December 15, 2008 / Vol. 33, No.

24 / OPTICS LETTERS

2913

High efciency multilevel phase-type fractal zone plates

Dong Wu, Li-Gang Niu, Qi-Dai Chen, Rui Wang, and Hong-Bo Sun*
State Key Laboratory on Integrated Optoelectronics, College of Electronic Science and Engineering, Jilin University, 2699 Qianjin Street, Changchun 130023, China *Corresponding author: hbsun@jlu.edu.cn Received September 8, 2008; revised October 29, 2008; accepted October 29, 2008; posted November 10, 2008 (Doc. ID 101284); published December 4, 2008 Three kinds of high efciency phase-type fractal zone plates were rapidly (in tens of minutes) fabricated by femtosecond laser two-photon photopolymerization. Their theoretical maximal diffractive efciencies of 24.5%, 12.52%, and 18.76% were predicted both by numerical simulation and analytical deduction and were veried by the measured values of 20.5%, 9.1%, and 13%, respectively. The characteristic of multifoci and the improved imaging ability for phase-type fractal lens was also demonstrated. Moreover, to further enhance the diffractive efciency, a four-level fractal phase lens, whose diffraction efciency reached as high as 37.6%, was proposed and realized. 2008 Optical Society of America OCIS codes: 220.4241, 050.1380, 050.6875, 130.1750, 130.3120, 140.3390.

Among various micro-optical components [13], fractal zone plates (FraZPs), owing to their potential in soft x-ray imaging where refractive optics is not available [4,5], are drawing increasing attention. They are constructed with a one-dimensional binary function with a regular Cantor fractal prole, followed by a rotation of the transformed function around one of its extremes. The period of a FraZP is associated with a Fresnel zone plate (FZP) by [6] p N,S = 2/ 2N 1 S . 1

In the rst stage S = 1, the initial plate is divided into an odd number of plates, 2N 1, and the segments in the even positions are removed. For the remaining N segments at the rst stage, this slicing and removing process is repeated in the second stage. The parameter N gives the number of segments forming the fractal structure, and S is the fractal level. Therefore a FraZP is considered as a modied FZP with some missing zones. Saavedra et al. [6] indicated theoretically its unique feature of multiple foci with internal fractal properties along the optical axis, and Davis et al. [7] reported experimentally FraZPs implemented in a liquid crystal display. Furlan et al. [8] found that FraZPs reduced chromatic aberration under white light imaging because they provided an extended depth of eld. These works would accelerate the practical use of the diffractive FraZPs. However, the diffraction efciency of FraZPs has not been systematically investigated from either theoretical or experimental points of view. Moreover, FraZPs reported so far belong to amplitude-type elements, whose diffractive efciency is low. This may limit their application to some extent. In this Letter, we proposed two methods to improve the diffractive efciency of the FraZPs. A phase-type structure is rst adopted, whose diffractive efciency is expected to be four times the amplitude-type one. On this basis, multilevel kinoforms are introduced, which further increases the diffraction efciency to a level as large as 37.6%. The high requirement to the fabrication accuracy is satised by femtosecond direct writing, a
0146-9592/08/242913-3/$15.00

powerful technique for three-dimensional (3D) micronanostructures fabrication [916]. Shown in Figs. 1(a) and 1(b) are the birds-eye view and locally magnied scanning electron microscopic (SEM) images of a phase-type FraZP fabricated by two-photon photopolymerization of the commercial resin SU-8, of which the refractive index is 1.56 at 532 nm wavelength. The femtosecond laser we used was a 120 fs pulse width, 1.3 W average power, and 790 nm wavelength Ti:sapphire pulsed laser (from Tsunami, Spectra-Physics), which operates at 80 MHz repetition. The laser beam was tightly focused by a high numerical aperture (NA= 1.35, 60 ) oil immersion objective lens. The focal spot was scanned laterally by steering a two-galvano-mirror set and along the optical axis by a piezo stage, both

Fig. 1. (Color online) Micro-phase-type FraZP (N = 2, S = 2) fabricated by femtosecond laser two-photon photopolymerization. (a) Top-view SEM image. (b) Its locally magnied SEM view. (c) Simulated intensity distribution of phase- (dashed curve) and amplitude-type (solid curve) FraZPs. The inset is AFM surface image of photopolymerized SU-8. (d) Measured focal spot. It is worth mentioning that here (a) and in Fig. 4(a), an elevated spot at the center of the FraZP is clearly visible. It is an experimental artifact caused by a failure of synchronization between the optical shutter and the two-galvano-mirror set. 2008 Optical Society of America

2914

OPTICS LETTERS / Vol. 33, No. 24 / December 15, 2008

Fig. 2. (Color online) Focusing and imaging characteristics of the FraZP (N = 2, S = 3) and FZP. (a) and (b) are SEM images of the FraZP and FZP. (c) and (d) are the images of the letters JLU by the FraZP and FZP. (e) and (f) are photographs of foci for the two lenses. The insets correspond to the simulated foci.

Fig. 3. Multiple foci of the phase FraZP (N = 3, S = 2). (a) SEM image of the entire FraZP. (b) Its locally magnied view. (c) Simulated intensity distribution along the optical axis. The six insets are the measured multifoci.

2NS 1 N,S,n = 2N 1
S

sinc2 1/n .

with high motion accuracy. The laser scanning was performed with an annular mode [10]. After femtosecond laser direct writing, the resin SU-8 was postbaked with a temperature ramp from 65 C to 95 C for 10 min, and was developed in SU-8 developer for 60 min to remove the unsolidied liquid resin, leaving a solid skeleton. The fractal lens was characterized with an optical setup comprising a 3D positioning stage and a charge coupled device (CCD) camera [10]. According to our previous systematic study [16] and atomic force microscopy (AFM) measurements conducted on the polymerized SU-8 platform region, the surface roughness in an approximately 10 m 10 m area is better than 10 nm [the inset of Fig. 1(c)]. The FraZP in Fig. 1(a) is of the phase type with N = 2, S = 2. In the phase-type diffractive lens, all the odd zones and even zones are transparent, but they have different refraction indices that induce a phase change over the zone plates thickness. Here, the thickness of the FraZP is 475 nm to induce the phase change of for light wavelength at 532 nm. The diameter of the inner ring is 12 m, much smaller than the ones reported previously: around 3.1 mm. Generally speaking, a fractal lens with a small scale has better resolution and imaging ability. The focal spot was shown in Fig. 1(d). The measured diffractive efciency, dened as the ratio of the power collected to the primary focal spot to the total incidence, is 20.5%. The issue of theoretical diffractive efciency, however, has not been deeply investigated, which is generally low for an amplitude-type device. We calculated the diffractive efciency of a FraZP, and obtained, for an n-level kinoform,

For N = 2, S = 2, the FraZP contains nine halfwavelength zones with two missing zones, giving rise to a diffractive efciency of 2 , 2 , 2 = 24.5%. On the other hand, based on the HuygensFresnel principle, the intensity distribution of the phase lens was simulated. The dashed and solid curves shown in Fig. 1(c) represent the simulated intensity distributions of phase- and amplitude-type FraZPs, respectively, in the focal plane perpendicular to the optical axis. It is obvious that the intensity for the phase-type is greater than the amplitude-type one. Their diffractive efciencies are 24.46% and 6.12%, respectively. Compared with the amplitude-type FraZP, the measured diffractive efciency for the phase FraZP, 20.5%, is distinctly enhanced and agrees well with the analytical value 24.5% and the numerical value 24.46%. The resolution, determined by the width of the outmost ring, is not ideal for the FraZP of N = 2, S = 2. For the higher fractal level S = 3, it has more rings and the width of the outmost ring is smaller, from which higher resolution and better imaging ability is expected. Shown in Fig. 2(a) is the SEM image of phase-type FraZP of (N = 2, S = 3). The width of the outmost ring is about 650 nm [Fig. 2(b)]. The FraZP reduces chromatic aberration under white light illumination because it shows multiple foci with internal fractal properties along the optical axis and provides an extended depth of eld. This advantage is evidently discerned in the photographs of the letters JLU in Figs. 2(c) and 2(d) imaged by the FraZP and FZP, respectively. The focal spot proles of the FraZP and FZP are shown in Figs. 2(e) and 2(f). The size of

December 15, 2008 / Vol. 33, No. 24 / OPTICS LETTERS

2915

= 2, S = 2). Its theoretical diffractive efciency is 2 , 2 , 4 = 49.0%. For an n-level lens, the thickness is dened by d= n nSU8 1 . 3

Fig. 4. Multilevel fractal diffractive phase lenses. (a) Topview SEM photograph of four-level phase-type fractal diffractive lenses. (b) Its locally magnied SEM view. (c) Photograph of the main focal spot.

For two- and four-level lenses, the single-level thicknesses are 475 and 238 nm, respectively, to induce the phase change of and / 2 at each subzone. The measured diffractive efciency reaches as high as 37.6%. In conclusion, phase-type and multilevel structures were proposed to enhance the diffractive efciency of the FraZP. The delicate devices were fabricated by femtosecond laser via two-photon polymerization. The measured diffraction efciencies are comparable to the theoretical maximum. This work shows that two-photon photopolymerization is a rapid and precise means for fabricating micro-optical and integrated optical components and systems. The improved diffractive capabilities may push wide applications of FraZPs in elds such as soft x-ray microscopy, terahertz imaging, tomography, and integrated circuits. This work was supported by the National Science Foundation of China (NSFC) under grant 60525412. References
1. E. Di Fabrizio, F. Romanato, M. Gentili, S. Cabrini, B. Kaulich, J. Susini, and R. Barrett, Nature 401, 895 (1999). 2. G. Roelkens, D. V. Thourhout, and R. Baets, Opt. Lett. 32, 1495 (2007). 3. S. Serak, N. Tabiryan, and B. Zeldovich, Opt. Lett. 32, 169 (2007). 4. W. L. Chao, B. D. Harteneck, J. A. Liddle, E. H. Anderson, and D. T. Attwood, Nature 435, 1210 (2005). 5. L. Kipp, M. Skibowski, R. L. Johnson, R. Berndt, R. Adelung, S. Harm, and R. Seemann, Nature 414, 184 (2001). 6. G. Saavedra, W. D. Furlan, and J. A. Monsoriu, Opt. Lett. 15, 971 (2003). 7. J. A. Davis, L. Ramirez, J. A. Rodrigo Martin-Romo, T. Alieva, and M. L. Calvo, Opt. Lett. 29, 1321 (2004). 8. W. D. Furlan, G. Saavedra, and J. A. Monsoriu, Opt. Lett. 32, 2109 (2007). 9. S. Jeon, V. Malyarchuk, J. A. Rogers, and G. Wiederrecht, Opt. Express 14, 2300 (2006). 10. Q. D. Chen, D. Wu, L. G. Niu, J. Wang, X. F. Lin, H. Xia, and H. B. Sun, Appl. Phys. Lett. 91, 171105 (2007). 11. S. Kawata, H. B. Sun, T. Tanka, and K. Takada, Nature 412, 697 (2001). 12. H. B. Sun, S. Matsuo, and H. Misawa, Appl. Phys. Lett. 74, 786 (1999). 13. M. Deubel, G. Freymann, M. Wegener, S. Pereira, K. Busch, and C. M. Soukoulis, Nature Mater. 3, 444 (2004). 14. K. K. Seet, V. Mizeikis, S. Matsuo, S. Juodkazis, and H. Misawa, Adv. Mater. 17, 541 (2005). 15. H. B. Sun, K. J. Takada, and S. Kawata, Appl. Phys. Lett. 79, 3173 (2001). 16. H. B. Sun and S. Kawata, Appl. Phys. Lett. 86, 071122 (2005). 17. K. Yamada, W. Watanabe, Y. Li, K. Itoh, and J. Nishii, Opt. Lett. 29, 1846 (2004).

the focal spot of the FraZP is the same as the one of the FZP, but there are more background light rings at the side of the main focal spot for the FraZP. The measured efciency is 9.1%, lower than the numerical value 12.48% and the analytical value 12.52%. One of the reasons for the deviation, in addition to the structural imperfections induced during fabrication, such as surface roughness, is that the cross section of the rings is not exactly the square shape as assumed in the analytical deduction and numerical calculation. To demonstrate along the optical axis multiple foci, the main reason for the FraZP to reduce chromatic aberration, the FraZP of N = 3 and S = 2, a structure not reported before, is studied [Figs. 3(a) and 3(b)]. From the simulated axial light intensity distribution [Fig. 3(c)] it is seen that there are more than one focal spots along the optical axis. Under 532 nm green laser illumination, the focal spot was detected with a CCD camera as a function of distance from the FraZP. The six insets are recorded focal plane images at different distances, which are in good agreement with theory. Three focal spots at distances of 57, 68, and 90 m were observed. Furthermore, another weak focal spot at the distance of 125 m was discovered. At the positions of 63 and 75 m the expected minima were found. The measured efciency is 13.00%, compared with the numerical value 18.73% and the analytical value 18.76%. An effective means to enhance the diffractive efciency of the FZP is to design a multilevel phase lens [17]. Likewise, a multilevel structure was adopted. Shown in Figs. 4(a) and 4(b) are the entire and locally magnied SEM images of the four-level FraZP (N