Metal delocalization and surface decoration in direct-write nanolithography by electron beam induced deposition

Article abstract of DOI:10.1063/1.1765736

Electrical and microstructural characterization of electron beam induced deposition (EBID) was presented providing an important benchmark for the direct-write nanoscale interconnects. Direct-write nanolithography by EBID was found to be simple, versatile

Full text of DOI:10.1063/1.1765736

Metal delocalization and surface decoration in direct-write nanolithography by electron
beam induced deposition
Vidyut Gopal, Eric A. Stach, Velimir R. Radmilovic, and Ian A. Mowat
Citation: Applied Physics Letters 85, 49 (2004); doi: 10.1063/1.1765736
View online: http://dx.doi.org/10.1063/1.1765736
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/85/1?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Optical field-enhancement in metal nanoparticle arrays contacted by electron beam induced deposition
Appl. Phys. Lett. 98, 243108 (2011); 10.1063/1.3599713
Fabrication of GaN-based nanoscale device structures utilizing focused ion beam induced Pt deposition
J. Appl. Phys. 100, 024306 (2006); 10.1063/1.2215354
Fabrication by electron beam induced deposition and transmission electron microscopic characterization of sub-
10 - nm freestanding Pt nanowires
Appl. Phys. Lett. 88, 213116 (2006); 10.1063/1.2206996
Investigation of a microtriode with a planar field emitter-extractor source fabricated by direct-write
nanolithography using electron beam induced deposition
J. Vac. Sci. Technol. B 23, 781 (2005); 10.1116/1.1875332
Probing intrinsic transport properties of single metal nanowires: Direct-write contact formation using a focused
ion beam
J. Appl. Phys. 96, 3458 (2004); 10.1063/1.1779972
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
128.114.34.22 On: Sat, 29 Nov 2014 05:58:08

APPLIED PHYSICS LETTERS
VOLUME 85, NUMBER 1
5 JULY 2004
Metal delocalization and surface decoration in direct-write nanolithography
by electron beam induced deposition
a)
Vidyut Gopal, Eric A. Stach, and Velimir R. Radmilovic
National Center for Electron Microscopy, Lawrence Berkeley National Laboratory, Berkeley,
California 94720
Ian A. Mowat
Charles Evans and Associates, Sunnyvale, California 94086
(
Received 10 December 2003; accepted 30 April 2004)
The ability to interconnect different nanostructures is crucial to nanocircuit fabrication efforts. A
simple and versatile direct-write nanolithography technique for the fabrication of interconnects is
presented. Decomposition of a metalorganic precursor gas by a focused electron beam resulted in
the deposition of conductive platinum nanowires. The combination of in situ secondary electron
imaging with deposition allows for the simultaneous identification and interconnection of nanoscale
components. However, the deposition was not entirely localized to the electron beam raster area, as
shown by secondary ion mass spectrometry measurements. The electrical impact of the metallic
spread was quantified by measuring the leakage current between closely spaced wires. The origins
of the spread and strategies for minimizing it are discussed. These results indicate that, while this
direct-write methodology is a convenient one for rapid prototyping of nanocircuits, caution must be
used to avoid unwanted decoration of nanostructures by metallic species. © 2004 American Institute
of Physics. [DOI: 10.1063/1.1765736]
Planar semiconductor fabrication technology is ap-
proaching its limits even as the aggressive scaling of transis-
tor dimensions continues. One-dimensional (1D) nanostruc-
tures, such as semiconductor nanowires and carbon
nanotubes, are currently subjects of intense research as build-
ing blocks for future nanoelectronics. Significant progress
has been made in the bottom-up synthesis and characteriza-
croscopes, depositing sub-10 nm features by limiting the sec-
ondary electron emission volume.
A systematic evaluation is necessary in order to use
EBID for nanocircuit interconnect applications. In this study,
we present detailed electrical and microstructural character-
ization of EBID, providing an important benchmark for these
direct-write nanoscale interconnects. A solid metalorganic
1
precursor,
trimethylcyclopentadienyl-platinum
tion of these building blocks. Less attention has been paid to
͓
͑CH ͒ CH C H Pt͔, was vaporized by heating to 50 °C.
3
3
3
5
4
the techniques of interconnecting them for the fabrication of
functional nanocircuits. Traditionally, 1D nanostructures are
connected to larger electrodes by a multistep electron beam
lithography process. This is a complex and time-consuming
process, often resulting in poor yield due to misalignment
and incomplete metal lift-off. Nanocircuit interconnect fabri-
cation can be greatly simplified by combining patterning and
metal deposition into a single step. One such technique,
direct-write nano-patterning by electron beam induced depo-
The gas was injected, by means of a 0.5-mm-diam needle,
into the path of a scanning electron beam in a FEI Strata
2
35 M dual (focused ion and electron) beam (FIB/SEM) sys-
tem. The electron beam spot size ranged from 8 nm at an
accelerating voltage of 5 kV to 5 nm at 20 kV. The probe
current ranged from 1.6 nA at 5 kV to 2.4 nA at 20 kV.
−5
The chamber pressure was ϳ10 Torr during deposition,
which was performed at room temperature. The wires were
deposited on oxidized silicon substrates with photolitho-
graphically pre-patterned gold electrodes in order to facilitate
electrical measurements. Figure 1(a) is a SEM image of a
typical Pt wire deposited to connect two adjacent Au elec-
trodes. The raster dimensions were 40 ␮m in length and
250 nm in width. Several wires were deposited by varying
the energy of the electron beam while keeping all other pa-
rameters constant. The wires had a Gaussian-type cross sec-
tion as shown in Fig. 1(b), which is an image of a section
through a wire created by FIB milling. The dimensions of the
wire cross section were obtained by such FIB sectioning,
after resistance measurements had been performed. The mi-
crostructure consisted of nanocrystallites of Pt embedded in
an amorphous carbon-containing matrix, as seen in Fig. 1(c),
which is a cross-section TEM image. The resistivity of the
wires as a function of beam energy is shown in Fig. 1(d).
Linear current–voltage characteristics were obtained in each
case. It is noteworthy that the resistivity was relatively insen-
sitive to the electron beam energy (resistivity decreased only
2
–8
sition (EBID), has recently received a lot of attention.
In
EBID, a metalorganic precursor is vaporized and injected
into the path of an electron beam. Precursor molecules ad-
sorbed on the substrate are decomposed by beam induced
surface reactions, resulting in localized deposition of a
metal-rich conductive material. It is generally accepted that
low energy secondary electrons emitted from the substrate
2
,3,6,9
are responsible for deposition.
This limits the minimum
feature size to be larger than the beam diameter. Monte Carlo
simulations have been employed to show that secondary
electrons are excited from an area far exceeding the primary
9
,10
beam size. A thrust of recent research has been to reduce
2
the minimum feature size. Mitsuishi et al. and Silvis-
3
Cividjian et al. have performed EBID on ultra-thin electron
transparent substrates in scanning transmission electron mi-
a)
Electronic mail: vgopal@lbl.gov
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
003-6951/2004/85(1)/49/3/$22.00 49 © 2004 American Institute of Physics
28.114.34.22 On: Sat, 29 Nov 2014 05:58:08
0
1

50
Appl. Phys. Lett., Vol. 85, No. 1, 5 July 2004
Gopal et al.
FIG. 1. (a) SEM image of Pt wire deposited by EBID. (b) SEM image of the
cross section of a Pt wire created by FIB milling. The cross-sectional area
necessary for obtaining resistivity was measured from such images. (c)
Cross-sectional TEM image of an EBID Pt wire. (d) Resistivity of Pt wires
as a function of electron beam energy.
FIG. 2. (a) Two Pt lines spaced 10 ␮m apart for leakage current measure-
ments. (b) Leakage resistance as a function of electron beam energy for
pairs of Pt wires spaced 10 ␮m apart. (c) Leakage resistance as a function of
gap spacing. The deposition was carried out at 20 keV beam energy.
The Gaussian-type profile is a reflection of the deposition
cross section [refer to Fig. 1(b)]. The 20 kV wire has a larger
full width at half maximum than the 5 kV deposited wire,
consistent with a larger secondary electron excitation vol-
ume. However, the EDS data do not shed light on the elec-
trical leakage between lines spaced several micrometers
apart. This was investigated by time-of-flight secondary ion
mass spectrometry (TOF-SIMS), which is far more surface
sensitive than EDS, as well as having lower detection limits.
Figure 3(b) is a composite positive ion map showing surface
decoration around a pair of wires spaced 10 ␮m apart. Fig-
ure 3(c) shows the corresponding Pt ion map. That the wires
cannot be distinguished clearly in this map indicates the ex-
tent of metallic spread. Figure 3(d) is a carbon ion map from
the same region. Carbon appears to be more localized than Pt
to the beam raster area.
by a factor of 3 with increase in deposition energy from
5
to 20 keV).
An important requirement for nanoscale interconnects is
that there be little or no crosstalk between closely spaced
metal lines. Any leakage pathway between two metal lines
interconnecting a nanostructure can easily lead to erroneous
interpretation of its transport properties. We investigated this
by patterning pairs of closely separated wires [Fig. 2(a)] and
measuring the leakage current between them. Leakage cur-
rent increased exponentially with beam energy, as plotted in
Fig. 2(b). Similarly, the leakage current was found to de-
crease exponentially with increased gap spacing, as shown in
Fig. 2(c). Several workers have reported broadening of EBID
metal deposits exceeding the dimensions described by the
3
,4,7
beam raster.
This is consistent with the mechanism of
secondary electron mediated decomposition. However, the
measurement of electrical leakage between lines spaced sev-
eral micrometers apart implies a widespread metal decora-
tion that cannot be explained by secondary electron induced
broadening.
Our results suggest that the delocalization of metal out-
side the electron beam exposed area occurs through two dis-
tinct mechanisms. The first is the broadening due to second-
ary electron assisted decomposition, which can be easily
seen in a SEM image of the deposit. The second is a wide-
spread delocalization that can only be detected by surface
sensitive techniques such as TOF-SIMS. To illustrate this, a
substrate was exposed to a 10 keV electron beam in spot
mode ͑diameterϳ7 nm͒ while the metalorganic gas flowed
for 60 s. A SEM image of the resultant deposit and the cor-
The deposit was chemically analyzed by energy disper-
sive spectroscopy (EDS). The intensity of the Pt–M line was
tracked as a semiquantitative measure of the Pt content in the
deposited material. Figure 3(a) is a comparison of the Pt–M
intensity for wires deposited at 5 and 20 kV. The spectra
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
were collected as line scans across the long axis of the wires.
responding TOF-SIMS Pt ion map are shown in Figs. 4(a)
1
28.114.34.22 On: Sat, 29 Nov 2014 05:58:08

Appl. Phys. Lett., Vol. 85, No. 1, 5 July 2004
Gopal et al.
51
FIG. 3. (Color) (a) EDS spectra of Pt wires deposited at
beam energies of 5 and 20 keV. TOF-SIMS ion maps
depicting the extent of spread: (b) total positive ion
map, (c) Pt ion map, and (d) C ion map.
and 4(b), respectively. The TOF-SIMS map shows a Pt
spread exceeding 10 ␮m in diameter, clearly larger than the
secondary electron excitation area. We believe that the large
spread results from thermally assisted diffusion of the depos-
ited Pt, with the primary beam locally heating the substrate
during exposure. This would explain the exponential increase
in leakage current with increased energy of the primary
beam, as well as the exponential dependence of leakage with
gap spacing. We attempted to deposit Pt on a sample
mounted on a cold stage to verify this hypothesis. However,
the precursor gas condensed on the substrate and electron
beam induced deposition did not occur.
In summary, direct-write nanolithography by EBID is
simple, versatile, and hence very attractive for prototyping
nanocircuit contacts and interconnects. However, as the leak-
age current and TOF-SIMS data show, there is considerable
metallic spread. This can decorate nanostructures of interest
and contribute to misleading electrical results. Care must be
exercised in interconnecting very closely spaced nanostruc-
tures. The spread is a strong function of the electron beam
energy, but the resistivity of the deposit is rather insensitive
to this parameter. Hence, it is desirable to deposit material at
low primary beam energy ͑5–10 keV͒. In its present form,
the technique is best suited for interconnecting widely
spaced nanoscale components. The most practical means of
extending the utility of EBID is to improve the conductivity
of the as-deposited metal. This will minimize deposition
time, and prevent the formation of a leakage pathway be-
tween adjacent lines. Recent reports of improved crystal
5
quality by water vapor incorporation during deposition are
very promising in this regard.
This work was supported by the Office of Science, Basic
Energy Sciences, Division of Materials Sciences of the U.S.
Department of Energy under Contract No. DE-AC03-
7
6SF00098. The authors thank Professor Peidong Yang, Uni-
versity of California at Berkeley for the use of the probe
station for electrical measurements. Dr. Warren Moberly-
Chan of Harvard University is acknowledged for performing
cold stage depositions.
1
C. M. Lieber, MRS Bull. 28, 486 (2003).
2
K. Mitsuishi, M. Shimojo, M. Han, and K. Furuya, Appl. Phys. Lett. 83,
2
064 (2003).
3
4
N. Silvis-Cividjian, C. W. Hagen, P. Kruit, M. A. J. v. d. Stam, and H. B.
Groen, Appl. Phys. Lett. 82, 3514 (2003).
K. Molhave, D. N. Madsen, A. M. Rasmussen, A. Carlsson, C. C.
Appel, M. Brorson, C. J. H. Jacobsen, and P. Boggild, Nano Lett. 3, 47
5(
2003).
D. N. Madsen, K. Molhave, R. Mateiu, A. M. Rasmussen, M. Brorson, C.
J. H. Jacobsen, and P. Boggild, Nano Lett. 3, 1499 (2003).
6
7
Y. M. Lau, P. C. Chee, J. T. L. Thong, and V. Ng, J. Vac. Sci. Technol. A
2
I. Utke, A. Luisier, P. Hoffman, D. Laub, and P. A. Buffat, Appl. Phys.
Lett. 81, 3245 (2002).
0, 1295 (2002).
8
9
L. Dong, F. Arai, and T. Fukuda, Appl. Phys. Lett. 81, 1919 (2002).
T. E. Allen, R. R. Kunz, and T. M. Mayer, J. Vac. Sci. Technol. B 6, 2057
(1988).
FIG. 4. (a) (Color) Secondary electron image of Pt deposited in spot mode
for 60 s. The primary beam voltage was 10 kV, the beam diameter was
ϳ7 nm. (b) TOF-SIMS Pt ion map from the same region illustrates wide-
1
0
N. Silvis-Cividjian, C. W. Hagen, L. H. A. Leunissen, and P. Kruit, Mi-
This a rst pi cr leea di ss cu or fpa yc re i gd he tceo dr a at i so ni n. dicated in the article. Reuse of AIP content is subjec ct r to oe l teh c et r ot en r. mE sn ga .t: 6h1 t –t p6 :2/ /, s 6c 9i t3a t( i 2o 0n 0. a2 )i p. .org/termsconditions. Downloaded to IP:
128.114.34.22 On: Sat, 29 Nov 2014 05:58:08