Etching of xerogel in high-density fluorocarbon plasmas

Etching of xerogel in high-density fluorocarbon plasmasT. E. F. M. Standaert,a) E. A. Joseph, and G. S. Oehrleinb)

Department of Physics, State University of New York at Albany, Albany, New York 12222

A. Jain, W. N. Gill, P. C. Wayner, Jr., and J. L. PlawskyRensselaer Polytechnic Institute, Troy, New York 12180

~Received 28 December 1999; accepted 11 July 2000!

The etching of various xerogel films has been studied in high-density fluorocarbon plasmas. Thexerogel etch rate is in part enhanced by the porosity. In discharges resulting in low surfacepolymerization, such as CF4 or oxygen-rich fluorocarbon plasmas, an additional enhancement up to60% is observed. When the polymerization of the discharge is increased, this additionalenhancement disappears and the xerogel etch rate becomes more suppressed. The suppression ismore pronounced for xerogel films with a higher porosity and a larger pore size. X-rayphotoelectron spectroscopy analysis on partially etched samples shows that the suppression in etchrate is accompanied by an increasing amount of fluorocarbon material at the xerogel surface,especially in the pores of the xerogel structure. Finally, a 30% porous xerogel film was patternedusing CHF3 as an etching gas. Slight bowing of the sidewalls was observed. ©2000 AmericanVacuum Society.@S0734-2101~00!00606-0#

th

mde

n-o

d

-e

siuse

ngua

alrtghanr-ms

hisena-ofs

por-s.red

erThe

w,

rly00nededions

ti-a

1

Inrk

I. INTRODUCTION

The current focus in interconnect technology is onintegration of copper with low dielectric constant~k! materi-als. In contrast to the conventional interconnect schebased on aluminum and SiO2, the combination of copper anlow-k dielectrics allows the device dimensions to shrink blow 0.18mm without losing device performance to interconect delays. Even though copper is certain to be the intercnect material, many low-k materials are introduced antested for intra- and interlayer dielectric applications.

Etching of vias and trenches in low-k dielectrics is essential for the ~dual! damascene approach. In this article wpresent a study on the etching of xerogels in high-denfluorocarbon plasmas. Xerogel is a spin-on glass with a table porosity. The dielectric constant of xerogel decrealinearly with porosity.1 Hence, xerogel offers the advantagthat the dielectric constant can be tailored over a wide radependent on the spin-on and subsequent curing techniqHowever, the porosity may also introduce problems, suchlow mechanical strength and moisture absorption.2 In thiswork we have studied three different xerogels for whichthese problems have been minimized. Some of the propeare outlined in Table I. The xerogels with the relatively hiporosity, 58% and 69%, were spin coated on 5 in. waferscured as described by Nittaet al.3 These films are characteized by a relatively large pore size on the order of 10–15 nThe film with the relatively low porosity of 30% is a poroumethyl silsesquioxane~MSQ!. It is an excellent candidateand is expected to enter the low-k market very soon. The

a!Current address: IBM Microelectronics, 2070 Route 52, Mail Stop AEHopewell Junction, New York 12533; electronic mail: [email protected]

b!Current address: Department of Materials & Nuclear Engineering andstitute for Plasma Research, University of Maryland, College PaMaryland 20742-2115; electronic mail: [email protected]

2742 J. Vac. Sci. Technol. A 18 „6…, Nov ÕDec 2000 0734-2101Õ200

e

e

-

n-

tyn-es

ees.s

lies

d

.

deposition technique is at the moment still proprietary. Tfilm has a relatively low pore size on the order of 2 nm. Evthough it is technically incorrect, we will label this film asxerogel for simplicity. In this work we will examine the effect of porosity and pore size on the etching behaviorthese films. It is important to point out that the xerogel filmcontain a significant amount of methyl (CH3) groups. Thepresence of these hydrophobic groups may have an imtant effect on the etching behavior of oxide-like dielectric4

The chemical compositions of the xerogel films aSiO2~CH3!0.3 for the 60%–70% porous films anSiO1.5~CH3! for the 30% porous film.

II. EXPERIMENTAL SETUP AND PROCEDURES

The reactor used for this work is a planar-coil transformcoupled plasma reactor and has been described in Ref. 5.etching of xerogel was studied in CF4, CHF3, C4F8, andC4F8/O2 plasmas maintained at 10 mTorr, 50 sccm gas floand 1400 W source power~13.56 MHz!. The self-bias volt-age was varied down to2200 V by applying a rf bias powe~13.56 MHz! to the substrate. Samples of approximate2 cm32 cm in size were positioned on the center of a 3mm anodized aluminum electrode. Etch rates were obtaiusing single wavelength ellipsometry. Partially etchsamples were transferred under ultrahigh vacuum conditto a Vacuum Generators ESCA MkP analysis chamber forx-ray photoelectron spectroscopy~XPS! analysis.

III. RESULTS

A. Patterning of xerogel

The 30% porous xerogel film was patterned with 0.6mmdeep ultraviolet photoresist and 75 nm of organic anreflection coating~ARC!. The pattern was transferred using

,

-,

27420Õ18„6…Õ2742Õ7Õ$17.00 ©2000 American Vacuum Society

rc

-r 4ees

toas

etea

e

o

heo

lmTh0%h

on-o beone

lsusompertheisof

rk

gre

00 W

xed

2743 Standaert et al. : Etching of xerogel 2743

CHF3 discharge maintained at 10 mTorr and 1400 W soupower. The self-bias voltage was set at2125 V. During thefirst 15 s of the process 50% O2 was admixed with the discharge to open the ARC layer. The etch was continued fos after which the underlying Si3N4 layer was reached. Thgas flow during the total etch was fixed at 50 sccm. Figurshows cross-sectional secondary electron micrographetched trenches~a! and vias~b! in the xerogel film. Eventhough the etch process is nonoptimized, it offers satisfacprofile control. The sidewalls are slightly bowed which halso been observed for hydrogen silsesquioxane~HSQ!.4,6

This suggests that the bowing is related to the presenchydrogen in the film. This effect and other issues associawith the patterning of xerogel films will be addressed inseparate publication.

B. Blanket etch rates

Blanket xerogel films were etched in CHF3 and C4F8 dis-charges maintained at 10 mTorr and 1400 W source powThe 30% porous xerogel etch rates~open symbols! are plot-ted as a function of self-bias voltage in Fig. 2. The SiO2 etchrates are indicated by the closed symbols for comparisThe minimum self-bias voltage for etching is around250 Vfor CHF3 and245 V for C4F8. Below this value fluorocar-bon deposition occurs. The 30% porous xerogel film etcfaster than SiO2. However, the enhancement in etch ratethe 30% porous xerogel film is less in C4F8 than in CHF3.The etch rates of the 58% and 69% porous xerogel fimeasured under identical conditions are shown in Fig. 3.etch rate of these films is relatively low compared to the 3porous film considering the fact that the latter is denser. Teffect will be highlighted below.

TABLE I. Some properties of the xerogel films investigated in this wo~The refractive index at 632.8 nm is denoted asn, k is the dielectric con-stant.!

Porosity Pore size~nm! Composition n ~632.8 nm! k

30% ;2 SiO1,5~CH3! 1.26 2.0–2.258% 10–15 SiO2~CH3!0.3 1.19 2.269% 10–15 SiO2~CH3!0.3 1.14 1.9

FIG. 1. Trenches~a! and vias~b! etched in 30% porous xerogel films usina CHF3 discharge maintained at 1400 W source power and 10 mTorr psure. The self-bias voltage was set at2125 V.

JVST A - Vacuum, Surfaces, and Films

e

0

1of

ry

ofd

r.

n.

sf

se

is

It is expected that a porous film etches faster than a nporous film, since less material per unit thickness needs tremoved as the porosity increases. To account for this,can define a normalized etch rateERnorm

ERnorm5~12P!ER, ~1!

whereER is the etch rate andP is the porosity of the xerogefilm. The normalized xerogel etch rates are plotted verself-bias voltage in Figs. 4 and 5 and were calculated frthe rates shown in Figs. 2 and 3, respectively. The uppanel in Fig. 4 shows that the normalized etch rate of30% porous film in CHF3 is enhanced. This enhancementreduced in C4F8 and may be attributed to the incorporation

.

s-

FIG. 2. Etch rates of blanket 30% porous xerogel~open symbols! as a func-tion of self-bias voltage in CHF3 and C4F8. SiO2 etch rates~closed symbols!are plotted for comparison. Source power and pressure were fixed at 14and 10 mTorr, respectively.

FIG. 3. Etch rates of blanket 60%–70% porous xerogel~open symbols! as afunction of self-bias voltage in CHF3 and C4F8. SiO2 etch rates~closedsymbols! are plotted for comparison. Source power and pressure were fiat 1400 W and 10 mTorr, respectively.

ud

fc

r

ckricig

di-.

heede

ehae

inialmtp

lm

.

nsys

-

4.7

doutfpon

ea-o-bontratee

sel.

rom

i-Q.gel

by

uo-etchuc-

in-gel

eto-nced

00ere


carbon and hydrogen in the xerogel matrix. In a recent ston the etching of methyl silsesquioxane~MSQ! and hydro-gen silsesquioxane~HSQ! it was found that incorporation ocarbon and hydrogen in these oxide-like materials enhanthe etch rate relative to SiO2 in low polymerizing discharges~for example CF4!, whereas the opposite was seen in mopolymerizing discharges~for example, C3F6/H2!.

4 An impor-tant parameter in the etching of MSQ and HSQ is the thiness of the fluorocarbon film that forms on the dielectsurface during etching. Once the film thickness allows a snificant portion of the ions to directly penetrate into theelectric material, an enhancement in etch rate is observedinvestigate the role of the fluorocarbon film formation in tetching of xerogel we performed XPS on partially etchxerogel samples. These results will be discussed in SIII C.

In contrast to the 30% porous xerogel film, the normalizetch rates for the 58% and 69% porous films are lower tthe SiO2 etch rate. This could in part be ascribed to the lowcarbon/silicon ratio of these xerogel films. Additionally,Sec. III C it will be shown that more fluorocarbon materaccumulates at the surface of the 58% and 69% porous fias a result of the relatively large pore size. It is interestingnote that the relative etch rate enhancement of the 69%rosity material with respect to the 58% porous xerogel fiin both CHF3 and C4F8 is correctly predicted by Eq.~1!, andthe normalized etch rates are essentially the same in Fig

C. XPS surface analysis

Xerogel films were partially etched at various conditioand transported under UHV conditions to a surface analchamber where XPS was performed. Panels~b!, ~c!, and~d!in Fig. 6 compare the surface chemistry of xerogel~58%porosity! with SiO2 after a partial etch in CF4, CHF3, andC4F8. For comparison, the C(1s) photoemission from an as

FIG. 4. Blanket etch rates of 30% porous xerogel~open dots! in CHF3 andC4F8 after correction for porosity@Eq. ~1!#. The SiO2 etch rates~solid dots!are shown for comparison. Source power and pressure were fixed at 14and 10 mTorr, respectively.

J. Vac. Sci. Technol. A, Vol. 18, No. 6, Nov ÕDec 2000

y

es

e

-

-

To

c.

dnr

ls

oo-

5.

is

received xerogel film is shown in Fig. 6~a!. The C(1s) pho-toemission for the as-received film is centered around 28eV and originates most likely from Si–CH3 groups. A pos-sible contribution from O–CH3 groups would have appeareat binding energies well above 285 eV. Figure 6 pointsthat, especially for C4F8, there is a significant amount ofluorocarbon material present at the xerogel surface uprocessing. It is important to point out that for CF4 the C(1s)intensity scale has been expanded by a factor of 4. The msured C(1s) intensities scale with the deposition rate of flurocarbon material that is observed for these fluorocarchemistries. The deposition rate on an unbiased subswas measured usingin situ ellipsometry. The deposition ratin C4F8 ~330 nm/min! is higher than in CHF3 ~250 nm/min!and CF4 ~5 nm/min!. The polymerization of the feed gas haa strong influence on the etching behavior of the xerogThe xerogel etch rate drops off with respect to the SiO2 etchrate in more polymerizing discharges. This can be seen fthe ratio of the normalized xerogel etch rate, Eq.~1!, and theSiO2 etch rate, indicated next to the panels in Fig. 6. A simlar behavior was observed for the etching of MSQ and HS4

The connection between fluorocarbon buildup at the xerosurface and the suppression in etch rate was first madeWang, Jin, and Park7 for a CHF3 etching chemistry. In thisarticle we provide a deeper understanding on how the flrocarbon accumulation at the surface and the xerogelrate vary with processing conditions and the physical strture of the xerogel film.

Angle resolved XPS was performed to obtain moresight into the surface topography of partially etched xerofilms ~58% porosity!. Figure 7 shows the integrated C(1s),Si(2p), and O(1s) photoemission intensities after blanketching in C4F8 as a function of self-bias voltage. The phtoemission intensities were acquired at normal incide~90°! and grazing incidence~22°!. The dashed lines labele

W

FIG. 5. Blanket etch rates of 60%–70% porous xerogels~open dots! in CHF3

and C4F8 after correction for their porosity@Eq. ~1!#. The SiO2 etch rates~solid dots! are shown for comparison. Source power and pressure wfixed at 1400 W and 10 mTorr, respectively.

s

as-th

m

r tigr-

an

te

dri

th

tionter-

urerderlIt

o-

ofeea-on

ize,aresanwuo-ur-

ro

els-

ed

pec-ities


‘‘as-received’’ show the integrated intensities for the areceived xerogel films. The dashed line labeled ‘‘CFx~1/2`!’’ indicates the integrated C(1s) intensity for a~.50 nm!thick fluorocarbon film deposited without applying a rf bito the substrate. The C(1s) intensities in the first panel suggest a significant amount of fluorocarbon material onsurface. The trend in the Si(2p) and O(1s) intensities, how-ever, suggests that a rather low amount of fluorocarbonterial is present on the surface at2200 V self-bias voltage.To explain these data consistently one has to account fophysical structure of xerogel, shown schematically in F8~a!. The high C(1s) intensities suggest that the fluorocabon material is being deposited inside the pores whichconsistent with surface studies performed by Wang, Jin,Park.7 The relatively large pore size, 10–15 nm,3 and thehigh porosity may even allow deposition inside pores locain the subsurface area.In situ ellipsometry did not indicate asignificant change in the bulk properties of the xerogel, ancan be concluded that the deposition of fluorocarbon mateis limited to the first 10–20 nm of the xerogel. Based on

FIG. 6. C(1s) photoemission spectra obtained at normal incidence for xegel and SiO2 films: ~a! xerogel as received;~b! xerogel and SiO2 partiallyetched in CF4 ; ~c! xerogel and SiO2 partially etched in CHF3 ; ~d! xerogeland SiO2 partially etched in C4F8. The ratio between the normalized xerogetch rate@Eq. ~1!# and the SiO2 etch rate is indicated. Source power, presure, and self-bias voltage were fixed at 1400 W, 10 mTorr, and2125 V,respectively. It is important to point out that for CF4 the vertical scale hasbeen expanded.


-

e

a-

he.

isd

d

itale

angle resolved XPS data we suggest a surface modificaas shown in Fig. 8~b!. It is important to note that the exactopographic details cannot be deduced from Fig. 7. Nevtheless, the invariance of the C(1s) signal with angle ofincidence and self-bias voltage is consistent with a pictwhere fluorocarbon material stretches to a depth on the oof 4–10 nm. The probing depth~in this case defined paralleto the detection angle! is smaller than the pore dimension.is expected that the C(1s) intensity is independent of thedetection angle when the pores are fully filled with fluorcarbon material. The Si(2p) and O(1s) intensities originateonly from the protruding xerogel features. The mergingthe Si(2p) and O(1s) intensities at a high self-bias voltagcan be explained by tapering of the protruding xerogel ftures which are in turn covered by a thinner fluorocarbfilm @see Fig. 8~c!#.

The 30% porous xerogel has a much smaller pore s;2 nm. The corresponding angle resolved XPS datashown in Fig. 9 for CHF3. Comparison with Fig. 7 suggesta different surface topography for the 30% porous film thfor the 58% porous film. As a result of the relatively lovalues in porosity and pore size, it can be expected that flrocarbon deposition only occurs in the pores at the very s

-

FIG. 7. Angle resolved XPS photoemission intensities from partially etchxerogel~58%! as a function of self-bias voltage. The etching gas was C4F8.Source power and pressure were fixed at 1400 W and 10 mTorr, restively. ~The dashed lines labeled As received show the integrated intensfor an as received xerogel film. The dashed line labeled CFx~1/2 `! indi-cates the integrated C(1s) intensity for a~.50 nm! thick fluorocarbon filmdeposited without applying a rf bias to the substrate.!

beoldau

enehe

gel

thethe

thetes

t toor

entthe

n-s

arelm

elsndndu-as

tchoreoly-theent

onive

heFpeiti


face. The fluorocarbon film formation at the surface willsuppressed more at higher ion energies in this weakly pmerizing gas chemistry. Based on the preceding and theshown in Fig. 9 we suggest a surface modification as olined in Fig. 10. The crossover of the 22° and the 90° intsities for Si(2p) and O(1s) can be understood when thxerogel surface exhibits more protruding features at higion energies as sketched in Fig. 10~c!.

FIG. 8. Suggested surface modification of xerogel~58%! in a C4F8 dischargeshown in cross section.

FIG. 9. Angle resolved XPS photoemission intensities from partially etcxerogel~30%! as a function of self-bias voltage. The etching gas was CH3.Source power and pressure were fixed at 1400 W and 10 mTorr, restively. @The dashed lines labeled As received show the integrated intensfor an as received xerogel film. The dashed line labeled CFx~1/2 `! indi-cates the integrated C(1s) intensity for a~.50 nm! thick fluorocarbon filmdeposited without applying a rf bias to the substrate.#


y-ta

t--

r

It can be concluded that the surface topography of xerofilms is significantly different from SiO2 during the fluoro-carbon etch. There is more fluorocarbon accumulation atxerogel surface and angle resolved XPS suggests thatfluorocarbon deposition occurs to a large extent insidepores at the surface of the xerogel film. Figure 6 indicathat the etch rate enhancement predicted by Eq.~1! is toohigh once the pores at the surface of the xerogel film starfill up with fluorocarbon material. On the other hand, fnonpolymerizing chemistries, such as CF4, the etch rate en-hancement is underestimated by Eq.~1!. This is analogous tothe etching of HSQ and MSQ, where either an enhancemor suppression of the etch rate is observed dependent onfluorocarbon surface film thickness, which in turn is depedent on the etching chemistry.4 The 58% and 69% porouxerogel films have a large pore size~10–15 nm!. Figures 4and 5 show that the normalized etch rates for these filmslower than the normalized etch rate of the 30% porous fiwhich has a smaller pore size~;2 nm!. This may be attrib-uted to the difference of carbon to silicon ratio of the xerogfilms ~see Table I!. Additionally, XPS surface analysis showmore fluorocarbon material at the surface of the 58% a69% porous films than the 30% films. A larger pore size aa higher porosity may allow for more fluorocarbon accumlation at the surface for selective etch conditions, suchC4F8. In this case, it may be expected that the xerogel erate drops in comparison to a denser film with a smaller psize. Hence, the xerogel structure, pore size, and the pmerization of the feed gas are important parameters inetching process. Finally, the loss in etch rate enhancemdue to polymer buildup may have important implicationsthe patterning of multilayer dielectric stacks using selectetch conditions.

d

c-es

FIG. 10. Suggested surface modification of xerogel~30%! in a CHF3 dis-charge shown in cross section.

agea

me

ouse1ithslm%

nrc

tcb

tceth

eo

-

fa

ogeletowth

a

p-

in

cananO8%

thes-

msx-ainulk

dling

or-her

iningef-

art

ce-lmes.PSthech-sity

aur-

thdTo

lly

and


D. Effect of oxygen addition

To optimize certain dielectric etch processes, oxygen mbe added to the fluorocarbon feed gas chemistry. Oxymay also be released from quartz material in contact withinductively coupled discharge. This effect is especially iportant for planar coil designs that have a relatively low oparea and suffer from a significant amount of capacitive cpling to the discharge. In this case, the oxygen flow releaat the quartz coupling window can be on the order ofsccm.8 The reactor that is used for this work is equipped wa coil that allows for highly inductively coupled dischargeFor example, the amount of oxygen in a fluorocarbon fideposited on an unbiased substrate was on the order of 1traced by XPS.

To study the effect of oxygen, xerogel and SiO2 etch rateswere measured in C4F8/O2 plasmas maintained at a constatotal gas flow of 50 sccm, 10 mTorr, and 1400 W soupower. The self-bias voltage was fixed at2125 V. The effectof O2 addition on the etch rate is shown in Fig. 11. The erate of 58% and 69% porous xerogel films are indicatedthe open and closed circles, respectively. The SiO2 etch ratesare indicated by the crossed circles. The normalized erates according to Eq.~1! are indicated by the triangles. Thxerogel etch rates increase by almost a factor of 4 whenO2 flow is increased to 30 sccm, whereas the SiO2 etch ratedecreases monotonically. The decrease in the xerogelrate at O2 flows above 30 sccm corresponds to a fluorocarbdepleted etching regime.

Figure 12 shows the C(1s) photoemission spectra obtained at normal incidence for partially etched xerogel~58%!samples. The amount of fluorocarbon material at the suris significantly reduced when O2 is admixed. At a flow ratioof 30/20 O2/C4F8, the C(1s) level is approximately threetimes lower than the level detected for an as received xerfilm. In this case, the fluorocarbon deposition is completsuppressed and oxygen can remove the exposed mgroups. Furthermore, this condition where the fluorocarbdeposition in the pores has stopped corresponds exactlythe maximum in the xerogel etch rate. The ratio betweennormalized etch rate of xerogel~58%! and the SiO2 etch rateis in this case 1.6. This value is very close to the etch r

FIG. 11. Blanket xerogel etch rates~closed circles 69%, open circles 58%!as a function of O2 addition in a C4F8 discharge. The total flow was fixed a50 sccm. The SiO2 etch rates are indicated by the crossed circles. Ttriangles and the dashed line indicate the blanket etch rate normalizeporosity. Source power and pressure were fixed at 1400 W and 10 mrespectively.


ynn-n-d

0

.

as

te

hy

h

e

tchn

ce

elyhylnithe

te

ratio of 1.4 found for a CF4 discharge~Fig. 6! in which thefluorocarbon deposition in the pores is also highly supressed.

The removal of methyl groups from porous silica filmsa pure O2 plasma has also been observed by Parket al.9

Their study showed that an extensive exposure of 15 minlead to a reduction in porosity by as much as 10% andincrease of 0.2–0.3 in the dielectric constant. In a pure2

plasma we measured an etch rate of 50–100 nm/min for 5and 69% porous xerogel films, butin situ ellipsometry didnot show a significant change in the optical properties ofxerogel films. The xerogel films for this study were depoited using a different method than was used for the filstudied by Parket al. In addition, we have used plasma eposures on the order of only a few minutes. This may explwhy we did not observe a significant change in the bxerogel during processing.

An additional study in a different inductively coupleplasma reactor showed that the effect of capacitive coupcan have a similar effect as the O2 addition in Fig. 11.2

Operating at similar conditions for CHF3 but with signifi-cantly more capacitive coupling, it was found that the nmalized etch rates of 58% and 69% xerogel films are higthan the SiO2 etch rates as opposed to the results shownFig. 4. Hence, it may be concluded that capacitive couplin inductively coupled discharges can have an importantfect on the xerogel etch rate.

IV. CONCLUSIONS

It can be concluded that the xerogel etch rate is in penhanced by the porosity as is expressed in Eq.~1!. In dis-charges characterized by little polymerization, such as CF4 oroxygen-rich fluorocarbon plasmas, an additional enhanment is observed up to a factor of 1.6. The fluorocarbon fiformation at the surface is relatively low in these dischargWhen the polymerization of the discharge is increased, Xshowed that fluorocarbon material is deposited insidepores at the xerogel surface. At this point, the xerogel eting is suppressed and the xerogel ettch rate after porocorrection@Eq. ~1!# falls below the SiO2 etch rate. The sup-pression is more pronounced for the xerogel films withhigher porosity and a larger pore size. The difference in s

ebyrr,

FIG. 12. C(1s) photoemission obtained at normal incidence for partiaetched xerogel films~58%! in C4F8 plasmas admixed with O2. The O2 /C4F8

flow ratio is indicated. Source power and pressure were fixed at 1400 W10 mTorr, respectively.

s

a

corXo

lid

y,

roc.

L.

. J.

T. J.

et,So-

ert,


face topography for these films has been suggested in Figand 10. Patterning of 30% porous xerogel in CHF3 sufferedfrom slight sidewall bowing, which will be addressed inseparate publication.

ACKNOWLEDGMENTS

This work was financially supported by the Semicondutor Research Corporation and SEMATECH. The authwould like to acknowledge C. Hedlund, E. Sanjuan, andLi for their assistance during the experiments. P. J. Matsuthanked for helpful discussions.

1J.-K. Hong, H.-S. Yang, M.-H. Jo, H.-H. Park, and S.-Y. Choi, Thin SoFilms 308–309, 495 ~1997!.

2A. Jain, S. Rogojevic, S. Nitta, P. C. Wayner, W. N. Gill, J. L. Plawsk


. 8

-s.is

T. E. F. M. Standaert, and G. S. Oehrlein, Mater. Res. Soc. Symp. P565, 29 ~1999!.

3S. V. Nitta, V. Pisupatti, A. Jain, P. C. Wayner, Jr., W. N. Gill, and J.Plawsky, J. Vac. Sci. Technol. B17, 205 ~1999!.

4T. E. F. M. Standaert, P. J. Matsuo, S. D. Allen, G. S. Oehrlein, and TDalton, J. Vac. Sci. Technol. A17, 741 ~1999!.

5T. E. F. M. Standaert, C. Hedlund, E. A. Joseph, G. S. Oehrlein, andDalton, J. Vac. Sci. Technol. A~submitted!.

6F. Vinet, P. Czuprynski, O. Joubert, D. M. Mariolle, and E. T. TabourAbstracts of the 44th National Symposium of the American Vacuumciety, San Jose, CA, 20–24 Oct. 1997, p. 215.

7S.-J. Wang, I.-S. Jin, and H.-H. Park, Surf. Coat. Technol.100–101, 59~1998!.

8M. Schaepkens, N. R. Rueger, J. J. Beulens, X. Li, T. E. F. M. StandaP. J. Matsuo, and G. S. Oehrlein, J. Vac. Sci. Technol. A17, 3272~1999!.

9H.-H. Park, M.-H. Jo, H.-R. Kim, and S.-H. Hyon, J. Mater. Sci. Lett.17,2083 ~1998!.

Etching of xerogel in high-density fluorocarbon plasmas

Documents

Transcript of Etching of xerogel in high-density fluorocarbon plasmas