Interface segregating fluoralkyl-modified polymers for high-fidelity block copolymer nanoimprint lithography

Article abstract of DOI:10.1021/ja1094292

Block copolymer (BCP) lithography is a powerful technique to write periodic arrays of nanoscale features into substrates at exceptionally high densities. In order to place these features at will on substrates, nanoimprint offers a deceptively clear path t

Full text of DOI:10.1021/ja1094292

COMMUNICATION
pubs.acs.org/JACS
Interface Segregating Fluoralkyl-Modified Polymers for High-Fidelity
Block Copolymer Nanoimprint Lithography
Vincent S. D. Voet, Teresa E. Pick, Sang-Min Park, Manuel Moritz, Aaron T. Hammack, Jeffrey J. Urban, D.
Frank Ogletree, Deirdre L. Olynick,* and Brett A. Helms*
The Molecular Foundry, Lawrence Berkeley National Laboratory, One Cyclotron Road, Berkeley, California 94720, United States
S Supporting Information
b
long-range ordered domains.4 Recent developments in nanoim-
ABSTRACT: Block copolymer (BCP) lithography is a
powerful technique to write periodic arrays of nanoscale
features into substrates at exceptionally high densities. In
order to place these features at will on substrates, nanoim-
print offers a deceptively clear path toward high throughput
production: nanoimprint molds are reusable, promote gra-
phoepitaxial alignment of BCP microdomains within their
topography, and are efficiently aligned with respect to the
substrate using interferometry. Unfortunately, when thin
films of BCPs are subjected to thermal nanoimprint, there is
an overwhelming degree of adhesion at the mold-polymer
interface, which compromises the entire process. Here we
report the synthesis of additives to mitigate adhesion based
on either PS or PDMS with short, interface-active fluoroalkyl
chains. When blended with PS-b-PDMS BCPs and sub-
jected to a thermal nanoimprint, fluoroalkyl-modified PS
in particular is observed to substantially reduce film adhe-
sion to the mold, resulting in a nearly defect-free nanoim-
print. Subsequent lithographic procedures revealed excellent
graphoepitaxial alignment of sub-10 nm BCP microdo-
mains, a critical step toward lower-cost, high-throughput
nanofabrication.
print tools and advanced mold fabrication strategies offer an
opportunity to translate templated BCP nanolithography into a
high-throughput, cost-effective process.5 Nevertheless, few ex-
amples of nanoimprint lithography (NIL) with BCPs have been
demonstrated due to technical challenges which include: (1)
difficult mold release from the imprinted film due to high
interfacial area and adhesion and (2) nonequilibrium BCP self-
assembly within the mold due to poor mobility at interfaces and
within confined geometries.6 Here we report the use of fluoro-
alkyl-modified homopolymers of either PS or PDMS as mold-
release agents when blended with PS-b-PDMS during BCP NIL
(Figure 1). These polymeric surfactant additives segregate to the
BCP-mold interface, which is directed by the fluoroalkyl coating
on the mold and the fluoroalkyl group on the homopolymer
surfactant. As a result, the low interfacial energy contacts show
considerably less adhesion than observed with the BCP alone,
and mold release is markedly improved. Concomitantly, the BCP
self-assembly is efficiently directed within the mold via graphoe-
pitaxy within one hour of the nanoimprint process. The fidelity of
the directed self-assembly of PS-b-PDMS and mold release
postimprint was readily demonstrated by etching the imprinted
film, thereby exposing the PDMS-domains previously encased in
the PS matrix. Sub-10 nm features are revealed with this procedure,
providing a key advance in BCP NIL toward a high-throughput,
robust tool for single-digit nanofabrication.
The synthesis of fluoroalkyl-modified PS (FA-PS) or PDMS
(FA-PDMS) proceeded via two distinct routes (Scheme 1).
FA-PS was prepared by controlled radical polymerization using
a fluoroalkyl-modified alkoxyamine unimer, which was afforded
from a well-described hydroxymethyl precursor first reported
by Hawker and co-workers.7 Conversion of 1 to the fluoroalkyl
alkoxyamine initiator 3 was accomplished by a modified Mitsu-
nobu reaction using ADDP and PBu3 with 1H,1H-perfluorono-
nanol 2 acting as a pseudocarboxylic acid in the transformation.
Polymerization of styrene using 3 led to well-defined FA-PS 4
across a broad range of molecular weights with polydispersities
less than 1.06 (Figure 2A). FA-PDMS 6 was prepared from 2
using living anionic ring-opening polymerization of 1,1,3,3,5,5-
hexamethylcyclotrisiloxane (D3) in the presence of triazabicy-
clodecene (TBD) as the organocatalyst.8 Prior to use in BCP
NIL, FA-PDMS-OH 5 was capped using trimethylsilylchloride
and triethylamine. Several different molecular weights over a
range of 3.8-10 kDa were prepared (Figure 2B).
lock copolymer (BCP) nanolithography is well poised to
B
become a powerful tool for patterning high-density periodic
features (typically 10-100 nm) (including arrays, gratings, or
other advanced architectures) into a variety of substrates, com-
plementing the ever-growing array of high-resolution nanofabri-
cation techniques.1 The technology is readily amenable to
generating masks, molds, and templates for fabricating compo-
nents used in micro- and optoelectronics, magnetic storage devices,
and nanoporous membranes.2 The true potential of BCP nano-
lithography, however, will be realized when the critical dimen-
sions of features reach the single-digit nanometer half-pitch
regime (thereby surpassing the resolution afforded by top-down
techniques)3 and when the processes by which features are
translated into functional device materials are cost effective.
Toward this end, strongly segregating BCPs such as polystyrene-
block-polydimethylsiloxane (PS-b-PDMS) are nearly ideal mate-
rials for obtaining well-ordered, sub-10 nm features via block
copolymer lithography. Unfortunately, their implementation has
been limited to strategies requiring expensive single-use, top-
down methods (e.g., e-beam patterned substrates) to template
Received: October 19, 2010
Published: February 15, 2011
r
2011 American Chemical Society
2812
dx.doi.org/10.1021/ja1094292 J. Am. Chem. Soc. 2011, 133, 2812–2815
|

Journal of the American Chemical Society
COMMUNICATION
Figure 2. Polymerization kinetics of (a) FA-PS or (b) FA-PDMS. For
FA-PS: [3]0 = 0.11 M; [styrene]0 = 8.35 M. For FA-PDMS: [2]0 =
[TBD]0 = 0.040 M; [D3]0 = 0.80 M. Living polymerizations were
demonstrated in both cases.
Contact angle measurements for FA-homopolymers with mo-
lecular weights (MWs) in excess of ∼3.2-3.8 kDa approached
those of their respective homopolymers with increasing MW.
Thus, in subsequent work, we employed only the smallest MW
FA-PS (Mn = 3.2 kDa) and FA-PDMS (Mn = 3.8 kDa) as these
presented the strongest case for segregation to the air-polymer
interface.
Figure 1. Fluoroalkyl-modified homopolymer (FA-polymer) additives
promote mold release and self-assembly in block copolymer thin films
during thermal nanoimprint processes. Subsequent lithographic steps
reveal underlying equilibrium phase segregated morphologies, which
show a high degree of graphoepitaxial alignment with the mold features
over the imprinted area.
XPS can be a valuable tool to distinguish surface constituents
from the bulk in thin films. In order to determine whether it
would be sensitive enough to detect fluoroalkyl-homopolymer
surfactant segregation to the air-polymer interface, XPS mea-
surements were carried out on a thermally annealed thin film of
FA-PS (Mn = 3.2 kDa, film thickness ∼11 nm) deposited onto a
silicon substrate modified with a PS brush (Figure S10). For PS,
the escape depth of photoelectrons is ∼1-2 nm for the 45°
glancing angle used here, similar to the radius of gyration for the
FA-PS.9 XPS data were acquired ∼1 min after the source was
initiated. Fluorine (F) and carbon (C) peaks were observed
within the first few minutes, with a F:C ratio of ∼0.06. A rapid,
exponential decay of this ratio to the noise floor (0.02) was also
observed within 5 min, owing to photoelectron-stimulated
desorption of fluorinated species during the measurement.
Extrapolating back to t = 0, the F:C ratio was very close to the
expected value of 0.075 for this polymer. The XPS-derived F:C
ratio of blends of FA-homopolymers (5% w/w) with either PS-
b-PDMS or PS did not present above the noise floor. These
data suggest that the interfacial segregation of FA-polymer
surfactants in blended BCP thin films most likely display
submonolayer coverage at the air-polymer interface at this
blending ratio and are thus below the detection limit. Never-
theless, as in the case of FA-PS, surface activity directed by the
fluoroalkyl chain can already be inferred from contact angle
measurements of blended films.
We next sought to determine whether a blend of these
fluoroalkyl-modified homopolymers with PS-b-PDMS BCPs
afforded well-behaved self-assembly in thin films. For our process
to be effective, we rely on the fluoroalkyl substituent to induce
macrophase separation from the bulk BCP film in favor of their
segregation to the air-polymer and mold-polymer interfaces,
as driven by the immiscibility of fluoroalkyl substituents with
either the PS or PDMS blocks and their lower interfacial energy
relative to PS and PDMS. We prepared thin films (32-35 nm) of
PS-b-PDMS (Mn,PS = 11 kDa, Mn,PDMS = 5 kDa) blended with
5% w/w of either FA-PS or FA-PDMS. After vacuum anneal-
ing the films at 170 °C for 20 h and subsequent cooling, the
morphology was revealed after a CF4/O2 plasma etching
sequence. As shown in Figures 3A, B, the cylinder segregating
Scheme 1. Synthesis of Fluoroalkyl-Modified Interface Seg-
regating Polymer Surfactants to Improve Mold Release during
BCP NIL: (a) Fluoroalkyl-Polystyrene via Nitroxide-Mediated
Controlled Radical Polymerization and (b) Fluoroalkyl-
Polydimethylsiloxane via Living Anionic Ring-Opening
Polymerization
In order to investigate the potential for low interfacial energy
fluoroalkyl groups to direct the oriented segregation of FA-PS or
FA-PDMS brushes at interfaces, we prepared thin films of these
materials or blends thereof and investigated their water contact
angles and their surface constituents via XPS. The contact angle
of a water droplet on a 35-nm film of PS (Mn = 10 kDa) on a silicon
substrate was θPS = 92°, while that for FA-PS (Mn = 3.2 kDa) was
θFA-PS = 103°. This deviation reflects the presentation of the
lower interfacial energy fluoroalkyl groups toward the polymer-
air interface. Water contact angles for PMDS (θPDMS = 103°)
and FA-PDMS (θFA-PDMS = 105°) indicated that PDMS and
fluoralkyl groups may have similar surface activity. In FA-PS
blends (5% w/w) with either a PS homopolymer (Mn = 10 kDa)
or a PS-b-PDMS (Mn,PS = 11 kDa, Mn,PDMS = 5 kDa), similar
deviations toward larger contact angles than the PS matrix
were obtained: θFA-PSþPS = 100° and θFA-PSþPS-b-PDMS = 103°.
2813
dx.doi.org/10.1021/ja1094292 |J. Am. Chem. Soc. 2011, 133, 2812–2815

Journal of the American Chemical Society
COMMUNICATION
50 W) followed by an O2 plasma (13 s, 90 W) to reveal the
underlying oxidized PDMS cylinders. Morphologies were ana-
lyzed by SEM.
In the absence of our FA-homopolymer additives, significant
adhesion of the polymer film to the mold was observed as was
poor registration of the BCP self-assembly with respect to the
boundaries imposed by trenches in the mold (Figure S13). Simi-
larly poor mold release was observed in the case of FA-PDMS
surfactants (Figure 3C). Notably, the domain spacing of the
PDMS features increased from L0 = 20 to 28 nm (Figure 3E).
Thus, we concluded that under pressure during the nanoimprint
with a fluoroalkyl-passivated mold, there were significant driving
forces for the incorporation of FA-PDMS into the PDMS
microdomains. While very little BCP residual layer was observed
between imprinted features, poor graphoepitaxial alignment was
noted, suggesting less mobility of BCPs in their presence within
the confined geometry of the mold.
By contrast, FA-PS surfactants gave markedly improved
release of the BCP polymer thin film from the mold, with
minimal BCP residual layer. Given that the optimum FA-PS
loading should depend on surface area and geometry of the mold
pattern and associated processing conditions, we tested several
FA-PS surfactant loadings. Imprinted BCP thin films at 1% w/w
loading FA-PS did not release efficiently, as was found with the
parent PS-b-PDMS, while at higher concentrations (e.g., 10% w/w)
it was difficult to obtain continuous thin films owing to dewetting
from the PS-passivated silicon substrate. Notably, a 5% w/w
loading of FA-PS with respect to the PS-b-PDMS provided
continuous films and excellent release post-nanoimprint with
no observed changes in the domain spacing within our measure-
ment error (Figures 3 and S15). Graphoepitaxial alignment was
significantly improved with FA-PS over FA-PDMS; thus, 8-nm
features arising from oxidized PDMS domains were readily
observed with excellent long-range ordering as defined by the
mold topology (Figure 3, D and F).
Beyond graphoepitaxy as a means to direct BCP alignment in
the nanoimprint mold, we also discovered shear flow played an
important role. At shorter imprint times (i.e., ∼20 min), the BCP
microdomains were generally aligned perpendicular to the mold
sidewalls as opposed to the parallel orientation expected from
graphoexpitaxy (Figure S16). The shear flow near the edges of
the trenches, on the other hand, was parallel to the trench long
axis. After longer imprint times (i.e., ∼1 h), the BCP micro-
domain alignment was predominantly parallel to the sidewall
(i.e., consistent with the thermodynamically most favorable
graphoepitaxial alignment), a process likely initiated at the ends
of the mold. A full investigation of the kinetics of this process is
underway.
The drive toward ever-smaller nanostructures written into
substrates for advanced device construction is relentless. To meet
the needs of this important endeavor, materials and nanofabrica-
tion techniques must converge within the framework of a
low-cost, high-throughput scheme. The results presented here de-
monstrate opportunities to that end brought about by nanoim-
print lithography with strongly segregating organic-inorganic
block copolymers. By blending our novel fluoroalkyl-polystyrene
with PS-b-PDMS during self-assembly, we are able to direct BCP
self-assembly within the imprint mold while concomitantly
establishing a mechanism for improved mold release, due to
the interfacial segregation of the fluoroalkyl-modified polymeric
surfactant. Thus, for the first time, nearly defect-free long-range
ordered sub-10 nm half-pitch features using BCP NIL provides a
Figure 3. Self-assembly of PS-b-PDMS thin films on PS-coated silicon
substrates in the presence of (a) FA-PDMS or (b) FA-PS (scale bars
represent 300 nm). Directed self-assembly of PS-b-PDMS in the
presence of (c) FA-PDMS or (d) FA-PS (scale bars represent 600 nm),
with magnified images thereof in (e) and (f), respectively (scale bars
represent 200 nm).
morphology was readily observed using either fluoroalkyl-mod-
ified interface-segregating homopolymer. Morphologies in the
blended films were identical to those observed for conventional
PS-b-PDMS BCP self-assembly in the absence of a fluoroalkyl
polymer surfactant. Specifically, a 20-nm pitch was determined in
all cases, with the oxidized PDMS features ∼8 nm. In addition,
the conservation of the domains’ sizes indicated that the FA-
homopolymers did not blend significantly with the individual PS
or PDMS microdomains.
Subsequently, we leveraged the FA-homopolymer surfactant
loaded BCP self-assembly process in a thermal nanoimprint
scheme for facilitating both mold release and rapid equilibrium
microphase separation of the BCP. For these experiments, we
prepared a grating mold on a silicon substrate by e-beam
lithography. A grating depth of 36 nm with 900 nm pitch (600
nm trench width) was used, thereby ensuring no more than a
single layer of cylinders in the imprinted film due to geometrical
constraints. The mold was passivated with a fluoroalkyl mono-
layer using well-described chlorosilane chemistries.6 Directed
self-assembly of BCPs in the presence or absence of FA-homo-
polymer interface segregating surfactants was investigated during
nanoimprint with the mold. Briefly, a solution of PS-b-PDMS
(1% w/w in toluene) with or without FA-homopolymer (1%, 5%,
or 10% w/w with respect to PS-b-PDMS) was spin coated onto a
silicon substrate modified with a 3-nm-thick PS-brush. The
fluoroalkyl-passivated mold was then applied using a hydraulic
press (200 psi). To ensure uniform pressure distribution, a
rubber compliance layer was added on both sides of the stack.
The press was heated at 170 °C for 1 h. After slowly cooling to
RT, the mold was carefully released from the substrate. The
imprinted substrate was then treated with a CF4 plasma (10 s,
2814
dx.doi.org/10.1021/ja1094292 |J. Am. Chem. Soc. 2011, 133, 2812–2815

Journal of the American Chemical Society
COMMUNICATION
significant step toward more robust tools for nanofabrication in
this important size regime.
(6) (a) Pina-Hernandez, C.; Guo, L. J.; Fu, P.-F. ACS Nano 2010, 4,
4776. (b) Honda, K.; Morita, M.; Masunaga, H.; Sasaki, S.; Takata, M.;
Takahara, A. Soft Matter 2010, 6, 870. (c) Alvine, K. J.; Ding, Y.; Douglas,
J. F.; Ro, H. W.; Okerberg, B. C.; Karim, A.; Lavery, K. A.; Lin-Gibson, S.;
Soles, C. L. J. Polym. Sci., Part B: Polym. Phys. 2009, 47, 2591. (d) Kim, S.;
Lee, J.; Jeon, S.-M.; Lee, H. H.; Char, K.; Sohn, B.-H. Macromolecules
2008, 41, 3401. (e) Segalman, R. A. Mater. Sci. Eng., R 2005, R48, 191.
(f) Li, L.; Yokoyama, H. Adv. Mater. 2005, 17, 1432. (g) Rolland, J. P.;
Hagberg, E. C.; Denison, G. M.; Carter, K. R.; DeSimone, J. M. Angew.
Chem., Int. Ed. 2004, 43, 5796. (h) Hong-Wei, Li, H.-W.; Huck, W. T. S.
Nano Lett. 2004, 4, 1633. (i) Deng, T.; Ha, Y.-H.; Cheng, J. Y; Ross,
C. A.; Thomas, E. L. Langmuir 2002, 18, 6719. (j) Segalman, R. A.;
Yokoyama, H.; Kramer, E. J. Adv. Mater. 2001, 13, 1152.
’ ASSOCIATED CONTENT
S
Supporting Information. Detailed experimental proce-
b
dures regarding the synthesis, characterization, and application of
fluoroalkyl-modified polymers for BCP NIL. Complete reference 5c.
This material is available free of charge via the Internet at
http://pubs.acs.org.
’ AUTHOR INFORMATION
(7) Bosman, A. W.; Vestberg, R.; Heumann, A.; Frꢀechet, J. M. J.;
Hawker, C. J. J. Am. Chem. Soc. 2003, 125, 715.
(8) Lohmeijer, B. G. G.; Dubois, G.; Leibfarth, F.; Pratt, R. C.;
Nederberg, F.; Nelson, A.; Waymouth, R. M.; Wade, C.; Hedrick, J. L.
Org. Lett. 2006, 8, 4683.
Corresponding Author
dlolynick@lbl.gov; bahelms@lbl.gov
(9) Ashley, J. A.; Anderson, V. E. IEEE Trans. Nucl. Sci. 1978, NS-26,
1566.
’ ACKNOWLEDGMENT
Bryan Cord for mold fabrication and Yeon Sik Jung, Xiaogan
Liang, and Andrꢀe Kirchner for helpful discussions. All work was
performed at the Molecular Foundry and was supported by the
Director, Office of Science, Office of Basic Energy Sciences, Division
of Materials Sciences and Engineering, of the U.S. Department of
Energy under Contract No. DE-AC02-05CH11231.
’ REFERENCES
(1) (a) Bang, J.; Jeong, U.; Ryu, D. Y.; Russell, T. P.; Hawker, C. J.
Adv. Mater. 2009, 21, 4769. (b) Cheng, J. Y.; Ross, C. A.; Smith, H. I.;
Thomas, E. L. Adv. Mater. 2006, 18, 2505. (c) Li, M.; Ober, C. K. Adv.
Polym. Sci. 2005, 190, 183. (d) Gates, B. D.; Xu, Q. B.; Stewart, M.; Ryan,
D.; Willson, C. G.; Whitesides, G. M. Chem. Rev. 2005, 105, 1171. (e)
Park, C.; Yoon, J.; Thomas, E. L. Polymer 2003, 44, 6725. (f) Cheng, J. Y.;
Ross, C. A.; Chan, V. Z. H.; Thomas, E. L.; Lammertink, R. G. H.;
Vancso, G. J. Adv. Mater. 2001, 13, 1174. (g) Park, M.; Harrison, C.;
Chaikin, P. M.; Register, R. A.; Adamson, D. H. Science 1997, 276, 1401.
(2) (a) Yang, X. M.; Wan, L.; Xiao, S.; Xu, Y.; Weller, D. K. ACS
Nano 2009, 3, 1844. (b) Jeong, S.-J.; Kim, J. E.; Moon, H.-S.; Kim, B. H.;
Kim, S. M.; Kim, J. B.; Kim, S. O. Nano Lett. 2009, 9, 2300. (c)
Stoykovich, M. P.; Kang, H.; Daoulas, K. C.; Liu, G.; Liu, C.-C.; de Pablo,
J. J.; Mueller, M.; Nealey, P. F. ACS Nano 2007, 1, 168. (d) Black, C. T.;
Ruiz, R.; Breyta, G.; Cheng, J. Y.; Colburn, M. E.; Guarini, K. W.; Kim,
H.-C.; Zhang, Y. IBM J. Res. Dev. 2007, 51, 605. (e) Guo, L. J. J. Phys. D:
Appl. Phys. 2004, 37, R123. (f) Thurn-Albrecht, T.; Schotter, J.; Kastle,
C. A.; Emley, N.; Shibauchi, T.; Krusin-Elbaum, L.; Guarini, K.; Black,
C. T.; Tuominen, M. T.; Russell, T. P. Science 2000, 290, 2126.
(3) (a) Cord, B.; Yang, J.; Duan, H.; Joy, D. C.; Klingfus, J.; Berggren,
K. K. J. Vac. Sci. Technol., B 2009, 27, 2616.(b) Semiconductor Research
Association International Technology Roadmap for Semiconductors; Semi-
conductor Research Association: San Jose, CA, 2009; (c) Guo, L. J. Adv.
Mater. 2007, 19, 495.
(4) (a) Yang, J. K. W.; Jung, Y. S.; Chang, J.-B.; Mickiewicz, R. A.;
Alexander-Katz, A.; Ross, C. A.; Berggren, K. K. Nat. Nanotechnol. 2010,
5, 256. (b) Jung, Y. S.; Lee, J. H.; Lee, J. Y.; Ross, C. A. Nano Lett. 2010,
10, 3722. (c) Ross, C. A.; Jung, Y. S.; Chuang, V. P.; Ilievski, F.; Yang,
J. K. W.; Bita, I.; Thomas, E. L.; Smith, H. I.; Berggren, K. K.; Vancso,
G. J.; Cheng, J. Y. J. Vac. Sci. Technol., B 2008, 26, 2489. (d) Jung, Y. S.;
Ross, C. A. Nano Lett. 2007, 7, 2046. (e) Bates, F. S.; Fredrickson, G. H.
Annu. Rev. Phys. Chem. 1990, 41, 525.
(5) (a) Williams, S. S.; Retterer, S.; Lopez, R.; Ruiz, R.; Samulski, E. T.;
DeSimone, J. M. Nano Lett. 2010, 10, 1421. (b) Choi, D. G.; Jeong, J. H.;
Sim, Y. S.; Lee, E. S.; Kim, W. S.; Bae, B. S. Langmuir 2005, 21, 9390. (c)
Bailey, T. C.; et al. Microelectron. Eng. 2002, 61, 461. (d) Dauksher, W. J.;
Nordquist, K. J.; Mancini, D. P.; Resnick, D. J.; Baker, J. H.; Hooper, A. E.;
Talin, A. A.; Bailey, T. C.;Lemonds, A. M.;Sreenivasan, S. V.;Ekerdt, J. G.;
Willson, C. G. J. Vac. Sci. Technol., B 2002, 20, 2857.
2815
dx.doi.org/10.1021/ja1094292 |J. Am. Chem. Soc. 2011, 133, 2812–2815