WO1990003597A1

WO1990003597A1 - Soft-bake treatment of photoresists

Info

Publication number: WO1990003597A1
Application number: PCT/US1989/003524
Authority: WO
Inventors: J. Grunwald John; Chava Gal; Marc T. Aronhime
Original assignee: Macdermid, Incorporated
Priority date: 1988-09-30
Filing date: 1989-08-18
Publication date: 1990-04-05
Also published as: EP0389593A1; JPH03502614A

Abstract

Soft-baking of photoresists at temperatures of 120°C or higher on a hot plate for a very short period gives a number of significant advantages. These include improved energy/focus latitudes, reduced film thickness loss, improved adhesion, improved contrast and especially increased resistance to thermal deformation of the subsequently produced photoresist image.

Description

SOFT-BAKE TREATMENT OF PHOTORESISTS

BACKGROUND OF THE INVENTION

1. Field of the Invention.

This invention relates to the preparation of photo- resist images and is more particularly concerned with improvements in the pre-treatment of photoresists prior to imagewise exposure.

2• Description of the Prior Art.

In the preparation of photoresist images on sub- strates the photoresist composition is first coated on the substrate and the coated substrate is then heated to remove solvent in a step which is commonly known as a "soft-bake". The photoresist film, free of solvent, is then exposed image- wise and developed using appropriate developer solvents to produce the required image. The latter is then generally subjected to a second heating step, commonly known as a "hard-bake", prior to subsequent f brication steps.

The soft-bake of the photoresist coating is generally carried out by heating in an oven at a temperature of the order of about 90 to 110"C. and for a period of time which is of the order of 30 minutes if an oven is used or 30 to 60 seconds if a conduction hot plate is used. The actual bake temperature and time employed in any given instance will vary depending upon the nature of the particular photoresist which is used. Typical of conditions employed in this step are those described by David J. Elliott, Integrated Circuit-Fabrication Technology, p.145-162, McGraw-Hill Book Company, New York 1982 and by W.S. DeForest, Photoresist:_. Materials and Processes, p.231, McGraw-Hill Book Company, New York, 1975. It has been considered hitherto that soft-bake temperatures in excess of about 110'C. are to be avoided because of excessive decomposition of the photo- senεitizer, particularly where the latter is of the diazo- naphthoquinone type as in most positive photoresists. Such decomposition greatly reduces the photosensitivity of the system and leads to the need for impractical increases in exposure times with corresponding detriment to the overall economics of the operation.

Contrary to the generally accepted wisdom and practice in the art, it has now been found that soft-bake temperatures higher than those hitherto employed, but using aggressive development conditions, can be employed success¬ fully. It has been found further that the use of such soft-bake conditions gives rise to unexpected and highly advantageous results in the subsequent processing of the photoresist.

SUMMARY OF THE INVENTION

It is an object of the invention to provide improved soft-bake processing conditions for photoresist composi¬ tions. It is a further object of the invention to provide processing conditions for photoresist compositions which give rise to photoresist images having reduced loss of film thickness, improved contrast and increased resistance to thermal deformation. These objects, and other objects which will become apparent from the description which follows, are achieved by the process of the invention. Thus, the inven- tion in its broadest aspect comprises an improved process for producing a photoresist image on a substrate comprising, in sequence, the steps of coating the substrate with a positive photoresist composition, soft-baking the coated substrate, exposing the coated substrate imagewise to actinic radiation and thereafter developing the image where¬ in the improvement comprises carrying out the soft-baking at a temperature above about 120*c. up to about 160*C. for a time not exceeding about 200 seconds.

The invention also comprises the photoresist images produced by the improved process of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGURE 1 is a graph of the percentage of absorption of radiation at various wavelengths versus temperature of soft-bake of a photoresist under conditions described in Example 1 below.

DETAILED DESCRIPTION OF THE INVENTION

The key step in the process of the invention centers on the processing conditions in the soft-bake stage in the fabrication of a positive photoresist image. As set forth above, the formation of a photoresist image on a substrate comprises a series of steps. In the first step a coating of a positive photoresist composition is applied to a desired substrate such as aluminum, copper, copper-reinforced epoxy resin laminates, quartz, silicon wafers and the like. Any of the coating techniques known in the art can be employed in this step. Illustrative of such techniques are spin-coat¬ ing, spraying, dipping, roller coating and the like. The surface of the substrate is generally cleaned by being pretreated chemically, depending upon the nature of the substrate material, using appropriate processes and composi¬ tions known in the art. The thickness of the coating applied to the substrate is generally of the order of about 0.5 to about 10 microns, but higher or lower thicknesses can be employed in any given instance.

The second step of the process comprises the soft-bake of the coating of photoresist on the substrate. The purpose of the soft-bake is principally to remove solvents present in the coating and also to promote adhesion of the photoresist to the substrate. Hitherto the soft-bake step in the process of producing a photoresist image has been carried out conventionally at temperatures not exceed¬ ing about 110'C, and, in many cases, at temperatures as low as about 70*C, and for periods of time which can be as high as about 60 minutes. In determining the optimum conditions within these limits it has been necessary to balance two effects. Thus temperatures in the higher end of the above range favor adhesion of the photoresist to the substrate, good contrast, good sidewall profiles, low film thickness loss, and resistance to thermal deformation in the photo¬ resist image ultimately produced. On the other hand, as the soft-bake temperature is increased under the conditions hitherto employed in the art, the photosensitizer in the photoresist composition has been reported to undergo thermolytic degradation with a consequent reduction in photosensitivity, an unaffordable increase in development time required to produce the image, and a tendency for "scumming" to occur. The latter is a phenomenon observed in the photoresist image subsequently produced in which traces of photoresist remain undissolved by the developer in areas not exposed to active radiation in the exposure step or traces of photoresist have redeposited after development.

In marked contrast to the above soft-bake conditions hitherto prevailing in the art it has been found, in accordance with the invention, that significantly higher soft-bake temperatures can be employed using soft-bake times that can be optimized by routine experimentation. It has been found further that such conditions are not only essen- tially free from the processing problems encountered in the prior art, but give rise to highly advantageous properties in the photoresists ultimately produced as will be described in more detail hereafter.

Thus, in accordance with the process of the inven¬ tion, the soft-bake step is carried out at a temperature in the range of about 120^*C. to 160 ' C . for a period of time of at least 5 seconds, but not exceeding about 200 seconds and, preferably, not exceeding about 60 seconds. The most appropriate time to be used at any given temperature within the above range in the case of any given photoresist can be determined readily by a process of trial and error. This soft-bake step can be carried out using equipment and procedures conventional in the art. Illustratively, the step can be carried out by placing the resist-coated substrate in a convection oven, preheated to the desired temperature, for the chosen period of time. Alternatively, and preferably, the step is carried out on a conduction hot plate or like heating means as the resist-coated substrate is being processed on in-line production equipment.

After the soft-bake step has been completed the coated substrate is subjected to imagewise exposure using actinic radiation (generally ultraviolet radiation) and employing procedures conventional in the art. The positive image is developed using appropriate developer solutions which, in the case of diazoquinone εensitizer systems, are aqueous alkaline solutions such as aqueous sodium or potassium hydroxide solution, sodium metasilicate, sodium orthophosphate, sodium hydrogen phosphate and the like. More aggressive developers comprise tetraalkylammonium hydroxides such as tetramethylammonium hydroxides in combination with surfactants and/or any of the above alkaline agents or as the sole or principal component of the developer.

The developed image, after washing with water to remove last traces of developer solution, is subjected to a post-bake also known as a hard-bake. Post-bake is employed to increase the resistance of the image to etching and other processes such as ion implantation and plasma etching to which the image and substrate may be subjected in subsequert steps in the formation of an integrated circuit. It has been found that, when the soft-bake has been carried out at temperatures and for times set forth above in accordance with the invention, the hard-bake can be carried out at significantly higher temperatures than hitherto employed without causing thermal deformation of the image such as rounding or flow of the image line profiles. Thus, for example, it is found that, when a soft-bake of the photo¬ resist has been carried out at 130'C. for 60 seconds on a conduction hot plate the hard-bake can be carried out at 125-130'C. without causing any thermal deformation of the image. In contrast, when the soft-bake of the same photoresist system has been carried out at 110'C. for 60 seconds on a hot plate in accordance with prior art procedures, a hard-bake temperature of 115*c. will cause rounding or contrast loss by flow of the profiles in the image. The ability to use higher hard-bake temperatures may also mean that the time for which the hard-bake step is carried out can be significantly reduced with consequent reduction of overall processing times. As will be apparent to one skilled in the art, the use of the process of the invention gives rise to significant reduction in processing times with resulting improvement in productivity.

The process of the invention is applicable principal- ly to the use of positive-acting photoresist compositions. Such systems generally comprise a base-soluble polymer such as a novolak resin and like phenolic resins and a photosensi- tizer containing naphthoquinone diazide sulfonate esters. Exposure of the composition to actinic radiation serves to convert the diazo-keto configuratibn of the sensitizer to a carboxylic acid thereby rendering the exposed composition soluble in alkaline developers. Illustrative of phenolic resins employed in such systems are those novolak resins prepared by acid condensation of formaldehyde and phenol or an alkyl-substituted phenol under conditions described, for example, in Chemistry and Application of Phenolic Resins, Knop et al., Chapter 4, Springer Verlag, New York 1979. Novolak resins derived from m-cresol alone or mixtures of m- and p-cresols and having average molecular weights in the range of about 600 to about 1600 have been found to be particularly suitable for use.

Illustrative of εensitizers containing the above diazo-keto configuration are the esters and amides of 2-diazo-l,2-naphthoquinone-4- and 5-sulfonic acids. The esters are generally those derived from polyhydric phenols such as 2,3,4-trihydroxybenzophenone, 2,3,4,4'-tetrahydroxy- benzophenone, 2,4-dihydroxybenzophenone, 4-decanoylresorcin- ol, 4,4-bis (4-hydroxyphenyl) valeric acid butyl ester, novolak resin and the like. The amides of the above two acids which can be employed in the compositions of the inven¬ tion are those derived from long chain aliphatic primary amines or from aromatic primary amines.

The photoresist compositions also generally comprise a solvent or mixture of solvents as well as optional components such as surfactants, adhesion promoters, finely- divided pigments and the like. Illustrative of solvents employed alone or in admixture are ketones such as methyl- ethyl ketone, methylisopropyl ketone, diethyl ketone and the like; chlorinated hydrocarbons such as trichloroethylene, 1,1,1-trichloroethane and the like; aliphatic alcohols such as ethanol, n-propyl alcohol, n-butyl alcohol, n-hexyl alcohol and the like; aliphatic esters such as n-butyl acetate, n-hexyl acetate, cellosolve acetate, ethyl lactate, and the like; glycol ethers such as ethylene glycol mono- ethyl ether, propylene glycol monomethyl ether and esters thereof such as the acetates, propionates and the like; and aromatic hydrocarbons such as toluene, xylene, anisole and the like. The amount of solvent or mixture of solvents employed in the compositions is generally such as to represent from about 50 to about 90 percent by weight, and preferably from about 60 to about 80 percent by weight, based on total weight of the composition. The proportion of phenolic resin in the compositions is generally of the order of acout 2 to 45 percent and usually about 2 to 30 percent by weight. The proportion of sensitizer is generally in the range of about 2 to about 20 percent by weight.

The use of the soft-bake conditions called for by the present invention gives rise to a number of highly εignifi- cant advantages both in the behavior of the photoresist film after the soft-bake step and in the properties of the photo¬ resist ultimately produced therefrom. Thus, in contrast to soft-bake conditions employed hitherto in the art, the soft-bake conditions of this invention result in significant- ly lesε thermolysis of the photosensitizer than reported in the prior art. This, in turn, results in significantly less loss of photosensitivity of the resist system during the soft-bake step leaving a sufficiently high content of photo¬ active components to achieve adequate image development. Further, it has been found that a photoresist system, which has been exposed to soft-bake conditions of the present invention, posseεses markedly improved energy/focus latitudes as compared to the same systems which have been exposed to soft-bake conditions conventionally used in the art hitherto. By energy/focus latitude is meant that substantial fluctuations in exposure energy and/or focus of the aerial image transfer will not εeverely affect the replication of the desired image as delineated by the mask. Photoresist εyεtems which have been subjected to the soft-bake conditionε of the preεent invention have been found to exhibit much better contrast or gamma than those which have been soft-baked under conventional conditions hitherto employed. "Gamma" is a measure of the contrast that the exposed photoresist exhibits between irradiated and non-irradiated reεist areas. The gamma of a system is des¬ cribed mathematically by the equation Q =

where E₀ is the energy for complete resist removal and E_j_ is the energy for first significant resist removal; see S.C.Elliott, supra, at p.179. This improved contrast in the systems exposed to soft-bake conditions of the invention is very important in forming high-resolution εtructures since the resiεt reproduceε more accurately the incoming image energy without responding to stray light especially that arising from reflected light during imagewise exposure. Thus, it is found that the images ultimately produced in accordance with the invention exhibit reduced standing wave effects even if a poεt-expoεure bake iε not carried out.

Photoresist systemε which have been subjected to the soft-bake conditions of the invention have also been found to suffer markedly less film thicknesε loss during develop¬ ment. Finally, as set forth previously, the images produced from photoresist systems which have been subjected to the soft-bake conditions of the invention possess excellent profiles and show markedly greater reεistance to thermal deformation than do images produced from systems which have been soft-baked under the lower temperature conditions hitherto employed in the art.

The following examples illustrate the process of the invention and show the best mode known to the inventors of carrying out the invention, but are not to be construed as limiting.

Example 1

The photoresist system employed in the following experiments was a commercially available positive photo¬ resist [EPA-914-27 : MacDer id, Inc. , Waterbury, CT] having a novolak resin binder, a mixture of isomers of 2-diazo-l,2 -naphthoquinonesulfonyl ester of trihydroxybenzophenone as the sensitizer in solution in propylene glycol monomethyl ether acetate.

A εerieε of silicon wafers was coated with a 2 micro- meter layer of the above photoresist composition. The wafers were soft-baked by heating on a conductive hot plate for 60 sees, at either BO'C, 90*C, 100'C, 110'C, 120'C, 130°C, 140*C. and 150^βC. The coating of resist on each treated wafer was then dissolved in acetone and the ultra- violet absorption spectra of each solution was determined at three wavelengths (350nm, 398nm, 436nm) corresponding to the G, H and I-line operating modes. The percentage of photosensitizer remaining in the soft-baked coatings at each temperature was thereby determined for each of the three wavelengths. The findings are charted in solid lines in the graph shown in Figure 1.

The above experiment was repeated using soft-bake temperatures of 80^βC, 90^βC, 100^βC, 110^βC, 120'C, 130'C, 140^βC. and 150^βc. for 30 minutes in a convection oven. The percentage of photosensitizer remaining in the coatings after soft-bake under these conditions was deter¬ mined using three different wavelengths as described above and is also charted in dotted lines in the graph shown in Figure 1.

It will be seen from the graph in Figure 1 that the coatings which had been soft-baked on the hot plate for 60 seconds sustained significantly less loss of photosensi¬ tizer than did those soft-baked in a convection oven for 30 minutes at comparable temperatures. It is to be noted further that, even after exposure on the hot plate at 150^βC, there is significant residual photoactive compo¬ nent preεent in the photoresist. There is clearly much less remaining in the specimens soft-baked in the oven at the same temperature. Example 2

A series of silicon wafers was coated with a layer of 1.2 micrometer thickness of the photoresist system employed in Example 1. The coated wafers were soft-baked for 60 seconds on a hot plate at temperatures of 90, 100, 110, 120, 130 or 140'C. Each of the wafers after soft-baking was exposed imagewise to ultraviolet light (300-450 nm) using a Perkin-Elmer Microalign 240. The amount of energy used in any particular exposure was that necessary to produce a clean image after development with an alkaline developer. The contrast or gamma for each coating was determined using a Perkin-Elmer Development Rate Monitor. The data so determined is summarized in Table I below.

Table I

140 10 20 40 60 80 95

The above experiment was repeated but carrying out soft-bake at 100"C. and 110°C. in a convection oven for 30 minutes. The data so derived is εu marized in Table II below.

A comparison of the data in Tables I and II shows that the energy required to clear the image at 100°C. and 110^βC. soft-bakes and comparable development times is markedly greater in the case of the coatings which were soft-baked for 30 minutes in an oven. Indeed the energy levels required to clear in the case of the coatings soft-baked in the oven almost approach those required in the case of the coatingε soft-baked on the hot plate at 140°C. The data in Table I also shows that soft-baking for 60 seconds at 120*C. and 130^*C. yields excellent contrast (4.60 and 5.56) at relatively low energy levels (15.9 and 18.4 mJ/cm², respectively). The sidewall profiles of the images produced at theεe εoft-bake temperatures were found to be excellent and close to vertical.

Example 3

A series of silicon wafers was coated with a layer of 1.3 micron thickness of the photoresist system employed in Example 1. The coated wafers were εoft-baked for 60 εeconds at temperatures of 100, 110, 120, 130, 140 and 150°C. The wafers were then exposed imagewiεe on an Oriel contact printer and developed using commercial alkaline developer [MacDermid 62-A : 46%]. The relative exposure energy needed to clear the image (referred to the energy E₀ needed at 100*C.) in each case was determined. After development of the images the temperature -at which rounding of the image occurred during post-bake for 60 secondε on a hot plate waε determined. The results are recorded in Table III.

Table III

Energy to clear Temp. °C. of

Soft-bake 'C./60 sees. (mJ/cm******. profile rounding

100 110 120 120 1. 1 E_c 125 130 1. 1 E_c 130 140 1. 5 E 130 150 2 E, 140 It will be seen from the above resultε that the energy to clear remainε essentially the same up to the 130*C. soft bake. The images produced exhibit markedly improved reεiεtance to thermal degradation. Thuε a comparable sample soft-baked at 110*C. for 30 minutes in a convection oven showed rounding of the image profile at 115'C.

Claims

What is claimed is:

1. In a process for producing a photoresist image on a substrate, said process comprising the steps of coating the substrate with a positive photoresist composition, soft-baking the coated substrate, exposing the coated substrate imagewise to actinic radiation and thereafter developing the image; the improvement which comprises carrying out the soft-bake step at a temperature in the range of at least about 120^βC. to about 160^βC. for a time of at least about 5 secondε, but not exceeding about 200 seconds.

2. A process in accordance with Claim 1 wherein the soft-bake is carried out for a time of approximately 60 seconds.

3. A proceεs in accordance with Claim 1 wherein the positive photoresist composition comprises a photo¬ sensitive diazonaphthoquinone sulfonate ester and a novolak resin.

4. A process in accordance with Claim 3 wherein the solvent employed in said photoresiεt composition comprises propylene glycol monomethyl ether acetate.