5
Simple technique for replication of micro-optical elements Philippe Nussbaum Ire ` ne Philipoussis Aline Husser Hans Peter Herzig University of Neucha ˆ tel Institute of Microtechnology Rue A. L. Breguet 2, CH-2000 Neucha ˆ tel Switzerland E-mail: [email protected] Abstract. A simple technique for the replication of micro-optical ele- ments is presented. Elastomeric material is used to realize a negative mold of the original optical element and UV-curing adhesive is used to make the replicated copy. Replicated elements, such as refractive and diffractive micro-optical elements, are produced. The refractive micro- lenses have a diameter of 970 mm and a height of 79 mm and the dif- fractive element is a multilevel blazed grating with a period of 64 mm. The characteristics of the replicated elements are measured using differ- ent methods. The deviation from a sphere of the original and the repli- cated refractive microlenses is 0.13l and 0.12l root mean square (rms), respectively. The replicated multilevel blazed grating has a diffraction efficiency of over 80%. An alternative method for realizing planoconcave microlenses from a planoconvex master is also presented. Subject terms: replication; micro-optical elements; microsystems; elastomeric material. 1 Introduction The replication of surface-relief micro-optical elements, such as refractive microlens arrays or diffractive optical elements ~DOEs!, is of great interest for optical microsystems. 1 Compared to the initial photoresist element, the replicated element offers high transmission and reduced material fluorescence. Standard replication techniques, such as hot embossing, molding, or casting, are key technologies for low-cost mass-production of micro-optical elements. 2 In contrast to these well-known techniques, we investigate a new method that offers a facility to replicate optical ele- ments without requiring complicated and expensive infra- structures, such as sputtering chambers, electroplating baths, embossing machines, or injection molding systems. This method was first applied to optical elements by Wilbur et al. 3 and Xia et al. 4 Their papers present the basic prin- ciple and some mechanical properties of the elements. In a more recent publication, Daly et al. 5 used the same tech- nique to replicate refractive microlens arrays with elasto- meric material for the mold. They used thermally curable epoxy for the replicated element. We report on the optical properties of replicated micro- optical elements. For the replication we use the same com- mercially available elastomeric material to realize the mold as in Ref. 5, but we use a UV-curing adhesive for the rep- lication. The elements were refractive microlens arrays, having features up to 1.4 mm diameter and 95 mm lens height, and multilevel diffractive blazed gratings, having up to eight levels and periods down to 2 mm. Different char- acterization methods, such as accurate interferometric analysis of the microlens shape or diffraction efficiencies of the blazed gratings, are used to analyze the performance of the replications. An alternative method to realize good quality planoconcave refractive microlenses is also pre- sented. First, we describe the simple replication method and then we present results for the realized elements. 2 Fabrication Method The fabrication of replicated elements is realized in two steps. The first step consists of making the negative mold of the original element, and the second step is the replication in UV-curing adhesive, as shown in Fig. 1. The original micro-optical element is first prepared by a cleaning proce- dure using a nitrogen jet, if the original element is made of photoresist, or using solvents, if it is made of a hard mate- rial such as glass or silicon. The elastomeric material used for the mold is commercially available from Dow Corning ~Sylgard 184!. This material is a two-component silicon elastomer developed for encapsulating and protecting elec- tronic circuits or solar cells against external influences such as shocks or dust particles. For the preparation of the moul- ding material, the manufacturer recommends a mixture of 1 part of catalyzer for 10 parts of elastomer. We also tried other mixtures with less success. After pouring the mixture onto the original master, both are put in an oven and baked at 50°C for one night ~12 h!. Once the elastomer is hard- ened, it is very simple to separate the mold from the origi- nal master. The adhesion of both parts is only by vacuum, thus the original master remains intact after the separation. This property can be an important issue if several molds must be realized from the same original master. For the replication step, a standard UV-curing adhesive is used. The NOA 61 from Norland Optical Adhesives shows very good results for refractive and diffractive micro-optical elements. The adhesive presents good optical quality, a refractive index of 1.56, and good transmission for visible and IR light. For the UV region below 400 nm, the adhesive has a high absorbance, as do all organic ma- Published in Optical Engineering 37, issue 7, 1804-1808, 1998 which should be used for any reference to this work 1

Simple technique for replication of micro-optical elements

  • Upload
    others

  • View
    8

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Simple technique for replication of micro-optical elements

Published in Optical Engineering 37, issue 7, 1804-1808, 1998which should be used for any reference to this work

1

Simple technique for replication of micro-optical elements

Philippe NussbaumIrene PhilipoussisAline HusserHans Peter Herzig

University of NeuchatelInstitute of MicrotechnologyRue A. L. Breguet 2,CH-2000 NeuchatelSwitzerland

E-mail: [email protected]

Abstract. A simple technique for the replication of micro-optical ele-ments is presented. Elastomeric material is used to realize a negativemold of the original optical element and UV-curing adhesive is used tomake the replicated copy. Replicated elements, such as refractive anddiffractive micro-optical elements, are produced. The refractive micro-lenses have a diameter of 970 mm and a height of 79 mm and the dif-fractive element is a multilevel blazed grating with a period of 64 mm.The characteristics of the replicated elements are measured using differ-ent methods. The deviation from a sphere of the original and the repli-cated refractive microlenses is 0.13l and 0.12l root mean square (rms),respectively. The replicated multilevel blazed grating has a diffractionefficiency of over 80%. An alternative method for realizing planoconcavemicrolenses from a planoconvex master is also presented.

Subject terms: replication; micro-optical elements; microsystems; elastomericmaterial.

1 Introduction

The replication of surface-reliefmicro-optical elements,such as refractive microlens arrays or diffractive opticalelements ~DOEs!, is of great interest for opticalmicrosystems.1 Comparedto theinitial photoresistelement,thereplicatedelementoffershigh transmissionandreducedmaterialfluorescence.Standardreplicationtechniques,suchashot embossing,molding,or casting,arekey technologiesfor low-costmass-productionof micro-opticalelements.2 Incontrastto thesewell-known techniques,we investigateanew methodthat offers a facility to replicateoptical ele-mentswithout requiring complicatedand expensiveinfra-structures, such as sputtering chambers,electroplatingbaths,embossingmachines,or injection molding systems.This methodwasfirst appliedto opticalelementsby Wilburet al.3 and Xia et al.4 Their paperspresentthe basicprin-ciple andsomemechanicalpropertiesof the elements.In amore recentpublication,Daly et al.5 usedthe sametech-nique to replicaterefractivemicrolensarrayswith elasto-meric material for the mold. They usedthermally curableepoxyfor the replicatedelement.

We reporton the optical propertiesof replicatedmicro-optical elements.For the replicationwe usethesamecom-merciallyavailableelastomericmaterialto realizethemoldasin Ref. 5, but we usea UV-curing adhesivefor the rep-lication. The elementswere refractive microlens arrays,having featuresup to 1.4 mm diameterand 95 mm lensheight,andmultilevel diffractive blazedgratings,havingupto eight levelsandperiodsdown to 2 mm. Different char-acterization methods, such as accurate interferometricanalysisof themicrolensshapeor diffraction efficienciesoftheblazedgratings,areusedto analyzetheperformanceofthe replications. An alternative method to realize goodquality planoconcaverefractive microlensesis also pre-

sented.First,wedescribethesimplereplicationmethodandthenwe presentresultsfor the realizedelements.

2 Fabrication Method

The fabrication of replicatedelementsis realized in twosteps.Thefirst stepconsistsof makingthenegativemold ofthe original element,andthe secondstepis the replicationin UV-curing adhesive,as shown in Fig. 1. The originalmicro-opticalelementis first preparedby a cleaningproce-dureusinga nitrogenjet, if theoriginal elementis madeofphotoresist,or usingsolvents,if it is madeof a hardmate-rial suchasglassor silicon. The elastomericmaterialusedfor the mold is commerciallyavailablefrom Dow Corning~Sylgard 184!. This material is a two-componentsiliconelastomerdevelopedfor encapsulatingandprotectingelec-tronic circuitsor solarcellsagainstexternalinfluencessuchasshocksor dustparticles.For thepreparationof themoul-ding material,themanufacturerrecommendsa mixtureof 1part of catalyzerfor 10 partsof elastomer.We also triedothermixtureswith lesssuccess.After pouringthemixtureonto theoriginal master,bothareput in anovenandbakedat 50°C for one night ~12 h!. Oncethe elastomeris hard-ened,it is very simpleto separatethe mold from the origi-nal master.The adhesionof both partsis only by vacuum,thusthe original masterremainsintact after the separation.This property can be an important issueif severalmoldsmustbe realizedfrom the sameoriginal master.

For the replicationstep,a standardUV-curing adhesiveis used. The NOA 61 from Norland Optical Adhesivesshows very good results for refractive and diffractivemicro-opticalelements.Theadhesivepresentsgoodopticalquality, a refractive index of 1.56, and good transmissionfor visible andIR light. For the UV regionbelow 400 nm,the adhesivehasa high absorbance,asdo all organicma-

Page 2: Simple technique for replication of micro-optical elements

2

terials.A small dropletof this NOA 61 adhesiveis pouredon the elastomericmold. Caremust be takento avoid airbubbles.A substrate~e.g.,glass,quartz,or plastic! is put ontop of this adhesivelayer to providea flat optical substratefor thereplicatedelement.For thepolymerization,a 500-Wflood exposuremercurylamp from Oriel is used.It is im-portantthat the adhesiveis well curedunderUV light. Inour case,an exposureof 20 min with an intensity of 10W/cm2 wasfoundto be ideal.Finally, theseparationof thereplicatedmicro-opticalelementfrom theelastomericmoldis againvery simple.

3 Experimental Results

3.1 Refractive Microlenses

Refractivemicrolensarrayswere usedas original mastersto verify thequality of this replicationmethod.Theoriginalmicrolensarraywasrealizedusingthe melting resisttech-nology,wherephotoresistpedestalswereobtainedby pho-tolithographicsteps.6,7 A melting step forms planoconvexmicrolenses.Using this technique,the realized elementswerequasi-perfectpartsof a sphere.8 The replicationtech-niquedescribedin the precedingsectionwasusedto makecopiesof thesemicrolenses.Thediameterof thelenseswas

Fig. 1 Replication of micro-optical elements using elastomericmolds and UV-curing adhesive.

970 mm andthe heightwas79 mm. A Twyman-Greenin-terferometer was used to characterize the replicatedelements.9 Theresultswerecomparedto themeasurementsof the original master.Figure 2 showsthe deviationfromsphereof the original resist microlensand the deviationfrom sphereof the microlensreplicatedin UV-curing ad-hesive.Theobserveddeviationfrom spherewas0.13l rootmeansquare~rms! and 1.11l peak-to-valley~p/v! for theoriginal masterand 0.12l ~rms! and 1.00l peak-to-valley~p/v! for the replicatedmicrolens.

An atomic force microscope~AFM! was usedto mea-surethe surfaceroughnessof the original photoresistmas-ter microlensandof the replicatedepoxymicrolens.A sur-face roughnessof 2 to 3 nm ~rms! was measuredfor theoriginal photoresistelementand about4 nm ~rms! for thereplicated structure.These surface roughnessvalues arenegligible for visible light and are of less importanceforUV light sincethe UV-curing adhesivehaslow transmis-sion in this wavelengthdomain ~30% transmissionat 300nm for NOA 61 adhesive!.

Replicationof micro-opticalelementsprovideshigh fi-delity of the geometricalshapeand dimension:diameter,lensheight,or radiusof curvatureof the original element.Due to thechangeof materialbetweentheoriginal andthereplicatedelement,the optical propertieschangealso. Achangein therefractiveindex,for example,directly affectsthefocal lengthof thereplicatedelement.This fact mustbetaken into account in the design of a replicatedmicro-optical element.

3.2 Planoconcave Replicated Microlens from aPlanoconvex Original Master

The melting resist technologyis limited to the fabricationof planoconvexlenses.Thereforeit is interestingto realizeplano-concavelenses by replication. The mold of a1.4-mm-diamand 95-mm-height microlens was realized.This planoconcavemold servedas a masterfor a second

Fig. 2 Deviation from sphere measured with a Twyman-Green interferometer: (a) original resist mas-ter and (b) replicated element. The diameter of the microlens is 970 mm and the height is 79 mm.

Page 3: Simple technique for replication of micro-optical elements

3

Fig. 3 Surface profiles measured with a stylus profilometer: (a) original planoconvex resist microlensmaster, (b) replicated planoconvex microlens, and (c) replicated planoconcave microlens. The diam-eter of the lens is 1.4 mm and the height is 95 mm.

molding.From this secondplanoconvexelastomericmold,a replication was madeusing UV-curing adhesiveas de-scribedpreviously.The achievedelementwasa replicatedplanoconcavemicrolens.Figure 3 showsthe surfacepro-file, measuredwith a Tencor stylus profilometer, of theoriginal planoconvexresistmicrolensmaster,thereplicatedplanoconvexmicrolens, and the replicatedplanoconcavemicrolens.

The replicatedplanoconvexandplanoconcaveelementshavethe samegeometricaldimensions.The measuredsur-faceprofilesshowthesameheightof 95 mm for a diameterof 1.4 mm for the threeelements~Fig. 3!. No significanterror is observedbetweenthe profiles. A Twyman-Greeninterferometeris well suitedto measurethe deviationfromsphereof a planoconvexmicrolens.However, it is moredifficult to getanaccurateresultfor planoconcavesphericalprofiles. In this work, it was not possibleto measurethedeviationfrom sphereof the planoconcavereplicatedele-ment. In the future, modificationsin the setupand in theinterferometer software must be realized to adapt themethodalsoto planoconcavesphericalshapes.

3.3 Multilevel Blazed Gratings

Multilevel blazedgratingsservedasmastersfor replicationof diffractive micro-optical elements.The original DOEwasrealizedin different sequencesusinga seriesof litho-graphicandetchingsteps.10 The obtainedstructureis a so-called multilevel blazedgrating. Replicationusing elasto-mericmaterialandUV-curing adhesiveswasperformedonthis type of element.We replicatedmultilevel blazedgrat-ings with periods from 64 to 8 mm having eight phaselevelswith goodresults.Figure4 showsscanningelectronmicroscopy~SEM! picturesof anoriginal masteranda rep-licated multilevel blazed grating ~64 mm period blazedgratinghavingeight levels!.

The diffraction efficiencies of the different elementswereanalyzed.The diffraction efficiency is definedas theratio betweenthepowerin thefirst diffraction orderandthepowerof theincidentbeam.Theelementswereilluminatedwith a He-Nelaserandthe intensity in the first diffractionorder was measured.For the original blazedgrating ~64-mm periodblazedgratinghavingeightlevels!, an efficiencyof 84.7% was obtained.The efficiency of the replicated

Page 4: Simple technique for replication of micro-optical elements

4

Fig. 4 SEM pictures of (a) the original master and (b) the replicated element. The period of theeight-level blazed grating is 64 mm.

multilevel blazed grating at the same wavelength was77.9%.The importantdifferencein thediffraction efficien-ciesof the original andthe replicatedelementsis dueto adifferentphasedelayfor the two elements.In this case,theoriginal quartz element was designedfor He-Ne wave-length of 633 nm with a refractive index of 1.457. Thereplicatedelementhasthe samegrating depthbut anotherphasedelaybecauseof the refractiveindexof 1.557at 633nm. The appropriatewavelengthfor the phasedelayof thereplicatedelementis around790nm.A tunableTi:Sapphirelaserat a wavelengthof 790 nm was usedto measurethediffraction efficiency.An efficiency in the first diffractionorderof 84.8%wasobtained.

4 Conclusions

We demonstratedthe fabricationof good-qualityreplicatedmicro-optical elements using elastomeric molds. Thissimple techniqueis very promising for small replicationseries.Goodresultsfor bothrefractivemicrolensarraysandDOEswereobtained.No significantdifferencebetweentheoriginal mastersand the replicatedelementswas noticed.The fact that theoriginal masteris still intactafter thepro-cessis anadvantageif severalmoldsmustberealizedfromthe sameoriginal element.We alsodemonstratedthat thistechniqueis suitablefor the realizationof invertedcopies.Planoconcavemicrolenseswereproducedfrom a planocon-vex original mastermicrolens.

Standardmicro-opticalelementsarealwaysrealizedona hardcarrier substratehavinga thicknessof severalhun-dredmicrometers.In somecases,suchasthe integrationofmicro-opticalelementsin microsystems,thesebulky sub-stratesare not acceptable.Using elastomericmolds, it ispossibleto use them as stampsto createisolatedmicro-opticalelements,which canbeintegrateddirectly in micro-systemswithout bulky carriersubstrates.

Acknowlegments

The work was supportedby the Swiss Priority ProgramOPTIQUE.The authorslike to thank R. Voelkel, C. Oss-man,A. Schilling of IMT-Uni Neuchatel, andK. J. Weibleof Weible Optechfor their help.

References1. H. P.Herzig,Ed.,Micro-Optics:Elements,Systems,andApplications,

Taylor & Francis,London ~1997!.2. M. T. Gale,‘‘Replication,’’ in Micro-Optics: Elements,Systems,and

Applications, H. P. Herzig, Ed., pp. 87–126, Taylor & Francis,Lon-don ~1997!.

3. J. L. Wilbur, R. J. Jackman,G. M. Whitesides,E. L. Cheung,L. K.Lee, and M. G. Prentiss,‘‘Elastomeric optics,’’ Chem. Mater. 8,1380–1385 ~1996!.

4. Y. Xia, E. Kim, X.-M. Zhao, J. A. Rogers,M. Prentiss,and G. M.Whitesides,‘‘Complex optical surfacesformed by replica moldingagainstelastomericmasters,’’Science273, 347–349 ~1996!.

5. D. J.Daly, R. A. Ferguson,andM. C. Hutley, ‘‘Replication of opticalcomponentsusing silicone rubberintermediatereplica,’’ Proc. SPIE3099, 83–88 ~1997!.

6. Z. D. Popovic,R. A. Sprague,andG. A. Neville-Connell,‘‘Techniquefor the monolithic fabrication of microlensarrays,’’ Appl. Opt. 27,1281–1284 ~1988!.

7. D. Daly, R. F. Stevens,M. C. Hutley, andN. Davies,‘‘The manufac-tureof microlensesby meltingphotoresist,’’J. Meas.Sci.Technol.1,759–766 ~1990!.

8. P. Nussbaum,R. Voelkel, H. P. Herzig,M. Eisner,andS. Haselbeck,‘‘Design, fabricationand testingof microlensarraysfor sensorsandmicrosystems,’’Pure Appl. Opt. 6, 1–20 ~1997!.

9. J. SchwiderandO. Falkensto¨rfer, ‘‘Twyman-Greeninterferometerfortestingmicrospheres,’’Opt. Eng.34, 2972–2975 ~1995!.

10. M. B. Stern,‘‘Binary opticsfabrication,’’ in Micro-Optics:Elements,Systems,and Applications, H. P. Herzig, Ed., pp. 53–86, Taylor &Francis,London ~1997!.

Philippe Nussbaum studied optics inFrance, where he received his diploma in1990. In 1992 he was in the Analytical Re-search Department of the Ciba-GeigyCompany in Basel, Switzerland, workingon the development of miniaturized opto-electronic sensors for chemical analysis.Since 1993, he has been with the AppliedOptics Group at the Institute of Microtech-nology of the University of Neuchatel,Switzerland. He is in charge of the refrac-

tive and diffractive micro-optical technology. Philippe Nussbaum ismember of the Swiss Society for Optics and Microscopy.

Page 5: Simple technique for replication of micro-optical elements

5

Irene Philipoussis did her apprentice-ship as a physics technician at the Euro-pean Organization for Nuclear Research(CERN) laboratories in Geneva, Switzer-land. From 1986 to 1990 she was with Os-cilloquartz in Neuchatel, Switzerland,working on atomic frequency standards.From 1990 to 1996, she was with the Liq-uid Crystal Display Group of the AsulabSMH Central Laboratories in Neuchatel,Switzerland. In 1997, she joined the Ap-

plied Optics Group at the Institute of Microtechnology of the Univer-sity of Neuchatel, Switzerland. She is in charge of the fabrication ofmicro-optical elements.

Aline Husser is a student at the Lyceed’Enseignment General et Technique(LEGT) in Saint-Louis, France, where sheis preparing her diploma in optics. In 1997she was with the Applied Optics Group atthe Institute of Microtechnology of the Uni-versity of Neuchatel, Switzerland, studyingthe performance of elastomeric materialsfor replication techniques.

Hans Peter Herzig received the diplomain physics from the Swiss Federal Instituteof Technology in Zurich, Switzerland, in1978. From 1978 to 1982 he was a scien-tist with the Optics Development Depart-ment of the Kern Company in Aarau, Swit-zerland, working on lens design andoptical testing. In 1983, he joined the Ap-plied Optics Group at the Institute of Mi-crotechnology of the University of Neucha-

tel, Switzerland, as a graduate research assistant, working in thefield of holographic optical elements, especially scanning elements.In 1987, he received his PhD degree in optics. He currently headsthe Micro-Optics Research Group and Privat-docent at the Univer-sity of Neuchatel. Dr. Herzig is member of OSA and EOS and aboard member of the Swiss Society for Optics and Microscopy.