**Soft UV Nanoimprint Lithography: A Versatile Tool for Nanostructuration at the 20nm Scale**

Andrea Cattoni1, Jing Chen1, Dominique Decanini1, Jian Shi 2 and Anne-Marie Haghiri-Gosnet1 *1Laboratoire de Photonique et de Nanostructures, LPN (CNRS-UPR20), Marcoussis 2Ecole Normale Supérieure de Paris, Paris France* 

#### **1. Introduction**

138 Recent Advances in Nanofabrication Techniques and Applications

Schift, H. (2008). Nanoimprint lithography: An old story in modern times? A review. *J. Vac.* 

Shank, C. V.; Yen, R. & Hirlimann, C. (1983). Time-resolved reflectivity measurements of

Xia, Q.; Keimel, K.; Ge, H.; Yu, Z.; Wu, W. & Chou, S. Y. (2003). Ultrafast patterning of

Xia, Q.; Morton K. J.; Austin, R. H. & Chou, S. Y. (2008). Sub-10 nm Self-Enclosed Self-

Xia, Q. & Chou, S. Y. (2010). Applications of excimer laser in nanofabrication, *Appl Phys A*

Yaws, C. L. (1998). Chemical Properties Handbook: Physical, Thermodynamics,

Yu, J. & Bulovic, V. (2007). Micropatterning metal electrode of organic light emitting devices using rapid polydimethylsiloxane lift-off. *Appl. Phys. Lett.* 91, 043102.

Spindt, C. A. (1968). J. Appl. Phys. 39 3504 A Thin-Film Field-Emission Cathode.

Limited Nanofluidic Channel Arrays, *Nano Lett.*, 8(11), 3830-3833.

Inorganic Chemicals (McGraw-Hill Professional).

femtosecond-optical-pulse induced phase-transitions in silicon. *Physical Review* 

nanostructures in polymers using laser assisted nanoimprint lithography. *Appl.* 

Environmental Transport, Safety & Health Related Properties for Organic &

*Sci. Technol. B*, 26 458–80.

*Phys. Lett.* 83(21), 4417-4419.

*Letters*, 50(6), 454.

98, 9–59.

#### **1.1 Why soft UV nanoimprint lithography?**

Since the pioneering work of Whitesides and coworkers on microContact Printing (mCP) and Soft Lithography (Kumar & Whitesides (1993)) (Xia & Whitsides (1998)), considerable progress has been made in the last years and Soft Lithography is now a well consolidated technology utilized in academic and industrial laboratories (Rogers & Nuzzo (2005)). These printing methods use a flexible elastomer material named PDMS (poly(dimethylsiloxane)) to transfer molecules on a surface thus creating localized chemical patterns (Cerf & Vieu (2010)). The PDMS stamp inked with the desired molecules is placed in contact with the substrate to perform the molecular transfer. mCP has received large attention for biological applications since this soft transfer occurs in a gentle manner which allows the biomolecules to be transferred without any damage. In addition, this powerful technique is cheap because the soft PDMS stamp can be replicated an indefinite number of times by simply pouring the PDMS prepolymer onto a single expensive silicon master mold that contains micro or nanostructures. Since the flexibility of the elastomeric stamp ensures a perfect conformal adhesion within the substrate, thus allowing replication on large areas up to several tens of cm2, the use of such flexible PDMS stamps was also efficiently applied to another low-cost and high-throughput manufacturing technique called Soft UV Nanoimprinting Lithography (Soft UV-NIL). This technique creates a thickness contrast by embossing thin polymeric films, highlighting the advantages of using a flexible PDMS stamp.

Historically, Nanomprint Lithography (NIL) in its original version was proposed by Stephen Chou in 1995 (Chou et al. (1995)) as an alternative technique for the embossing of high resolution patterns in thermoplastic materials. The patterning of features as small as 10 nm has been demonstrated from the beginning (Chou (1997)). This nanoimprint process, usually referred to as thermal-assisted NIL (T-NIL), is based on the use of a hard mold, namely a silicon wafer. As schematically shown in Figure 1, this hard mold containing nanoscale surface-relief features is pressed at high pressure (50-100 bar) onto a thin polymeric resist film. The resist is held some 90-100 ◦C above its glass-transition temperature (Tg) for few minutes to allow the flowing of the polymer in the mold nanocavities. The thin residual layer of polymer intentionally left to prevent the direct contact between the substrate and the rigid mold is

Although there are no apparent resolution limitations for T-NIL due to its purely mechanical embossing nature, this technology is unable to meet the stringent requirements of semiconductor IC manufacturing. In particular, the high pressure and temperature required during the process limit its use to few applications. UV-NIL (Haisma et al. (1996); Ruchhoeft et al. (1999)), which appeared quickly after T-NIL, is considered to be the most attractive variant for semiconductor IC manufacturing (Figure 1). With this variant, a transparent quartz mold is pressed at room temperature with a lower pressure on a liquid precursor that can be cured by UV light before the release of the mold. Industry quickly adopted UV-NIL, particularly Step and Flash Imprint Lithography (SFIL) which replicates nanostructures on a large area by patterning several fields (Bailey et al. (2002)). This industrial approach consists of dispensing a drop of the low viscosity photopolymerizable liquid resist on each field surface prior to the imprinting process. This allows for the patterning on the whole wafer area, much

<sup>141</sup> Soft UV Nanoimprint Lithography:

As an alternative to UV-NIL, Soft UV-NIL has recently been proposed as a means to reduce the cost of master fabrication. Soft UV-NIL uses a flexible transparent stamp normally made of poly(dimethylsiloxane) (PDMS) or other flexible polymers that can be easily replicated from a single Silicon master mold. This Silicon master mold is fabricated with conventional lithography techniques, like EBL, which are more time consuming and expensive. The flexibility of the Polymeric stamp ensures contact with the surface substrate on large surfaces at low pressures, even on curved or flexible substrates. In particular, PDMS exhibits attractive properties like low interfacial free energy (∼21.6 dyn/cm), chemical stability and high optical transparency. Moreover, its permeability prevents problems caused by trapped air bubbles in the resist layer when imprinting at ambient pressure. Soft UV-NIL is considered today to be the most promising variant to replicate patterns in the sub-50 nm range for mass production at low cost. Soft UV-NIL is currently being applied in a large spectrum of emerging area like

In this paper we will discuss the ability to use PDMS based stamps, as candidates for nanoimprinting lithography in the sub-20nm scale on large surfaces and at very low pressure (< 1 atm). In terms of resolution, the replication of features with sizes as small as 2 nm has been demonstrated by using a unique carbon nanotube as a template and a modified PDMS material for the stamp (Hua et al. (2004)). These recent experiments demonstrate that PDMS properties, and specifically the density of cross-links, are important parameters that can influence the ultimate resolution. In this context, testing the resolution limits for soft UV-NIL

In the frame of the European Project TERAMAGSTOR (EU funded FP7 STREP project) for the production of the next generation of perpendicular magnetic storage media with an areal density greater than 1 Tbit/inch2, our group has been working on the feasibility to apply Soft UV-NIL for the nanofabrication of bit pattern media at ultra high resolution in the sub-10 nm scale. Our optimized approach for the fabrication of the silicon master mold using classical electron beam lithography will be presented with different fabrication techniques and resists. Due to the high cost of the master mold fabrication by EBL, we will propose new routes for the replication and/or inversion of the master mold at the whole wafer scale based on T-NIL. Once the hard mold with highly resolved patterns is obtained, the second challenge concerns the replication of the PDMS stamp. It should exhibit enough local hardness for the replication of nanometric features, and at the same time, preserve a global flexibility to ensure conformal contact at low imprinting pressures. The description of our bilayer hard-PDMS/PDMS stamp will be detailed and we will demonstrate the replication of 20 nm features under specific

Flexible Electronics, Photonics, Biotechnology and Nanomedicine.

with flexible stamps and for real applications appears very challenging.

like a stepper lithography tool does.

A Versatile Tool for Nanostructuration at the 20nm Scale

then removed by reactive ion etching (RIE) to complete the resist pattern transfer. T-NIL uses a very simple experimental set-up, which results in a very short process time, from seconds to minutes.

In contrast to conventional lithography methods based on exposure and development of a resist, nanoimprint lithography is based on mechanically embossing a thin polymer film under a mold that contains micro or nanopatterns. For this reason, limitations imposed by light diffraction or beam scattering in conventional lithography techniques can be overcome. As a second advantage, NIL provides parallel processing with high throughput, which is suitable for large-scale patterning with very high resolution. Another important benefit is the low cost of the NIL equipment, which compared with the processing cost of classical lithography techniques like deep UV optical lithography or Electron Beam Lithography (EBL). Nanoimprinting lithography has rapidly received a lot of attention from both the research community and industry, so much so that MIT's 2003 Technology Review named nanoimprint lithography as one of the "ten emerging technologies that will change the world." Additionally, NIL has been added into the International Technology Roadmap for Semiconductors (ITRS) for the next 22 nm node.

Fig. 1. Left panel: schematic of the T-NIL process proposed by S.Y. Chou in 1995 with a SEM image of 60 nm deep holes imprinted into PMMA which have a 10 nm minimum diameter and a period of 40 nm. Reprinted with permission from (Chou (1997)). Copyright 1997 American Vacuum Society. Central panel: schematic of the UV-NIL process and relative SEM image of the imprinted resist after UV curing with feature sizes as small as 5 nm. Reprinted with permission from (Austin et al. (2004)). Copyright 2004 American Vacuum Society. Right panel: SEM image of imprinted Amonil resist (AMO GmbH Aachen, Germany) of 100 nm square dots with 300 nm pitch after Soft UV NIL with PDMS stamp replicated by a Silicon master mold.

2 Will-be-set-by-IN-TECH

then removed by reactive ion etching (RIE) to complete the resist pattern transfer. T-NIL uses a very simple experimental set-up, which results in a very short process time, from seconds to

In contrast to conventional lithography methods based on exposure and development of a resist, nanoimprint lithography is based on mechanically embossing a thin polymer film under a mold that contains micro or nanopatterns. For this reason, limitations imposed by light diffraction or beam scattering in conventional lithography techniques can be overcome. As a second advantage, NIL provides parallel processing with high throughput, which is suitable for large-scale patterning with very high resolution. Another important benefit is the low cost of the NIL equipment, which compared with the processing cost of classical lithography techniques like deep UV optical lithography or Electron Beam Lithography (EBL). Nanoimprinting lithography has rapidly received a lot of attention from both the research community and industry, so much so that MIT's 2003 Technology Review named nanoimprint lithography as one of the "ten emerging technologies that will change the world." Additionally, NIL has been added into the International Technology Roadmap for

UV-NIL Soft UV-NIL

• Low Pressure (< 1 bar) • Room Temperature

• Flexible/not planar substrates

• Cheap

quartz/glass mold polimeric stamp

• Low pressure (0 – 5 bar) • Room Temperature

Fig. 1. Left panel: schematic of the T-NIL process proposed by S.Y. Chou in 1995 with a SEM image of 60 nm deep holes imprinted into PMMA which have a 10 nm minimum diameter and a period of 40 nm. Reprinted with permission from (Chou (1997)). Copyright 1997 American Vacuum Society. Central panel: schematic of the UV-NIL process and relative SEM image of the imprinted resist after UV curing with feature sizes as small as 5 nm. Reprinted with permission from (Austin et al. (2004)). Copyright 2004 American Vacuum Society. Right panel: SEM image of imprinted Amonil resist (AMO GmbH Aachen, Germany) of 100 nm square dots with 300 nm pitch after Soft UV NIL with PDMS stamp replicated by a Silicon

minutes.

Semiconductors (ITRS) for the next 22 nm node.

T-NIL

• High Pressure (50 – 100 bar) • High Temperature

rigid mold

substrate

master mold.

Although there are no apparent resolution limitations for T-NIL due to its purely mechanical embossing nature, this technology is unable to meet the stringent requirements of semiconductor IC manufacturing. In particular, the high pressure and temperature required during the process limit its use to few applications. UV-NIL (Haisma et al. (1996); Ruchhoeft et al. (1999)), which appeared quickly after T-NIL, is considered to be the most attractive variant for semiconductor IC manufacturing (Figure 1). With this variant, a transparent quartz mold is pressed at room temperature with a lower pressure on a liquid precursor that can be cured by UV light before the release of the mold. Industry quickly adopted UV-NIL, particularly Step and Flash Imprint Lithography (SFIL) which replicates nanostructures on a large area by patterning several fields (Bailey et al. (2002)). This industrial approach consists of dispensing a drop of the low viscosity photopolymerizable liquid resist on each field surface prior to the imprinting process. This allows for the patterning on the whole wafer area, much like a stepper lithography tool does.

As an alternative to UV-NIL, Soft UV-NIL has recently been proposed as a means to reduce the cost of master fabrication. Soft UV-NIL uses a flexible transparent stamp normally made of poly(dimethylsiloxane) (PDMS) or other flexible polymers that can be easily replicated from a single Silicon master mold. This Silicon master mold is fabricated with conventional lithography techniques, like EBL, which are more time consuming and expensive. The flexibility of the Polymeric stamp ensures contact with the surface substrate on large surfaces at low pressures, even on curved or flexible substrates. In particular, PDMS exhibits attractive properties like low interfacial free energy (∼21.6 dyn/cm), chemical stability and high optical transparency. Moreover, its permeability prevents problems caused by trapped air bubbles in the resist layer when imprinting at ambient pressure. Soft UV-NIL is considered today to be the most promising variant to replicate patterns in the sub-50 nm range for mass production at low cost. Soft UV-NIL is currently being applied in a large spectrum of emerging area like Flexible Electronics, Photonics, Biotechnology and Nanomedicine.

In this paper we will discuss the ability to use PDMS based stamps, as candidates for nanoimprinting lithography in the sub-20nm scale on large surfaces and at very low pressure (< 1 atm). In terms of resolution, the replication of features with sizes as small as 2 nm has been demonstrated by using a unique carbon nanotube as a template and a modified PDMS material for the stamp (Hua et al. (2004)). These recent experiments demonstrate that PDMS properties, and specifically the density of cross-links, are important parameters that can influence the ultimate resolution. In this context, testing the resolution limits for soft UV-NIL with flexible stamps and for real applications appears very challenging.

In the frame of the European Project TERAMAGSTOR (EU funded FP7 STREP project) for the production of the next generation of perpendicular magnetic storage media with an areal density greater than 1 Tbit/inch2, our group has been working on the feasibility to apply Soft UV-NIL for the nanofabrication of bit pattern media at ultra high resolution in the sub-10 nm scale. Our optimized approach for the fabrication of the silicon master mold using classical electron beam lithography will be presented with different fabrication techniques and resists. Due to the high cost of the master mold fabrication by EBL, we will propose new routes for the replication and/or inversion of the master mold at the whole wafer scale based on T-NIL. Once the hard mold with highly resolved patterns is obtained, the second challenge concerns the replication of the PDMS stamp. It should exhibit enough local hardness for the replication of nanometric features, and at the same time, preserve a global flexibility to ensure conformal contact at low imprinting pressures. The description of our bilayer hard-PDMS/PDMS stamp will be detailed and we will demonstrate the replication of 20 nm features under specific

**Process 1: PMMA resist**

A Versatile Tool for Nanostructuration at the 20nm Scale

2) Deposition of Ni (20nm) and lift-off of PMMA

**Process 1(B)**

silicon

<sup>143</sup> Soft UV Nanoimprint Lithography:

2) Post bake at 400°C:

4 HSiO3/2 ї 3 SiO4/2 + 2 SiH4 SiH4 ї 3 Si + 2 H2 or cleavage or Si-H

silicon

**Process 2: HSQ resist**

1) EBL with HSQ (negative-tone)

3) RIE of Silicon

4) Removal of Ni in HNO3

Fig. 2. Schematic of the Process 1, based on positive-tone PMMA resist and of the Process 2,

solution at 20 ◦C for 35 sec, rinsed in IPA for 10 s and then dried with a nitrogen gun. Pattern transfer in the Silicon substrate is then performed with a highly anisotropic RIE process based on CHF3/SF6/O2, and the PMMA resist is finally removed with acetone. This process allows for the fabrication of high resolution and low sidewall roughness structures, as shown in the

Since PMMA is a positive tone resist, an inverted pattern can be obtained either by exposing the complementary area, which is time consuming, or by a multi-step process with metal lift-off. In order to test the PMMA performance and the feasibility of this second process, we have carried out a challenge exposure of nanodots in a hexagonal pattern with a nominal diameter of 20 nm and a pitch of 60 nm, in a 80 nm thick PMMA resist (950PMMA A2, MicroChem Corp.). After the sample was developed as described in process 1 and lifted off of the 15 nm Ni mask, the dot patterns were transferred into the silicon wafer by reactive ion etching. The Ni mask was then removed by chemical wet etching in a HNO3 acid solution. The Figure 3 (c) shows a SEM image of the Si master molds after the Ni mask was removed. The inset shows a distribution graph of the area of the dots, including the standard deviation and the relative average diameter of the dots. In general, this process is suitable for the fabrication of structures up to the 100 nm scale, but it is not reproducible on sub-20 nm scale structures. This is due to the multiple intermediate steps which introduce broadening of the structures, defects and sidewall roughness. In order to improve shape uniformity and sidewall roughness, we have developed an alternative process based on negative-tone resist, Hydrogen Silsesquioxan (HSQ). HSQ is a new non-chemically amplified resist for EBL

1) EBL with PMMA (positive-tone)

silicon PMMA

**Process 1(A)**

2) RIE of Silicon

3) Removal of PMMA

based on negative-tone HSQ resist.

Figure 3 (a and b).

experimental conditions. To conclude with real application, we will present the fabrication by Soft UV NIL of plasmonic nanostructures for biosensing applications with superior optical performances directly related to their high quality and uniformity on the whole pattern area.
