Epitaxial electrodeposition of freestanding large area single crystal substrates

Article abstract of DOI:10.1063/1.2752531

The authors report on a method for producing freestanding single crystal metal films over large areas using electrodeposition and selective etching. The method can be turned into an inexpensive continuous process for making long ribbons or a large area of single crystal films. Results from a 5×5 mm2 Ni single crystal film using electron backscattering pattern pole figures and x-ray diffraction demonstrate that the quality of material produced is equivalent to the initial substrate without annealing or polishing.

Full text of DOI:10.1063/1.2752531

Epitaxial electrodeposition of freestanding large area single crystal substrates
Jae Wook Shin, Adam Standley, and Eric Chason
Citation: Applied Physics Letters 90, 261909 (2007); doi: 10.1063/1.2752531
View online: http://dx.doi.org/10.1063/1.2752531
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/90/26?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
FePt films fabricated by electrodeposition
J. Appl. Phys. 101, 09K519 (2007); 10.1063/1.2711810
In situ observation of the grain growth of the copper electrodeposits for ultralarge scale integration
Appl. Phys. Lett. 89, 161924 (2006); 10.1063/1.2364119
Structure and magnetic properties of electrodeposited, ferromagnetic, group 3-d element films grown onto GaAs
(011) substrate
J. Appl. Phys. 93, 7634 (2003); 10.1063/1.1543916
Sharp ferromagnet/semiconductor interfaces by electrodeposition of Ni thin films onto n- GaAs (001) substrates
Appl. Phys. Lett. 82, 2853 (2003); 10.1063/1.1571229
Deformation field in single-crystal fields semiconductor substrates caused by metallization features
Appl. Phys. Lett. 74, 2352 (1999); 10.1063/1.123848
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:
129.120.242.61 On: Sun, 23 Nov 2014 00:14:23

APPLIED PHYSICS LETTERS 90, 261909 ͑2007͒
Epitaxial electrodeposition of freestanding large area single crystal
substrates
a͒
Jae Wook Shin, Adam Standley, and Eric Chason
Brown University, Box D, Providence, Rhode Island 02912
͑
Received 5 March 2007; accepted 5 June 2007; published online 26 June 2007͒
The authors report on a method for producing freestanding single crystal metal films over large areas
using electrodeposition and selective etching. The method can be turned into an inexpensive
continuous process for making long ribbons or a large area of single crystal films. Results from a
2
5
ϫ5 mm Ni single crystal film using electron backscattering pattern pole figures and x-ray
diffraction demonstrate that the quality of material produced is equivalent to the initial substrate
without annealing or polishing. © 2007 American Institute of Physics. ͓DOI: 10.1063/1.2752531͔
In many thin film applications, single crystals can im-
prove performance, e.g., by inducing preferred orientation or
eliminating grain boundaries. However, polycrystalline films
are often used because single crystal substrates ͑such as Si
wafers͒ are too expensive or difficult to work with. In this
letter, we describe a method for creating single crystal metal
thin films that are freestanding, i.e., not attached to the ma-
terial from which they are formed. The method can be turned
into a process for creating long continuous ribbon or large
area single crystal substrates that use comparatively little ma-
terial and cost less than equivalent bulk-fabricated substrate.
Improvements in factors such as cost, ease of fabrication,
mechanical properties, and weight may make single crystal
substrates usable in a wide range of technologies that were
previously not viable.
material. To create an epitaxial layer that can be separated
from the starting substrate, sacrificial layers can be deposited
between the substrate and final layer and then removed by
1
6–19
etching,
lift-off.” An alternative method for creating freestanding
a process sometimes referred to as “epitaxial
2
0
crystalline layers is the “Smart-Cut” process in which ion
implantation creates a highly defected layer of bubbles in the
substrate that fractures and separates the layer above the im-
planted region from the starting substrate.
Other techniques have been developed to fabricate
highly oriented polycrystalline substrates. The rolling-
2
1
assisted biaxially textured substrate process uses mechani-
cal rolling followed by annealing to produce preferred tex-
2
2
ture. Ion beam assisted deposition
creates preferred
orientation by using an energetic ion beam in conjunction
with film deposition. Directional solidification processes pro-
mote the regrowth of polycrystalline films with larger ori-
There are numerous applications that can potentially
benefit from these types of substrates. In high temperature
superconducting ͑HTS͒ wires, the critical current of films
grown on single crystal substrates can be higher than those
2
3
ented grains on the surfaces of substrates without an epi-
taxial orientation. Recrystallization and abnormal grain
growth can also lead to the formation of large grained thin
film material. Although these different methods are well
suited for many applications, the continuous process that we
describe has unique differences in several respects, i.e., pro-
ducing large areas or continuous long ribbons, high speed of
fabrication, potentially low cost, low weight, and processing
in a nonvacuum environment.
1
grown on textured polycrystalline substrates due to reduced
2
in-plane misorientation. Solar cells made of single crystal
material typically have higher performance than polycrystal-
3
line or amorphous material. In addition to higher perfor-
mance, thin single crystal substrates on plastic could signifi-
cantly decrease the weight of photovoltaic systems for space-
based applications. In fabrication of microelectromechanical
systems ͑MEMs͒ using the lithography, electroplating and
The method for producing freestanding single crystal
films is based on a combination of electrodeposition and
electrochemical etching steps shown schematically in Fig. 1.
In step 1, we start with a clean, flat single crystal substrate
material. In the second step, we electrodeposit a thin layer of
material that is epitaxial with the starting substrate. This
layer is referred to as the sacrificial layer because it will
4
molding process ͑LIGA͒, lower residual stress in the Ni
layers can significantly reduce deformation in the final
5
device. Many more uses can be envisioned for single crystal
substrates if they were inexpensive and commercially
available.
Single crystal substrates can be produced by several
standard methods with different advantages and drawbacks.
6
Material produced by bulk growth techniques ͑Bridgman,
7
8
float zone, or Czochralski growth͒ must be oriented, cut,
and polished to form substrates and the area of the wafer is
limited to the size of the boule from which it is cut. Single
crystal layers can be grown epitaxially on suitable single
9
–12
13
14
crystal substrates
by vapor phase, liquid phase, or
1
5
electrochemical deposition. However, the final epitaxial
layer is generally attached to the substrate so that the cost of
fabrication and area are still limited by the starting substrate
FIG. 1. ͑Color online͒ Schematic of electrochemical deposition and etching
steps used to form freestanding single crystal films.
a͒Electronic mail: jaeគwគshin@brown.edu
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:
003-6951/2007/90͑26͒/261909/3/$23.00 90, 261909-1 © 2007 American Institute of Physics
29.120.242.61 On: Sun, 23 Nov 2014 00:14:23
0
1

261909-2
Shin, Standley, and Chason
Appl. Phys. Lett. 90, 261909 ͑2007͒
FIG. 2. ͑Color online͒ Schematic of continuous electrodeposition and etch-
ing process to form freestanding single crystal films.
ultimately be removed. In step 3, we electrodeposit another
epitaxial single crystal layer of the material that will become
the final product. In the final step, the sacrificial layer is
removed by selective etching so that the final layer is re-
leased from the substrate and becomes a freestanding film.
The initial substrate material is not destroyed in the process
and may be used again for producing another film.
The deposition and etching processes described above
can be used as the basis of a process for producing continu-
ous ribbons or sheets, as shown in Fig. 2. For the starting
substrate, a long loop of single crystal material is used in-
stead of a single piece. The loop passes through several elec-
trochemical baths in which the processes described in Fig. 2
take place sequentially. In the last stage, the final layer is
etched-free from the substrate, resulting in a continuous rib-
bon of single crystal, as shown in the figure. At this point, a
backing material may be applied to the thin epitaxial layer, or
another layer of material may be deposited before the etching
process to increase the mechanical strength of the ribbon.
Although the single crystal template must initially be fabri-
cated by other means ͑such as bulk growth and polishing͒,
subsequent loops can be produced from the method de-
scribed here.
In order to develop the continuous fabrication process,
we must first develop the individual processes needed to epi-
taxially electrodeposit the layers and etch the sacrificial
layer. In the following, we report on the fabrication of indi-
vidual freestanding Ni single crystal films with Cu as the
sacrificial layer. Characterization of these samples ͑described
below͒ demonstrates that good crystalline quality can be ob-
tained in the final film, equivalent to the starting Ni substrate.
Single crystal Ni ͑001͒ disks with a diameter of 1.2 cm
were used as the starting substrates. They were mechanically
polished up to 0.03 ␮m alumina powder, followed by elec-
tropolishing in a concentrated sulfuric acid solution. After
electropolishing, the surface roughness was approximately
FIG. 3. Characterization of freestanding single crystal nickel samples
2
͑
5ϫ5 mm , 10 ␮m thick͒ showing quality of material. ͑a͒ EBSD pole fig-
ures of ͕001͖, ͕111͖, and ͕110͖ planes. FWHMs of EBSD pole distribution in
b͒ normal and ͑c͒ in-plane directions are shown.
͑
texture is developed within the first few nanometers of depo-
sition. The relationship between the deposition conditions
and the surface morphology is complex, but smooth surfaces
͑
rms roughness of 2–3 nm͒ could be obtained with proper
control of the conditions.
The top Ni layer was grown in 1.6M
Ni͑SO NH ͒ XH O/0.5M H BO to a thickness of 10 ␮m
3
2 2
2
3
3
in order to give sufficient mechanical strength to the final
freestanding material. Thinner samples were found to
wrinkle upon their release from the multilayer. The first
5 ␮m of the Ni layer was electrodeposited at the same rate as
the Cu layer ͑7–8 nm/s͒. The remaining 5 ␮m was elec-
trodeposited at twice this rate while maintaining equally
good crystallinity ͑shown by XRD͒. Alternatively, a backing
material such as a polymer or other electrodeposited layers
may be added to the epitaxial layer before the sacrificial
layer is etched out to provide mechanical support without
growing such a thick epitaxial layer.
Selective etching of the Cu layer is the most difficult part
of the process because of the small ratio of Cu/Ni exposed
surface area ͑1/50 000͒. The etching of the sacrificial layer
must be fast to release the layer in a reasonable time, but it
cannot be so aggressive that it damages the final film or the
initial substrate material. Using a nitric acid based solution, it
was possible to obtain rapid Cu dissolution with no Ni dis-
solution by the formation of a protective passive layer on the
1
.8 nm ͑rms͒, as measured by atomic force microscopy
2
͑
AFM͒. After rinsing in Milli-Q/DI water, a 5ϫ5 mm re-
2
4
gion for growth was patterned using a masking tape. Epitax-
ial growth of Cu and Ni multilayers by electrodeposition is
Ni surface. These etching conditions were sufficient to
2
separate a 5ϫ5 mm sample in approximately 10 min with
1
5
well known; the small lattice mismatch between Ni and Cu
2.5%͒ leads to good epitaxial growth, and allows the crystal
no significant damage to the Ni film and substrate. To en-
hance the separation rate, mechanical assistance ͑e.g., pulling
on the Ni film͒ may be beneficial. More rapid opening of the
interfacial region would provide fresh etchant to the
diffusion-limited Cu-solution interface and increase the Cu
etch rate. Alternatively, it may be possible to deposit a highly
tensilely stressed layer on top of the Ni that would cause the
released layer to “curl up” after the Cu is removed and en-
hance the separation of the two layers.
͑
structure of the Ni substrate to transfer to the top Ni layer
through the thin Cu buffer layer. An epitaxial 100-nm-thick
Cu layer was grown in 0.45M CuSO /2.0M H SO at the
4
2
4
rate of approximately 7–8 nm/s. The crystallinity of the Cu
layer was confirmed by electron backscattered patterns
͑
EBSP͒ and x-ray diffraction ͑XRD͒; the Cu ͑002͒ x-ray
rocking curve peak had a full width at half maximum
FWHM͒ of only 0.275°. The surface morphology varied de-
͑
A 5 mmϫ5 mmϫ10 ␮m thick Ni freestanding film
was produced by this process and characterized by multiple
techniques to determine its quality. EBSP measurements of
the Ni surface ͑on the side corresponding to the original
pending on deposition parameters such as voltage, current
density, concentration, temperature of the solution, etc. The
roughness of the substrate surface and sacrificial layer can
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:
have a large effect on the texture of the final film because the
Cu–Ni interface͒ are shown in Fig. 3. The pole figures of
129.120.242.61 On: Sun, 23 Nov 2014 00:14:23

261909-3
Shin, Standley, and Chason
Appl. Phys. Lett. 90, 261909 ͑2007͒
FIG. 4. ͑a͒ X-ray diffraction pattern
from ␪-2␪ scan of 10 ␮m thick Ni
film. ͑b͒ Rocking curve of Ni ͑002͒
peak in out-of-plane direction; ͑c͒
AFM image of surface of freestand-
ing Ni ͑001͒ film; rms roughness
=
2.8 nm.
͕
001͖, ͕111͖, and ͕110͖ planes ͓Fig. 3͑a͔͒ were obtained from
sive single crystal metal films over large areas which could
have an impact on numerous technological applications such
as HTS wires, solar cells, MEMS, and others.
2
2
500 Kikuchi patterns over a 50ϫ50 ␮m area; the very
narrow distribution indicates the good crystallinity of the
film. Several different regions ͑50ϫ50 ␮m ͒ were randomly
2
The authors gratefully acknowledge useful conversations
with Clyde Briant and the support of the Brown University
Materials Research Science and Engineering Center
chosen over the entire sample and gave the same results. To
quantify the extent of grain misorientation, the angular dis-
tributions of the ͕001͖ poles in the surface normal ͓Fig. 3͑b͔͒
and the in-plane ͓Fig. 3͑c͔͒ directions were calculated. The
FWHM of the distribution is narrow and essentially identical
in the normal ͑0.3°͒ and the in-plane ͑0.33°͒ directions.
X-ray diffraction was used to measure the orientation
through the entire thickness of the film ͓Figs. 4͑a͒ and 4͑b͔͒.
In ␪-2␪ scans, only one ͑002͒ peak was observed indicating
the film had a single ͑001͒ normal orientation. The Ni ͑002͒
rocking curve peak around the surface normal is also very
sharp ͑FWHM of ϳ0.77°͒. The morphology of the Ni sur-
face ͑on the side corresponding to the original Ni–Cu inter-
face͒ was measured with scanning electron microscopy and
AFM ͓Fig. 4͑c͔͒. The alignment of step edges parallel or
normal to each other indicates the epitaxial growth of the
layer and may be a consequence of spiral growth around
dislocation cores. The rms roughness of the surface is mea-
sured to be 2.8 nm.
͑DMR0079964͒ for this research.
1
D. Dimos, P. Chaudhari, and J. Mannhart, Phys. Rev. B 41, 4038 ͑1990͒.
2
A. Goyal, N. Rutter, C. Cantoni, and D. F. Lee, Physica C 426-431, 1083
3͑
2005͒.
M. A. Green, K. Emery, D. L. King, Y. Hisikawa, and W. Warta, Prog.
Photovoltaics 14, 45 ͑2006͒.
S. M. Spearing, Acta Mater. 48, 179 ͑2000͒.
A. Rajamani, B. W. Sheldon, E. Chason, and A. Bower, Appl. Phys. Lett.
4
5
8
1, 1204 ͑2002͒.
6
K. Hoshikawa, H. Nakanishi, K. Kohda, and M. Sasaura, J. Cryst. Growth
9
4, 643 ͑1989͒.
7
8
S. Kimura and K. Kitamura, J. Am. Ceram. Soc. 75, 1440 ͑1992͒.
J. Evers, P. Klüfers, R. Staudigl, and P. Stallhofer, Angew. Chem., Int. Ed.
42, 5684 ͑2003͒.
9
0
Y. Kamada, Y. Hitomi, T. Kingetsu, and M. Yamamoto, J. Appl. Phys. 90,
5
104 ͑2001͒.
1
N. M. Markovic and P. N. Ross, Jr., J. Vac. Sci. Technol. A 11, 2225
1͑
1993͒.
1
In summary, we have developed a process using elec-
trodeposition and selective etching processes to fabricate
high quality single crystal thin films. We have demonstrated
that the process can produce freestanding Ni films ͑5
F. J. Lamelas, C. H. Lee, Hui He, W. Vavra, and Roy Clarke, Phys. Rev. B
40, 5837 ͑1989͒.
2
1
1
C. H. Lee, Hui He, F. Lamelas, W. Vavra, C. Uher, and Roy Clarke, Phys.
Rev. Lett. 62, 653 ͑1989͒.
R. J. Molnar, W. Götz, L. T. Romano, and N. M. Johnson, J. Cryst.
3
2
ϫ5 mm , 10 ␮m thick͒ with crystalline quality equivalent to
Growth 178, 147 ͑1997͒.
that of the starting substrate ͑as confirmed by EBSP pole
figures and x-ray diffraction͒. The film was separated from
the starting Ni͑001͒ substrate by selectively etching a Cu
sacrificial layer without damaging the original single crystal
Ni template. The electrodeposition and etching processes that
we have developed can serve as the basis for a continuous
process to produce long ribbons or large areas. Scaling of
this process to production levels would produce material at
the rate of 5–50 m/h, though improvements in the deposi-
tion and etching rates ͑for instance by mechanical assistance
in the layer separation͒ could significantly increase these
rates. Although the process has been demonstrated for Ni
films, it could be extended to other materials systems ͑e.g.,
Pt for catalytic applications or Co and Fe for magnetic ma-
terials͒. Because it is based on inexpensive and scalable elec-
trochemical methods, the technique can be readily integrated
with other methods to create a continuous process that pro-
duces the entire material from the substrate up to the finished
product. This process has the potential for creating inexpen-
14
F. Kawamura, T. Iwahashi, K. Omae, M. Morishita, M. Yoshimura, Y.
Mori, and T. Sasaki, Jpn. J. Appl. Phys., Part 2 42, L4 ͑2003͒.
T. P. Moffat, J. Electrochem. Soc. 142, 3767 ͑1995͒.
E. Yablonovitch, T. Gmitter, J. P. Harbison, and R. Bhat, Appl. Phys. Lett.
1
1
5
6
5
1, 2222 ͑1987͒.
1
7
M. Konagai, M. Suginoto, and K. Takahashi, J. Cryst. Growth 45, 277
18͑
1978͒.
M. G. Lagally, MRS Bull. 32, 57 ͑2007͒.
1
9
M. M. Robert, L. J. Klein, D. E. Savage, K. A. Slinker, M. Friesen, G.
Celler, M. A. Eriksson, and M. G. Lagally, Nat. Mater. 5, 388 ͑2006͒.
U. Gösele, Y. Bluhm, G. Kästner, P. Kopperschmidt, G. Kräuter, R.
Scholz, A. Schumacher, St. Senz, Q.-Y. Tong, L.-J. Huang, Y.-L. Chao,
and T. H. Lee, J. Vac. Sci. Technol. A 17, 1145 ͑1999͒.
2
0
21
A. Goyal, F. A. List, J. Mathis, M. Paranthaman, E. D. Specht, D. P.
Norton, C. Park, D. F. Lee, D. M. Kroeger, D. K. Christen, J. D. Budai,
and P. M. Martin, J. Supercond. 11, 481 ͑1998͒.
2
2
S. R. Foltyn, P. N. Arendt, Q. X. Jia, H. Wang, J. L. MacManus-Driscoll,
S. Kreiskott, R. F. DePaula, L. Stan, J. R. Groves, and P. C. Dowden,
Appl. Phys. Lett. 82, 4519 ͑2003͒.
23
J. S. Im, R. S. Sposili, and M. A. Crowder, Appl. Phys. Lett. 70, 3434
4͑
1997͒.
2
L. Sun, C. L. Chien, and P. C. Searson, Chem. Mater. 16, 3125 ͑2004͒.
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:
29.120.242.61 On: Sun, 23 Nov 2014 00:14:23
1