CA2265617A1 - Reactive ion etching of silica structures - Google Patents

Abstract

The invention relates to a method for etching of silica-based layers/substrates by reactive ion etching system (10) using an etching gas mixture of CHF3/AR through a photoresist mask. Reactive ion etching is carried out under conditions of simultaneous isotropic deposition of a carbon-based polymer where the polymer deposition rate is controlled by adjusting process control parameters of RF power, sample temperature, O2 and CF4 additions.

Description

W0 98/ 15504101520253035CA 02265617 1999-03-08PCT/AU97/006631Reactive Ion Etching of Silica StructuresField of the InventionThe present invention relates to creation of silicastructures and in particular to the reactive ion etching ofsuch structures.Background of the InventionSilica—based channel waveguides, fabricated on siliconor silica wafer substrates, are potential building blocks ofplanar lightwave circuits (PLCs) that are becomingincreasingly important for telecommunications systems.While there are a number of thin film techniques that havebeen used to deposit silica waveguiding layers (flamehydrolysis, chemical vapour deposition and plasma enhancedchemical vapour deposition (PECVD)), almost all reportedwaveguide fabrication schemes have used reactive ion etching(RIE) to delineate the channel waveguide (core) geometry.RIE is also commonly used in integrated optics, and inplanar waveguide fabrication in particular, for etchinglight turning mirrors. It further finds use in the creationof other Micro Electro Mechanical Systems (MEMS).RIE of silica glass in integrated circuit (IC)manufacture is a well established and routine process with,for example, CHE} based mixtures being used to obtain highselectivity over photoresist. Although basically similar,the silica films used in planar waveguides have severalunique differences which influence the development ofsuitable RIE processes. Firstly, the thicknesses of silicain waveguide devices can be as much as 5 to 10 um, asopposed to typically less than 1 um in IC technology. Thisplaces extra demands on mask thickness and/or materialselectivities, as well as on the silica etch rate whichshould be high enough to obtain reasonable throughput.Different masking materials for waveguide etching such asphotoresist, amorphous silicon (a—Si) and chromium have beenreported. Generally, the use of non—photoresist masksallows for larger etching depths and silica etch rates.SUBSTITUTE SHEET (RULE 26)101520253035WO 98115504CA 02265617 1999-03-08PCT/AU97/006632The roughness of the etched walls of the waveguidestructures or light turning mirrors should ideally be assmall as possible in order to reduce the loss due to lightscattering. A number of works on sidewall roughnessreduction for etching with photoresist masks have beenreported. In these cases, however, the etching depth of aEtched profilecontrol is also important and some slope in the etchedSiog layer was restricted to around l um.profile is sometimes desirable in order to facilitatefilling of the gaps between closely spaced waveguides duringcladding deposition. Profile slope is normally achieved bycontrolled photoresist mask erosion. Despite a number ofpublished works on different aspects of the RIE of silicafor waveguide fabrication, it is unclear as to the effect ofall relevant parameters, such as etch rates, sidewallroughness, profile slope and the relationship between them.Summary of the InventionIt is an object of the present invention to provide forthe development of a high—rate silica RIE process suitablefor low temperature waveguide fabrication.In accordance with the first aspect of the presentinvention there is provided a method for etching of silica-based glass layers or substrates comprising reactive ionetching through a mask executed under conditions ofsimultaneous isotropic deposition of a carbon based polymer.Preferably, the polymer deposition rate or/and itssteady-state thickness on different surfaces of the etchedstructure is controlled by adjusting one or several processcontrol parameters in order to control etched profile,dimension loss, sidewall and bottom etched surfaceroughness, and etching selectivity between the silica—basedlayer and mask material.A gas or a mixture of gases containing fluorine orcarbon atoms is used and a photoresist mask or other form ofmask such as amorphous silicon. Adjustable parameters caninclude RF power and substrate temperature. The temperatureSUBSTITUTE SHEET (Rule 26)101520253035W0 98/ 15504CA 02265617 1999-03-08PCT/AU97/006633can be adjusted to achieve low sidewall roughness and lowdimension loss at the same time. Further, resputtering ofany metal present wihtin or/and in contact with thedischarge zone can be prevented.The invention is ideally suited wherein reactive ionetching is performed in a high plasma density hollow cathodeetching system and the etching gas mixture is CH3F andArgon.Various products made utilizing the previous methodsare also disclosed.In accordance with a further aspect of the presentinvention there is provided a high plasma density hollowcathode etching system is disclosed which has been shown toprovide higher etch rates than those achievable inpreviously known standard RIE systems. Etching was carriedout in a CHF5[Ar mixture with additions of O2 and CF4. Theeffects of the different chemistries as well as the use ofdifferent masks (photoresist and amorphous silicon) and theeffects the substrate temperature on etching rates, sidewallroughness and etch profiles have been investigated. Using aphotoresist mask generally results in greater sidewallroughness compared to an amorphous silicon mask.Importantly, polymer deposition during the etching processcan exacerbate the development of roughness but is stilldesirable to a certain extent in the prevention of the lossof line width during etching. Two mechanisms of polymerthe addition ofvarying amounts of 02 or CF}, and elevating the temperatureof the substrate.deposition control are disclosed, namely,The latter was found to give a goodcompromise between control over the line width loss and thesidewall roughness. In order to explain the variety ofexperimental results obtained, a simple phenomenologicalmodel based on a polymer etching/deposition rate equilibriumon etched surfaces is proposed and examined.Brief Description of the DrawingsNotwithstanding any other forms which may fall withinSUBSIHIHISSHEETGRMQZS)101520253035W0 98/ 15504CA 02265617 1999-03-08PCT/AU97/006634the scope of the present invention, preferred forms of theinvention will now be described, by way of example only,with reference to the accompanying drawings in which:Fig. 1 is a schematic illustration of the basic layoutof hollow cathode discharge chamber used in the preferredembodiment.Fig. 2a to Fig. 2g illustrate graphs of etch rates andselectivities over mask material for etching with a—Si (Fig.2a) and photoresist (Fig. 2g) masks as a function of RFpower. Pressure is 12 Pa. Gas flow rates: 60 sccm of Ar,15 sccm of CHF3. Sample temperature is 80°C.Fig. 3a to Fig. 3d illustrate graphs of etching profileslope (Fig. 3a) 3b),(Fig. 3C) (Fig. 3d)dimension loss (Fig. sidewall roughnessand polymer deposition rate as afunction of RF power for etching with a—Si and photoresistThe pressure is 12 Pa. 60 sccm15 sccm of CHFymasks.of Ar,The gas flow rates:The sample temperature is 80°C.The dimension loss was normalized to an etching depth of 5um. The polymer deposition rate was measured in the areashielded from ion bombardment.Fig. 4a to Fig. 4h are Electron microscope images ofetching profiles as a function of RF power for etching witha—Si (Fig. 4a to Fig. 4d)Fig. 4h). The RF power was as follows: Fig. 4a— unetcheda-Si mask, Fig. 4b— 250W, Fig. 4c-500W, Fig. 4d—65OW, Fig.4e—unetched photoresist mask, Fig. 4f—3OOW, Fig. 4g—4OOW,Fig. 4h 500W.of Ar, 15 sccm of CHF3.Fig. 5a to Fig. 5e illustrate graphs of etch rates andand photoresist masks (Fig. 4e toPressure is 12 Pa. Gas flow rates: 60 sccmSample temperature 80°C.selectivities with Fig. 5a to 5c illustrating S102 etchrates and selectivities over an a—Si mask and Fig. 5d toFig. 5e illustrating vertical and lateral etch rates of a—Simask as a function of sample temperature, for a 02 flow rateand a CF} flow rate, respectively. RF power is 500W.Pressure is 12 Pa. Gas flow rates: 60 sccm of Ar, 15 sccmof CHF3. Sample temperature 80°C unless varied.SUBSTITUTE SHEET (RULE 26)101520253035WO 98/15504CA 02265617 1999-03-08PCT/AU97/006635Fig. 6a to Fig. 6j illustrate graphs of the etching6c)and polymer deposition rateprofile slope6d to Fig. 6f)6j) as a function of sample temperature, 02 flow rate and CF4(Fig. 6a to Fig. sidewall roughness (Fig.(Fig. 6g to Fig.flow rate,500W.respectively using an a—Si mask. RF power isPressure is 12 Pa. Gas flow rates: 60 sccm of Ar, 15sccm of CHF3. Sample temperature 80°C unless varied.Polymer deposition rate was measured in the area shieldedfrom ion bombardment.Fig. 7a and Fig. 7b illustrate electron microscopeimages for etched sidewalls for sample temperature 80° (Fig.7a) and 320°C (Fig. 7b).RF power is 500W. Pressure is 12 Pa.15 sccm of CHFLThe a—Si mask is still in place.Gas flow rates: 60sccm of Ar,Fig. 8 illustrates a sectional view of a wafer showingthe four possible etched surfaces with respect to intensityof ion bombardment and the possibility of polymer filmformation: (i)mask 31, the sidewalls of the mask 32,sidewalls of Sioz 33 and,the surfaces include the top surface of thethethe bottom surface of the Siog(ii) (iii)(iv)34. A steady state thickness of polymer film can be presenton (i-iii), whereas (iv) is assumed polymer free under theetching conditions used in this study.Fig. 9a to Fig. 9c are electron microscope images oftime evolution of the etched profile: with Fig. 9a showingan unetched a—Si mask, Fig. 9b after 3 minutes etching, Fig.RF power is 500W. Pressure is15 sccm of CHFy9c after 6 minutes etching.12 Pa. Gas flow rates: 60 sccm of Ar,Sample temperature is 80°C. Etching selectivity over the a~Si mask is approximately 14:1. A “negative undercut” isshown to be developed without mask width reduction. Themore vertical profile of the a—Si mask is “buried” underpolymer formed at a steady state angle determined by thepolymer etching and deposition equilibrium.Fig. 10 is a schematic illustration of a mechanism ofsloped profile formation. A steady state profile angle 0°SUBSTITUTE SHEET (RULE 26)W0 98/15504101520253035CA 02265617 1999-03-08PCT/AU97/006636is formed under conditions where ErmflWBr= For thisDrpolymer -angle the steady state thickness of polymer film on thesidewall is enough to prevent etching. The no—zero polymeretch rate at O = 90° is due to ions scattered in plasmasheath.Fig. 11a and 11b illustrate etched sidewalls foretching with a photoresist mask lla) and a a-Si mask(Fig. 11g).Ar, 15 sccm of CHFy(Fig.Pressure is 12 Pa. Gas flow rates: 60 sccm ofSample temperature is 80°C. RF poweris 500W for etching with the photoresist mask and 650W foretching with the a-Si mask.Description of Preferred EmbodimentsA first embodiment of the present invention relies uponthe utilisation of plasma enhanced chemical vapourdeposition (PECVD) in a hollow cathode discharge chamber.Turning initially to Fig. 1 there is shown a suitable vacuumchamber assembly 10 including a top electrode 11 and abottom electrode 12 connected as shown to RF source 13 whichcomprised a 13.56 MHz RF source. In use for the.purposes ofPECVD, the chamber 14 is evacuated via pump port 15 andgases such as CH4/SF6 mixtures, CHE}/Ar mixtures areintroduced Via corresponding ports e.g. 16, 17, so as tocause controlled etching on wafers or substrates 19 locatedin the RF field induced plasma located between electrodes11, 12. This apparatus 10 is utilised to perform thecontrolled ion etching operation as discussed in detailbelow.The high plasma—density hollow cathode dischargeetching system suitable for use is described in C.M.J. Vac. Sci.The twoHorwitz, S. Boronkay, M. Gross and K.E. Davies,Technology A6, at pages 1837 to 1844 (1988).opposing RF powered parallel circular electrodes 11, 12 aresurrounded by a grounded chamber 21. A conventional diodedischarge is produced between each of the electrodes 11, 12and the grounded chamber 21 but a high density plasma isgenerated between the two RF powered electrodes 11, 12 dueSUBSTITUTE SHEET (RULE 26)1015.20253035WO 98/15504CA 02265617 1999-03-08PCTlAU97I006637to the “electron mirror" effect. Both the upper and lowerelectrodes were water—cooled and covered with 100 mmdiameter silicon wafers 18, 19. The latter is to preventresputtering of the electrode material (A1) which can resultin metal contamination and subsequent surface roughness andthe formation of sloped etching profiles due to metal-basedpolymer deposition. Examination of the polymer deposited inthe ion-shielded areas (as described below) using wavelengthdispersive X—ray spectroscopy (WDS) showed that no traces ofA1 or other vacuum chamber materials could be detected atthe 0.01 % level, thus confirming that metal contaminationis not an issue. lThe silica films used in the etching experiments had athickness of 8 um and were deposited on silicon substrates19 using the hollow-cathode PECVD technique. Masking layersof 1 um of PECVD a—Si or 2 um of photoresist were thenapplied to the wafer 19. The photoresist mask was patternedusing conventional photolithography, while patterning of thea—Si layer was carried out using conventionalphotolithography followed by etching in a CE}/SFg mixture.In each run the samples 19 were etched to a depth ofaround 4-5 um, with the rates determined by surfaceprofilometry. The etched profile, sidewall roughness and- dimension loss were further determined by SEM examination.The dimension loss was calculated or defined as thedifference between the line width measured at the bottom ofthe mask before etching and the width at the top of theetched ridge. When determining the sidewall roughness, theSEM photos were taken from the top side of the ridge, at aglancing angle to the sidewall. Sidewall roughness figuresset out hereinafter are the average amplitudes of the RIE~induced corrugations measured over a distance of a fewmicrons. The initial (unetched) roughness of both thephotoresist and a—Si mask edge was not larger than 0.02 um.The silica films were etched in a 20 % CHF;_in Armixture with various additions of CF} or 02. The pressure wasSUBSTITUTE SHEET (RULE 25)101520253035W0 98/ 15504CA 02265617 1999-03-08PCTIAU97/006638kept at 12 Pa in all experiments. All gases had a statedpurity of 99.95 % or better. Etching in CHE;wasaccompanied by some polymer deposition. The polymerdeposition was estimated using a shadowing technique,whereby the polymer deposition is assumed to be isotropic.An overhanging structure consisting of two overlappingsilicon wafers was used and the thickness of polymerdeposited under the overhang (and so shielded from ionbombardment) was measured using surface profilometry.The temperature of the samples was controlled byvarying the thermal contact between the sample 19 and thecooled electrodes eg. 12. For an Arsi mask, three caseswere characterized by thermocouple measurements: (i) nothermal contact between the sample and the electrode; (ii)partial thermal contact through several point contacts ofVacuum grease; (iii) good thermal contact through vacuumGood thermalcontact was used in all experiments where the temperaturegrease spread on the back of the sample.was held constant and for all photoresist masked samples.ResultsThe Effect of RF PowerEtch rates: The resulting etch rates as a function ofRF power (at 13.56 MHZ) coupled into the discharge are shownin Fig. 2a for an a—Si mask and in Fig. 2b for a photoresistmask. It is seen that the SiO2 etch rate is slightly higher(around 10 %) in the case of the a-Si mask for similar powerlevels. With the a—Si mask the SiO2 etch rate increases byalmost a factor of three over the investigated power rangeThe a-Si etch rate increases with power faster than the SiO2 etchreaching a value of 0.8 um/min at the maximum power.rate thus causing an overall decrease in selectivity over a-Si from 20:1 to 12:1. Similarly, the selectivity overphotoresist also decreases with the power.Etched profile and dimension loss: The etched sidewallslope angle, as a function of power for a-Si and photoresistmasks, is shown in Fig. 3a. The SEM photographs of theSUBSTITUTE SHEET (RULE 26)101520253035W0 98/ 15504CA 02265617 1999-03-08PCTIAU97/006639corresponding etched profiles are shown in Figs. 4a to 4h.For both mask materials the angle of the profile slope wasfound to increase with the power, being greater for an a-Simask for similar power levels. The dimension loss (Fig. 3b)(normalised to a depth of 5 pm) for the a-Si mask was foundnot to exceed 0.2 um (although it seemed to increaseslightly with power), at the same time the dimension lossfor the photoresist mask was quite significant (>1 um) andclearly increased with power as illustrated in Fig. 3b.This difference probably points to different mechanismsbeing responsible for the sloped profile formation in bothcases. It may be noted from Figs. 4a to 4h, where theinitial mask profiles before etching are shown, that a facetis developing on the photoresist mask sidewall, thuspossibly contributing to the observed dimension loss.Sidewall roughness: As illustrated in Fig. 3c, thesidewall roughness appears to be consistently higher for aphotoresist mask than for an a—Si mask. In both cases,however, it was found to increase with power and, as can beseen in Fig. 3c, the sidewall roughness for etching with ana-Si mask at the highest power level is comparable with thesidewall roughness obtained with the photoresist mask atlower power levels.Polymer deposition rate: The polymer deposition ratein the area shadowed from ion bombardment was found to giveresults as indicated in Fig. 3d. It was found to increaseby about 30 % over the whole power range. Also, it may benoted that at the minimum power, the polymer deposition ratewas around 3 times smaller than the Siog etch rate, whichmeans that in ion bombarded areas, during etching of l um ofSiO;, around 0.35 um of polymer is simultaneously removed.As the power increases this portion of removed polymer isreduced to around 20 % or 0.2 um for 1 um of SiobThe Effect of O2 and CF} Additions, and Sample TemperatureVariationAs polymer deposition was found to play an importantSUBSTITUTE SHEET (RULE 26)101520253035W0 98/15504CA 02265617 l9’99-03-08PCTIAU97/0066310role in the etching mechanism, different methods ofcontrolling it were investigated. These include (i) 02additions, (ii) CF} additions and, (iii) elevated substratetemperature.Etch rates: Figs. 5a to 5f show SiO2 and a-Si etchrates and selectivity plotted on the same scales for allIt is seen that the Siog etch ratedecreases with temperature and O2 additions, but increaseswith CF} additions.three varied parameters.At the same time the selectivity overa-Si decreases in all three cases. The a-Si mask etch rateshown in Figs. 5a to 5f has been separated into twocomponents, a vertical component, which is related to themask thickness decrease, and a lateral component, which isrelated to the mask width decrease. It can be seen fromthat under the conditions of no 02 or CF4 and a low sampletemperature (80°C), the lateral etch rate is essentiallyzero (<8OA/min). At elevated temperature and with 02 addedit stays initially at around zero but then rises, ultimatelyapproaching the vertical etch rate values, which impliesisotropic etching of the-a-Si mask. For CF} the behaviouris different, the lateral etch rate increases gradually butWithincreases in all three parameters the vertical a-Si etchthe mask etching remains basically anisotropic.rate increases which, together with a decrease in the SiO;etch rate at higher temperatures and O3 additions, and aslower rate of increase with CF} additions, results in anoverall decrease in selectivity.Etched profile: As illustrated in Fig. 6a, the slopeof the etching profile was found to first increase withtemperature and then decrease below the initial value. Asshown in Fig. 6b, 0; additions caused a small initialincrease in the slope followed by a gradual decrease. Asillustrated in Fig. 6c, the slope was found to beessentially independent of the CF4 flow rate.sidewall roughness. As shown in the sidewall roughnesswas found to decrease with both temperature (Fig- 6d) and 0;SUBSTITUTE SHEET (RULE 26)101520253035W0 98/ 15504CA 02265617 1999-03-08PCT/AU97/00663llflow (Fig. 6e) but was not effected by the CF4 6f).It is seen that the sidewall roughness can be reduced to0.02 um,adding 02 to the gas mixture.with Fig. 5a and Fig. 5d,(Fig.either by elevating the sample temperature or byHowever, by comparing Fig. 6dit can be seen that, usingtemperature as a control parameter, minimum sidewallroughness can be achieved while maintaining the anisotropyof the a—Si mask etching. 02 can also be used to reduceroughness (Fig. 6e), but the same minimum roughness can onlybe achieved at the expense of dimension loss, since, at therequired 0; flow rates, the etching of a-Si becomesessentially isotropic (Fig. 5e). From a practical point ofview this suggests that the sample temperature is a moreuseful control parameter for reducing sidewall roughnesscompared to the addition of 02. The improvement in thesidewall roughness can be seen in Figs. 7a and 7b, whichshows SEM images of two sidewalls etched at differenttemperatures.Polymer deposition rate: The polymer deposition rateon a shadowed surface as a function of sample temperature,is shown in Fig. 6g, Fig. 6hthat by02 flow rate and CF1 flow rate,and Fig. 6j, respectively. It is seen (Fig. 6g)increasing the temperature, the polymer deposition is firstreduced and then, with a further increase in thetemperature, is totally suppressed, which means that thereis no polymer deposition above a certain sample temperature(Fig. 6h)cause only a small decrease in the polymer deposition rate,even in absence of ion bombardment. O2 additionswhich is noteworthy given the similar effect of temperatureand O; on the sidewall roughness. This may indicate adifference in the mechanisms by which the two parametersreduce the sidewall roughness. Finally it is seen from Fig.6j that the polymer deposition rate increases with Cqflow byabout 25 % of its initial value.It is evident from the aforegoing analysis thatessentially isotropic polymer deposition occursSUBSTITUTE SHEET (RULE 26)l0l520253035WO 98/15504CA 02265617 1999-03-08PCTIAU97/0066312simultaneously with etching. Furthermore, the formation ofpolymer films under similar etching conditions, withthicknesses depending on an equilibrium between the polymeretch and deposition rates, has been demonstrated in a numberof other investigations.Turning to Fig. 8, in the case of the etched structuresdescribed it is possible to specify four surfaces 31 - 34 onwhich such a film may exist. Previous investigations usingsimilar conditions have shown that Sioz surfaces 34 are freeof polymer film for RF bias voltages above 100V (at apressure of 0.13 Pa), and that the threshold bias voltagebetween polymer etching and deposition decreases withTherefore, using 12 Pa and 400V — 600V bias, aThispressure.polymer free Siog bottom surface 34 generally results.is supported by the fact that Siog etch rates do notincrease with polymer suppression, either by O; additions,or increasing sample.(Fig. 5a and 5b). Thus, a polymer film can32 and 33.Temperaturebe present only on surfaces 31, The polymer onasurface 3;and 33 willdetermines the etching selectivity, whereas 32effect the etching profile and sidewallroughness. The presence of a finite thickness of polymerimplies that both etching species and reaction products mustdiffuseetchedthrough the polymer on their way to or from thesurface, a mechanism which has previously beensuggested by others. Here, in addition to etching of thepolymer film by normal surface process, it is assumed thatetching species diffusing through the polymer film have acertain probability of reaction with the polymer, which isproportional to the film thickness. Porosity in the polymerfilm can contribute to thispolymercomponent also increases,etching mechanisnm As thefilm thickness increases, this “diffusion” etchingthus increasing the total polymerremoval rate and preventing continuos film growth. For aconstant polymer deposition rate, these effects will giverise to a certain equilibrium polymer thickness, which willSUBSTITUTE SHEET (RULE 26)1015202530W0 98/ 15504CA 02265617 1999-03-08PCTIAU97/0066313determine the etch rates of the underlying surfaces.While not wishing to be bound by theory, a simplifiedphenomenological model describing this mechanism can bewritten as follows:ER ocla (1—-ad) s1 (EQ.l)ER,.,,,., cc c.1,,ad +c.1,Cos(¢) Y(1,,,E ,,¢),ads1 E0-2DR,,,,,,,,,,,sc;/ (T)1,, (120.3)where ER is the etch rate of the surface under the polymerfilm, 15 is the flux of active etching species at thepolymer film surface, a is the probability of polymeretching by diffusing active species per unit of filmthickness, d is the thickness of the polymer film, ERWfiw@r isthe polymer etch rate, C; and C2 are empirical constants, I;is the ion flux, ¢ is the sidewall slope or effective ionangle of incidence, }<]a,[;,¢) is the reactive sputteringyield as a function of active species flux Id, ion energy E;and effective ion angle of incidence ¢. I9 is the flux ofpolymer forming species and 7”‘ is the sticking probabilityof the polymer forming species as a function of the surfaceand SiO; etchresults of Figs. 2 and 3 can be explained using this model.temperature. The photoresist, a—Si rateThe observed decrease in etching selectivity over bothphotoresist and a~Si with power of Fig. 2a and Fig. 2b canbe explained by a decrease in the steady—statepolymerthickness on both mask surfaces 31, 32 (the SiO2 surface isassumed polymer free). This occurs in spite of theincreasing polymer deposition rate with. power (Fig. 3d).According to Eq. (2) above this means that the increase inpolymer deposition rate is overshadowed by the increase inbothcomponents,reactive sputter etching and “diffusion” etchingthus requiring a smaller polymer thickness toSUBSTITUTE SHEET (RULE 25)101520253035W0 98/15504CA 02265617 1999-03-08PCTIAU97/0066314maintain the polymer etching/deposition equilibrium.Increasing temperature, 0; flow and CF4 flow all reducethe etching selectivity over a—Si,(Fig. 3).surface 34 is assumed to be polymer free,mainly through a greatervertical a—Si etch rate Again, since the SiO;this implies adecrease in the thickness on a—Si.polymertemperature reduces the sticking probability of the polymerIncreasingforming species with causes a reduction in the polymerdeposition rate according to Eq. 3. This is also confirmedby Fig. 6g where the polymer deposition rate in the areashielded from ion bombardment is shown to decrease withtemperature. The polymer thickness then decreases (Eq. 2)causing and increase in a—Si etch rate through Eg. I. In6h).Similarly, CF4 additions lead to a reduction in the polymerthe case of O; additions is not significant (Fig.thickness despite a small increase in the polymer deposition6j). the a—Si etch ratecan increase,rate (Fig. In this case, however,not only because of a reduction in. polymerthickness, but also due to an increase in the active speciesflux, resulting from CF;dissociation.While the vertical a-Si etch rate increases in responseto all three factors (temperature, 0; and CF}), the laterala—Si etch rate behaves differently for the temperature and(Fig. 5d toand O3 flow the0: cases on one side and CF; case on the otherFig. 5f). For increasing temperaturelateral etch rate increases, approaching the vertical etchrate, thus indicating isotropic etching of time a—Si mask.In the case of CF; additions, however, the lateral a-Si etchrate increase is small and the anisotropy remains unchangeddue to a proportional increase the vertical etch rate.According to Eq. 1, the difference in vertical andlateral etch rate of the a—Si mask is due to the differentsteady—state polymer film thickness on its top surface andsidewalls. The less ion ‘bombardmentsidewalls receiveduring etching which, according to Eq. 2, reduces thereactive sputtering component of polymer etching and causesSUBSTITUTE SHEET (RULE 25)l01520253035W0 98/ 15504CA 02265617 1999-03-08PCTIAU97/0066315an increase in its steady—state thickness to the point wherelateral etching of the mask ceases, as seen in the first few5d to Fig. 5f.rate with temperature and O2 additions is due to a reductionpoints in Fig. The increasing lateral etchin polymer thickness on the sidewalls. In the temperaturecase, this can be attributed to reduced polymer depositionFig. 6g. In the 02 case,datawhere the polymer deposition rateFig. 6g, thereduction in sidewall polymer thicknesspolymerpolymer on the sidewalls for high 02 flow rates,show only a small decrease similaris due to higherremoval rate by active oxygen. The absence ofas opposedto its presence in the areas shielded from ion bombardmentFig. 6j is likely to be due to some ion bombardment on thesheath10 Pamask sidewalls by ions scattered in the(characteristic for the operating pressure aroundalthoughinitiate polymer etching by the products of O;employed which is, relatively small, apparentlyenough todissociation).The addition of CF“ although causing an increase inthe vertical a-Si etch rate in a way similar to increasedtemperature and 0;, does not alter the ratio of the verticalto lateral etch rates. According to the model this impliesthat some polymer film remains on both the sidewalls and topsurface of the a—Si mask, with its steady state thickness onboth surfaces reduced proportionally with the CE} additions.Finally,we note that, despite there being no polymeron the a-Si mask surface at high temperature or high 0;flows, its vertical etch rate, although increased, stillremains around 5 ‘times smaller than the Siog rate. Thissuggests that the well accepted mechanism of SiO2 etching,where the SiO2/Si selectively is due to selective polymerremoval by oxygen released from the SiO2 during etching, isThatbombardment, theproduction of CHF3 dissociation apparently etch SiO2 fasterthan Si,accompanied by an additional mechanism in this case.is, under conditions of intense ioneven when there is no protective polymer film onSUBSTITUTE SHEET (RULE 26)101520253035CA 02265617 l9’99-03-08WO 98/15504 PCTIAU97l0066316the Si surface.sidewall angleA sloped profile in silica, a material with knownintrinsically anisotropic etching characteristics may beproduced in two ways. The first mechanism, mask erosion,produces a sloped profilewelldeliberately induced by,through lateral mask etching.This is documented in the literature and may befor example, addition. of 02 whenusing a photoresist mask. The considerable dimension loss(Fig. 3b)that mask erosion is the cause of the sloped profile in thisobserved when using a photoresist mask suggestscase. The reason for high lateral photoresist mask etching,in the light of similar selectively to S102 as a—Si, isprobably the result of faceting of the photoresist maskedge. 4h where theThis can be seen in Fig. 4a to Fig.photoresist mask is shown before and after etching. From(Fig. 4e)(Fig. 4h) at theAn estimate of the lateral etching ratearound 1000A/min, which is more than twice the vertical etch rate. Thewithattributed to thethe round shape before etching the mask sidewallbecomes flat and sloped preferentialsputtering angle.of the photoresist mask due to faceting isincrease in slope angle power observed for aphotoresist mask can be preferentialincrease in the SiO; sidewall etch rate compared to theincrease in the lateral mask etch. rate. Here, the sio,sidewall. is not protected tux polymer, since the angle isless than the steady-state value required for polymer filmformation.The effect of the second mechanism of sloped profileformation can be seen in Fig. 9a to Fig. 9c where the timeevolution of an etched silica profile is shown starting with9a).the width of the mask has not changed, but the effectivelinewidth hasan unetched a—Si mask (Fig. Here, it can be seen thatrather than decreased. as is theThis effect,undercut” orincreased,case of mask erosion. which has been termed“negative mask “overcut” can occur underSUBSTITUTE SHEET (RULE 26)101520253035W0 98/15504CA 02265617 1999-03-08PCT/AU97/0066317conditions of anisotropic etching of the substrate and maskin the presence of a simultaneous isotropic depositionprocess.The Siog sidewall etch zero when all theetching species are consumed in the sidewall polymer film(Eq. 1).thickness required to satisfy thisrate isbefore reaching the SiO2 surface, or when dd = 1The steady—state polymercondition is found whenboth the ion flux to theHowever, sinceERpolymer = DrpoJ.ymer-sidewall and the angular dependenceof the reactive sputtering yield depend on the sidewallslope, the etching rate of the sidewall polymer will dependon the slope (¢ in Eq. 2). Thus, assuming isotropic polymerdeposition,(¢J =the angle were greater than this equilibrium value then thethere is a sidewall angle ¢m at which ERmnW,DREQMM, and hence zero SiO2 sidewall etch rate. Ifreactive component of polymer etching would be lower, DRWHWH> ERmgmm/ and net deposition would occur. Conversely, ifthe angle were less than the equilibrium value, then thereactive component of polymer etching would be higher,This mechanismtheDRmuWE,< ERmuww, and net etching would occur.10.isotropicis shown graphically Fig. The point 41polymercrosses the line 42 representing the angular dependence ofwheresemicircle 40 representing depositionthe polymer etch rate corresponds to the angle at which thepolymer etching/deposition balance is achieved; This anglesidewall 44. Theangular dependence of the polymer etch rate in Fig. 10 isis the steady—state angle of the Siogdrawn schematically and includes both the angular dependenceof the ion flux on the sidewall. IiCos(¢) and. the angulardependence of the sputtering yield, Y(ImE;, ¢)As the polymer deposition ratethebetween it and the angular dependence curve 42shiftsSimilarly,the latterhaving a maximum around 60°.(the semicircle radius 45)increases, intersection(the P%lymeretching/deposition equilibrium point) upwards thusdecreasing the sidewall angle. if the polymerSUBSTITUTE SHEET (RULE 26)l0l520253035WO 98115504CA 02265617 1999-03-08PCTIAU97/0066318etch rate increases, the intersection shifts downwards andthe angle increases. This mechanism can be applied toexplain the experimental data.The etched profile observed using an a-Si mask is seen3b and Fig. 9c). Thereactiveto be “overcut” (Fig. 9a to Fig.effect of higher power is to increase thesputtering component of polymer etching due to higher ionenergy and density.According to the above mechanism, this establishes anew polymer etching/deposition equilibrium at a highersidewall angle, as observed. The increase in polymerdeposition rate, which also occurs with power (Fig. 3d) isapparently less than the increase in its etch rate.One can explain the profile slope dependences ontemperature, O2 and CF4 additions in similar terms, keepingin mind the a—Si mask lateral etch rate tendencies (Fig. 5ato Fig. 5f). The profile slope initially increases withtemperature and with O3 flow (although only slightly) and6b). Theinitial increase in the slope profile versus temperature isthen decreases in both cases Fig. (6a and Fig.due to a decrease in polymer deposition through a reducedsticking probability of polymer forming species. The newdeposition/etching equilibrium occurring at ea higher angleaccording to Fig. 10. In the case of 0; flow, the polymerdeposition rate decreases only slightly but its reactivesputtering rate increases due to active oxygen produced inthe discharge thus increasing the equilibrium sidewallangle. Further increases in temperature and 0; flow causetotal polymer removal from the sidewalls of the a—Si mask,resulting in lateral etching of the a—Si and a smallersidewall angle due to mask erosion.Of CF4The sidewall slope isrelatively independent additions, which indicatesthat the increase in polymer deposition rate due to CF4 flowis balanced by the simultaneous increase in its etchingrate, probably due to an increase in the fluorine flux.Sidewall RoughnessSUBSTITUTE SHEET (RULE 26)1015920253035W0 98/15504CA 02265617 1999-03-08PCT/AU97l0066319Since the initial mask~edge roughness, determined. bySEM examination of both a-Si and photoresist masks, wassignificantly smaller than the post—etch sidewall roughness,this can be eliminated as a source of the observed sidewallroughness. Thus, either the mask edge is roughened duringetching,itself.or the roughness is formed on the silica sidewallBoth these occurrences can result from micromaskingin the presence of ion bombardment and polymer deposition.Since the photoresist masked samples were cooled,photoresist reticulation is not an issue.Fig. lla shows a etched sidewall with a photoresistmask still in place. It is seen that roughness has beengenerated in the photoresist during the process and thentransferred to the silica sidewall where the mask edge hasbeen thinned by the faceting which is evident. The increasein silica sidewall roughness with power can be explained byboth ion bombardment andan increase in micromasking aspolymer deposition rate increase.In the case of an a—Si mask (Fig. llb), some of theobserved sidewall roughness may be produced by maskroughening, as can be deduced from Fig. llb, which shows anetched sidewall with the a-Si mask still in place. However,while there is some faceting of the top corner of the mask,the upper part of the sidewall close to.the mask is smootherthan the lower part, suggesting that a larger part of theroughness has not been transferred from the mask edge, butrather has formed on the sidewall during etching. Thereason for this additional roughness is likely to be thesidewall polymer which can act as a Hucromasking material.The sidewall roughness increase with power can be explainedby an increase in Hucromasking, in this case in both themask edge and the sidewall itself, as both ion bombardmentand polymer deposition rate increase.Both increased sample temperatureSe).is likely to be the result of sidewall polymer suppression.and O; additionsreduce roughness (Fig. 5d and Fig. In both cases thisSUBSTITUTE SHEET (RULE 26)101520253035W0 98/ 15504CA 02265617 1999-03-08PCT/AU97/0066320In the temperature case, as a result of a reduction of thepolymer deposition rate to zero and in the OgcaseIn the(Fig- 69):through an increase in the polymer etch rate.polymer, thismicromasking induced roughness is eliminated.absence of sidewall contribution toThe sidewall6f),which is consistent with the fact that the profile slopelatterunperturbed balance between polymer etching and depositionroughness does not change with CE} additions (Fig.also does not change, since the implies anand therefore a constant polymer thickness on the sidewalls.Trade off between sidewall roughness and dimension lossUsing the sample temperature as a control parameterallows smooth sidewalls to be obtained without dimensionalloss, whereas using 0; additions does not allow for aprocess window where both dimension control and smoothsidewalls can be achieved. In the 02Case, active oxygenenhances the polymer etching rate on both the a—Si mask andSiO3 sidewalls and therefore, together with an improvementis sidewall roughness, it brings about dimension loss due toisotropic etching of the mask. In the case of increasingsample temperature, the flux of polymer forming species fromthe plasma remains unchanged, but their sticking probabilityis reduced, thus decreasing the effective polymer depositionrate, which then results in reduced roughness. However thethat withsticking probabilityresults obtained here suggest increasingtemperature the of polymerthan to theThis can allow for polymer free SiO2 sidewalls andformingspecies to silica is reduced faster a-Sisurface.simultaneously, sufficient polymer remaining on the a—Simask sidewalls to prevent dimension loss.It can be seen from the forgoing, the application ofreactive ion etching of silica in a high plasma densityhollow cathode etching system to the fabrication of silicabased integrated optic devices can be effectively utilised.This application imposesspecific requirements on theetching depth, sidewall roughness and profile slope control.SUBSTHUTESHEET(RULE25)101520253035W0 98/15504CA 02265617 1999-03-08PCT/AU97I0066321Due to the hollow(over O.5um for a-high plasma density produced in thecathode discharge, high silica etch ratesSi) and significant dimension lossesdepth).(>lum for Sum etchingThe disadvantages with photoresist are believed tooriginate from strong faceting which occurs on the masksidewall. In a-Si mask the faceting is negligible.Increasing RF power results in a decrease in selectivity andan increase in sidewall roughness for both photoresist anda-Si masks.The effects of sample temperature an the addition ofCF, and O; on the etching characteristics for an a-Si casehave been investigated. The etching selectivity is reducedwith all three parameters.and O20.02um.Increasing sample temperaturecontent permit sidewall roughness reduction toHowever, in the 0; case, the reduction in sidewallroughness is accompanies by lateral etching of the a—Si maskcausing dimension loss and a decrease in the profile slope.A similar effect observed with the sample temperature,however in this case there appears to be a temperature rangewhere smooth sidewalls can be obtained, together with almostvertical sidewalls and without dimension loss.Finally based on polymer deposition rate measurements,a model explaining the variety of experimental data isexplained. The model is based on a balance betweenisotropic polymer deposition and etching. A polymer film ofa certain steady—state thickness is formed as a result ofthis balance on (i) the top surface of the mask, (ii) thethe sidewalls of the SiopThe polymer thickness on the topsidewalls of the mask and (iii)surface determines theetching selectivity, whereas the polymer thickness on themask sidewalls and SiO2 sidewalls determines the profileslope and sidewall roughness.In conclusion, the silica reactive ion etching processof the preferred embodiment satisfies all the requirementsof planar waveguide fabrication and can also be used forother integrated optics applications or MEMS applicationsSUBSTITUTE SHEET (RULE 26)CA 02265617 1999-03-08WO 98/15504 PCTIAU97/0066322where deep etching of silica is required along with smoothetched sidewalls and vertical or sloped etching profiles.It would be appreciated by a person skilled in theart that numerous variations and/or modifications may bemade to the present invention as shown in the specificembodiments without departing from the spirit or scope ofthe invention as broadly described. The presentembodiments are, therefore, to be considered in allrespects to be illustrative and not restrictive.SUBSTITUTE SHEET (RULE 26)

Claims (13)

We Claim:

1. A method for etching of silica-based glass layers or substrates comprising reactive ion etching through a mask executed under conditions of simultaneous isotropic deposition of a carbon based polymer.

2. The method of claim 1 wherein the polymer deposition rate or/and its steady-state thickness on different surfaces of the etched structure is controlled by adjusting one or several process control parameters in order to control etched profile, dimension loss, sidewall and bottom etched surface roughness, and etching selectivity between the silica-based layer and mask material.

3. The method of any previous claim wherein gas or a mixture o gases is used, which contain fluorine and carbon atoms.

4. The method of any previous claim wherein a photoresist mask is used.

5. The method of claims 1 to 3 wherein a non-photoresist mask is used.

6. The method of claim 5 wherein amorphous silicon is the mask material.

7. The method of any previous claim wherein the RF
power coupled into the discharge is the adjusted parameter.

8. The method of any previous claim wherein the substrate temperature is the adjusted parameter.

9. The method of claim 8 where the temperature is adjusted to achieve low sidewall roughness and low dimension loss at the same time.

10. The method of any previous claim wherein a resputtering of any metal present within or/and in contact with the discharge zone is prevented.

11. The method of any previous claim wherein reactive ion etching is performed in a high plasma density hollow cathode etching system.

12. The method of any previous claim wherein the etching gas mixture is CH3F and Argon.

13. A product made by the method of any previous claim.