May 1, 2009 / Vol. 34, No. 9 / OPTICS LETTERS 1393

Process equipped with a sloped UV lamp for thefabrication of gradient-refractive-index lenses

Jui-Hsiang Liu* and Yi-Hong ChiuDepartment of Chemical Engineering, National Cheng Kung University Tainan, Taiwan 70101

*Corresponding author: jhliu@mail.ncku.edu.tw

Received February 9, 2009; revised March 30, 2009; accepted April 1, 2009;posted April 2, 2009 (Doc. ID 106508); published April 23, 2009

In this investigation, a method for the preparation of gradient-refractive-index (GRIN) lenses by UV-energy-controlled polymerization has been developed. A glass reaction tube equipped with a sloped UV lamp wasdesigned. Methyl methacrylate and diphenyl sulfide were used as the reactive monomer and nonreactivedopant, respectively. Ciba IRGACURE 184 (1-hydroxy-cyclohexyl-phenyl-ketone) was used as the initiator.The effects of initiator concentration, the addition of acrylic polymers, and the preparation conditions on theoptical characteristics of the GRIN lenses produced by this method were also investigated. Refractive indexdistributions and image transmission properties were estimated for all GRIN lenses prepared.© 2009 Optical Society of America

OCIS codes: 060.0060, 220.0220.

A gradient-refractive-index (GRIN) rod lens is a cy-lindrical medium with a parabolic refractive distribu-tion, in which the refractive index is the highest onthe rod’s optical axis and decreases toward the pe-riphery at a rate proportional to the square of the ra-dial distance from the optical axis. GRIN elementswith imaging and light-focusing properties have beenwidely used in image-transmission systems [1–3] andoptical communication systems, such as copy ma-chines, facsimile lens arrays, and optical circuit net-works [4–8]. Glasses and polymers are the two mostcommonly applied materials in optics. Glasses haveexcellent transparency and low optical attenuation,but brittleness and high processing costs are signifi-cant disadvantages. Polymers have higher opticallosses but offer excellent mechanical properties,lighter weight, superior flexibility, simplified process-ing, and lower cost. As a result, several polymericGRIN lenses have been developed [9,10].

Recently, GRIN lenses with quadratic refractive in-dex distributions that vary continuously from the op-tical axis to the periphery have been widely studiedbecause of their potential application in self-focusingimaging and optical communications. Several meth-ods have been used to prepare GRIN lenses, includ-ing two-step copolymerization [11,12], extrusion [13],interfacial-gel copolymerization [14], and photopoly-merization [15].

In previous studies, we have reported the prepara-tion of large-core bubble-free GRIN lenses viaswollen-gel polymerization and centrifugal polymer-ization [16–19]. In the present Letter, we propose amethod to fabricate GRIN plastic rods. A monomermixture in a glass tube was irradiated with con-trolled UV light. Owing to the use of controlled UVenergy, light-induced polymerization occurred fromthe bottom up, leading to the formation of a bubble-free GRIN plastic rod.

Ye et al. previously reported a rapid “on-demand”laser irradiation method to fabricate GRIN lenses [9].In this system, flexible solids with active components

were designed to achieve diffusion-driven photopoly-

0146-9592/09/091393-3/$15.00 ©

merization. In this Letter, we used a UV lampequipped with a small sloped angle to fabricate GRINrod lenses. As seen in Fig. 1, the energy distributionis achieved using a gradient energy-controlled slopedUV lamp. GRIN lenses could thus be fabricated viaone-step UV exposure. A liquid monomer mixtureconsisting of the monomer, a photoinitiator, and non-reactive diphenyl sulfide was used in this process,with the advantage that the presence of a liquidphase effectively compensated for polymer matrixshrinkage due to the polymerization of reactivemonomers.

In the controlled UV light polymerization process,a mixture of methyl methacrylate (MMA) and diphe-nyl sulfide (DS) with a specified amount of acrylicpolymer and Ciba IRGACURE 184 (1-hydroxy-cyclohexyl-phenyl-ketone) initiator was poured into aglass tube. The tube containing the monomer mix-ture was then capped and vertically rotated at20 rpm. As shown in Fig. 1, a germicidal UV lamp(10 W, �max=254 nm) equipped with a small slope

Fig. 1. (Color online) Schematic representation of theequipment for UV-controlled polymerization of plastic rods.The UV lamp is equipped with a small slope angle. Thehighest energy exists at the bottom and decreases gradu-ally from the bottom to the top, as indicated by the dottedlines. This is consistent with gel zone formed in the reac-

tion tube between A and B.

2009 Optical Society of America

1394 OPTICS LETTERS / Vol. 34, No. 9 / May 1, 2009

angle (�=7.6° to vertical) provided the UV lightsource. Owing to the incident light distribution, thebottom of the reactor tube receives a higher energydose. The incident energy distribution can be mod-eled as a V curve between A and B, with the highestenergy intensity occurring at the bottom of the reac-tor and decreasing gradually from bottom to top, asindicated by the dotted lines in Fig. 1. During the ini-tial stage, monomers at the bottom accept higher en-ergy and polymerize rapidly, leading to the formationof a polymer gel at the bottom of the glass tube. Asshown in Fig. 1, a V-shaped gel distribution occurs asthe reactor continues to receive UV irradiation. Thisdistribution results in the gradual formation of apolymer from the periphery to the center of the rod,with the polymerization geometry “pushing” most ofthe nonreactive DS dopant toward the center of therod as well. The final distribution of the DS dopantfixed in the polymer matrix produces the desired gra-dient refractive index profile. Shrinkage due to thepolymerization of monomers in the matrix was offsetby the presence of an upper layer of liquid monomers.Specifically, because polymerization involves thebinding of monomers, this process usually drawsmonomer molecules closer together, resulting inshrinkage of the polymer matrix. If monomers do notfill the A–B zone shown in Fig. 1, a gel rod withbubbles will be produced by the polymerization reac-tion. In this process, a bubble-free gel rod was formedvia primary controlled UV polymerization and thenheated at 60°C for 24 h to completely polymerize themonomers contained in the rod. In this investigation,a Pyrex glass tube was used as a reactor. After poly-merization, the glass tube was scored with a diamondcutter. The touch of a melted glass ball at the scorededge could break the glass tube, and the GRIN lenscould then be separated from the glass tube.

Figure 2 shows the dependence of the initiator con-

Fig. 2. (Color online) Dependence of the initiator concen-tration on the �n distribution of the GRIN rods. Initiatorconcentrations: circles, 1.0 wt.%; diamonds, 1.5 wt.%; tri-angles, 2.0 wt.%. Increased initiator concentration de-

creases the refractive index difference.

centration on the �n distribution of the GRIN lens,where np and Rp denote the refractive index at theperiphery and the radius of the plastic rod, respec-tively. The results of the refractive index distributionsuggest that the concentration of high-refractive-index dopant decreases from the center axis to theperiphery of the gel rod, resulting in a GRIN distri-bution. Higher initiator concentrations were found toincrease the polymerization rate. By contrast, nonre-active DS was fixed more easily in the polymer ma-trix, leading to a decrease in refractive index at thecentral axis. As shown in Fig. 2, unlike plastic rodsprepared by the swollen-gel technique [16], the avail-able refractive index distribution area produced bythis method covers the rod almost entirely. These re-sults suggest that the DS material is effectively“pushed” inward throughout the entire course of thepolymerization process. The refractive index profileof the prepared GRIN lenses was measured using aYork-P102 profile analyzer with an accuracy of±0.001. The refractive index of the matching oil usedin the system was 1.458. Images viewed through theGRIN polymer rods were recorded using a cameraequipped with an enlarging lens.

The effects of the addition of PMMA on the opticalcharacteristics of GRIN lenses were also studied. Theresults are summarized in Table 1. The parametersof the fabricated GRIN lenses were estimated usingthe following equations:

n0 sin �max = NA, �1�

NA = n0�2�n�1/2, �2�

L = 2�/A, �3�

where n0 denotes the refractive index at the rod cen-ter, L corresponds to the pitch of the rod, �n is therefractive index difference, NA is the numerical aper-ture, and A is the transmitting constant. The addi-tion of PMMA into the plastic rod was proposed as ameans to inhibit the shrinkage caused by polymeriza-tion. However, as seen in Table 1, the increase ofPMMA leads to a decrease in the NA, suggesting thatthe addition of PMMA may increase the viscosity ofthe monomeric mixture, thereby decreasing the mo-bility of DS in the polymerization system. This maythus lead to a decrease in the refractive index differ-ence and the NA values. Our results suggest that a

Table 1. Effects of the Addition of PMMA on theOptical Characteristics of GRIN rodsa

PMMAb

(wt.%) 0% 5% 10%

�n 0.0267 0.0237 0.0230A 0.0478 0.0317 0.0456

NA 0.2841 0.2668 0.26282�max 33.07 30.95 30.47

aMMA /DS=3 /1, Ciba IRGACURE 184=1 wt.%.bAmount of PMMA added.

gradient refractive index plastic rod could be success-

May 1, 2009 / Vol. 34, No. 9 / OPTICS LETTERS 1395

fully fabricated without the addition of PMMA andthat under these conditions, no significant voids andbubbles were observed inside the fabricated lens.These results suggest that any shrinkage caused bythe polymerization of the monomers was compen-sated for by the presence of liquid monomers at thetop of the reaction tube.

The reproducibility and thermal stability of thefabricated GRIN plastic rods were confirmed. WhileDS is a nonreactive compound that is fixed by thePMMA matrix at normal operating temperatures, theplastic rod may not be usable at higher tempera-tures. After a series studies, only a small decay in therefractive index gradient is observed after a heattreatment at 60°C for 24 h. The heat tolerance ofGRIN lenses produced by this method may be im-proved by the addition of cross-linking multifunc-tional monomers [17,18]. The results obtained sug-gest that this UV-controlled polymerization methodmay be suitable for larger-scale fabrication of GRINplastic rods.

Figure 3 shows a color (online) image transmittedthrough a freshly fabricated GRIN lens with a 15 mmdiameter and 80 mm length. An inverted virtual im-age was obtained through the plastic rod fabricatedin this investigation. The distance between the imageand the face of the GRIN lens was 60 mm in thistrial. The quality of these results suggests that GRINlenses can be successfully fabricated via UV-controlled polymerization to produce GRIN lenseswith excellent optical properties.

Theoretically, controlled energy polymerization

Fig. 3. (Color online) Recorded color image transmittedthrough a freshly fabricated GRIN lens with 15 mm diam-eter and 80 mm length.

could also be achieved using parallel UV lamps sepa-

rated by a grayscale filter. In such an arrangement,transmitted UV energy should increase from the bot-tom to the top of the grayscale filter. Thus, an energy-controlled polymerization similar to the reaction de-scribed in Fig. 1 should be possible, with a V-shapedgel distribution resulting from UV irradiation.

In conclusion, we have demonstrated that a UV-controlled polymerization can be used to fabricateGRIN plastic rods. This method is a technique that iseasy to perform and requires relatively inexpensiveequipment and materials.

The authors thank the National Science Council ofthe Republic of China (Taiwan) and National ChengKung University for financially supporting this re-search under contracts 96-2221-E006-009 and B0040,respectively.References

1. K. Iga and N. Yamamoto, Appl. Opt. 16(5), 1305 (1977).2. H. Tsuchida, T. Nagaoka, and K. Yamamoto, Jpn. J.

Appl. Phys. Part 1 37, 3633 (1998).3. W. A. Reed, M. F. Yan, and M. J. Schnitzer, Opt. Lett.

27, 1794 (2002).4. F. Okano, J. Arai, and M. Okui, Opt. Lett. 31, 1842

(2006).5. G. Stewart, A. Mencaglia, W. Philip, and W. Jin, J.

Lightwave Technol. 16, 43 (1998).6. S. H. Song, S. Park, C. H. Oh, P. S. Kim, M. H. Cho,

and Y. S. Kim, Opt. Lett. 23, 1025 (1998).7. T. Fukushima, T. Yokota, and T. Sakamoto, J.

Lightwave Technol. 15, 1938 (1997).8. F. J. Marinho and L. M. Bernardo, Opt. Lett. 31, 1723

(2006).9. C. Ye and R. R. McLeod, Opt. Lett. 33, 2575 (2008).

10. X. Chen and N. George, Appl. Opt. 47, 6190 (2008).11. Y. Ohtsuka and Y. Terao, J. Appl. Polym. Sci. 26, 2907

(1981).12. Y. Ohtsuka and T. Sugaho, Appl. Opt. 22, 413 (1983).13. B. C. Ho, J. H. Chen, W. C. Chen, and Y. H. Chang,

Polym. J. (Tokyo, Jpn.) 27, 310 (1995).14. Y. Koike, T. Ishigure, and E. Nihei, J. Lightwave

Technol. 13, 1475 (1995).15. J. H. Liu, J. H. Wang, and M. H. Chu, Angew.

Makromol. Chem. 174, 1 (1990).16. J. H. Liu, H. Y. Wang, and C. H. Ho, J. Polym. Res. 10,

13 (2003).17. J. H. Liu, D. S. Wu, and K. Y. Tseng, Macromol. Chem.

Phys. 205, 2205 (2004).18. J. H. Liu, P. C. Yang, and Y. H. Chiu, J. Polym. Sci.,

Part A: Polym. Chem. 44, 5933 (2006).

19. J. H. Liu and H. T. Liu, Opt. Lett. 22, 668 (1997).