Access to this full-text is provided by AIP Publishing.
Content available from AIP Advances
This content is subject to copyright. Terms and conditions apply.
AIP ADVANCES 7, 125125 (2017)
Soft nanoimprint lithography on SiO2sol-gel to elaborate
sensitive substrates for SERS detection
Fr´
ed´
eric Hamouda,1,aJean-Franc¸ois Bryche,1,3 Abdelhanin Aassime,1
Emmanuel Maillart,4Valentin Gˆ
at´
e,2Silvia Zanettini,2J´
er´
emy Ruscica,2
Daniel Turover,2and Bernard Bartenlian1
1Centre de Nanosciences et de Nanotechnologies, CNRS, Univ. Paris-Sud,
Universit´
e Paris-Saclay, C2N – Orsay, 91405 Orsay cedex, France
2SILSEF, 382 rue Louis Rustin-Archamps Technopole, 74160 Archamps, France
3Laboratoire Charles Fabry - Institut d’Optique Graduate School, CNRS,
Universit´
e Paris-Saclay, 91127 Palaiseau, France
4HORIBA Europe Research Center, Avenue de la Vauve, Passage Jobin Yvon, 91120 Palaiseau,
France
(Received 11 September 2017; accepted 18 December 2017;
published online 29 December 2017)
This paper presents a new alternative fabrication of biochemical sensor based on sur-
face enhanced Raman scattering (SERS) by soft nanoimprint lithography (S-NIL) on
SiO2sol-gel. Stabilization of the sol-gel film is obtained by annealing which simpli-
fies the manufacturing of these biosensors and is compatible with mass production at
low cost. This detector relies on a specific pattern of gold nanodisks on a thin gold
film to obtain a better sensitivity of molecules’ detection. Characterizations of SERS
devices were performed on a confocal Raman microspectrophotometer after a chem-
ical functionalization. We report a lateral collapse effect on poly(dim´
ethylsiloxane)
(PDMS) stamp for specific nanostructure dimensions. This unintentional effect is
used to evaluate S-NIL resolution in SiO2sol-gel. © 2017 Author(s). All arti-
cle content, except where otherwise noted, is licensed under a Creative Com-
mons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
https://doi.org/10.1063/1.5004122
I. INTRODUCTION
Raman Spectroscopy is an efficient analysis technique to characterize chemical composition
of molecules, this optical characterization is relied on inelastic light scattering1,2though for low
concentration it gives a weak signal. To overcome this limitation, rough metallic surfaces or metal
nanostructures or both can be used.3,4This method refers to surface enhanced Raman scattering
(SERS) and thus allows to identify precisely the molecules with their specific peaks and high enhance-
ment factors. The interest to develop new SERS-based spectroscopic sensors combined with other
spectroscopic detections for chemical or biological applications were increased during these past few
years.5,6Sample design used in this work with thin gold film under nanostructures has been already
reported for his interest in SERS and Surface Plasmon Resonance Imaging (SPRI).7–9
To pattern such device two kinds of lithographic processes can be used. The first one, refer-
enced as conventional like electron beam lithography (EBL) requires long writing times for large
areas and a significant cost for large sample productions. The second one, considered as a next
generation lithography is nanoimprint lithography (NIL) which is rapid, with low cost fabrication,
and allows large area structuring. The principle is to pattern a polymer with a mold which may be
rigid or flexible. Two main ways in NIL have been developed, thermal process (T-NIL), commonly
known as hot embossing lithography (HEL), demonstrated for the first time by S. Y. Chou in 199510
and UV-NIL developed by J. Haisma from Philips research laboratories in 1996.11 Thermal NIL
aElectronic mail: frederic.hamouda@c2n.upsaclay.fr
2158-3226/2017/7(12)/125125/6 7, 125125-1 ©Author(s) 2017
125125-2 Hamouda et al. AIP Advances 7, 125125 (2017)
allows high resolutions but needs high pressures (between 20 and 100 bar) and high temperatures
to reach glass transition temperature (Tg) of the imprinted polymer.12 While UV-NIL resists are
liquid and UV-curable, advantages are patterning at low pressures and carried out at ambient tem-
perature. It requires a transparent mold or substrate. With this method, mainly two techniques can
be distinguished: Hard UV-NIL and Soft UV-NIL. The first uses a rigid transparent stamp, generally
in quartz,13,14 which permits to achieve high resolutions close to those obtained by EBL. How-
ever, with a non-planar surfaces, critical resolutions cannot be reached. In this case, Soft UV-NIL
is useful thanks to the use of a flexible stamp which can be conforming to the shape of the sur-
face. A common polymer material used as flexible stamp is poly(dim´
ethylsiloxane) (PDMS) in early
works15–19 for its interesting physical properties such as flexibility, elasticity, low surface energy, and
UV-transparency.
To improve performances of Soft UV-NIL, technological developments have mainly focused
on new processes for mold’s manufacture with new polymers,20–24 new resists and anti-sticking
treatment.25–27
Among recent developments for NIL resists, sol-gel resist appears as an alternative.28–30 Sol-gel
is based on the polymerization of inorganic monomers to obtain partially or totally mineral films.
The polymerization uses a soft chemistry method which relies on controlled reaction mechanisms,
taking place in liquid solution and usually at ambient temperature. These mechanisms are based
on the transformation of a liquid solution (sol) into a solid material (a dry gel called xerogel) via
an inorganic polymerization process. Then sol-gel thin film can be stabilized and cured either by
annealing31,32 to permit the evaporation of residual solvent still present inside the xerogel thin film or
by photocuring.33 The interest to use such inorganic polymers as material for nanoimprint lithography
has been growing over the last decade.28–30 In another way, studies have shown the benefit of using
sol-gel films directly to obtain rough surfaces which promote the production of SERS signals.34
In our study, we have developed a new method for the manufacturing of SERS sensors with
different sizes and pitches of gold nanodisks on gold film. This alternative technique is based on a
soft nanoImprint lithography (S-NIL) in SiO2thermo-curable sol-gel deposited on a sacrificial resist
layer and a lift-off process to obtain gold nanodisks. The added value of this fabrication method
with respect to classical UV-NIL process is the replacement of an UV-curable resist by a thermo-
curable one. This offers the possibility for other applications requiring a stamp and substrate together
opaque to UV light. Thereby this method makes it possible to produce easily sensor with large
quantity comparable to previous soft method.35 Hence, we have patterned various holes’ diameters
(150-400 nm) with a periodicity of 400 nm and 600 nm on the bilayer resists (SiO2sol-gel/PolyMethyl
Methacrylate A2) with a flexible stamp and without external pressure. PMMA resist is used for the
lift-off process after gold deposition. Then we report on optical characterizations of the device with
thiophenol molecules (C6H6S) by SERS measurements. For specific dimensions, we obtained a lateral
collapse of pillars on PDMS stamp. This unintentional effect has been used to evaluate print resolution
in SiO2sol-gel.
II. EXPERIMENTAL
A. PDMS stamp fabrication
First step to obtain soft NIL stamp is the silicon master mold fabrication using an electron beam
lithography system (NB4 from NanoBeam Limited, UK). Exposure was performed on PMMA A4
resist (200 nm) using an accelerating voltage of 80 kV and a current of 2 nA. After development
in methyl isobutyl ketone/ isopropanol (MIBK/IPA) (1/3, 1min 30sec) and a rinse in IPA (30 sec)
a reactive ion etching (RIE) process has been optimized to transfer nanoholes into a silicon sub-
strate. To compare SERS intensities versus structures’ density several holes’ patterns were designed
(300 µm∗300 µm of lateral dimensions) with diameters (D) ranging from 150 nm to 350 nm for two
periods (P) 400 nm and 600 nm on the same master mold. Before cast molding, an anti-adhesion layer
treatment based on trimethylchlorosilane (TMCS) was used to reduce surface energy of the silicon
and make easier the demolding step. Then to reduce PDMS’s viscosity and improve the penetration
of the polymer in nanoholes, we used 5% hexane diluted standard PDMS (RTV615). The mixing
125125-3 Hamouda et al. AIP Advances 7, 125125 (2017)
solution with his curing agent (1 curing agent : 5 PDMS) ration by weight, is deposited on Si master
mold and cured at 60◦C during 4 h.
B. Gold nanodisk fabrication and thiophenol molecules deposition
Gold nanodisks on continuous film were obtained by a nanoimprint lithography in SiO2sol-gel
and lift-off process with the bilayer resist method. As sol-gel is an inorganic thermo-curable polymer
we have used a sacrificial PMMA A2 layer spinned on a glass substrate previously covered with 2 nm
of Ti and 30 nm Au layers. These metallic layers were obtained using a Plassys evaporator equipment.
The sol-gel used in these researches provides from SILSEF Company. 90 nm thick layer was deposited
by spin coating on 120 nm PMMA layer before manual imprint with the flexible stamp. An annealing
at 110◦C during 10 sec was done before demolding. Etching process of the sol-gel and sacrificial layer
have been performed by a RIE equipment (STS). RIE process for sol-gel was based on 50 sccm CHF3,
at low pressure (15 mTorr) and 325 W power. For PMMA resist, etching was adjusted with 10 sccm O2,
a chamber pressure of 4.7 mtorr, and 10 W power and stopped when the level of gold film is achieved.
Then, an Au layer (30 nm) is evaporated and sol-gel was removed by a lift-off process in acetone
thanks to the PMMA underlayer. To test SERS sensitivity of this substrate, functionalization was
performed with thiophenol (C6H6S) molecules before characterizations. Samples were immersed in
0.1 mM solution of thiophenol for 2.5 h and rinsed in ethanol for 5 min before dried up with nitrogen.
III. RESULTS AND DISCUSSION
A. Flexible stamp
As describe above the flexible stamp was obtained by a molding process on a silicon master mold
on which several holes networks with different diameters have been patterned by E-beam lithography.
Figure 1(a) shows an AFM view of PDMS stamp with dots. They have a conical shape with a diameter
(D) of 200 nm at half of their height (h= 210 nm) and a periodicity of 600 nm. This shape is the result
of an optimized etching process in Si master mold.
In Figure 1(b) AFM view for diameter 200 nm and periodicity of 400 nm shows a lateral collapse
effect where PDMS pillars stick to each other randomly. Dilution has permitted to obtain aspect ratio
(h/D) 1.4 but greater than critical aspect ratio.23 This unintentional collapse effect was particularly
interesting because it allowed us to evaluate ultimate imprint resolution in sol-gel resist.
B. Nanoimprint in sol-gel and gold nanodisks
Printing tests were performed by placing gently the flexible stamp on top of SiO2sol-gel layer,
without any additional pressure.17 An annealing at 110◦C during 10 sec was used to cure the resist.
After annealing, stamp was released leaving the sol-gel patterned. Figure 2shows SEM views of the
imprint results.
Figure 2(a) is a SEM tilted view of the imprint in SiO2sol-gel. The residual layer thickness
obtained was lower than 10 nm. When diameter increases to 200 nm with the same period, we obtain
figure 2(b) result which corresponds to the imprint with the PDMS stamp having lateral collapse
FIG. 1. AFM images of stamp PDMS: with diameter 200 nm pitch 600 nm (a), diameter 200 nm pitch 400 nm (with collapse
effect of simultaneously 2 or 4 dots) (b).
125125-4 Hamouda et al. AIP Advances 7, 125125 (2017)
FIG. 2. SEM images of the imprint in SiO2sol-gel: diameter 350 nm and periodicity 600 nm (a), diameter 200 nm and
periodicity 400 nm with a collapse effect (b), diameter 150 nm and periodicity 400 nm (c).
effect. Figure 2(c) shows an imprint with diameter 150 nm and pitch 400 nm. Diameter 150 nm is
the lower limit that we can obtain in a reproducible manner with a PDMS. We can notice a consistent
transfer with the stamp. Etching processes of sol-gel and PMMA was performed as described in
“EXPERIMENTAL” section and stop at gold film. A gold deposit (30 nm) and lift-off process in
solvent with ultra sonic were performed to obtain gold nanodisks as shown in Figure 3. Figure 3(a)
shows regular dot arrays with diameter 230 nm and pitch 600 nm. A certain roughness around gold
nanodisks is attributed to the RIE process of SiO2sol-gel, this was confirmed by a SEM observation
after etching step. This process step also induces also the widening of the nanostructures.
Figure 3(b) shows gold nanodisks for diameter 200 nm and pitch 400 nm. We can notice a gold
nanostructure randomization due to the lateral collapse effect in PDMS which has been transferred
until this last step. Some discs can touch each other with a random distribution. As shown in black
square, others are very close with a gap smaller than 50 nm. Hence, this result shown in Figure 3(b)
makes possible to evaluate a print resolution in SiO2sol-gel around this value.
C. Optical characterizations
To characterize the performance of substrate as a sensor, SERS measurements were performed
on several nanostructured areas functionalized by thiophenol molecules. Some Raman characteristic
peaks of thiophenol molecules are 419, 1000, 1075, 1575 cm-1. These Raman spectra were recorded
using a XploRA spectrophotometer from Horiba Scientific. The acquisition time was fixed to 30 s.
A 638 nm laser was used for these measurements with a power of 2.6 mW. The laser excitation was
focused on the substrate using a microscope objective (x20, N.A.= 0.7). The same objective was
used to collect the Raman signal from the SERS substrates in a backscattering configuration. SERS
spectra were recorded with a spectral resolution under 4 cm-1.
Figure 4shows an increasing SERS signal versus nanostructures density as expected.
We calculated enhancement factor (EF) defined by the following equation:
EF =
Isers
IRaman
×NRaman
Nsers
(1)
FIG. 3. SEM image of gold nanodisks on gold film after lift-off process: (a) diameter 230 nm and periodicity 600, (b) diameter
200 nm and periodicity 400 nm.
125125-5 Hamouda et al. AIP Advances 7, 125125 (2017)
FIG. 4. SERS measurements with pitch 600 nm and diameters 230-360 nm (a).
Where Isers and IRaman are respectively the SERS and Raman intensities. NRaman and Nsers are
defined as the number of excited molecules in Raman and SERS experiments, respectively NRaman is
obtained from Ref. 35 and estimated to 4.22 ×1015.Nsers is evaluated from the following equation:
Nsers =NA×Scollected ×σSurf ×Sstructures
P2with NAthe Avogrado’s number (mol-1), Scollected the illumi-
nated area around 7 µm2,σSurf is the surface coverage of thiophenol molecules and is approximately
around 0.544 nmol/cm2,36,37 Sstructures is the sum of the lateral and top surface of nanodisks. P
corresponds to the periodicity of the nanostructures array.
SERS EF are between 2.6x106and 3.1x106depending of the nanodisks’ sizes for the period
600 nm. The order of magnitude of 106agrees with characterization of gold nanostructures function-
alized on gold film.8,9SERS measurements have also been performed for 400 nm period nanodisks
and shown due to collapse effect lower peak intensities.
IV. CONCLUSIONS
In this work we have presented an original fabrication process for biochemical surface enhanced
Raman scattering sensor based on soft nanoimprint lithography (S-NIL) on thermo-curable SiO2
sol-gel at low temperature. This soft nanoimprint method, mainly manual, and the lift-off process
with sacrificial layer under SiO2sol-gel permit to product easily sensors at low cost, thereby this
method is compatible with mass production. Otherwise for a periodicity of 400 nm we have obtained
a collapse effects of the pillars on PDMS stamp. We used this effect to evaluate the resolution of the
imprint in sol-gel. SERS characterization with thiophenol were performed with spectrophotometer.
An evaluation of SERS intensities on various structures’ density has been shown with the periode
600 nm. This new alternative manufacturing provided devices with specific coupling properties due
to the periodicity and the size of gold nonodisks on a gold film underlayer. These results consistent
with previous studies validate this new sol-gel-based process.
ACKNOWLEDGMENTS
This work was done with the C2N facility and partly supported by the RENATECH network and
the General Council of Essonne. More over the authors acknowledge ANR-12-NANO-0016-03.
1C. V. Raman, Journal of Physics 2, 387 (1928).
2C. V. Raman and K. S. Krishnan, Nature 121, 501 (1928).
3M. Fleischman, P. J. Hendra, and A. J. Mac Quillan, Chem. Phys. Lett. 26, 163 (1974).
4K. Kneipp, Y. Wang, H. Kneipp, L. T. Perelman, I. Itzkan, R. R. Dasari, and M. S. Feld, Phys. Rev. Lett. 78, 1667 (1997).
5A. Stefan, S. A. Meyer, E. C. Ru, and P. G. Etchegoin, Anal. Chem. 83(6), 2337 (2011).
6F. Fei Le, D. W. Brandl, Y. A. Urzhumov, H. Wang, J. Kundu, N. J. Halas, J. Aizpurua, and P. Nordlander, ACS Nano 2(4),
707 (2008).
125125-6 Hamouda et al. AIP Advances 7, 125125 (2017)
7M. Sarkar, J. F. Bryche, J. Moreau, M. Besbes, G. Barbillon, B. Bartenlian, and M. Canva, Optics Express 23, 27376 (2015).
8J. F. Bryche, R. Gillibert, G. Barbillon, M. Sarkar, A. L. Coutrot, F. Hamouda, A. Aassime, J. Moreau, M. Lamy de la
Chapelle, B. Bartenlian, and M. Canva, J Mater Sci 50, 6600 (2015).
9J. F. Bryche, R. Gillibert, G. Barbillon, P. Gogol, J. Moreau, M. Lamy de la Chapelle, B. Bartenlian, and M. Canva,
Plasmonics 11, 601 (2015).
10 S. Y. Chou, P. R. Krauss, and P. J. Renstrom, Appl. Phys. Lett. 67, 3114 (1995).
11 J. Haisma, M. Verheijen, K. van den Heuvel, and J. van den Berg, J. Vac. Sci. Technol. B 14, 4124 (1996).
12 C. Gourgon, C. C. Perret, J. Tallal, F. Lazzarino, S. Landis, O. Joubert, and R. Pelze, J. Phys. D: Appl. Phys. 38, 70 (2005).
13 M. Bender, M. Otto, B. Hadam, B. Spangenberg, and H. Kurz, Microelectron. Eng. 53, 233 (2000).
14 P. Voisin, M. Zelsmann, C. Gourgon, and J. Boussey, Microelectron. Eng. 84, 916 (2007).
15 B. Bender, U. Plachetka, J. Ran, A. Fuchs, B. Vratzov, H. Kurz, T. Glinser, and F. Linder, J. Vac. Sci. Technol. B 22, 3229
(2004).
16 Y. Xiangdong, L. Hongzhong, D. Yucheng, L. Hansong, and L. Bingheng, Microelectron. Eng. 86, 310 (2009).
17 F. Hamouda, G. Barbillon, F. Gaucher, and B. Bartenlian, J. Vac. Sci. Technol. B 28(82), 82 (2010).
18 N. Koo, M. Bender, U. Plachetka, A. Fuchs, T. Wahlbrink, J. Bolten, and H. Kurz, Microelectron. Eng. 84, 904 (2007).
19 F. Hamouda, H. Sahaf, S. Held, G. Barbillon, P. Gogol, E. Moyen, A. Aassime, J. Moreau, M. Canva, J. M. Lourtioz,
M. Hanb¨
ucken, and B. Bartenlian, Microelectron. Eng. 88, 2444 (2011).
20 H. Schmid and B. Michel, Macromolecules 33, 3042 (2000).
21 U. Plachetka, M. Bender, A. Fuchs, T. Wahlbrink, T. Glinsner, and H. Kurz, Microelectron. Eng. 83, 944 (2006).
22 M. M¨
uhlberger, I. Bergmair, A. Klukowsk, A. Kolander, H. Leichtfried, E. Platzgummer, H. Loeschner, Ch. Ebm,
G. Gr¨
utzner, and R. Sch¨
oftner, Microelectron. Eng. 86, 691 (2009).
23 A. Finn, B. Lu, R. Kirchner, X. Thrun, K. Richter, and W. J. Fischer, Microelectron. Eng. 110, 112 (2013).
24 A. Ferchichi, F. Laariedh, I. Sow, C. Gourgon, and J. Boussey, Microelectron. Eng. 140, 52 (2015).
25 M. Vogler, S. Wiedenberg, M. M¨
uhlberger, I. Bergmair, T. Glinsner, H. Schmidt, E. B. Kley, and G. Gr ¨
utzner, Microelectron.
Eng. 84, 984 (2007).
26 A. Fuchs, M. Bender, U. Plachetka, L. Kock, N. Koo, T. Wahlbrink, and H. Kurz, Current Applied Physics 8, 669 (2007).
27 N. Koo, J. W. Kim, M. Otto, C. Moormann, and H. Kurz, J. Vac. Sci. Technol. B 29, 06FC12-1 (2011).
28 J. S. Kang, C. J. Lee, M. S. Kim, and M. S. Lee, Bull. Korean Chem. Soc. 24(11), 1599 (2003).
29 C. Perroz, V. Chauveau, E. Bartel, and E. Sondergard, Adv. Mater. 21, 555 (2009).
30 D. Shuxi, W. Yang, Z. Dianbo, H. Xiao, S. Qing, W. Shujie, and D. Zuliang, Journal of Sol-Gel Science and Technol. 60,
17 (2011).
31 V. Gat ´
e, G. Bernaud, C. Veillas, A. Cazier, F. Vocanson, Y. Jourlin, and M. Langlet, Opt. Eng. 52, 091712 (2013).
32 N. H. Sulaiman, M. J. Ghazali, B. Y. Majlis, J. Yunas, and M. Razali, Bio-Medical Materials and Engineering 26, S103
(2015).
33 V. Gat ´
e, Y. Jourlin, F. Vocanson, O. Dellea, G. Vercasson, S. Reynaud, D. Riasseto, and M. Langlet, Optical Materials 35(9),
1706 (2013).
34 M. Volkan, D. L. Stokes, and T. Vo-Dinh, Sens. Actuators, B 106, 660 (2005).
35 M. Cottat, N. Lidgi-Guigui, I. Tijunelyte, G. Barbillon, F. Hamouda, P. Gogol, A. Aassime, J. M. Lourtioz, B. Bartenlian,
and M. Lamy de la Chapelle, Nanoscale Research Letters 9, 623 (2014).
36 D. A. Stern, E. Wellner, G. N. Salaita, L. Laguren-Davidson, F. Lu, N. Batina, D. G. Frank, D. C. Zapien, N. Walton, and
A. T. Hubbard, J. Am. Chem. Soc. 110, 4887 (1988).
37 J. D. Caldwell, O. Glembocki, F. J. Bezares, N. D. Bassim, R. W. Rendell, M. Feygelson, U. Ukaegbu, R. Kasica, L. Shirey,
and C. Hosten, ACS Nano 5, 4046 (2011).
Available via license: CC BY 4.0
Content may be subject to copyright.
