Submicrometer 3D structures fabrication enabled by one-photon absorption direct laser writing

Submicrometer 3D structuresfabrication enabled by one-photon

absorption direct laser writing

Mai Trang Do,1 Thi Thanh Ngan Nguyen,1,2 Qinggele Li,1

Henri Benisty,3 Isabelle Ledoux-Rak,1 and Ngoc Diep Lai1,∗

1Laboratoire de Photonique Quantique et Moleculaire, UMR CNRS 8537,Institut D’Alembert, Ecole Normale Superieure de Cachan,

61 avenue du President Wilson, 94235 Cachan cedex, France2Institute of Materials Science, Vietnam Academy of Science and Technology,

18 Hoang Quoc Viet, Cau Giay, Hanoi, Vietnam3Laboratoire Charles Fabry de l’Institut d’Optique, UMR 8501 CNRS,

Universite Paris-Sud, F-91127 Palaiseau cedex, France∗[email protected]

Abstract: We demonstrate a new 3D fabrication method to achieve thesame results as those obtained by the two-photon excitation technique, byusing a simple one-photon elaboration method in a very low absorptionregime. Desirable 2D and 3D submicrometric structures, such as spiral,chiral, and woodpile architectures, with feature size as small as 190 nmhave been fabricated, by using just a few milliwatts of a continuous-wavelaser at 532 nm and a commercial SU8 photoresist. Different aspects ofthe direct laser writing based on ultralow one-photon absorption (LOPA)technique are investigated and compared with the TPA technique, showingseveral advantages, such as simplicity and low cost.

© 2013 Optical Society of America

OCIS codes: (220.3740) Lithography; (220.4000) Microstructure fabrication; (140.3390)Laser materials processing; (110.0180) Microscopy; (010.1030) Absorption; (170.6900) Three-dimensional microscopy.

References and links1. P. Torok and F. J. Kao (eds.),Optical Imaging and Microscopy,2nd ed. (Springer Series in Optical Sciences

2007).2. J. B. Pawley (ed.),Handbook of Biological Confocal Microscopy,3rd ed. (Berlin, Springer 2006).3. W. Denk, J. H. Strickler, and W. W. Webb, “Two-photon laser scanning fluorescence microscopy,” Science248,

73–76 (1990).4. B. H. Cumpston, S. P. Ananthavel, S. Barlow, D. L. Dyer, J. E. Ehrlich, L. L. Erskine, A. A. Heikal, S. M.

Kuebler, I.-Y. S. Lee, D. McCord-Maughon, J. Qin, H. Rckel, M. Rumi, X.-L. Wu, S. R. Marder, and J. W. Perry,“Two-photon polymerization initiators for three dimensional optical data storage and microfabrication,” Nature398, 51– 54 (1999).

5. S. Kawata, H.-B. Sun, T. Tanaka, and K. Takada, “Finer features for functional microdevices,” Nature412, 697–698 (2001).

6. M. Deubel, G. Von Freymann, M. Wegener, S. Pereira, K. Busch, and C. M. Soukoulis, “Direct laser writing ofthree-dimensional photonic-crystal templates for telecommunications,” Nat. Mater.3, 444–447 (2004).

7. M. Farsari and B. N. Chichkov, “Materials processing: two-photon fabrication,” Nat. Photonics3, 450–452(2009).

8. W. Haske, V. W. Chen, J. M. Hales, W. Dong, S. Barlow, S. R. Marder, and J. W. Perry, “65 nm feature sizesusing visible wavelength 3-D multiphoton lithography,” Opt. Express15, 3426–3436 (2007).

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20964

9. D. A. Parthenopoulos, and P. M. Rentzepis, “Three-dimensional optical storage memory,” Science245, 843–845(1989).

10. W. R. Zipfel, R. M. Williams, and W. W. Webb, “Nonlinear magic: multiphoton microscopy in the biosciences,”Nat. Biotechnol.21, 1369–1377 (2003).

11. C. Rensch, S. Hell, M. V. Schickfus, and S. Hunklinger, “Laser scanner for direct writing lithography,” Appl.Opt.28, 3754–3758 (1989).

12. B. Richards and E. Wolf, “Electromagnetic diffraction in optical systems. II. Structure of the image field in anaplanatic system,” Proc. Roy. Soc. A253, 358–379 (1959).

13. G. Witzgall, R. Vrijen, E.Yablonovitch, V. Doan, and B. J. Schwartz, “Single-shot two-photon exposure of com-mercial photoresist for the production of three-dimensional structures,” Opt. Lett.23, 1745–1747 (1998).

14. T. Kondo, S. Matsuo, S. Juodkazis, V. Mizeikis, and H. Misawa, “Multiphoton fabrication of periodic structuresby multibeam interference of femtosecond pulses,” Appl. Phys. Lett.82, 2758–2560 (2003).

15. M. Campbell, D. N. Sharp, M. T. Harrison, R. G. Denning, and A. J. Turberfield, “Fabrication of photonic crystalsfor the visible spectrum by holographic lithography,” Nature404, 53–56 (2000).

16. X. Wang, J. F. Xu, H. M. Su, Z. H. Zeng, Y. L. Chen, H. Z. Wang, Y. K. Pang, and W. Y. Tam, “Three-dimensionalphotonic crystals fabricated by visible light holographic lithography,” Appl. Phys. Lett.82, 2212–2214 (2003).

17. N. D. Lai, W. P. Liang, J. H. Lin, C. C. Hsu, and C. H. Lin, “Fabrication of two- and three-dimensional periodicstructures by multi-exposure of two-beam interference technique,” Opt. Express13, 9605–9611 (2005).

18. H.-B. Sun, T. Tanaka, and S. Kawata, “Three-dimensional focal spots related to two-photon excitation,” Appl.Phys. Lett.80, 3673–3675 (2002).

19. C.-H. Lee, T.-W. Chang, K.-L. Lee, J.-Y. Lin, and J. Wang, “Fabricating high-aspect-ratio sub-diffraction-limitstructures on silicon with two-photon photopolymerization and reactive ion etching,” Appl. Phys. A79, 2027–2031 (2004).

20. S. Maruo and K. Ikuta, “Three-dimensional microfabrication by use of single-photon-absorbed polymerization,”Appl. Phys. Lett.76, 2656–2658 (2000).

21. M. Thiel, J. Fischer, G. V. Freymann, and M. Wegener, “Direct laser writing of three-dimensional submicronstructures using a continuous-wave laser at 532 nm,” Appl. Phys. Lett.97, 221102 (2010).

22. M. Malinauskas, P. Danileviius, and S. Juodkazis, “Three-dimensional micro-/nano-structuring via direct writepolymerization with picosecond laser pulses,” Opt. Express19, 5602–5610 (2011).

23. W. Bogaerts, R. Baets, P. Dumon, V. Wiaux, S. Beckx, D. Taillaert, B. Luyssaert, J. V. Campenhout, P. Bienstman,and D. V. Thourhout, “Nanophotonic Waveguides in Silicon-on-Insulator Fabricated With CMOS Technology,”J. Lightwave Technol.23, 401–412 (2005).

1. Introduction

Optical microscopy and nanoscopy play a major role in modern science and technology, withnumerous applications including optical imaging, confocal scanning microscopy, optical datastorage, micro- and nano-fabrication, etc [1]. Conventional high-aperture microscopy (up to NA= 1.4 with oil immersion) is nowadays close to reach its maximum with large wide aplanetic (=diffraction limited) fields available in the best objectives lenses (OLs). A focusing spot with aneffective volume of less than 1µm3 is then achieved. Such high NA OLs allow addressing orelaborating an object with a submicrometer resolution through a large volume. For this purpose,the focused laser beam is usually raster-scanned across the target sample in all dimensions.Furthermore, for imaging applications, the use of a confocal laser scanning microscope (CLSM)provides depth discrimination and improves the longitudinal resolution within the focal plane[2].

Two excitation ways have been implemented up to now for photo-induced fabrication of sub-micrometric structures, namely one-photon and two-photon absorption (OPA, TPA), involvingdifferent excitation mechanisms and aiming at specific applications. Indeed, when using a thinfilm, the OPA excitation method is a very convenient technique, based on a simple and lowcost laser source operating at a wavelength located within the absorption band of the thin filmmaterial. However, due to a strong linear absorption, in the 104

−105 cm−1 range, this methodis limited to the surface of the sample,i. e. to a two-dimensional (2D) scanning. High spatialresolution three-dimensional (3D) imaging or fabrication is therefore impossible with the OPAmethod applied to strongly absorbing samples. In contrast, TPA excitation provides intrinsic 3Daddressing, thanks to a quasi absence of linear absorption and a local nonlinear absorption [3],

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20965

which occurs only for the very high local intensity provided by tight focusing in a CLSM com-bined with a femtosecond laser operating at a wavelength around 800 nm for which the materialis kept fully transparent [4]. This TPA nowadays allows many potential applications in differentdomains, in particular, for 3D imaging and 3D fabrication of submicrometric structures [4–10],but requires expensive laser sources and delicate specialized optics.

In this work, we propose a new way to benefit from the advantages ofbothexcitation tech-niques, by using a simple, low cost one-photon elaboration method operating in an ultralowabsorption regime. This technique employs a continuous laser only, as in the case of OPA,while allowing 3D addressing as in the case of TPA

2. Working principle of LOPA microscopy

In the optical confocal microscope system, the propagation of light towards the focusing regionis formally described by a nonlinear wave equation for the optical electric field. The intensityincreases nonlinearly and reaches a maximum value at the focal plane. For a high NA OL witha cm-size pupil, the light intensity at the focusing plane can be about 108 times higher than thatat the input beam. Inside a material, the excitation efficiency and the resulting effect,e.g.flu-orescence or photopolymerization, is strongly dominant at the focusing region. In some cases,the photo-induced effect is only observed at the limited size focusing spot where the intensity ishigh enough to induce physical or mechanical or chemical changes. A good example is the useof an infrared pulsed laser to induce TPA in the focusing region. In an overwhelming majorityof cases, only such TPA techniques have been applied to fluorescence 3D microscopes [3, 10],

Fig. 1. (a) Sketch of the experimental setup used to fabricate desired sub-micrometer 3Dstructures. PZT: piezoelectric translator; OL: oil immersion microscope objective (×100,NA = 1.3); DM: dichroic mirror; M: mirror; QWP: quarter-wave plate; S: electronic shutter;L1-L3: lenses; PH: pinhole; F: 580 nm long-pass filter; APD: silicon avalanche photodiode.(b) Simulation of intensity distribution at the focusing region using vector Debye method,including the absorption of the material (σ = 723 m−1). The two curves represent theintensities distributions along transverse (x- or y-axis) and longitudinal (z-axis) directions,respectively, at a 10λ depth.

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20966

to 3D data storage [9], and to 3D optical lithography [4–8]. Until now, it was not possible to useOPA to achieve the same results as those obtained by TPA technique. In fact, in the case of OPA,the excitation wavelength is often chosen at high absorption regions of the studied material, tominimize exposure and time requirements. As a consequence, most incoming photons are ab-sorbed beyond a propagation length of a fewµm. The intensity then exponentially decreasesalong the propagation direction of the light and may even vanish at the focusing region. TheOPA has been therefore used only to deal with a thin material layer, which is repositioned inthe focal plane of the optical system [11]. However, we may notice that in this OPA technique,the material presents a strong linear absorption at the excitation laser wavelength. Therefore,optically induced physical or chemical effects are observed even if the light intensity is weak,not only at the focusing spot but within the whole irradiation region.

We propose to use the specific advantage of OPA technique,i. e., a simple and low cost laser,to achieve 3D addressing by a combination of an ultralow absorption effect and a tightly focus-ing spot. Indeed, the absorption of photosensitive materials has marked spectral features. At thered edge of the absorption band, the absorption is very weak, or close to zero. If a laser beam,whose wavelength is positioned in this range, is applied to CLSM, the light intensity distribu-tion remains almost the same as in the absence of material,i. e., the intensity is increased by afactor of 109 at the focusing spot. Although the absorption is ultra weak, the effective photo-induced effect in focusing region is therefore comparable to that obtained by a laser beam witha high absorption effect. Of course, nothing happens at other out-of-focus areas due to ultralowabsorption effect. Figure 1(a) illustrates a typical CLSM, in which a simple continuous and lowpower laser is used. The laser beam is tightly focused into a submicrometer spot by using a highNA OL (NA = 1.3, oil immersion). Figure 1(b) shows the intensity distribution at the focusingregion, simulated by using the vector Debye method [12]. In this calculation, we have includedthe absorption of the material to be used (σ , absorption coefficient) and investigated the effectof this parameter on the intensity distribution. The result shown in Fig. 1(b) was obtained witha very low absorption coefficient (σ = 723 m−1), corresponding to the value measured in ourexperiment. In the presence of a low absorption coefficient, the focal spot shape remains almostthe same within a 300µm depth inside the material. The typical size of the focal spot is 0.4λ(transverse) and 1.33λ (longitudinal), whereλ is the wavelength of the excitation laser. Due tothe ultra-low absorption of the material at this wavelength, the excitation intensity (accordinglythe laser power) must be high. However, thanks to the use of a high NA OL, just a few milliwatts

Fig. 2. (a) Absorption spectrum of SU8 photoresist, plotted on a logarithmic scale. (b)Fluorescence photon number versus excitation intensity. Square dots are experimental dataand continuous line is a linear fit.

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20967

are enough to achieve a huge light intensity at the focusing area, and to induce photochemicaleffects in that region. Furthermore, since we are dealing with a linear absorption effect, only asimple continuous laser is needed, and the results obtained by this low OPA (called hereafterLOPA) will be same as those obtained by TPA method, in which an expensive femtosecondlaser and accompanying optical components are required.

3. Demonstration of LOPA based microfabrication

This LOPA technique can obviously be used for all 3D applications, including 3D imaging and3D fabrication. In this paper, we present the use of this technique for the fabrication of somerelevant 2D and 3D submicrometer structures, as a proof of principle.

For such a demonstration, two conditions are required: a photoresist that presents an ultralowabsorption at the wavelength of the excitation laser, and a high focusing CLSM. For the firstcondition, we have found that SU8, a commercial photoresist, is an ideal candidate, thanks toits ultralow absorption in the visible range, in particular at 532 nm, which is the wavelength ofa very popular and low cost solid state frequency-doubled neodymium laser. Figure 2(a) showsthe absorption spectra of SU8, on a logarithmic scale, highlighting the very low absorption co-efficient,σ = 723 m−1, atλ = 532 nm. Note that the absorption of commercial SU8 photoresisthas been remeasured [13,14] but little or not attention was paid to its ultralow absorption in thevisible range. The second condition is fulfilled by using a CLSM presented in Fig. 1(a). In such

Fig. 3. Fabrication of voxels at differentz-positions and with different exposure times. (a)Identical voxels array realized at differentz-positions. (b) Example of a voxels array atz4showing the effect of exposure time,i.e., dose. The yellow color curves indicate the shape ofthe voxel as a function of the exposure time (t1, t2 andt3, etc.). (c) Theoretical calculationof the contour plot of light intensity at the focusing region of the used OL (NA= 1.3,n= 1.515,λ = 532 nm). This explains the evolution of the voxel shape as a function of theexposure dose as shown in (b). (d) An array of touching voxels fabricated with low dose(short exposure time,t1), showing an ellipsoidal form similar to results obtained by theTPA method.

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20968

setup, an avalanche photodiode is used to detect very low numbers of photons, which can be areflection or emission light from the focusing spot.

We first demonstrated, by analyzing fluorescence emission, that the SU8 photoresist linearlyabsorbs the excitation laser at 532 nm wavelength. Figure 2(b) shows the fluorescence intensityof the SU8 photoresist as a function of the excitation intensity, obtained by using the CLSMshown in Fig. 1(a). The fluorescence spectrum has been also measured, but not shown here. Thesame measurement is realized on a glass substrate in order to verify the fluorescence emissionof this photoresist. It is clear that the fluorescence emission linearly depends on the excita-tion at 532 nm, confirming thus an OPA effect. We notice that the intensity is on the order of107 W/cm2 at the focusing region, and is only 10−2W/cm2 at the input of the optical system.The LOPA CLSM therefore allows the excitation and fluorescence detection of the focal spotvolume only. The use of an avalanche photodiode (APD) is also a key point in fluorescencemeasurements in which the emission is quite low and cannot be detected by common p-i-nphotodiodes.

4. Submicrometer 3D structures fabrication enabled by LOPA direct laser writing

By using a standard fabrication process, we then demonstrated that polymerization is achievedonly at the focusing spot of the microscope objective, where the excitation intensity is largelysufficient to compensate the low linear absorption of the resist. Note that, in the case of TPAfabrication, there exists a first threshold related to light intensity, above which two photonsare simultaneously absorbed, and a second threshold related to dose, above which completephotopolymerization is achieved. However, it should be borne in mind that resist exposure isalso per se a highly nonlinear process. Indeed, there exists a dose threshold above which thepolymerization process can be fully completed. An example of the complete polymerizationcondition is the use of the interference technique with a low power continuous-wave laser tofabricate 2D and 3D structures [15–17]. Therefore, in the case of LOPA, there exists only onethreshold related to dose. Furthermore, thanks to a high intensity of the focusing spot, thecomplete photopolymerization is only achieved in this region. Figure 3 shows an example ofexperimental results, imaged by a scanning electron microscope (SEM). For each exposure,a solid structure, called “voxel”, corresponding to a focusing spot, is obtained. By changing

Fig. 4. (a) SEM image of a voxels array fabricated at different exposure times and withP = 2.5 mW. (b) Exposure time dependence of voxel size, with different laser power values,P = 2.5 mW; 5 mW; 7.5 mW, respectively. The continuous curves are obtained by a tentativefit using the diameter-dose relationship for one-photon absorption [18, 19]. Insert shows aSEM image of a small voxel obtained with an exposure time of 0.5 second at a laser powerof 2.5 mW.

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20969

either the excitation power or the exposure time,i. e. the dose, the voxel size and shape canbe adjusted, as shown in Figs. 3(a) and 3(b). Note that, in order to obtain and to evaluate thesesmall voxels on a glass substrate, we fabricated the same voxel array at differentz-positionsand with different exposure times. Only 2.5 mW of a continuous green laser is needed to createthese structures, for an exposure time of about 1 second per voxel. Figure 3(b) shows, forexample, a voxel array realized atz4-position, with three different exposure times. The dosedependence can be explained theoretically by calculating the iso-intensity of the focusing spotat different levels, as shown in Fig. 3(c). Three kinds of voxels obtained witht1, t2 andt3 asshown in Fig. 3(b) correspond to three different iso-intensities, namely 0.94, 0.63 and 0.44,respectively. The evolution of voxel size and shape observed experimentally fully confirms thisoperation in the OPA regime. Indeed, in the case of TPA, the creation of bone-like voxel shape,corresponding to that obtained witht3, for example, requires very high excitation intensity andcould not be easily realized due to the TPA intensity threshold, as already mentioned above.In the case of LOPA, all these voxels shapes are obtained by simply adjusting the exposuretime while keeping a low laser power. To obtain small voxels as shown in Fig. 3(d), shorterexposure times,t1, for example, should be applied. Of course, the exposure time required tocreate submicrometer structures varies as a function of the laser power.

Figure 4 shows the voxels size as a function of exposure time for three values of laser power.The exposure time is changed finely around timet1 as shown in Fig. 3(d). A voxel with asize as small as 190 nm is obtained with, for example, a power of 2.5 mW and an exposuretime of 0.5 second, as shown in Fig. 4(b). The fabrication of smaller voxels is possible, butin this first demonstration this performance is limited, due to the quality of the instrumentsused (piezoelectric translation system, SEM, etc.) and also due to the poor adhesion of thesmall polymerized voxel on a glass substrate. The size of individual voxels is quite small whenconsidering the wavelength (532 nm) used for the writing process. As for the linear dependencewith intensity (OPA vs. TPA) [18, 19], the diameter-dose relationship agreement confirms thisbehaviors, as we have used the intensityI0 for our fit, instead ofI2

0 in the case of TPA.With the above evidence that the LOPA polymerization is only realized at the focusing spot

where the excitation intensity is very high, we scanned this focal spot in 3D to fabricate dif-ferent desired structures. Figures 5 and 6 show, as examples, two kinds of submicrometer 3Dstructures fabricated by this LOPA direct laser writing (DLW) method. Clearly, 3D chiral orspiral structures are well created, similar to the results obtained by induced TPA polymeriza-tion. The structures features are well separated, layer by layer, in horizontal and in verticaldirections. The feature sizes are about 300 nm (horizontal) and 650 nm (vertical), as shown in

Fig. 5. SEM image of a chiral structure fabricated with the following parameters: distancebetween rods = 2µm; distance between layers = 0.75µm; number of layers = 20; laserpower = 2.8 mW. (a) View of the whole structure, (b) Zoom in on the top surface of thestructure, and (c) Side view of the structure.

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20970

Fig. 5(c). For the fabrication of these structures, the focusing spot was scanned continuouslywith a scanning speed of about 1.34µm/s. This LOPA technique is easily applicable to anydesired 3D structure with sufficient void fraction. The limit in terms of void fraction is, likely,rather high, as we can guess for instance, from the absence, in the present attempt, of any diffi-culty related to the excess exposure of the regions above the main exposed regions, as apparentalso in the voxel array (Fig. 4(a)). Furthermore, the use of a shorter laser wavelength (532 nmversus 800 nm) allows a reduction of the voxel size, which is related to the diffraction limit ofthe optical system.

We note that Maruo and Ikuta [20] have employed a continuous-wave UV laser to fabricate3D structures, but the mechanism is not understood in their work and the structure size isvery large. Recently, two teams reported on the fabrication of SU8 structures, in which theauthors claimed a TPA polymerization of SU8 photoresist by using a continuous-wave laserat 532 nm [21, 22]. However, as evidenced by our theoretical calculations and experimentalresults, the fabrication of SU8 structure with a continuous laser is actually due to the LOPAeffect. Indeed, such a continuous laser emitting only a few milliwatts (≈ 107 W/cm2) cannotinduce a decent third-order nonlinear effect,i. e. TPA, which usually requires a high peakintensity such as that from a femtosecond laser (≈ 1012 W/cm2). Of course, TPA on SU8 mayoccur at this 532 nm wavelength by using a femto or pico-second laser [22], but its efficiency ishowever very weak as compared with that of the linear absorption even if the OPA coefficientis very low.

5. Energy accumulation effect in LOPA direct laser writing

We have further investigated the fabrication of submicrometer structures when two voxels ortwo lines structures are created closely. In fact, the intensity distribution at the focusing regionof a high NA OL in the focal plane, for example, is represented by an Airy spot surrounded bydiffraction rings of weaker intensity. When the separation between two focal spots is reduced,approaching to the second- or first-order diffraction ring of the Airy spot (focal spot), typically1 µm, a proximity effect is observed. Indeed, in the case of OPA, photons could be absorbedanywhere if they are present, with an efficiency depending on the linear absorption cross-sectionof the irradiated material. The absorbed energy is gradually accumulated as a function of ex-posure time. If a single exposure is made, a single voxel could be created corresponding to avolume in which the accumulated energy is larger than the polymerization threshold [15–17].The area where the accumulated energy is lower than this polymerization threshold will bewashed out during the development process. When multiple exposures are applied, the energy

Fig. 6. SEM image of a spiral structure fabricated with the following parameters: diameterof a spiral = 2 µm; period of spiral inz direction = 2µm; distance between centers of twoclose spirals = 2.5µm; spiral height = 15µm; laser power = 2.6 mW. (a) View of the wholestructure, (b) Zoom in the top surface of the structure, and (c) Side view of the structure.

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20971

of this exposure will contribute to others, if they are very close, and the effective voxel sizeconsequently increases. Figure 7 shows the theoretical calculations and experimental resultsof the distance dependence of the voxels size. A clear accumulation effect is observed whentwo voxels are separated by a distance shorter that 1µm, resulting in a voxel of larger size.According to the same model as that used above, the full width at half maximum of each voxelis increased from 190 nm to 300 nm when the separation changes from 2µm to 0.5µm. More-over, for a short separation, the voxels array is not uniform from its center to the edge part, ascan be seen in case of 0.5µm. Such a consequence is well-known for the proximity effectsin electron-beam lithography for instance and was recently considered for deep UV used forphotonic chips in a microelectronic multichip-project foundry [23]. Small voxels separated byless than 0.5µm-separation are possible. But in this case, a proximity correction,i.e., a controlof exposure time or power and a compensation of the dose between different voxels, should beapplied.

Finally, as mentioned above, this LOPA technique can be applied for any 3D optical ad-dressing. The fabrication of arbitrary 3D structures by this simple technique is just a particularbut also very important application. Indeed, the LOPA CLSM could be used to image sub-micrometer structures in 3D by using a very modest laser power, which allows avoiding thedestruction of studied materials, in particular in biology. An other application could be alsoenvisioned is 3D data storage, for which LOPA allows minimize the optical system by using avery simple diode laser at suitable wavelengths.

Fig. 7. Dependence of voxel sizes on separation distance between voxels, showing the in-fluence of energy accumulation from this focusing spot to others. Insets show simulatedimages (blue background) and SEM images (black background) of 2D submicrometer vox-els array fabricated with different distances between two voxels: 2µm; 1 µm; 0.5µm.

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20972

6. Conclusion

As aconclusion, we have developed a simple and low cost 3D laser writing process based onOPA phenomena in a weakly absorbing photoresist material. This novel technique enables thefabrication of objects as small as 190 nm, and of submicrometer 2D and 3D arbitrary struc-tures, similar to those obtained by TPA technique. The demonstration of this LOPA direct laserwriting technique is realized by a continuous-wave laser at 532 nm with only a few milliwattsand a commercial SU8 photoresist. Our work also brings a clear understanding about the fabri-cation of 3D structures using a continuous laser. We could notably evaluate that the proximityeffects arise only in relatively “full” structures, such effect being not observed in “sparse” struc-tures. The idea of using LOPA also opens a new and inexpensive way to address optically 3Dstructures, namely 3D fluorescence imaging.

Acknowledgments

This work has been supported by the “Laboratoire d’Excellence NanoSaclay”, in the frameworkof the project “ONE-FAB-3D”, and by the “Triangle de la Physique”, in the framework ofthe project “PUTTON”. Special thanks are devoted to J.-F. Roch for various equipment helps.M. T. Do and T. T. N. Nguyen acknowledge the fellowship from the Vietnam InternationalEducation Development “322 program” and Q. Li acknowledges the fellowship from the ChinaScholarship Council.

#191839 - $15.00 USD Received 5 Jun 2013; accepted 28 Jun 2013; published 30 Aug 2013(C) 2013 OSA 9 September 2013 | Vol. 21, No. 18 | DOI:10.1364/OE.21.020964 | OPTICS EXPRESS 20973