phys. 21 (1988) L209.
[64] B. Van Wees, H. Van Houten, C. Beenakker, J.G. Williamson, L. Kouwenhoven,
D. Van der Marel, C. Foxon, Phys. Rev. Lett. 60 (1988) 848.
[65] A. Cho, J. Arthur, Prog. Solid State Chem. 10 (1975) 157–191.
[66] Y. Chen, D. Symons, M. Lakrimi, A. Salesse, G. Houston, R. Nicholas, N. Mason,
P. Walker, Superlattices Microstruct. 15 (1994) 41.
[67] D. Spirito, G. Frucci, A. Di Gaspare, L. Di Gaspare, E. Giovine, A. Notargiacomo,
S. Roddaro, F. Beltram, F. Evangelisti, J. Nanopart. Res. 13 (2011) 5699–5704.
[68] S. Tsubaki, N. Fujita, F. Wakaya, Y. Yuba, K. Gamo, Superlattices Microstruct.
27 (2000) 363–367.
[69] G. Engels, M. Tietze, J. Appenzeller, M. Hollfelder, T. Schäpers, H. Lüth,
Superlattices Microstruct. 23 (1998) 1249–1255.
[70] K. Bluthner, M. Gotz, W. Krech, H. Muhlig, T. Wagner, H.J. Fuchs, D. Schelle, L.
Fritzsch, B. Nachtmann, A. Nowack, J. Phys. IV 6 (1996) 163–167.
[71] H. Grabert, M.H. Devoret, Single Charge Tunneling: Coulomb Blockade
Phenomena in Nanostructures, Plenum Pub Corp, 1992.
[72] T. Fulton, G. Dolan, Phys. Rev. Lett. 59 (1987) 109.
[73] Y. Chen, P. Hadley, C. Harmans, J.E. Mooij, G.I. Ng, S.F. Yoon, A new process for
nanometer scale devices, Spie – Int Soc Optical Engineering, Bellingham,
1997.
[74] Y. Nakamura, D. Klein, J. Tsai, Appl. Phys. Lett. 68 (1996) 275–277.
[75] G.L. Chen, W.T. Lai, D.M.T. Kuo, P.W. Li, Ieee, Room-temperature Observation
of Large Coulomb-blockade Oscillations from Germanium Quantum-dot
Single-hole Transistors with Self-aligned Electrodes, Ieee, New York, 2007.
[76] M. De Vittorio, M.T. Todaro, V. Vitale, A. Passaseo, T.K. Johal, R. Rinaldi, R.
Cingolani, S. Bernardi, Microelectron. Eng. 61–2 (2002) 651–656.
[77] H.C. George, T.A. Orlova, A.O. Orlov, G.L. Snider, J. Vac. Sci. Technol., B 29
(2011).
[78] Y.C. Lee, A.O. Orlov, G.L. Snider, J. Vac. Sci. Technol., B 29 (2011).
[79] A.A. Tseng, A. Notargiacomo, T.P. Chen, J. Vac. Sci. Technol., B 23 (2005) 877–
894.
[80] J.M. Yuk, K. Kim, Z. Lee, M. Watanabe, A. Zettl, T.W. Kim, Y.S. No, W.K. Choi, J.Y.
Lee, ACS Nano 4 (2010) 2999–3004.
[81] W. Lee, P. Su, Nanotechnology 20 (2009).
Please cite this article in press as: Y. Chen, Microelectron. Eng. (2015), http://dx.doi.org/10.1016/j.mee.2015.02.042
[82] M.F. Gonzalez-Zalba, D. Heiss, G. Podd, A.J. Ferguson, Appl. Phys. Lett. 101
(2012).
[83] N.D. Akhavan, A. Afzalian, C.W. Lee, R. Yan, I. Ferain, P. Razavi, R. Yu, G. Fagas,
J.P. Colinge, I.E.E.E. Trans, Electron Devices 58 (2011) 26–32.
[84] A.T. Tilke, F.C. Simmel, R.H. Blick, H. Lorenz, J.P. Kotthaus, Prog. Quantum
Electron. 25 (2001) 97–138.
[85] J. Guttinger, F. Molitor, C. Stampfer, S. Schnez, A. Jacobsen, S. Droscher, T. Ihn,
K. Ensslin, Rep. Prog. Phys. 75 (2012).
[86] W. Fan, R.Q. Zhang, Front. Phys. China 4 (2009) 315–326.
[87] J.E. Fischer, A.T. Johnson, Curr. Opin. Solid State Mater. Sci. 4 (1999) 28–33.
[88] S. Iwasaki, M. Maeda, T. Kamimura, K. Maehashi, Y. Ohno, K. Matsumoto, Jpn.
J. Appl. Phys. 47 (2008) 2036–2039.
[89] M. Ahlskog, R. Tarkiainen, L. Roschier, M. Paalanen, P. Hakonen, in: S. Saito, T.
Ando, Y. Iwasa, K. Kikuchi, M. Kobayashi, Y. Saito (Eds.), Nanonetwork
Materials: Fullerenes, Nanotubes and Related Systems, Amer Inst Physics,
Melville, 2001, pp. 141–144.
[90] K. Ishibashi, M. Suzuki, D. Tsuya, Y. Aoyagi, Microelectron. Eng. 67–8 (2003)
749–754.
[91] R. Tarkiainen, M. Ahlskog, M. Paalanen, A. Zyuzin, P. Hakonen, Phys. Rev. B 71
(2005).
[92] D. Tsuya, M. Suzuki, Y. Aoyagi, K. Ishibashi, Jpn. J. Appl. Phys. Part 1 – Regul.
Pap. Brief Commun. Rev. Pap. 44 (2005) 1588–1591.
[93] V.A. Fedotov, P.L. Mladyonov, S.L. Prosvirnin, A.V. Rogacheva, Y. Chen, N.I.
Zheludev, Phys. Rev. Lett. 97 (2006).
[94] A.S. Schwanecke, V.A. Fedotov, V.V. Khardikov, S.L. Prosvirnin, Y. Chen, N.I.
Zheludev, J. Opt. A Pure Appl. Opt. 9 (2007) L1–L2.
[95] A. Tsiatmas, A.R. Buckingham, V.A. Fedotov, S. Wang, Y. Chen, P.A.J. de Groot,
N.I. Zheludev, Appl. Phys. Lett. 97 (2010).
[96] M.V. Bashevoy, F. Jonsson, A.V. Krasavin, N.I. Zheludev, Y. Chen, M.I.
Stockman, Nano Lett. 6 (2006) 1113–1115.
[97] W. Zhang, A. Potts, D. Bagnall, B. Davidson, J. Vac. Sci. Technol., B 24 (2006)
1455–1459.
[98] A. Potts, A. Papakostas, D.M. Bagnall, N.I. Zheludev, Microelectron. Eng. 73–4
(2004) 367–371.
[99] A. Potts, A. Papakostas, N.I. Zheludev, H.J. Coles, R. Greef, D.M. Bagnall, J.
Mater. Sci. Mater. Electron. 14 (2003) 393–395.
[100] M.X. He, J.G. Han, Z. Tian, J.Q. Gu, Q.R. Xing, Optik 122 (2011) 1676–1679.
[101] Y.Q. Ye, S. He, Appl. Phys. Lett. 96 (2010).
[102] Y.F. Chen, A.S. Schwanecke, V.A. Fedotov, V.V. Khardikov, P.L. Mladyonov, S.L.
Prosvirnin, A.V. Rogacheva, N.I. Zheludev, E. Huq, Microelectron. Eng. 86
(2009) 1081–1084.
[103] Y.F. Chen, X.D. Wang, S. Banu, A.S. Schwanecke, H. Morgan, N.I. Zheludev,
Ieee, Nano Metamaterials and Photonic Gratings by Nanoimprint and Hot
Embossing, Ieee, New York, 2006.
[104] E. Plum, V.A. Fedotov, A.S. Schwanecke, N.I. Zheludev, Y. Chen, Appl. Phys.
Lett. 90 (2007).
[105] M. Bayindir, E. Cubukcu, I. Bulu, E. Ozbay, Phys. Rev. B 63 (2001).
[106] F.M. Huang, T.S. Kao, V.A. Fedotov, Y.F. Chen, N.I. Zheludev, Nano Lett. 8
(2008) 2469–2472.
[107] F.M. Huang, Y. Chen, F.J.G. de Abajo, N.I. Zheludev, J. Opt. A Pure Appl. Opt. 9
(2007) S285–S288.
[108] F.M. Huang, N. Zheludev, Y.F. Chen, F.J.G. de Abajo, Appl. Phys. Lett. 90 (2007).
[109] V.A. Fedotov, A.V. Rogacheva, N.I. Zheludev, P.L. Mladyonov, S.L. Prosvirnin,
Appl. Phys. Lett. 88 (2006).
[110] R. Chantiwas, S. Park, S.A. Soper, B.C. Kim, S. Takayama, V. Sunkara, H. Hwang,
Y.K. Cho, Chem. Soc. Rev. 40 (2011) 3677–3702.
[111] D.Y. Xia, J.C. Yan, S.F. Hou, Small 8 (2012) 2787–2801.
[112] C.J. Bettinger, R. Langer, J.T. Borenstein, Angew. Chem. Int. Edit. 48 (2009)
5406–5415.
[113] D. Guduru, M. Niepel, J. Vogel, T. Groth, Int. J. Artif. Organs 34 (2011) 963–
985.
[114] C.M. Kolodziej, H.D. Maynard, Chem. Mater. 24 (2012) 774–780.
[115] D.E. Discher, L. Sweeney, S. Sen, A. Engler, Biophys. J. (2007). 32A–32A.
[116] A.J. Engler, M.A. Griffin, S. Sen, C.G. Bonnetnann, H.L. Sweeney, D.E. Discher, J.
Cell Biol. 166 (2004) 877–887.
[117] A.J. Engler, S. Sen, H.L. Sweeney, D.E. Discher, Cell 126 (2006) 677–689.
[118] M. Lanniel, B.R. Lu, Y.F. Chen, S. Allen, L. Buttery, P. Williams, E. Huq, M.
Alexander, J. Pharm. Pharmacol. 62 (2010) 1346–1347.
[119] M. Lanniel, B.R. Lu, Y.F. Chen, S. Allen, L. Buttery, P. Williams, E. Huq, M.
Alexander, Thin Solid Films 519 (2011) 2003–2010.
[120] B.R. Lu, M. Lanniel, M. Alexandar, R. Liu, Y.F. Chen, E. Huq, J. Vac. Sci. Technol.,
B 29 (2011).