Microelectronics Technology - American Chemical Society


Microelectronics Technology - American Chemical Societyhttps://pubs.acs.org/doi/pdf/10.1021/bk-1995-0614.ch003Figure 1...

1 downloads 76 Views 2MB Size

Chapter 3

Structure—Property Relationship of Acetal-and Ketal-Blocked Polyvinyl Phenols as Polymeric Binder in Two-Component Positive Deep-UV Photoresists 1

1

1

1

1

C. Mertesdorf, N. Münzel , P. Falcigno , H. J. Kirner , B. Nathal , H. T. Schacht , R. Schulz , S. G. Slater , and A. Zettler 1

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

1

1

2

1

OCG Microelectronic Materials AG, Klybeckstrasse 141, Basel, CH-4002, Switzerland OCG Microelectronic Materials Inc., 200 Massasoit Avenue, East Providence, RI 02914 2

Acetals have been used as protecting groups in chemically amplified positve DUV-photoresists in order to render phenol derivatives aqueous base insoluble. In the present study acetals and ketals were investigated as pendant blocking groups in polyvinyl phenols. The impact of the protecting group structure on the protecting group decomposition and the glass transition was studied. Mechanistic considerations were used to explain the protecting group stabilities observed. Two-component resist formulations containing these polymers and a photo acid generator were prepared and exposed on a high NA GCA and a low NA Canon excimer laser stepper. Resolution, DOF, post exposure delay latitude, thermal flow stability and resist decomposition were determined. Secondary reactions of the protecting groups, released in the resist film, were briefly discussed. Some resists exhibited 0.23 μm line/space resolution and a good focus latitude. Approximately two hours post exposure delay latitude and a thermal flow stability of 140°C were obtained with an optimized system. It is believed that IC device manufacturers will switchfromconventional to deep-UV lithography at design rules below 0.3 μπι. Most commercial positive deep-UV resists are based on partially protected polyvinyl phenols. Unlike conventional diazonaphthoquinone/novolak technology, here an acid is liberated in a primary photochemical event which catalyzes deblocking of the masked phenols in a subsequent reaction. The technical hurdles typically associated with such chemically amplified resists, such as delay time sensitivity, seemed to be insurmountable. Widely reported problems included dimensional changes of the latent resist image (1,2) and the formation of an insoluble surface layer (5). These profile deteriorations emerged 0097-6156/95/0614-0035$12.25/0 © 1995 American Chemical Society

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

36

MICROELECTRONICS TECHNOLOGY

R" OH

Figure 1. Synthesis of acetal (R'-H) and ketal protected polyvinyl phenols. The degree of protection in mol% is given by (b/a+b) χ 100 with n=a+b. Reproduced with permission from reference 37.

Figure 2. Selection of the enol ethers used as protecting groups. Reproduced with permission from reference 37.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

3.

MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phenols 37

upon prolonged intervals between distinct process steps, such as exposure and post exposure bake. However, considerable progress has been made recently, based on a better understanding of the mechanisms causing these deficiencies. Basic material properties, as, for example, thefreevolume of the resist binder (4, 5) and the diffusion length of the acid within the polymer matrix (6,7) play a decisive role in terms of surface inhibition and dimensional change, respectively. The deleterious influence of these factors becomes critical as the dimensions shrink. Manufacturing of 256 MB DRAM, for example, requires accurate control of 0.25 μπι features throughout the process. A variety of different protecting groups known in organic synthesis (8) have been applied to positive deep-UV photoresists. Among these, carbonates (9-14), ethers (15), silylethers (16) and acetals (2, 17-19) are the most frequently used phenol blocking groups. Acid labile esters were attached onto the phenol side group via an ether linkage (20-22). Upon deprotection the carboxylic acid is released. In this study protecting groups of moderate stability, such as acetals and ketals, were considered. These groups were attached as pendant protecting groups onto a polyvinyl phenol backbone. The chemical structure of the protecting groups was varied. Polymer properties and lithographic performance of two-component resist mixtures containing these polymers and a photo acid generator, were evaluated. Experimental Materials. Acetal and ketal protected resins were prepared according to a literature procedure (23) by an acid catalyzed addition reaction of enol ethers and polyvinyl phenol (PVP) as shown in Figure 1. Enol ethers were obtained commercially. 3,4Dihydro-6-methyl-2H-pyran (MDHP) was prepared in house according to Perkin (24). Figure 2 gives a selection of the enol ethers applied. Different degrees of protection were obtained according to Figure 1 by selecting the appropriate ratio of the starting materials. Upon conversion of the enol ethers listed in the upper box of Figure 2, ketal protected PVP's were derived. Acetal blocked polymers were obtained by treatment of PVP with the vinyl ethers summarized in the lower box. The structural difference between the families of ketals and acetals is confined to the nature of the substituent R " (Figure 1). Disulfones (25) and sulfonium salts (26) were used as the photoacid generators (PAG's). Synthesis of the appropriate PAG's was performed according to literature procedures. Thermal properties of the polymers were recorded on a Perkin Elmer 7 series thermal analyzer. Degrees of protection were obtained either from the weight loss observed in the thermogravimetric run (TGA), or from inverse gated decoupling ^ C NMR (Bruker AM-400). Good agreement was found between the two methods. Molecular weights were determined by gel permeation chromatography (GPC) using a Hewlett-Packard-liquid chromatograph (HP 1090) equipped with three PSS-gel GPC columns (Polymer Standards Service: ÎO^Â, 10' Â, 10" Â, 5 μπι). Calibration was based on polystyrene standards obtainedfromPSS. UV detection operating at 254 nm was applied. 3

2

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

MICROELECTRONICS TECHNOLOGY

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

38

Figure 3. TGA traces of acetal (EVE) and ketal (MP) protected PVP (same degree of protection): a) isothermal at 90°C, b) heating rate 10°C/min. Reproduced with permission from reference 37.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

3. MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phénols39

Lithography. Resist formulations were prepared by dissolving the polymer along with the PAG in propylene glycol monomethyl ether acetate (PGMEA) or in methyl-3methoxypropionate (MMP). All solutions were filtered through 0.2 μπι teflon membrane filters. The substrates were vapor primed with HMDS either in a Y.E.S. oven at 125°C or by liquid dispense prior to resist coating. Resist solutions were spin coated onto silicon substrates. Film thickness was measured interferometrically after soft bake (SB) using a refractive index of 1.59 (@633 nm). Different film thicknesses between 0.6 and 0.95 μπι, corresponding to maximum and minimum light incoupling were evaluated. Exposures were performed on a Canon FPA 4500 (NA=0.37, 5x reduction optics) or on a GCA XLS (NA=0.53, 4x reduction) DUV stepper. Both steppers were equipped with KrF excimer lasers, operating at 248.4 nm. After a post exposure bake (PEB), the wafers were developed with OPD 4262 (0.262 Ν tetramethyl ammonium hydroxide) by a 60 s immersion, or by a spray/puddle process (3-5 s/20-50 s). A 20 srinsewith deionized water and a spin dry were applied after development. Results and Discussion Impact of the Protecting Group Structure on Thermal Decomposition. The chemical structure of the protecting groups, as outlined in Figures 1 and 2, is one of the factors which governs their rate of decomposition. Another factor, which will be addressed later, is the amount of residual unprotected phenols, or vice versa, the degree of protection (see Figure 1). If one thinks in terms of the residual free phenolic hydroxyl groups in the copolymer acting as a weakly acidic environment, the situation can be considered as low molecular weight acetal or ketal compounds dissolved in a weakly acidic solution. Data concerning the hydrolysis of acetals and ketals in an acidic aqueous media have been known for over 70 years (27). The authors found that acetone diethyl ketal hydrolyzed 2500 times faster than acetaldehyde diethyl acetal, which proceeds opposite to their rate of formation. Formation is driven by the electrophilic activity of the carbonyl carbon, which is known to be higher in the case of the aldehyde. Acidolysis of acetals versus ketals corresponds also to the general observation that the hydrolysis rates of ethers increase with an increasing number of carbons linked to the α-carbon adjacent to the ether oxygen (28, 29). This goes with the propensity of the alkyl group to be released as a cationfromthe ether oxygen via the intermediate oxonium ion state. Due to inductive effects, stability of tertiary cations is superior to secondary and primary ones. Steric effects play a further role. Strain is relieved by protonation of the oxygen atom and subsequent cleavage of the oxygen-carbon bond (30, 31). The same stability sequence is observed for acetals (EVE) and ketals (MP), when attached to the PVP backbone, as measured by TGA (Figure 3). The ketal protected polymer loses the majority of its protecting groups, when stored for 15 hours at 90°C. In contrast to this result, the weight loss observed for the acetal resin is only 0.4% under the same conditions (Figure 3a). This weight loss can be attributed to loss of adsorbed water. In TGA runs recorded with a heating rate of 10°C/min, decomposition onset of the acetal is approximately 75 °C higher than the onset recorded for the ketal analog (Figure 3b). Due to the impact of residual phenolic sites Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

MICROELECTRONICS TECHNOLOGY

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

40

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

3. MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phenols

on the polymer stability (32, 33% the same degrees of protection were considered. EVE and MP have a similar structure with respect to the nature of R' and R . A mechanism of proton assisted decomposition of EVE and MP protected PVP, which accounts for these observations, is depicted in Figure 4. Protonation can either occur at the alkyl phenyl ether oxygen or at the aliphatic ether oxygen site, which depends on their relative base strengths. In any of the two cases, the intermediate oxocarbenium ion is considerably more stabilized in the case of the ketal (A, B) than in the case of the acetal (C, D). Moreover, it is reasonable to assume that ketal oxygens are more basic than acetal oxygens, according to a base strength study of some aliphatic ethers in aqueous sulfuric acid (30). As mentioned above, protecting group decomposition is further affected by the amount of residual unprotected phenols in the copolymer. This is illustrated in Figure 5 for the MP-PVP system. The temperature at which deprotection occurs increases with increasing degree of protection. This dependency has already been described in detail for the t-BOC-PVP system (32, 33) and was also found for the acetal and ketal protected resins investigated in this study. This seems to be a general relationship which is independent from the nature of the protecting group. More likely, it is due to a common structural element within such polymers, as the weakly acidic phenols, which act as a deprotection catalyst (auto-catalytic effect (32, 33)). Phase separations, due to immiscible polymers formed during the protection reaction, were not observed in the present study. Irregularities of decomposition temperatures within a series of partially protected t-BOC-PVP's were caused by such phase separations (32). The TGA weight loss in the temperature range given in Figure 5 corresponds well with the theoretically expected weight loss, if a total thermolysis of the protecting groups is considered. Although somewhat higher values were obtained by NMR, NMR and TGA correlate well. The small difference can be ascribed to some non volatile decomposition products formed upon heating. The values given were derived from TGA. Isothermal recording at 90°C (Figure 5b) reveals that the rate of blocking group thermolysis is not linearly dependent on the degree of protection. MP-PVP's with 73, 67, 59 and 52 mol% protection exhibit a linear relationship between weight loss and thermolysis time, with increasing slope for decreasing blocking. In contrast to this, 34 mol% protected MP-PVP shows already a strongly nonlinear behavior with 40 % of its total protecting groups lost after 200 min. In the case of 67 mol% protection, only 7.5 % of the total blocking groups are released under the same conditions. Figure 6 shows a plot of the ratio of released and total blocking groups versus the degree of protection for respective MP-PVP's stored for 200 min at 90°C. The curve is fitted well by an exponential function. At degrees of protection below 50 mol% rapid decomposition occurs and the stability required is not provided. If this auto-acceleration is due to a hyperacidity effect, as known from high ortho novolaks (34), is very unlikely. In PVP's with an all-para structure, phenols are radially distributed from the backbone and intramolecular hydrogen bonding is less likely to occur. However, weaker consecutive hydrogen bonds between individual chains may develop upon deprotection, which may increase acidity and thus the deprotection rate in a non linear fashion. Release of the ketal-carbon-oxygen or acetal-carbon-oxygen bonds, as depicted in Figure 4, explains the stability difference observed between ketals and acetals.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

fIf

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

41

MICROELECTRONICS TECHNOLOGY

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

42

Figure 5. TGA traces of MP-PVP with different degrees of protection: a) heating rate 10°C/min, b) isothermal at 90°C. Reproduced with permission from reference 37.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

3. MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phenols 43

Cleavage of the O-R' bond could be competitive when R' is a tertiary radical, in accordance with the high stability of the intermediate tertiary carbon cation of R formed (29). In fact, the decomposition temperature drops, if a primary carbon radical R' (EVE) is substituted by a tertiary one (TBVE). In the case of TBVE-PVP, decomposition onset was shifted to lower temperatures by roughly 50°C with respect to EVE-PVP (Figure 7). A mechanistic consideration of this competitive path is given in Figure 8. Proton catalyzed decompostion yields the considerably stabilized tertiary butyl cation intermediate and the hemi acetal. The hemi acetal is known to hydrolyze very fast to acetaldehyde (28) releasing, in this particular case, the phenol moiety. Six membered cyclic structures can adopt a strain free chair or boat configuration. Taking this into consideration, Tdec's of MDHP protected PVP's are expected to be in the range of non-cyclic ketals with primary alkoxy carbons R', as for example MP-PVP. However, lower decompositon onsets were found with the cyclic protecting group (MDHP), as shown in Figure 9. It is likely that the ring attached to the phenol via its 6-position is not able to be completely strainfree,due to restrictions implied by the backbone. The same tendency was observed, when the respective cyclic and non-cyclic acetals, such as DHP-PVP and EVE-PVP, were compared, with EVEPVP being the more stable system. Five membered rings of the tetrahydrofuran type are rather planar and therefore strained, which makes them much more prone to proton catalyzed ring opening ether cleavage than the tetrahydropyran ring. Arnett at al. discussed the contribution of strain effects in cyclic ethers to their base strengths (31). An interpretation of their observation of tetrahydrofuran being considerably more basic than tetrahydropyran is given in terms of electron correlation forces between non-bonded electron pairs on the oxygen and the electrons engaged in bonding the hydrogens to the adjacent carbons. The repulsion of the oxygen lone pair for nearby bonding pairs is reduced upon coordination with a proton. This effect is supposed to be considerably larger in the case of tetrahydrofuran whose unshared electrons are constrained to orbitals that are eclipsed by the adjacent bonds to hydrogen. In the case of tetrahydropyran, the orbitals are staggered to the adjacent carbon-hydrogen bonds. Bonding of the cyclic ethers to the PVP-backbone changes their conformation and thus the base strengths of the endocyclic oxygens. Also introduction of the methyl group - when considering ketals with respect to acetals - affects the basicities of furan and pyran type protecting groups to a different extent. However, the relative basicities are maintained in the ketal series, as deducedfromthe protecting group stabilities observed. Thus, MDHF protected PVP is less stable than its pyran analog (MDHP-PVP). According to Hesp et al., who investigated the acetal analogs (DHF-PVP and DHP-PVP), a higher photospeed can be expected in the case of the more basic furan protecting group (18, 19). In the present study photospeeds of the cyclic ketals were not determined, due to fast decomposition of MDHF and MDHP protected PVP occurring in the TGA experiments which gives rise to a low stability of the corresponding resist formulations. Crosslinking between individual polymer chains has to be considered in the protection reaction outlined in Figure 1. Crosslinking occurs, when a polymer immobilized hydroxy group adds, for example, to the oxocarbenium ion intermediate (see Figure 4: A,B,C and D) which formsfromthe protected phenol moiety. Further routes can be imagined, as the foregoing formation of a phenyl vinyl ether group via

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

1

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

44

MICROELECTRONICS TECHNOLOGY

OUpS

0.50

0.40 ^

CD

cσ> *·+-* u ω t

leased/'Total

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

~o



0.20 ^

0.10 ^

ω 0.00

•ιιIIIIIIIIIIIIIIII

40

I I I I I I I I I IJ I I I I I I

50

60



Protection Degree (mole %) Figure 6. Thermolysis of MP-PVP dependent on the degree of protection; values derived from Figure 4b (200 min at 90°C). Reproduced with permission from reference 37.

70 Η 50

100

150 200 250 Temperature [°C]

300

Figure 7. TGA traces of EVE and TBVE protected PVP (same degree of protection): heating rate 10°C/min. Reproduced with permission from reference 37.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

3. MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phenols 45

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

H

Figure 8. Mechanism of competitive proton assisted decomposition path of TBVE protected PVP.

Figure 9. TGA traces of MDHP and MP protected PVP (same degree of protection): heating rate 10°C/min. Reproduced with permission from reference 37.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

46

MICROELECTRONICS TECHNOLOGY

Molecular weight

I

I

Figure 10. GPC traces of TBVE-PVP dependent on the amount of catalyst used in the synthesis (χ 10" mol%): a) 31, b) 256, c) 784; values given are relative to PVP. 4

Figure 11. Dependence of Tg and Tdec (onset and peak) on the protection degree of MF-PVP; values derived from DSC and TGA: heating rate 10°C/min. Reproduced with permission from reference 37.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

3. MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phenols 47

proton releasefromthe oxocarbenium ion and its subsequent conversion with phenols from adjacent polymer chains. The extent of any route depends on the rate constants of the individual reaction step. According to our results, the acidolytic lability of the OR' linkage is one of the factors which determines the rate of crosslinking. This would mean that acidolysis of the O-R' linkage with formation of the hemi-acetal is the rate determining step, as depicted, for example, in Figure 8. The GPC traces given in Figure 10 illustrate crosslinking as occurring in the synthesis of TBVE-PVP. An increase of the molecular weight is observed at long reaction time (60 h) and higher catalyst concentration. This cannot be explained by simply an addition of the vinyl ethers to some of the phenol side groups. Rough conditions (prolonged reaction time, high amount of catalyst) can cause gelation of the reaction mixture. It is interesting to note that DHP does not show branching under the conditions applied. Presumably, this is due to the O-R' subunit being linked to the acetal carbon (R' and R'" together form a trimethylene bridge) which makes the back reaction (reformation of pyran-ring) likely to occur. Impact of the Protecting Group Structure on the Glass Transition. The glass transition (Tg) governs the flow stability of the profiles generated in a resistfilm.A lack inflowstability causes failures during pattern transfer in the plasma etch. On the other hand, it is desirable to have a process window available between Tg and decomposition, to be able to effectively anneal the resist film at temperatures high enough above Tg, without decomposing the material. Annealing of the resist film reduces thefreevolume and thus the penetration of ambient contaminants (4). Further, the diffusion length of the acid should be reduced, if thefreevolume is decreased. Protecting groups with large andflexiblecarbon radicals (R* and R ) are expected to affect packing of the polymer backbone more than small, compact radicals do. This results in a larger mobility of the chain segments for a given temperature, due to less interchain interactions, which translates to a lower Tg. How Tg depends on the degree of protection is given in a plot in Figure 11 for MP-PVP together with the respective decomposition temperatures (see preceding section). Upon increasing the ratio of protected phenol units, the Tg decreases. This Tg dependency is known for partially protected PVP's and has already been published for t-BOC-PVP (55). In the case of 35 mol%, MP-PVP decomposes before the Tg is reached. In order to have a process window available between Tg and decomposition, degrees of protection of at least 50 mol% are required for MP-PVP (see also preceding section). An example of the lithographic performance derived with a two-component resist based on MP-PVP (55 mol% protection) and 3% PAG is given in the next section (Figure 14). The resist shows high resolution and vertical wall profiles. However, as already pointed out in the previous section, fast decomposition of MPPVP limits shelf life of the resist composition. Partial decomposition was also observed on the hotplate during processing. The degree of protection was decreased in the unexposed regions after the post exposure bake. It turned out that, despite the considerably high degree of protection of MP-PVP and its low Tg (see Figure 11), thermalflowstability of the resist was higher than 130°C. This can be ascribed to the ,M

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

48

MICROELECTRONICS TECHNOLOGY

fact that the resist decomposes on the hotplate and the Tg gradually increases upon volatilization of MPfromthe resist film (see next section). A large window between Tg and decomposition was found, when protecting groups of higher stability, such as acetals, were attached to the PVP backbone. Figure 12 shows decomposition and Tg versus degree of protection-plot obtained for the EVE-PVP system. In the case of compact protecting groups, such as MP, Tg changes only slightly when the degree of protection is varied over a wide range. From Figure 11 it can be deduced that this range is roughly 20°C between 50 mol% and 73 mol% protection. In contrast to this, Tg of the copolymer is more degraded by the EVE protecting group than by an equal amount of MP. This tendency is continued when applying IB VE which distorts packing of the polymer backbone even more. A considerably small Tg degradation was observed with the cyclic protecting groups, such as DHP. This is due to the number of possible conformations being limited in the case of cyclic hydrocarbons with respect non-cyclic systems composed of same number of carbons. The results is a smaller distortion of backbone packing with the conformation limited cyclic protecting groups.

0 4

1

1

1

1

1

20

30

40

50

60

70

Protection [mol%] Figure 12. Dependence of Tg and Tdec (onset and peak) on the protection degree of EVE-PVP; values derived from DSC and TGA: heating rate 10°C/min. Reproduced with permissionfromreference 37.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

3. MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phenols 49

Secondary Reactions of the Protecting Groups. Since acetal/ketalformation/decomposition is an equilibrium, secondary reactions of the protecting groups liberated in the resistfilmhave to be considered. Secondary reactions were only sparingly discussed in the literature and correlated with the lithographic performance (55, 36). They depend on the volatility of the protecting group fragment(s) and the concentration of further reactants affecting equilibrium (acid, PVP-phenol, water). In the case of cyclic blocking groups (DHP, MDHP and MDHF) only one fragment is released upon acidolysis, since the alcohol component (O-R) is covalently linked to the acetal or ketal carbon atom. A study of how DHP-decomposition products affect resist behavior has been published by Sakamizu et al. for DHPPVP/Novolak (55). Non-cyclic protecting groups (MP, TBVE, IBVE and EVE) can decompose into two fragments, e.g. acetone and methanol in the case of MP. The primaryfragmentis determined by the oxygen site, which is protonatedfirst,and by the direction of cleavage. This, in turn, is dependent on factors, such as the basicity of the oxygens, the relief of strain and the stability of the intermediate oxonium ion or carbocation etc.. The situation is outlined in Figure 13 for MP-PVP, with MP being the primary fragment. According to GC-MS measurements, MP (1) is released upon pyrolyzing MPPVP (Mertesdorf, C ; Mareis, U.; Schaedeli, U. unpublished data.). MP (1) is the primary decomposition product if protonation and subsequent cleavage occurs at the phenyl alkyl ether oxygen site with elimination of a protonfromthe oxocarbenium ion intermediate (Figure 13a). Under the conditions present within the exposed parts of the resistfilm,MP (1) is in an equilibrium state with acetone (4). Acetone (4) can undergo self-condensation in an aldol-type reaction. Acetone (4) and its condensates (6), (7) were detected by GC-MS of an exposed resist film (Mertesdorf, C ; Mareis, U.; Schaedeli, U. unpublished data.). Of course, MP must not nessessarily be the primary decomposition product in the resistfilm,as it is the case when pyrolyzing the pure resin. However, secondary reactions of the protecting groups released have been outlined in Figure 13a, starting with MP for simplicity. MP (1) undergoes rapid polymerization in the presence of strong acids, e.g. sulfonic acids. In the absence of water, cationic polymerization (Figure 13b) has to be considered as a further possible reaction, due to the MP-acetone equilibrium (Figure 13a) being shifted to the MP side (i). Lithography The SEM micrograph of 0.25 μιη line/space features, generated with a two component resist containing MP-protected polyvinyl phenol and 3% PAG, is shown in Figure 14. Strong standing waves are apparent and are found to increase with the degree of protection. Standing waves are characteristic for a diffusion limited system. The standing wave phenomenon could be caused by a portion of the MP molecules formed being consumed in an oligomerization process which is initiated by excess of the photo generated acid. As long as a growing chain (8) is not terminated by a protic reagent containing nucleophilic sites (e.g. water, alcohol), the proton consumed is not

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

so

MICROELECTRONICS TECHNOLOGY

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

a)

OMe

n

MeO

"

ι

χ

-

1

(8)

Figure 13. Acid-catalyzed-secondary reactions of MP-PVP. Reproduced with permission from reference 37.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

3.

MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phenoh 51

liberated. The diffusion of the growing chain decreases with increasing chain length, which could be the reason for the limited diffusion. It can be expected that the MPoligomers suggested do not act as strong dissolution inhibitors in the exposed regions of the resist film. Although high resolution (0.23 μιη), appreciable DOF and delay time robustness was achieved with MP-PVP based resists, tradeoffs with respect to shelf life and decomposition during processing limited their applicability. These tradeoffs were surmounted by using acetal chemistry, which is much more tolerant with respect to thermal decomposition. Processing and lithographic performance of acetal based resists, designated Acetal-1 and Acetal-2 are given below. These resists are compatible with process conditions used for conventional photoresists. Table I summarizes the conditions applied for Acetal-1. Table I. Process conditions used for Acetal-1 Process Step

Process Conditions

Softbake (SB) Film thickness Exposure

60 s 130°C, vacuum hot plate 0.82 μιη GCA XLS DUV stepper, NA = 0.53, ~18 mJ cm"2 60 s 110°C, vacuum hotplate OPD 4262 (0.262 Ν TMAH), 5 s spray/30 s puddle

Post exposure bake (FEB) Development

Soft bake (SB) is performed above Tg and below the thermal decomposition of the resist. A bake window of approximately 30°C was used for SB optimization. Although the resists are characterized by a large post exposure bake (PEB) latitude, optimization of the PEB temperature is critical in order to minimize proximity effects and CD shrinkage during post exposure delay (PED). The resists are compatible with 2.38 % aqueous TMAH developers. Less than 2 nm of dark erosion were observed after a 60 s development. Track development was not yet optimized, however, good results were obtained with a 5 s spray/30s single puddle process vnih Acetal-1. The coded mask line width of bright and dark field line space (1/s) features and the isolated lines and trenches were printed down to 0.23 μπι on silicon with the NA 0.53 exposure tool. As shown in Figure 15 and 16, 0.23 μπι 1/s were linearly resolved within a ± 10% CD with Acetal-1 and Acetal-2. Figure 17 gives the depth of focus (DOF) obtained with Acetal-1 which is 0.8 μπι at 0.25 μπι 1/s (bright field), with the non-optimized spray/puddle development process. PED experiments were performed at different sites with different steppers. Although deblocking proceeds without PEB at room temperature, a PEB step is required as a reset in order to minimize CD change. Roughly 2 h PED latitude were obtained within a ± 10% CD with the two resists. Acetal-1 provides a thermal flow stability of 140°C on a vacuum hotplate, as shown in Figure 18.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

52

MICROELECTRONICS TECHNOLOGY

Figure 14. SEM micrograph of 0.25 μιη line/space features (NA=0.53) obtained with a 2-component resist containing MP-PVP (55 mol% protection) and 3% PAG. Reproduced with permission from reference 37.

Figure 15. Resolution/linearity of Acetal-1. Reproduced with permission from reference 37.

Figure 16. Resolution/linearity of Acetal-2:filmthickness = 0.82 μπι, NA = 0.53, Dev.: 0.262 TMAH im..

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

3.

MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phenoh

Figure 17. DOF of Acetal-1: 0.25 μπι 1/s (bright field). Reproduced with permission from reference 37.

Figure 18. Thermal flow stability of Acetal-1: 120 s on a vacuum hotplate. Reproduced with permissionfromreference 37.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

53

54

MICROELECTRONICS TECHNOLOGY

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

Conclusion Different enol ethers yielding acetal and ketal protected PVFs were investigated as pendant blocking groups in positive DUV-resists. As expected, ketal polymers decompose much faster than their structural acetal analogs. Within a ketal or acetal family, decomposition rates vary according to their structure which can be ascribed to factors, such as the base strength of the oxygens, the relief of strain and the stability of cationic intermediates. Tg of PVP is degraded by indroduction of the pendant protecting group. Degradation is dependent on the degree of protection and on the conformational freedom of the respective ketal or acetal moiety. Secondary reactions of the protecting groups released in the resist film have to be considered in the structural design of acetal and ketal protected PVFs. Although ketals showed appreciable resolution, a considerable tradeoff with respect to shelf life and process stability was determined. Preoptimized resist formulations based on acetal chemistry exhibit 0.23 μπι line/space resolution, 0.8 μπι focus latitude at 0.25 μπι resolution and approximately two hours post exposure delay latitude. Acknowledgments. We would like to acknowledge C. De Leo, D. Frey and J.P. Unterreiner for lithographic work. Thank you to O. Nalamasu (AT&T Bell Labs) and A.G. Timko (AT&T Bell Labs) for exposures on the high NA stepper and U . Schaedeli (Ciba) and T.X. Neenan (AT&T Bell Labs) for helpful discussions. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Padmanaban, M . ; Kinoshita, Y.; Kudo, T.; Lynch, T.; Masuda, S.; Nozaki, Y.; Okazaki, H.; Pawlowski, G.; Przybilla, K.J.; Roeschert, H.; Spiess, W.; Suehiro, N.J. Photopolym. Sci. Technol. 1994, 7, 461. Hattori, T.; Schlegel, L.; Imai, Α.; Hayashi, N.; Ueno, T. J. Photopolym. Sci. Technol. 1993, 6, 497. Hinsberg, W.D.; MacDonald, S.A.; Clecak N.J.; Snyder, C.D. Proc. SPIE, Adv. Resist Technol. Process.IX1992, 1672, 24. Ito, H.; England, W.P.; Sooriyakumaran, R.; Clecak, N.J.; Breyta, G.; Hinsberg, W.D.; Lee, H.; Yoon, D.J. J. Photopolym. Sci. Technol. 1993, 6, 547. Ito, H.; Breyta, G.; Hofer, D.; Sooriyakumaran, R.; Petrillo, K.; Seeger, D. J. Photopolym. Sci. Technol. 1994, 7, 433. Yoshimura, T.; Nakayama, Y.; Okazaki, S. J. Vac. Sci. Technol. Β 1992, 10, 2615. Asakawa, K. J. Photopolym. Sci. Technol. 1993, 6, 505. Greene, T.W. Protecting Groups in Organic Synthesis; John Wiley & Sons: New York, 1981. Eib, N.K.; Barouch, E.; Hollerbach, U.; Orszag, S.A. Proc. SPIE, Adv. Resist Technol. Process. X 1993, 1925, 186. Sturtevant, J.; Holmes, S.; Rabidoux, P. Proc. SPIE, Adv. Resist Technol. Process.IX1992, 1672, 114. Ito H.; Willson, C.G. Polym. Eng. Sci. 1983, 23, 1012. Münzel, N.; Holzwarth, H.; Falcigno, P.; Schacht, H.T.; Schulz, R.; Nalamasu, O.; Timko, A.G.; Reichmanis, E.; Kometani, J.; Stone, D.R.; Neenan, T.X.; Chandross, E.A.; Slater, S.G.; Frey M.D.; Blakeney, A. Proc. SPIE, Adv. Resist Technol. Process. XI 1994, 2195, 47.

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

3.

13. 14. 15. 16. 17.

Downloaded by UNIV OF IOWA on May 22, 2016 | http://pubs.acs.org Publication Date: May 5, 1995 | doi: 10.1021/bk-1995-0614.ch003

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

MERTESDORF ET AL.

Acetal- and Ketal-Blocked Polyvinyl Phenols

55

Nalamsu, O.; Timko, A.G.; Cheng, M.; Kometani, J.; Galvin, M.; Heffner, S.; Slater, S.G.; Blakeney, A.J.; Münzel, N.; Schulz, R.; Holzwarth, H.; Mertesdorf, C.; Schacht, T. Proc. SPIE, Adv. Resist Technol. Process. Χ 1993, 1925, 155. Nalamasu, Ο.; Cheng, M.; Timko, A.G.; Pol, V.; Reichmanis, E.; Thompson, L.F. J. Photopolym. Sci. Technol. 1991, 4, 299. Crivello, J.V.; Conlon, D.A.; Lee, J.L. Polym. Mater. Sci. Eng. 1989, 61, 422. Murata, M.; Kobayashi, E.; Yamachika, M.; Kobayashi, Y.; Yumoto, Y.; Miura, T. J. Photopolym. Sci. Technol. 1992, 5, 79. Hattori, T.; Schlegel, L.; Imai, Α.; Hayashi, N.; T. Ueno, T. Proc. SPIE, Adv. Resist Technol. Process.X1993, 1925, 146. Hesp, S.A.M.; Hayashi, N.; Ueno, T. J. Appl. Polym. Sci. 1991, 42, 877. Hayashi, N.; Hesp, S.M.A.; Ueno, T.; Toriumi, M.; Iwayanagi, T.; Nonogaki, S. Polym. Mater. Sci. Eng. 1989, 61, 417. Onishi, Y.; Oyasato, N.; Niki, H.; Hayase, R.H.; Kobayashi, Y.; Sato, K.; Miyamura, M . J. Photopolym. Sci. Technol., 1992, 5, 47. Sinta, R.; Hemond, R.C.; Medeiros, D.R.; Rajaratnam, M.M.; Thackeray, J.W.; Canistro, D.; United States Patent 1993, 5,258,257. Thackeray, J.W.; Adams, T.; Fedynyshyn, T.H.; Georger, J.; Hemond, R.; Madeiros, D.; Mori, J.M.; Orsula, G.W.; Sinta, R.F.; Small, R.D. J. Photopolym.Sci.Technol. 1993, 6, 645. Shiraishi, H.; Hayashi, N.; Ueno, T.; Sakamizu, T.; Murai, F. J. Vac. Sci. Technol. Β 1991, 9, 3343. Perkin, W.H. J. Chem. Soc (London) 1887, 51, 702. Aoai, T.; Aotani, Y.; Umehara, Α.; Kokubo, T. J. Photopolym.Sci.Technol. 1990, 3, 389. DeVoe, R.J.; Olofson, P.M.; Sahyun, M.R.U. Adv. Photochem. 1992, 17, 313. Skrabal, Α.; Mirtl, K.H. Z. Phys. Chem. (Leibzig) 1924, 111, 98. Skrabal, Α.; Schiffrer, Α. Z. Phys. Chem. (Leibzig) 1921, 99, 290. Tronow, B.W.; Ladigina, L.W. Ber. Dtsch. Chem. Ges. 1929, 62, 2844. Arnett, E.M.; Wu, C. Y. J. Am. Chem. Soc. 1962, 84, 1680. Arnett, E.M.; Wu, C. Y. J. Am. Chem. Soc. 1962, 84, 1684. Ito, H. J. Polym. Sci., Part A 1986, 24, 2971. Paniez, P.J.; Rosilio, C.; Mouanda, B.; Vinet, F. Proc. SPIE, Adv. Resist Technol. Process.XI1994, 2195, 14. Dammel, R. In Diazonaphthoquinone-based Resist, O'Shea, D.C., Ed.; SPIE Tutorial Texts in Optical Engineering; SPIE Optical Engineering Press: Bellingham, Washington 1993, Vol. TT 11; 41-48. Sakamizu, T.; Shiraishi, H.; Yamagushi, H.; Ueno, T.; Hayashi, N. Jpn. J. Appl. Phys. 1992, 31, 4288. Funhoff, D.J.H.; Binder, H.; Dijkstra, H.J.; Goethals, M.; Krause, Α.; Moritz, H.; Reuhman-Huisken, M.E.; Schwalm, R.; Van Driessche, V.; Vinet, F. Proc. SPIE, Adv. Resist Technol. Process. X 1993, 1925, 53. Mertesdorf, C.; Münzel, N.; Holzwarth, H.; Falcigno, P.; Schacht, H.T.; Rohde, O.; Schulz, R.; Slater, G.S.; Frey, D.; Nalamasu, O.; Timko, A.G.; Neenan, T.X., "Structural Design of Ketal and Acetal Blocking Groups in 2-Component Chemically Amplified Positive DUV-Resists" in Advances in Resist Technology and Processing XII, Robert D. Allen, Editor, Proc. SPIE 2438, 84-98 1995.

RECEIVED

August 9, 1995

Reichmanis et al.; Microelectronics Technology ACS Symposium Series; American Chemical Society: Washington, DC, 1995.