Organometallic photoresist developer compositions and processing methods

KR103025631B1Active Publication Date: 2026-09-29INPRIA CORP +1
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
KR1020247028119
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Priority Date
2019-04-12
Filing Date
2020-04-10
Publication Date
2026-09-29
Estimated Expiration
2040-04-10

Smart Images

  • Figure R1020247028119_ABST
    Figure R1020247028119_ABST
Patent Text Reader

Abstract

A developer composition is described based on a blend of solvents, wherein the developer is particularly effective for EUV patterning using organometallic-based patterning compositions. A method of using such a developer composition is described. The blend of solvents can be selected according to Hansen solubility parameters. Generally, one solvent has low polarity, expressed as the sum of δP + δH, and the second solvent component of the developer has a higher value of δP + δH. The corresponding solvent composition is described.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] Cross-reference regarding related applications

[0002] This application claims priority to U.S. provisional patent application No. 62 / 833,343, filed on April 12, 2019, by Jiang et al., titled “Organometallic Photoresist Developer Compositions,” which is incorporated herein by reference.

[0003] Technology field

[0004] The present invention relates to an improved treatment of an organometallic radiation patterning composition for reducing defects. In particular, an improved chemical development composition for reducing microbridges and similar defects is described. A rinsing after development for reducing defects is also described. Background Technology

[0005] To form semiconductor-based devices, other electronic devices, or other complex microstructures, materials are typically patterned to integrate the structure. Thus, structures are generally formed through a repetitive process of sequential deposition and etching steps, through which patterns are formed in various materials. In this way, a large number of devices can be formed in a small area. Some advancements in this technology may include a reduction in the footprint of the device, which can be desirable for improving performance.

[0006] Since organic compositions can be used as radiation-patterned resists, radiation patterns are used to alter the chemical structure of the organic composition corresponding to the pattern. For example, a process for patterning a semiconductor wafer may involve lithographically transferring a desired image from a thin film of an organic radiation-sensitive material. Patterning of the resist generally involves several steps, such as exposing the resist to a selected energy source via a mask to record a latent image, and developing and removing selected areas of the resist. In the case of a positive tone resist, the exposed areas are modified to allow for the selective removal of these areas, whereas in the case of a negative tone resist, unexposed areas can be removed more easily.

[0007] Generally, the pattern can be developed using radiation, reactive gas, or liquid solutions to remove selectively sensitive parts of the resist, while other parts of the resist act as a protective anti-etching layer. Liquid developers can be particularly effective for developing latent images. The substrate can be selectively etched through windows or gaps in the remaining areas of the protective resist layer. Alternatively, material can be deposited into the exposed areas of the underlying substrate through the developed windows or gaps in the remaining areas of the protective resist layer. Finally, the protective resist layer is removed. This process can be repeated to form additional layers of the patterned material. The material can be deposited using chemical vapor deposition, physical vapor deposition, spin coating, or other desired methods. Additional processing steps, such as the deposition of conductive materials or dopant implantation, may be used. In the fields of micro and nano fabrication, the feature size of integrated circuits has been reduced to very small, achieving high integration densities and improving circuit functionality. The problem to be solved

[0008] The compositions and materials disclosed in this specification enable the formation of high-fidelity patterns and structures with improved, i.e., reduced, pattern defect density, such as microbridging, in the lithographic processing of organometallic photoresists.

[0009] In a first aspect, the present invention relates to a method for developing a radiation-exposed organometallic patterning layer to form a developed patterned layer comprising an organometallic oxide / hydroxide network, the method comprising the step of contacting the radiation-exposed organometallic patterning layer with a developer composition comprising a solvent blend. Each solvent blend independently has a sum of Hansen solubility parameters δH + δP of about 16 (J / cm³). 1 / 2 With at least 55 volume% of one or more solvents less than or equal to, and each independently, the sum of Hansen solubility parameters δH + δP is approximately 16 (J / cm³) 1 / 2 It may include two or more solvents having about 0.25 volume% to about 45 volume% of one or more solvents.

[0010] In a further embodiment, the present invention relates to a method for developing a radiation-exposed organometallic patterning layer comprising an organometallic oxide / hydroxide network, the method comprising the step of contacting the radiation-exposed organometallic patterning layer with a developer composition comprising about 0.25 volume% to about 45 volume% of water, alcohol, glycol ether, pyrrolidone, lactone, carboxylic acid, or a combination thereof, and 55 volume% or more of a solvent blend of a ketone, ether, ester, or a combination thereof. Brief explanation of the drawing

[0011] The drawings incorporated in this application are incorporated into the specification and form part of the specification. They serve to illustrate embodiments of the disclosure and, together with the description, explain the principles of the disclosure. The drawings are merely examples of specific embodiments and do not limit the disclosure. Figure 1 shows six scanning electron microscope (SEM) images of 32 nm pitch lines with 16 nm spacing patterned by EUV lithography, with the bottom left, bottom middle, and bottom right images showing microbridge defects, and the top left, top middle, and top right images showing magnified views of the microbridge defects in the corresponding bottom images. Figure 2 shows the gel dose (D) for the sum of Hansen solubility parameters δH + δP for the solvents in Table 1. g It is the plot of ), and D g is the EUV dose measurement for the gel when each solvent is used as a developer for organotin oxide hydroxide photoresist treated at a baking (PEB) temperature after exposure to 160 °C. FIG. 3 is a flowchart of a photolithography process according to an embodiment of the present disclosure. Figure 4 shows three for 2-heptanone (top), 10 vol% formic acid in 2-heptanone at time t=0 (middle), and 10 vol% formic acid in 2-heptanone at time t=1 week (bottom). 1 This is a set of H NMR spectra. The results show the stability of the formic acid solution over a period of one week. FIG. 5 is a plot of microbridge density (Def / Die 16p32) as a function of baking temperature after exposure for organometallic resists patterned via EUV lithography to form 32 nm pitch lines for three developer compositions, namely 2 volume% formic acid in 2-heptanone, 10 volume% formic acid in 2-heptanone, or 2-heptanone. FIG. 6 shows the equivalent dose density (16p32 E) for an organometallic resist patterned via EUV lithography to form a 32 nm pitch line for three developer compositions, namely 2 vol% formic acid in 2-heptanone, 10 vol% formic acid in 2-heptanone, and 2-heptanone. size This is a plot of microbridge density (Def / Die 16p32) as a function of ). FIG. 7 shows the equivalent dose to size (16p32 E) for three developer compositions, namely 2 vol% formic acid in 2-heptanone, 10 vol% formic acid in 2-heptanone, and 2-heptanone, for an organometallic resist patterned via EUV lithography to form a 32 nm pitch line. size This is a plot of nanometer-scale linewidth roughness (LWR) as a function of ). FIG. 8 is a plot of microbridge density (Def / Die 16p32) at a target CD of 16 nm as a function of baking temperature after exposure for organometallic resists patterned via EUV lithography to form a 32 nm pitch line for three developer compositions, namely 2 vol% 2-ethylhexanoic acid (EHA) in 2-heptanone, 10 vol% EHA in 2-heptanone, and 2-heptanone. FIG. 9 shows the equivalent dose density (16p32 E) for an organometallic resist patterned via EUV lithography to form a 32 nm pitch line for three developer compositions, namely 2 vol% 2-ethylhexanoic acid (EHA) in 2-heptanone, 10 vol% EHA in 2-heptanone, and 2-heptanone. size This is a plot of microbridge density (Def / Die 16p32) as a function of ). FIG. 10 is an array of 12 scanning electron microscope (SEM) images of a silicon substrate with an organometallic resist patterned via EUV lithography to form a 32 nm pitch line, each image in the field having an average line CD closest to 16 nm, the four SEM images in the top row correspond to a baking (PEB) temperature after exposure of 140 °C, the four SEM images in the second row correspond to treatment at a PEB temperature of 160 °C, the four SEM images in the bottom row correspond to treatment at a PEB temperature of 180 °C, and the four images in each row from left to right correspond to the development by 2-heptanone (HFO), PGMEA (PFO), 10 vol% formic acid in 2-heptanone (HF2), and 10 vol% formic acid in PGMEA (PF2), respectively. Fig. 11 is an array of 10 scanning electron microscope (SEM) images of a silicon substrate with an organometallic resist patterned via EUV lithography to form 32 nm pitch lines, each image in the field having an average line CD closest to 16 nm, and the 5 SEM images in the top row from left to right correspond to phenomena caused by 2-heptanone (HF0), 5 vol% formic acid in 2-heptanone (HF1), 10 vol% formic acid in 2-heptanone (HF2), 15 vol% formic acid in 2-heptanone (HF3), and 20 vol% formic acid in 2-heptanone (HF4), and the 5 images in the bottom row from left to right correspond to phenomena caused by PGMEA (PF0), 5 vol% formic acid in PGMEA (PF1), 10 vol% formic acid in PGMEA (PF2), 15 vol% formic acid in PGMEA (PF3), and 20 vol% in PGMEA Each corresponds to a phenomenon caused by formic acid (PF4). FIG. 12a is a plot of nanometer linewidth roughness (LWR) as a function of nanometer line CD at baking (temperature) after exposure at 140 °C for four developer compositions, namely PGMEA, 2-heptanone, 10 vol% formic acid in PGMEA, and 10 vol% formic acid in 2-heptanone. FIG. 12b is a plot of nanometer line width roughness (LWR) as a function of nanometer line CD at baking (temperature) after exposure at 160 °C for four developer compositions, namely PGMEA, 2-heptanone, 10 vol% formic acid in PGMEA, and 10 vol% formic acid in 2-heptanone. FIG. 12c is a plot of nanometer linewidth roughness (LWR) as a function of nanometer line CD at baking (temperature) after exposure at 180 °C for four developer compositions, namely PGMEA, 2-heptanone, 10 vol% formic acid in PGMEA, and 10 vol% formic acid in 2-heptanone. FIG. 13 shows the developer composition at three post-exposure baking (PEB) temperatures. HF0-HF 4 and PF0-PF4 Equivalent dose for size (16p32 E) for an organometallic resist patterned via EUV lithography using size This is a log-linear plot of the estimated number of defects per 150 images (defects(18p32) / 150 images) for the 18p32 pattern as a function of ). Fig. 14 is an array of 18 scanning electron microscope (SEM) images of a silicon substrate with an organometallic resist patterned via EUV lithography to form 32 nm pitch lines, each image in the field having an average line CD closest to 16 nm, the four SEM images in the top row correspond to a baking after exposure (PEB) temperature of 170 °C, the four SEM images in the second row correspond to treatment at a PEB temperature of 180 °C, and the four SEM images in the bottom row correspond to treatment at a PEB temperature of 190 °C, and the four images in each row from left to right correspond to PGMEA (PFO), 1 vol% acetic acid in PGMEA (PA1), 2 vol% acetic acid in PGMEA (PA2), 5 vol% acetic acid in PGMEA (PA3), 7.5 vol% acetic acid in PGMEA (PA4), and 10 vol% in PGMEA, respectively. This corresponds to a phenomenon caused by acetic acid (PA5). FIG. 15 shows the developer composition at three post-exposure baking (PEB) temperatures. PF0 and PA1-PA5 Equivalent dose for size (16p32 E) for an organometallic resist patterned via EUV lithography using size This is a log-linear plot of the estimated number of defects per 150 images (defects(18p32) / 150 images) for the 18p32 pattern as a function of ). FIG. 16 shows the developer composition at three post-exposure baking (PEB) temperatures. PF0 and PA1-PA5 Equivalent dose for size (16p32 E) for an organometallic resist patterned via EUV lithography using size This is a log-linear plot of the estimated number of defects per 150 images (defects(16p32) / 150 images) for the 16p32 pattern as a function of ). FIG. 17 shows the developer composition at three post-exposure baking (PEB) temperatures. HW1-HW4Equivalent dose (16p32 E) for the size of an organometallic resist patterned via EUV lithography using size This is a plot of the estimated number of defects per 150 images (defects(16p32) / 150 images) for the 16p32 pattern as a function of ). FIG. 18 is a plot of the estimated number of defects per 150 images (defects (16p32) / 150 images) for a 16p32 pattern as a function of rinsing conditions for an organometallic resist patterned by EUV lithography, developed with 2 volume% 1,2-hexanediol in 2-heptanone and rinsed under various rinsing conditions, namely (A) no rinsing, (B) rinsing with 2-heptanone for 15 seconds, (C) rinsing with 2 volume% 1,2-hexanediol in 2-heptanone for 15 seconds, or (D) rinsing with 2 volume% 1,2-hexanediol in 2-heptanone for 15 seconds after rinsing with 2 volume% 1,2-hexanediol in 2-heptanone. Figure 19 is a plot showing representative SEM images and performance data for organometallic resists patterned via EUV lithography to form a hexagonal columnar pattern of 23 nm diameter columns and treated with a baking temperature after exposure of 170 °C and a 2-heptanone developer (left), a baking temperature after exposure of 190 °C and a 2 vol% acetic acid developer in PGMEA (middle), and a baking temperature after exposure of 190 °C and a 5 vol% acetic acid developer in PGMEA (right). FIG. 20 is an array of six scanning electron microscope (SEM) images of a silicon substrate with an organometallic resist patterned via EUV lithography, each image in the field has an average line CD closest to 16 nm, the three SEM images in the top row correspond to a baking after exposure (PEB) temperature of 160 °C, the three SEM images in the bottom row correspond to treatment at a PEB temperature of 180 °C, and the three images in each row from left to right are each 2-heptanone ( HFO ), PGMEA( PF0), PGME( PG0 ), and n-propanol ( NP0 It corresponds to a phenomenon caused by ). FIG. 21 shows 2-heptanone (baking (temperature) after exposure at 160 °C and 180 °C) HF0 ), PGME( PG0 ), and n-propanol ( NP0 ) Equivalent dose for size (16p32 E) for an organometallic resist patterned via EUV lithography to form a 32 nm pitch line for a developer composition size This is a log-linear plot of the estimated number of defects per 150 images (defects(18p32) / 150 images) for the 18p32 pattern as a function of ). Specific details for implementing the invention

[0012] An improved solution has been discovered as a developer for organometallic photoresists that reduces pattern defect density. The developer solution composition is generally selected for its ability to enable high solubility contrast between the exposed and unexposed regions of the photoresist. That is, the developer can effectively convert the latent image resulting from exposure within the photoresist coating into a physical pattern. The developer compositions described herein mainly comprise an organic solvent that provides baseline negative tone development of the exposed organometallic resist and also comprise selected additives discovered to improve pattern defectability. In some embodiments, an amount of water, generally less than 1 volume%, may be optionally added to further aid in defect reduction. Organic additives preferred for the developer can generally be identified by Hansen solubility parameters and / or their functional groups.

[0013] To produce commercially desirable patterning results, it is generally desirable for the patterned material to be sufficiently defined and possess high fidelity to enable specific functions of the integrated device, such as conductivity or capacitance. Defects within the pattern, such as holes or breakage, can be detrimental to device function and reduce overall device yield. Therefore, for the patterned structure to be successfully integrated into a device or circuit, it is desirable for the photoresist pattern after development to have high fidelity and be free of defects, such as unwanted bridging between adjacent features. Consequently, there is a demand for development methods and compositions that generate high-fidelity patterns, i.e., low defect rates. It has been found that the presence of specific additive compositions can improve patterning performance, particularly in mitigating patterned defects such as microbridges or scum between features. Linewidth roughness (LWR) can also be improved, as shown in the example below.

[0014] A new class of radiation-based resists may be based on metal oxide chemistry (metal oxo / hydroxo compositions) using radiation-sensitive ligands to control the stability and processability of the resist. Generally, these resist compositions function as negative-tone photoresists when developed with organic solvents. Generally, the patterning of these metal-based compositions can benefit from the developers described herein. Related resist compounds are discussed in U.S. Patent No. 8,703,386B2 by Bass et al., titled "Metal Peroxo Compounds With Organic Co-ligands for Electron Beam, Deep UV and Extreme UV Photoresist Applications," which is incorporated herein by reference. While tin compositions are exemplified herein and the data presented focus on tin-based resists, the developer compositions described herein are expected to be effective for other metal-based resists described below.

[0015] Organometallic photoresists, such as organotin oxide hydroxides, have been shown to possess excellent properties as photoresists for use in lithographic photopatterning. Suitable organometallic photoresists include organotin materials as described in U.S. Patent No. 9,310,684B2 by Meyers et al., titled “Organometallic Solution Based High Resolution Patterning Compositions”; published U.S. Patent Application No. 2016 / 0116839A1 by Meyers et al., titled “Organometallic Solution Based High Resolution Patterning Compositions and Corresponding Methods”; and U.S. Patent No. 10,228,618B2, titled “Organotin Oxide Hydroxide Patterning Compositions, Precursors, and Patterning”, all of which are incorporated herein by reference. Other organometallic patterning compositions based on various metals are described in published U.S. Patent Application No. 2002 / 0076495 by Maloney et al., titled "Method of Making Electronic Material," and U.S. Patent No. 9,372,402B2 by Freedman et al., titled "Molecular Organometallic Resists for EUV," all of which are incorporated herein by reference. Resists containing metal oxide particles having an organic coating are described in published U.S. Patent Application No. 2015 / 0234272A1 by Sarma et al., titled "Metal Oxide Nanoparticles and Photoresist Compositions," all of which are incorporated herein by reference. The applicant has developed highly advanced organotin patterning materials, some of which are the exemplified compositions.

[0016] Suitable organotin materials are generally of formula R z SnO (2-(z / 2)-(x / 2)) (OH) x A radiation-sensitive patterning composition represented by (wherein 0 < z ≤ 2 and 0 < (z + x) ≤ 4, where R is a hydrocarbyl having 1 to 31 carbon atoms) or a blend having distinct R groups (for N individual compositions, R N It is based on chemicals of (which can be written as ). In the coating layer, the composition can be incorporated into a common oxo / hydroxo network. In particular, branched alkyl ligands are compounds R 1 R 2 R 3 It may be desirable for some patterning compositions that can be represented as CSn(NR')3, where R 1 and R 2 is independently an alkyl group having 1 to 10 carbon atoms, and R 3 is hydrogen or an alkyl group having 1 to 10 carbon atoms. As mentioned below, the alkyl ligand R thus represented is generally R 1 R 2 R 3 It can be similarly applied to other embodiments having CSn(X)3, where X corresponds to a trialkoxide or triamide moiety. In some embodiments, R 1 and R 2 can form a cyclic alkyl moiety, and R 3 It can also link other groups within the cyclic moiety. Suitable branched alkyl ligands are, for example, isopropyl(R 1 and R 2 is methyl and R 3 is hydrogen), tert-butyl(R 1 , R 2 and R 3 is methyl), tert-amyl(R 1 and R 2 is methyl and R 3is -CH2CH3), sec-butyl(R 1 is methyl, and R 2 is -CH2CH3, and R 3 is hydrogen), neopentyl (R 1 and R 2 is hydrogen, and R3 is -C(CH3)3), may be cyclohexyl, cyclopentyl, cyclobutyl, and cyclopropyl. Examples of suitable cyclic groups include, for example, 1-adamantyl (-C(CH2)3(CH)3(CH2)3 or tricyclo(3.3.1.13,7)decane bonded to a metal of a tertiary carbon) and 2-adamantyl (-CH(CH)2(CH2)4(CH)2(CH2) or tricyclo(3.3.1.13,7)decane bonded to a metal of a secondary carbon). In other embodiments, the hydrocarbyl group may comprise an aryl or alkenyl group, for example, benzyl or allyl, or an alkynyl group. In other embodiments, the hydrocarbyl ligand R may comprise any group composed only of C and H and containing 1 to 31 carbon atoms. In summary, some examples of suitable alkyl groups bonded to tin include, for example, linear or branched alkyl (i-Pr ((CH3)2CH-), t-Bu ((CH3)3C-), Me (CH3-), n-Bu (CH3CH2CH2CH2-)), cyclo-alkyl (cyclo-propyl, cyclo-butyl, cyclo-pentyl), olefin groups (alkenyl, aryl, allyl), or alkynyl groups, or combinations thereof. In further embodiments, suitable R groups may comprise hydrocarbyl groups substituted with heteroatom functional groups comprising cyano, thio, silyl, ether, keto, ester, or halogenated groups or combinations thereof.

[0017] A precursor composition for forming an organotin oxo / hydroxyl coating composition integrated into a common oxo / hydroxo network may comprise one or more soluble organotin oxo / hydroxo compounds, or a corresponding compound having a hydrolyzable ligand that forms an oxo and / or hydroxo ligand upon hydrolysis. In the case of a precursor composition having multiple compounds, the compounds may have distinct organic ligands having metal-carbon bonds and identical or distinct hydrolyzable ligands. Thus, a precursor composition for forming a radiation-sensitive coating R n SnX 4-n It may comprise a solution of one or more compounds represented by, wherein n = 1 or 2, R is a hydrocarbyl group having 1 to 31 carbon atoms as described above, and X is a ligand having a hydrolytic MX bond. Suitable hydrolytic ligands are, for example, alkinides RC≡C, alkoxides RO - Carboxylate RCOO - It may include halides, dialkylamides, or combinations thereof. In particular, the organotin trialkoxide composition may include the formula RSn(OR 0 It can be expressed as )3, where R 0 The group may be one of the same moiety as described above with respect to R. In some embodiments, the aforementioned organotin precursor composition is MX4 and / or MO ((m / 2)-l / 2) (OH) l It may additionally include a composition represented by, where 0 < z ≤ 2, 0 < (z + w) ≤ 4, m = M m+The formal valence is 0 ≤ l ≤ m, and M = M' or Sn, where M' is a non-tin metal of groups 2 to 16 of the periodic table. Generally, organotin photoresists exhibit both high resolution and high etching resistance capable of forming small features and patterns. Hydrolysis in situ can be used during the coating process or after the coating step to hydrolyze hydrolytic MX bonds, thereby forming an oxo / hydroxo network in the coating before patterning. The precursor compound may also form clusters in solution by appropriate ligand rearrangement, wherein at least some of the hydrolyzable ligands may be replaced with oxo bridges or hydroxyl groups having, for example, three tin atoms, as described in published U.S. Patent Application No. 2019 / 0053001 by Cardineau et al., titled “Organotin Clusters, Solutions of organotin Clusters, and Application to High Resolution Patterning,” and published U.S. Patent Application No. 2019 / 0308998 by Cardineau et al., titled “Tin Dodecamers and Radiation Patternable Coatings With Strong EUV Absorption,” the two documents are incorporated herein by reference.

[0018] The formation of a photosensitive organotin coating can be achieved through various means known to those skilled in the art, such as spin coating. For solution deposition of the precursor, in the case of the tin-based resist described above, the tin concentration may generally be in the range of about 1 mM to about 1 M in terms of the amount of tin, about 2 mM to about 750 mM in additional embodiments, and about 5 mM to about 500 mM in other embodiments. In some embodiments, the photosensitive organotin coating may be formed by vapor deposition techniques such as atomic layer deposition (ALD) or chemical vapor deposition (CVD), as described in U.S. Patent No. 10,228,618B2 by Meyers et al., titled “Organotin Oxide Hydroxide Patterning Compositions, Precursors, and Patterning,” and U.S. Patent No. 9,996,004B2 by Smith et al., titled “EUV Photopatterning of Vapor-Deposited Metal Oxide-Containing Hardmasks,” respectively, the two documents incorporated herein by reference.

[0019] The thickness of the coating may generally be a function of the precursor solution concentration, viscosity, and rotation speed. For other coating processes, such as vapor deposition, the thickness may also generally be adjusted through the selection of coating parameters. In some embodiments, it may be desirable to use a thin coating to facilitate the formation of small and highly resolved features. In some embodiments, the coating material may have an average dry thickness of about 1 micrometer or less before development, about 250 nanometers (nm) or less in additional embodiments, about 1 nanometer (nm) to about 50 nm in additional embodiments, about 1 nm to about 40 nm in other embodiments, and about 1 nm to about 25 nm in some embodiments. The range of coating thickness after development for the exposed area is generally within the same range as presented above, provided that development can remove a relatively small amount of the exposed material. Those skilled in the art will recognize that additional ranges of solution concentration and thickness are considered within the specified ranges and are contained in this disclosure. Thickness can be evaluated using non-contact methods of X-ray reflectance and / or ellipsometry based on the optical properties of the film.

[0020] After the deposition and formation of an organotin coating, an edge bead removal (EBR) rinsing step is typically used. EBR treatment typically occurs after photoresist deposition and prior to any heat treatment or baking, and involves rinsing the peripheral edges of the wafer or substrate with a solvent to remove photoresist within selected areas. EBR and back-side rinsing involve applying an edge bead rinsing solution to the edge and the back-side of the wafer, as described in U.S. Patent No. 10,627,719 by Waller et al., titled "Methods of Reducing Metal Residue in Edge Bead Region from Metal-Containing Resists," which is incorporated herein by reference.

[0021] Soft baking or post-application baking (PAB) is generally performed prior to radiation exposure to hydrolyze hydrolytic bonds of the precursor composition and / or further remove the solvent and promote densification of the coating material. In some embodiments, PAB may be performed at a temperature of about 25 °C to about 250 °C, in further embodiments about 50 °C to about 200 °C, and in further embodiments about 80 °C to about 150 °C. Post-exposure heating may generally be performed for at least about 0.1 minutes, in further embodiments about 0.5 minutes to about 30 minutes, and in further embodiments about 0.75 minutes to about 10 minutes. Those skilled in the art will recognize that additional ranges of PEB temperature and time within the specified ranges are considered and are within the scope of this disclosure. The coated material generally comprises a polymeric metal oxo-hydroxo network based on oxo-hydroxo ligand bonding to a metal, where the metal also has some alkyl ligands, or a molecular solid composed of a multinuclear metal oxo-hydroxo species having alkyl ligands.

[0022] Generally, organotin photoresist coatings can be patterned using radiation. Suitable radiation sources include extreme ultraviolet (EUV), ultraviolet (UV), or electron beam (EB) radiation. For semiconductor device manufacturing, EUV radiation is generally preferred because it offers higher resolution than UV radiation and higher throughput compared to EB-based processing. Radiation can generally be directed onto the substrate material through a mask, or the radiation beam can be controlledly scanned across the entire substrate to form a latent image within the resist coating.

[0023] According to the international standard ISO 21348 (2007), incorporated herein by reference, ultraviolet light extends between wavelengths of 100 nm and less than 400 nm. A krypton fluoride laser can be used as a source of 248 nm ultraviolet light. The ultraviolet range can be subdivided in several ways according to accepted standards, such as extreme ultraviolet (EUV) from 10 nm to less than 121 nm and far ultraviolet (FUV) from 122 nm to less than 200 nm. 193 nm rays from an argon fluoride laser can be used as a radiation source for FUV. EUV light was used for lithography at 13.5 nm, and this light is generated from Xe or Sn plasma sources excited using high-energy lasers or discharge pulses. Commercial sources of EUV photons include scanners manufactured by ASML Holding NV Netherlands. Soft X-rays may be limited to wavelengths of 0.1 nm to less than 10 nm. Light is directed through a mask to form a latent image within a radiation-sensitive coating that has exposed and unexposed regions.

[0024] The amount of electromagnetic radiation can be characterized by the fluence or dose obtained by the integrated radiation plus according to exposure time. In some embodiments, a suitable radiation fluence is about 1 mJ / cm². 2 Up to about 200 mJ / cm² 2 , in additional implementation examples, approximately 2 mJ / cm² 2 Up to about 150 mJ / cm² 2 , in an additional implementation example, about 3 mJ / cm² 2 Up to about 100 mJ / cm² 2 It may be. In one embodiment, the EUV radiation is approximately 150 mJ / cm². 2 It can be performed with a dose below or the electron beam is approximately 2 mC / cm at 30 kV. 2It may be performed at the following doses. Those skilled in the art will recognize that additional ranges of radiation fluence are considered within the specified ranges and are within the scope of this disclosure.

[0025] In electron beam lithography, the electron beam typically induces secondary electrons, which generally deform the irradiated material. In materials where higher resolution is generally believed to result from a shorter secondary electron range, the resolution may be a function of at least some part of the secondary electron range. Based on the high resolution achievable by electron lithography using the organometallic coating materials described herein, the range of secondary electrons in organometallic materials is limited. The electron beam may be characterized by the energy of the beam, and suitable energies may be in the range of about 5 eV to about 200 keV and, in further embodiments, about 7.5 eV to about 100 keV. The proximity-corrected beam dose at 30 keV is about 0.1 microcoulombs per square centimeter (μC / cm²). 2 ) to about 5 millicoulombs (mC / cm²) per square centimeter 2 It may be in the range of ), and in additional embodiments, about 0.5 μC / cm 2 to about 1 mC / cm 2 and in other embodiments, about 1 μC / cm 2 to about 100 μC / cm 2 It may be within the range. Those skilled in the art can calculate corresponding doses at different beam energies based on the teachings of this specification and will recognize that additional ranges of electron beam characteristics within the specified ranges are considered and are within the scope of this disclosure.

[0026] After exposure to radiation and the formation of a latent image, subsequent post-exposure baking (PEB) is typically performed. In some embodiments, PEB may be performed at a temperature of about 45 °C to about 250 °C, in further embodiments about 50 °C to about 190 °C, and in further embodiments about 60 °C to about 175 °C. Post-exposure heating may generally be performed for at least about 0.1 minutes, in further embodiments about 0.5 minutes to about 30 minutes, and in further embodiments about 0.75 minutes to about 10 minutes. Those skilled in the art will recognize that additional ranges of PEB temperature and time within the specified ranges are considered and are within the scope of this disclosure. PEB may be designed to further incorporate the exposed area without decomposing the unexposed area into a metal oxide.

[0027] It has been shown that both positive tone and negative tone patterning can be achieved in organotin oxide hydroxide systems due to a composition containing both metal oxides and organic ligands. For example, when an organic solvent is used as a developer, negative tone patterning is realized in which the unexposed material dissolves and the exposed material remains. In contrast, when an aqueous acid or base solution, for example, containing tetraalkylammonium hydroxide, is used as a developer, positive tone patterning can be realized in which the exposed material dissolves and the unexposed material remains. The improved developer described herein relates to the formation of negative tone images.

[0028] In the case of negative tone imaging, the developer may include an organic solvent, such as the solvent used to form the precursor solution. Additives may be added to the base organic solvent to form an improved developer. The additive may be a different organic solvent, and the characteristics of the developer components are described to identify and distinguish them from other components. Effective additives have been found to include compositions having specific values ​​of Hansen solubility parameters (HSP) developed to characterize solvent properties. Hansen solubility parameters include δD (dispersion), δP (polarity), δH (hydrogen bonding), and δT (total, where (δT) 2 = (δD) 2 + (δP) 2 + (δH 2 ...is included. Regarding the notation for Hansen solubility parameters, the literature uses "δ" or "d" interchangeably, and the "d" convention is used for provisional priority application. The values ​​for Hansen solubility parameters were first presented in Charles M. Hansen's 1967 Ph.D. thesis and have since been extensively discussed in the technical literature. Generally, Hansen solubility parameters for compositions are determined empirically, and the values ​​presented herein are from the literature [Knovel Solvents - A Property Database ((2008, 2012) Author(s) / Editor(s): Wypych, George, Publisher: ChemTec Publishing Copyright Date2008; 2012; 2019ISBNN / AElectronic ISBN978-1-59124-533-9)], which is incorporated herein by reference.

[0029] Although we do not wish to be limited by theory, compositions having higher δH and / or δP parameters compared to solvents such as 2-heptanone are believed to be better able to solubilize partially hydrolyzed and / or concentrated materials suspected of containing specific pattern defects, such as microbridges and / or scum between features. As described below, the selection of the base solvent can be generalized around 2-heptanone, and additives can be more effectively identified by a relatively large sum of δH and δP. The enhanced effect of the developer having composition blends may also be attributed to chemical interactions between at least part of the developer or rinse composition and the organotin material, namely complexation and / or coordination. In general, developer compositions having higher δH and δP parameters than 2-heptanone are expected to achieve improved results, namely lower defect densities. By selecting a solvent blend, the intensity of the developer can be controlled to generate desired features at an appropriate dose. A base solvent may be selected to effectively dissolve the unexposed patterning composition. An additive solvent increases the developer strength to help dissolve the partially exposed or partially concentrated patterning composition that may be present along the edges of the pattern or at random locations on the pattern, thereby enabling sharper edges (i.e., edges with lower edge roughness).

[0030] Generally, the selection of an appropriate developer solvent composition can be influenced by solubility parameters with respect to coating materials (both irradiated and non-irradiated), as well as the developer's volatility, flammability, toxicity, viscosity, and potential chemical interactions with other process materials. In particular, base solvents suitable for developers include, for example, aromatic compounds (e.g., benzene, xylene, toluene), esters (e.g., propylene glycol monomethyl ether acetate, ethyl acetate, ethyl lactate, n-butyl acetate, butyrolactone), ketones (e.g., methyl ethyl ketone, acetone, 2-butanone, cyclohexanone, 2-heptanone, 2-octanone), ethers (e.g., tetrahydrofuran, dioxane, anisole), 4-methyl-2-pentanol (and other weakly polar alcohols), and blends thereof. Generally, base solvents have a sum of Hansen solubility parameters δH + δP of approximately 16.0 (J / cm²). 3 ) 1 / 2 The details are as follows. Additionally, suitable solvents are generally identified within the chemistry identified above, provided that all members of these groups may possess the identified solubility characteristics, particularly when containing multiple functional groups. In some contexts, solvent blends can be conveniently characterized by chemical species and their functional groups without mention of solubility parameters.

[0031] In the case of embodiments using a blend of solvents, the developer generally comprises about 55 volume% or more, about 60 volume% to about 99.75 volume% in additional embodiments, about 65 volume% to about 99.5 volume% in additional embodiments, about 70 volume% to about 99.25 volume% in additional embodiments, and about 75 volume% to about 99 volume% in other embodiments, and may comprise one or more solvent compounds having selected solubility parameters. Those skilled in the art will recognize that additional ranges of base solvent concentrations within the specified ranges are considered and are within the scope of this disclosure. Volume percentage values ​​are calculated based on the volume of solvent before mixing. Volume percentage values ​​may be converted to weight percentage values ​​according to density if desired.

[0032] As described herein, one or more highly polar additional solvents, referred to as polar solvents, may be added to the solvent blend to form an improved developer. In some embodiments, the developer may have 0.25 volume% to about 45 volume% polar solvent, in further embodiments about 0.4 volume% to about 30 volume%, in further embodiments about 0.5 volume% to about 25 volume%, and in other embodiments about 0.75 volume% to about 22 volume% polar solvent. In some embodiments, the polar solvent is about 16.0 (J / cm²). 3 ) 1 / 2The sum of the above Hansen solubility parameters δH + δP may be present. Suitable polar solvents include, for example, water, acetone, polar monohydroxyl alcohols (e.g., methanol, ethanol, propanol, isobutanol, pentanol, and mixtures thereof), polyhydroxyl compounds (e.g., ethylene glycol, propylene glycol, glycerol), pyrrolidone (e.g., 2-pyrrolidone, 1-ethyl-2-pyrrolidone, N-methyl-2-pyrrolidone), glycol ethers (ethylene glycol monomethyl ether, etc.), carboxylic acids (e.g., formic acid, oxalic acid, 2-ethylhexanoic acid), diols (e.g., 1,2-hexanediol, 1,2-propanediol, 1,3-propanediol), and mixtures thereof. Those skilled in the art will recognize that additional ranges of polar solvent concentrations are considered and are within the scope of the present disclosure within the specified ranges.

[0033] Water may be present in the initial solvent, and such inherent water must be taken into account in the assessment of water content. For convenience, to distinguish developers with a potential significant contribution of water, developers may be classified as having 0.1 wt% (1000 ppm) or more, or less than 0.1 wt% (1000 ppm) of water; for convenience, these may be referred to herein as having water, or having no water or substantially having water, respectively. In the case of water, the weight percentage is converted to a volume percentage according to density to allow the calculation of the volume percentage of water consistent with the general discussion herein. Unless otherwise specified, ppm here denotes parts per million by mass. In some embodiments, water is used as an additive alone in small amounts within the base solvent. Accordingly, in these embodiments, the developer has water of at least about 2000 ppm, about 2500 ppm to about 10 wt% in additional embodiments, about 3500 ppm to about 5 wt% in some embodiments, and about 4000 ppm to about 3 wt% in other embodiments. Those skilled in the art will recognize that additional compositional ranges are considered within the specified ranges and are within the scope of this disclosure.

[0034] In addition to the primary developer solvent composition, the developer may include additional compositions that facilitate the development process. Suitable additives include, for example, cations selected from the group consisting of ammonium, d-block metal cations (hafnium, zirconium, lanthanum, etc.), f-block metal cations (cerium, lutetium, etc.), p-block metal cations (aluminum, tin, etc.), alkali metals (lithium, sodium, potassium, etc.), and combinations thereof, and dissolved salts having anions selected from the group consisting of fluorides, chlorides, bromides, iodides, nitrates, sulfates, phosphates, silicates, borates, peroxides, butoxides, formates, ethylenediamine-tetraacetic acid (EDTA), tungstates, molybdates, etc., and combinations thereof. A surfactant may be added to lower surface tension to facilitate the application of the developer. Suitable surfactants may include, for example, ionic surfactants (e.g., alkyl ether sulfates, benzyl sulfonates, and phosphate esters, etc.) and nonionic surfactants (e.g., ethoxylated and alkoxylated fatty acids, ethoxylated amines, ethoxylated alcohols, alkyl and nonyl-phenol ethoxylates, etc.). Other suitable optional additives include, for example, phase transition agents (e.g., tetraalkylammonium salts, polyethylene glycol, and crown ethers). Where optional additives are present, the developer may contain about 5 weight% or less of the additive, about 2.5 weight% or less of the additive in further embodiments, and about 1 weight% or less of the additive in further embodiments. Those skilled in the art will recognize that additional ranges of additive concentrations within the specified ranges are considered and are within the scope of this disclosure. Additives may be selected to improve contrast, sensitivity, and linewidth roughness. Additives in the developer may also inhibit the formation and precipitation of metal oxide particles.

[0035] The developer can be applied to the patterned coating material using an appropriate approach. For example, the developer can be sprayed onto the patterned coating material, or the structure can be immersed in the developer or otherwise submerged. Additionally, spin coating can be used. For automated processing, a puddle method can be used to pour the developer onto the coating material in a fixed manner. If desired, the development process can be completed using spin rinsing and / or drying. After the image is developed, the coating material is placed on the substrate as a pattern.

[0036] Development can be performed using any suitable processing method, such as spray coating, puddle dipping, etc. For commercial production, development is generally adapted to the provided process equipment. Development may be performed for about 2 seconds to about 30 minutes, about 3 seconds to about 15 minutes in additional embodiments, about 4 seconds to about 10 minutes in other embodiments, and about 5 seconds to about 5 minutes in additional embodiments. Those skilled in the art will recognize that additional scopes are considered within the specified scopes and are within the scope of this disclosure.

[0037] Depending on the design of the coating material, there may be a large contrast in material properties between the irradiated area where the coating material is concentrated and the unirradiated area of ​​the coating material containing substantially intact photosensitive ligands, such as organic and / or carboxylates. This high contrast in material properties further facilitates the formation of high-resolution lines with smooth edges in the pattern after development, as described in the examples.

[0038] After completing a developing step including any optional rinsing, the coating material may be heat-treated to further concentrate the material, further dehydrate, densify, or remove residual developer from the material. This heat treatment may be particularly desirable in embodiments where the oxide coating material is incorporated into the final device, but may also be desirable in some embodiments where the coating material is used as a resist and ultimately removed if stabilization of the coating material is required to facilitate further patterning. In particular, baking of the patterned coating material may be performed under conditions where the patterned coating material exhibits a desired level of etching selectivity. In some embodiments, the patterned coating material may be heated to a temperature of about 100 °C to about 600 °C, in further embodiments to about 175 °C to about 500 °C, and in further embodiments to about 200 °C to about 400 °C. Heating may be performed for at least about 1 minute, for about 2 minutes to about 1 hour in other embodiments, and for about 2.5 minutes to about 25 minutes in additional embodiments. Heating may be performed in air, vacuum, or an inert gas atmosphere such as Ar or N2. Those skilled in the art will recognize that additional ranges of temperature and time for heat treatment are considered and are within the scope of the disclosure within the specified ranges. Likewise, non-heat treatments, including blanket UV baking or exposure to an oxidizing plasma such as O2, may also be used for similar purposes.

[0039] In some embodiments, adjacent linear segments of neighboring structures may have an average pitch (half pitch) of about 60 nm (30 nm half pitch) or less, in some embodiments an average pitch (half pitch) of about 50 nm (25 nm half pitch) or less, and in additional embodiments an average pitch (half pitch) of about 34 nm (17 nm half pitch) or less. Pitch can be evaluated by design and can be verified by scanning electron microscopy (SEM), such as top-down images. As used herein, pitch refers to a spatial period or the center-to-center distance of repeating structural elements, and as is commonly used in the art, a half-pitch is half of the pitch. The feature dimensions of the pattern may be described in relation to the average width of the feature and are generally evaluated away from edges, etc. Additionally, the feature may refer to gaps between material elements and / or material elements. In some embodiments, the average width may be about 25 nm or less, in additional embodiments about 20 nm or less, and in additional embodiments about 15 nm or less. Those skilled in the art will recognize that additional ranges of pitch and average width are considered within the specified ranges and are contained within this disclosure. Based on this process, patterning can generally be adapted for the formation of various devices, such as electronic integrated circuits, through a repeated patterning process to form a suitably layered structure, such as a transistor or other component.

[0040] Wafer throughput is a factor that substantially limits the implementation of EUV lithography in high-volume semiconductor manufacturing and is directly related to the dose required to pattern a given feature. However, although chemical strategies exist to reduce imaging dose, a negative correlation between the imaging dose required to print target features and feature size uniformity (e.g., LWR) is typically observed in EUV photoresists with feature sizes and pitches < 50 nm, which limits end-device operability and wafer yield. Photoresist sensitivity can be expressed as dose-to-gel values, and imaging dose requirements can be evaluated by forming an array of exposed pads with exposure times stepwise from pad to pad to change exposure dosing. The film can then be developed, and the thickness of the remaining resist can be evaluated for all pads, for example, using spectroscopic ellipsometry. The measured thickness can be normalized to the maximum measured resist thickness and plotted against the logarithm of the exposure dose to form a characteristic curve. The maximum slope of the normalized thickness versus log dose curve is relative to the photoresist ( γ The dose value where the tangent drawn through this point is 1, defined as ) is the gel dose of the photoresist ( D g It is defined as ). In this way, the common parameters used for photoresist characterization are [Mack, C. Fundamental Principles of Optical Lithography It can be approximated according to John Wiley & Sons, Chichester, UK; pp 271-272, 2007.

[0041] In some embodiments, the average linewidth roughness may be about 5.5 nm or less, in some embodiments about 5 nm or less, and in additional embodiments about 4.5 nm or less. As shown in the examples, linewidth roughness may be evaluated as a function of critical dimensions. The evaluation of linewidth roughness is performed by top-down SEM image analysis to derive the 3σ deviation from the average linewidth. The average includes high-frequency roughness and low-frequency roughness, i.e., short correlation length and long correlation length, respectively. While the linewidth roughness of the organic resist is mainly characterized by long correlation length, the organometallic coating material of the present invention exhibits significantly short correlation length. In the pattern transfer process, short correlation roughness is smoothed during the etching process, which can produce a pattern of much higher fidelity. Those skilled in the art will recognize that additional ranges of linewidth roughness are considered within the specified ranges and are contained within the present disclosure. Based on this process, patterning can generally be adapted for the formation of various devices, such as electronic integrated circuits, through a repeated patterning process to form appropriately layered structures, such as transistors or other components.

[0042] In some embodiments, it may be advantageous to perform an additional rinsing process immediately after the developing process. The rinsing process may generally involve contacting the developed material with an additional composition comprising a solvent or a solvent and an additive. As shown in the examples, a rinsing step using only a solvent tends to result in a greater number of defects. A rinsing step using the improved developing composition described above, which includes an additive, may result in a reduction of defects and / or greater uniformity of the process results. In some embodiments, the rinsing composition may be one of the same compositions described above with respect to the developer. For example, the rinsing composition may include a solvent selected from ketones, esters, ethers, or mixtures thereof, and the additive may include a carboxylic acid, a monohydroxyl alcohol, a polyol, such as a diol, pyrrolidone, a lactone, or a mixture thereof. Specifically, for example, the rinsing composition may include a mixture of 2-heptanone and 1,2-hexanediol. Generally, the rinsing process may be similar to the developing process, except that significant coating removal occurs during the developing stage, whereas generally little material is removed during the rinsing stage. The rinsing process may be performed for about 1 second to about 20 minutes, in additional embodiments about 2 seconds to about 12 minutes, and in additional embodiments about 4 seconds to about 6 minutes. Those skilled in the art will recognize that additional scopes are considered within the specified scopes and are within the scope of this disclosure.

[0043] An alternative rinsing approach was described by the applicants in U.S. Patent Application No. 16 / 654,080 by Kocsis et al., incorporated herein by reference under the title “Patterned Organometallic Photoresists and Methods of Patterning.” The rinsing in the aforementioned '080 application was based on an aqueous rinsing agent that tended to dissolve the exposed portions of the pattern to some extent, but the rinsing process was controlled to limit the dissolution of these exposed portions of the pattern. This rinsing was effective in removing defects and also altered the pattern dimensions. This treatment can be considered an alternative to the current improved development treatment. In the rinsing described in the examples, defects generally tended to increase in the rinsing process using only the base solvent. Rinsing using the improved developer resulted in a further reduction of defects.

[0044] Current processing methods are designed to reduce or eliminate defects, such as microbridging defects as well as larger defects. For the uniformity of data comparison, bridge defects, particle defects, and large cluster defects are not considered in the evaluation of microbridging defects. Bridging differs from microbridging in that it can be determined by the critical number of pixels in the y-direction—if the size threshold is exceeded, it is considered a bridge and excluded. In most, though not all, cases of manual review of these images reveal other mechanisms, such as fall-on type particles. This contributes relatively little to the total number of defects compared to the number of microbridges (<10% for high-defect samples, such as control wafers). An automated process is used for the evaluation of microdefects as described in the examples below. Basically, a series of critical dimension scanning electron microscope images are taken. The CD-SEM images are analyzed using analysis software that counts the number of microbridges. An example of a microbridge is shown in the SEM micrograph in Fig. 1.

[0045] To aid in the evaluation of development, wafers were patterned to assess pattern formation as a function of EUV dose. Primarily, imaging is considered as a step function of the illuminated and unilluminated regions. For the patterning of very fine features, such as lines and spaces, and for development to achieve low feature roughness using selected developers, Esize is determined by evaluating the actual feature width and pitch for a specific dose; this can be defined as the dose at which the target feature is generated using a mask designed to achieve the feature width. Since strong developers can etch more along the edges for a specific illumination dose, and weak developers can correspondingly etch less along the edges, the developed linewidth varies depending on the developer. Although wafers were exposed across the EUV dose range to generate a field grid of nominally identical features, they were illuminated with different EUV doses to produce correspondingly different feature sizes. The dose at which the target linewidth is achieved is called Esize, and accordingly, the dose unit (mJ / cm²) 2 It has ). The value of Esize is presented in an example evaluating the performance of a developer under specified process conditions.

[0046] Generally, since illumination time can vary depending on the dose, it is desirable to lower the EUV dose to improve throughput through EUV scanning components. However, the trade-off between patterning dose and feature roughness and / or ultimate resolution is well known, in which pattern quality generally improves as the dose increases. As can be seen in the examples below, the developer, process conditions, and related methods described herein can enable a larger process window than conventional ketone developers, namely a dose range with sufficiently low LWR and / or low defects. Furthermore, according to the present invention, specific process conditions and developer compositions can be selected to reduce LWR and / or patterned defects without an incidental increase in dose. The patterning composition itself is designed to have lower dose requirements. Therefore, the design of the improved patterning composition and method, such as that used herein, achieves excellent pattern quality with appropriate dose values.

[0047] After patterning, the patterned material can be used for further processing, such as deposition of material into the gaps of the patterned material and / or etching to remove the substrate material between the gaps of the patterned material. Afterward, the patterned resist material can be removed after further processing using a suitable etchant composition, such as a diluted base or BCl3 plasma. Functional components are formed by frequently repeating the above process to form a stack of patterned layers.

[0048] In the following examples, data demonstrating the improved performance of the composition of the present invention are presented. A representative organotin photoresist composition described in U.S. Patent No. 10,228,618B2 by Meyers et al. was used to examine the efficacy of the composition in this disclosure. The performance of the present invention was found to be improved compared to standard ketone solvent developer compositions. Using the newly discovered developer, patterns and structures with improved pattern defect density, particularly microbridging defects, can be formed at the same dimensions.

[0049] Examples

[0050] Example 1. Patterning using various developer compositions

[0051] This example presents gel dose values ​​obtained by exposing an organotin oxide hydroxide photoresist to EUV radiation and developing it in an organic solvent having various Hansen solubility parameters.

[0052] A tert-butyltin oxide hydroxide photoresist film was deposited from a precursor coating solution prepared by combining separately prepared solutions of t-BuSn(NEt2)3 and Sn(NMe2)4 in dry 4-methyl-2-pentanol, as described in Example 4 of the '618 patent cited above, to obtain a final solution containing a mixture of 0.04 M t-BuSn(NEt2)3 and 0.01 M Sn(NMe2)4. A thin film for EUV contrast curves was deposited on a silicon wafer (100 mm diameter) with a natural oxide surface. The Si wafer was treated with HMDS (hexamethyldisilazane) vapor priming prior to deposition. The precursor coating solution was spin-coated in air at 1000 rpm and baked on an air hot plate at 100 °C for 2 minutes to remove residual solvent and volatile hydrolysis products. The film thickness after coating and baking was measured to be ~25 nm using the ellipsometric method.

[0053] A linear array of 50 circular pads with a diameter of ~500 μm was exposed to each wafer with EUV light using the Lawrence Berkeley National Laboratory Micro Exposure Tool. Pad exposure times were set in exponential 8% steps, with EUV doses delivered to each pad ranging from 0.63 to 27.39 mJ cm⁻¹. -2 It was adjusted stepwise. After exposure, the wafer underwent post-exposure baking (PEB) on an air hot plate at 160 °C for 2 minutes. The exposed film was then immersed in one of the developer solvents listed in Table 1 for 30 seconds to form a negative tone image, i.e., to remove unexposed portions of the coating. The film was removed from the developer solvent and dried using a nitrogen gun. After development, a final hot plate baking was performed in air at 150 °C for 5 minutes. The residual thickness of the exposed pads was measured using a JA Woollam M-2000 spectroscopic ellipsometer. The measured thickness for each pad was then normalized by the maximum measured resist thickness and plotted against the logarithm of the exposure dose to form characteristic curves for each developer solvent. The maximum slope of the normalized thickness versus log dose curve was relative to the photoresist ( γ Defined as ) and dose value D 1 It occurs in. Photoresist gel dose ( D g ) is slope γ D 1 The tangent drawn through is defined as a capacitance where it is 1. In this way, common parameters used for photoresist characterization are described in the literature [Mack, C. Fundamental Principles of Optical Lithography It was approximated according to John Wiley & Sons, Chichester, UK; pp 271-272, 2007.

[0054] D g The actual interpretation is the dose required to achieve the maximum development rate change for the unexposed material. For the same organotin oxide hydroxide photoresist film developed in each solvent of Table 1 D g The values ​​are plotted in Figure 2 for the total polarity term for each solvent, defined as the sum of the Hansen δP and δH parameters. The Hansen solubility parameter is based on the literature [Wypych, G. Knovel Solvent Database Taken from [ , ChemTec Publishing, 2008, 2012]. D g A clear positive correlation is observed between the total polarity term—therefore, highly polar developer solvents are generally stronger developers and have higher polarity for fixed resists and processes. D g It was revealed to be approaching.

[0055]

[0056] Example 2: Preparation of organotin coating resist and evaluation of patterning performance

[0057] In this example, the developer composition t Prepared by in situ hydrolysis of a precursor solution containing BuSnX3 and MeSnX3 compounds t BuSnO (3 / 2-(x / 2)) (OH) x and MeSnO (3 / 2-(x / 2)) (OH) x It is described in the context of organotin oxide hydroxide photoresist films containing a mixture of, and the evaluation of these films is generally described to provide context for the results of the following examples.

[0058] The photoresist precursor coating solution was prepared using a method similar to that described in published U.S. Patent Application No. 2019 / 0391486 by Jiang et al. 20 mol% MeSn(O t Amyl)3 and 80 mol% t BuSn(O t Amyl)3(here O tA mixture containing amyl (2-methylbutane-2-oxide) was prepared under an inert atmosphere and subsequently dissolved in 4-methyl-2-pentanol. The H2O concentration was pre-adjusted to ~300 ppm. The final Sn concentration of the precursor coating solution was 0.05 M.

[0059] A flowchart of the procedure for analyzing wafer patterning for defects is presented in Fig. 3. The precursor solution described above was used to coat an organotin oxide hydroxide photoresist film using a TEL CLEAN TRACK LITHIUS Pro Z coater / developer. A silicon wafer (300 mm diameter) was coated with a 10 nm SOG (spin-on-glass) underlayer (ISX328, JSR) and baked in air at 220 °C for 1 minute prior to photoresist coating. A solvent pre-wetting step using a PGME / PGMEA mixture was used to improve photoresist coating uniformity. The resist precursor coating solution was spin-coated onto the underlayer-coated substrate in air at 1394 rpm. After edge bead rinsing (EBR) and back rinsing, the coated wafer was subjected to post-coating baking (PAB) on an air hot plate at 100 °C for 1 minute. Edge bead rinsing is generally described in Waller et al.’s published U.S. patent application No. 2018 / 0046086, titled “Methods of Reducing Metal Residue in Edge Bead Region From Metal-Containing Resists,” which is incorporated herein by reference.

[0060] Following PAB, the resist-coated substrate was exposed to extreme ultraviolet light using an ASML NXE:3300B scanner with a numerical aperture of 0.33 and dipole 90x illumination. A pattern of identical vertical 16 nm lines and spatial (16p32) was projected onto the coated wafer with a fixed focus, and a field of approximately annular arrays was generated around the center of the wafer in steps after each exposure at various exposure doses. Then, the exposed resist film and substrate were returned to the coater / developer track and processed by hot plate PEB in air at a specified temperature for 1 minute. After PEB, the exposed film was developed using a puddle development recipe set for 15 seconds with a selected developer, followed by dynamic rinsing with a rinse solution for an additional 10 seconds to form a negative tone image, i.e., the unexposed portion of the coating was removed. After development, a final hot plate baking was performed in air at 150 °C for 1 minute.

[0061] Using a Hitachi CG5000 CD-SEM, the average linewidth (critical dimension, CD) and 3-sigma linewidth roughness (LWR) for each field were measured along the scanned dose range. From these data, the dose for size (Esize, i.e., the dose required to print a CD with the same line as 1 / 2 pitch) and the four fields spanning the target CD selected for microbridge defect analysis were calculated by fitting to the CD-dose curve.

[0062] Microbridge defect analysis was performed by collecting 150 CD-SEM images from identical intrinsic locations within each of the four selected line-space fields on each wafer. Images were acquired in TV scans at a magnification of approximately 164,000X and an FOV of 0.823 x 0.823 μm with an acceleration voltage of 800 V and a beam current of 8 pA. 16 frames were averaged per image. Using Stochalis image analysis software (developed by the Interuniversity Microelectronics Centre, Belgium and available), the total number of microbridge defects detected in each set of 150 images (operated in "space" mode, parameter settings: SumLines = 4, Smoothing = 7; CD_Threshold = 61; DThres_nok_Space = 90; and Lnok_min_Space = 4) and each field were tabulated. EUV image analysis was performed [P. De Bisschop and E. Hendrickx "Stochastic effects in EUV lithography", Proc. SPIE 10583, Extreme Ultraviolet (EUV) Lithography IX, 105831K (19 March 2018); https: / / doi.org / 10.1117 / 12.2300541] is further described, and the said document is incorporated herein by reference. By fitting an exponential function to the defect density for CD curves constructed from four data points per wafer, the microbridge defect density at the target CD (16 or 18 nm as specified in each example) across the measured range was estimated and reported as defects per 150 images or "Def / Die 16p32". The total sampling area for each 150 images is approximately 101.6 square micrometers.The results are presented for a developer containing a formic acid additive in FIGS. 5, 6, and 13; for a developer containing an EHA additive in FIGS. 8 and 9; for a developer containing an acetic acid additive in FIGS. 15; for a developer containing a diol additive in FIGS. 16; for a developer containing water as an additive in FIGS. 17; and for an alcohol developer composition in FIGS. 21. FIGS. 13 illustrates the defect density of the 18p32 pattern for a developer containing a formic acid additive. FIGS. 15 illustrates the defect density of the 18p32 pattern for a developer containing an acetic acid additive. FIGS. 18 illustrates the effect of rinsing conditions on the defect density. The disclosed additives and process conditions were shown to generally reduce the defect density of patterned organotin oxide hydroxide photoresists.

[0063] Example 3. 2-Heptanone and formic acid developer composition

[0064] This example shows improved results obtained using a developer that uses a mixture of 2-heptanone and formic acid, where a significant reduction in microbridge defects is obtained.

[0065] Conventional 2-heptanone developer composition ( HFO ) was used as a control developer and rinsing solvent.

[0066] 98% (w / w) pure formic acid and 2-heptanone 2 volume% formic acid ( D1 ) or 10 volume% formic acid ( D2A set of developer solutions was prepared by mixing at the concentration of ). Before formulation, all glassware was rinsed three times with isopropyl alcohol and baked at 100°C for 24 hours. Preparation was carried out in a fume hood. A known volume of 2-heptanone (EUVR developer, TOK America, 100-300 ppm H2O) was dispensed into glass vials. Then, the volume required to achieve the target volume% of formic acid (obtained from EMD Millipore) was added to the vials using a graduated cylinder or automatic pipette, the mixture was closed, and the vials were placed on a shaker table to be thoroughly mixed.

[0067] Wafers were coated, exposed, developed, rinsed, and analyzed according to Example 2. The rinsing solvent was 2-heptanone. All samples were tested at baking temperatures after exposure of 140 °C, 160 °C, and 180 °C. Samples ( D1 ) and ( D2 ) was additionally tested at a baking temperature after exposure to 120 ℃. The next sample ( HD0 ), ( D1 ) or ( D2 ) was developed. After development, a final 60-second hot plate baking was performed in air at 150°C.

[0068] Assuming the formic acid stock contains 2 wt% water, a 10 volume% formic acid developer solution contains about 3000 ppm of water, and a 2 volume% formic acid developer solution contains about 600 ppm of water. Although the individual effect of water in the 2-heptanone solution is examined in Example 9, the effect of water in these developer samples is much less than the effect of formic acid.

[0069] Developer composition ( D2 )of 1 H NMR analysis was performed at the time of preparation and after 1 week, and the 2-heptanone control developer of Example 2 ( HFOIt was compared with ). See Fig. 4. The NMR results suggest that the developer composition was stable over a period of one week.

[0070] Figure 5 provides microbridge density as a function of baking temperature after exposure. D1 ) and ( D2 2-heptanone control developer at all tested PEB temperatures for ) HFO A decrease in microbridge density is observed compared to ).

[0071] Figure 6 shows the microbridge density as a function of equivalent dose density. Although the dose appears to increase at the set PEB, defects can be reduced at the equivalent dose for size (Esize) by changing the process PEB temperature.

[0072] Figure 7 provides nanometer-scale linewidth roughness (LWR) as a function of equivalent dose with respect to size. Although the dose was shown to increase at the set PEB, the equivalent dose density (mJ / cm²) was increased by changing the process PEB temperature. 2 LWR can be reduced in ).

[0073] The data in Figures 5 through 7 suggest that a developer composition containing 2-10% (volume) of formic acid additive in 2-heptanone can reduce microbridge density in organotin resist patterns having 16 nm lines with a 32 nm pitch compared to a 2-heptanone control developer composition. The reduction in microbridge density is approximately 10-fold at the equivalent dose density and up to 100-fold with a significant dose penalty. Increased additive concentrations were shown to reduce microbridging density at a given PEB temperature or dose density. Increased additive concentrations were generally shown to reduce linewidth roughness at a given dose density.

[0074] Example 4. 2-Heptanone and 2-Ethylhexanoic Acid (EHA) Developer Composition

[0075] This example shows improved results obtained with a developer solution using a blend of 2-heptanone and ethylhexanoic acid, where a significant reduction in microbridge defects is obtained.

[0076] Conventional 2-heptanone developer composition ( HFO ) was used as a control developer and rinsing solvent.

[0077] Pure ethylhexanoic acid (EHA) and 2 volume% EHA ( D3 ) or 10 volume% EHA( D4 A set of developer solutions was prepared by mixing 2-heptanone of the following concentrations. Before formulation, all glassware was rinsed three times with isopropyl alcohol and baked at 100°C for 24 hours. Preparation was carried out in a fume hood. A known volume of 2-heptanone (EUVR developer, TOK America, 100-300 ppm H2O) was dispensed into glass vials. Then, the volume required to achieve the target volume% of EHA (Alfa Aesar) was added to the vials using a graduated cylinder or automatic pipette, the mixture was closed, and the vials were placed on a shaker table to be thoroughly mixed.

[0078] The wafers were coated, exposed, developed, rinsed, and analyzed according to Example 2. The rinsing solvent was 2-heptanone. All samples were tested at baking temperatures after exposure of 140 °C, 160 °C, and 180 °C. The samples were then ( HD0 ), ( D3 ) or ( D4 ) was developed. After development, a final 60-second hot plate baking was performed in air at 150°C.

[0079] Figure 8 provides the microbridge density at a target CD of 16 nm as a function of baking temperature after exposure. D3 ) and ( D4) 2-heptanone control developer at various tested PEB temperatures ( HFO A decrease in microbridge density was observed compared to ).

[0080] Fig. 9 shows the control group developer ( HF0 ) and ( D3 ), ( D4 It provides microbridge density as a function of equivalent dose density for ). Although the dose at the set PEB was found to increase, defects at the equivalent dose density can be reduced by moving the process PEB temperature.

[0081] The data in Figures 8 and 9 suggest that a developer composition containing 2-10% (vol) of EHA additive in 2-heptanone can reduce the microbridge density in organotin oxide hydroxide photoresist patterns having 16 nm lines with a 32 nm pitch compared to a 2-heptanone control developer composition. A clear decrease in microbridge density is observed for the EHA-containing developer at baking temperatures after exposure of 160 °C and 180 °C. In Figure 7, the developer ( D4 The microbridge density at a PEB temperature of 180 ℃ using ) is the control developer ( HF0 It can be seen that it is about 1 / 3 of the microbridge density observed in ). Additionally, the decrease in microbridge density at the baking temperature after exposure to 120 °C is due to a developer having a higher EHA concentration (10 v volume% EHA) ( D4 It is observed in ). Figure 8 shows that increasing the additive concentration from 2 vol% to 10 vol% reduces the microbridge density at a given dose density. While an increase in additive concentration generally correlated with a decrease in defect density, the results demonstrate that the performance of the developer compared to the control developer is influenced by the PEB temperature and dose density. 140 °C or approximately 42 mJ / cm² 2 In this study, 2 volume% EHA did not demonstrate better performance than the control developer. The positive effects of the EHA additive are particularly pronounced at low dose densities. For example, 2 volume% EHA was 30 mJ / cm². 2Approximately 1.7 times the control developer at a dose density of ( HFO Provides defect reduction compared to ).

[0082] Example 5. Organic solvent and carboxylic acid developer composition

[0083] This example describes the preparation of a developer solution composed of 2-heptanone or PGMEA containing formic acid or acetic acid as an additive.

[0084] Before formulation, all glassware was rinsed three times with isopropyl alcohol and baked at 100°C for 24 hours. Preparation was carried out in a fume hood. Known volumes of 2-heptanone (EUVR developer, TOK America, 100-300 ppm H2O) or propylene glycol methyl ether acetate (PGMEA, Sigma Aldrich, 99.5%) were dispensed into glass vials. Then, the volume required to achieve the target volume% of acetic acid (JT Baker, 99.5-100.5%) or formic acid (98-100%, EMD) was added to the vials using a graduated cylinder or automatic pipette, the mixture was closed, and placed on a shaker table to be thoroughly mixed. Solutions listed in Table 2 were produced by varying the volume percentage of the added additives relative to the total volume of the added acid and added solvent.

[0085]

[0086] Examples 6 and 7 present performance data for the developer compositions of Table 2.

[0087] Example 6. Microbridging analysis from formic acid containing the developer of Example 5

[0088] This example demonstrates improved patterning performance obtained by using formic acid as a developer additive. Organotin oxide hydroxide photoresist patterns are developed with a developer solution using a mixture of 2-heptanone or PGMEA and formic acid prepared as in Example 5. 16p32 images were analyzed for dose and LWR relative to size, and patterns for four fields were evaluated for the number of defects based on a CD of 18 nm, as described in Example 2. For the developer containing the formic acid additive, generally improved performance was observed compared to the control developer, and significant improvement was observed as the line CD increased.

[0089] An organotin oxide hydroxide photoresist precursor solution and a photoresist film were prepared and treated as described in Example 2 to form a line-space pattern with a pitch of 32 nm. Developer solution of Example 5 HF0-HF4 and PF0-PF4 It was used in the development stage, and all wafers were rinsed with the corresponding developer solution for 10 seconds. While all 8 developer solutions were tested on wafers treated with PEB at 180 °C, the developer solution ( HF0 ), ( PF0 ), ( HF2 ) and ( PF2 Tested only on wafers treated with PEB at 140°C and 160°C.

[0090] Representative SEM images of 32 nm pitch patterns generated in a field having an average line CD closest to 16 nm are presented in Figs. 10 and 11 for each wafer. At the same PEB temperature, the developer solution containing the formic acid additive is E sizeIt was found to cause a decrease in LWR compared to the control solvent, which correlates with an increase in field dose nearly identical to that and an increase in the volume percentage of formic acid in the developer solution. This trend is evident even to the naked eye, as the patterns developed in solutions containing higher concentrations of formic acid additives show reduced scumming and microbridge formation even in small areas sampled in a single image. As shown in Fig. 11, a developer solution composed of pure PGMEA at 180 °C PEB ( PF0 ) cannot consistently separate line-space patterns, and a large number of particle, bridge, and scum defects are evident. However, when a formic acid additive is present in the same solvent, image fidelity is significantly improved at all tested additive levels.

[0091] For all 12 wafers presented in FIG. 10, the LWR values ​​for all fields having a measurable line CD are plotted as a function of CD in FIG. 12a to FIG. 12c for PEB temperatures of 140 °C, 160 °C, and 180 °C, respectively. For each PEB temperature at low CD (and simultaneously low dose), all developer solutions are the aforementioned control developer ( PF0 Except for ), it appears to have a similar effect on LWR. In general, a slight decrease in LWR at fixed CD is due to the control developer ( PF0 ) and ( HF0 Compared to the phenomenon using ) ( PF2 ) and ( HF2 It is observed at ). However, strong divergence is observed as line CD approaches 16 nm. PF0 ) and ( HF0 While the LWR of the manifested pattern among ) begins to increase rapidly, ( PF2 ) and ( HF2The LWR of the developed pattern among them slowly continues to decrease, and then eventually begins to increase as the CD exceeds ~18 nm. Within the range where the lithography process window is constrained by LWR and edge placement error (as there may be many advanced node applications), larger line sizes printable before this "LWR take-off" present an expansion of the effective process window of the photoresist when developed with a formic acid-containing developer solution as disclosed herein.

[0092] The microbridge defect analysis methodology described in Example 2 was similarly applied to the 18 wafers described above. A target CD of 18 nm was selected, and 150 SEM images were collected and analyzed in each of the four fields using a CD near the target value for each wafer. After image analysis, defect counting, and curve fitting, estimates of the number of defects per 150 images for the 18p32 pattern were generated for each process condition. In Fig. 13, the logarithm (base 10) of this number is E size Plot against (16p32). The developer solution used for each wafer is indicated by a data point label. For 180 ℃ PEB conditions, the control developer ( PF0, HF0 The lowest volume% of formic acid tested compared to ) PF1, HF1 A strong decrease (~50x) in the number of microbridge defects is clearly observed even in ). The number of defects is (volume % of formic acid) HF1 ) and ( PF1 )at ( HF4 ) and ( PF4 It continues to decrease as ) increases. In all tested PEBs, ( HF2 ) and ( PF2 )Is ( HF0 ) and ( PF0 It was found that defects decreased by nearly a similar amount (~500x) compared to ). In particular, this significant defect reduction in fixed PEB was due to the increased dose (E sizeWhile it occurs under ), under nearly identical dose conditions (i.e., HF0 , 160 ℃ and HF1 , 180 ℃ or HF0 , 140 ℃ and HF2 A comparison of data points and trend lines at , 180 ℃) shows that using developer additives with different process conditions can substantially reduce defects at the same nominal dose.

[0093] Example 7. Analysis of microbridging from the acetic acid-containing developer of Example 5

[0094] This example demonstrates improved patterning performance obtained by using acetic acid as a developer additive. An organotin oxide hydroxide photoresist pattern is developed with a developer solution using a blend of PGMEA and acetic acid. High-resolution imaging is used to demonstrate that microbridge defects are significantly reduced.

[0095] An organotin oxide hydroxide photoresist precursor solution and a photoresist film were prepared and treated as described in Example 2 to form a line-space pattern with a pitch of 32 nm. Developer solution PA1-PA5 was prepared by mixing acetic acid and PGMEA as described in Example 5. All wafers were rinsed with the corresponding developer solution. Each of the five developer solutions was tested on wafers treated with PEB at 170 °C, 180 °C, and 190 °C, and ( PF0 It was compared with ).

[0096] Representative SEM images of 32 nm pitch patterns generated in a field having an average line CD closest to 16 nm are shown for each wafer in Fig. 14. At the same PEB temperature, the developer solution containing the acetic acid additive is E size It was found to cause a decrease in LWR compared to the control solvent, which is correlated with an increase in field dose almost identical to that and an increase in the volume percentage of acetic acid in the developer solution.

[0097] The microbridge defect analysis methodology described in Example 2 was applied to line-space patterns generated using an 18 nm target CD. The logarithm of the number of fitted defects per 150 images for the 18p32 field of each wafer is E in Fig. 15. size Plot against (16p32). The developer solution used for each wafer is indicated by the data point label. For all PEB conditions, 1 volume% acetic acid in PGMEA ( PA1 In ) as well, the control group developer ( PF0 A strong reduction (~50x) in the number of microbridge defects is observed again compared to ). As seen in formic acid, the volume % of acetic acid is 10% ( PA5 As it increases by ), the number of defects continues to decrease. Likewise, under nearly identical dose conditions (i.e., PF0 , 170 ℃ and PA1 A comparison of data points and trend lines at , 190 ℃) shows that using acetic acid additives with other process conditions can substantially reduce defects without dose penalty.

[0098] Example 8. 2-Heptanone and Diol Developer Composition

[0099] This example demonstrates improved patterning performance obtained by using a diol as a developer additive. An organotin oxide hydroxide photoresist pattern is developed with a developer solution using a mixture of 2-heptanone and 1,6-hexanediol, 1,2-hexanediol, or 1,2-propanediol. High-resolution imaging is used to demonstrate that microbridge defects are significantly reduced.

[0100] Before formulating the developer solution, all glassware was rinsed three times with isopropyl alcohol and baked at 100°C for 24 hours. Preparation was carried out in a fume hood. 2 mass% 1,6-hexanediol solution ( HD1) was prepared by adding 10 g of 1,6-hexanediol (TCI) to a 500 mL volumetric flask containing ~400 mL of 2-heptanone (TOK). The flask was stirred until dissolution was complete, and then diluted to a total volume of 500 mL using additional 2-heptanone.

[0101] A known volume of 2-heptanone (EUVR developer, TOK America, 100-300 ppm H2O) was dispensed into a glass vial. Then, the volume required to achieve the target volume% of diol was added to the vial using a graduated cylinder or automatic pipette, each solution was stopped and placed on a shaker table to be thoroughly mixed.

[0102] Developer solution ( HD2 ) was prepared by mixing 1,2-hexanediol (TCI, 96%) with 2-heptanone at 2 volume%.

[0103] A 10 volume% 1,2-propanediol solution was prepared by mixing 450 mL of 2-heptanone (TOK) and 50 mL of 1,2-propanediol (Sigma-Aldrich) in a 500 mL Kimax bottle ( HD3 Once fully combined, the solution was divided into two 240 mL amber glass bottles.

[0104] The wafer was coated, exposed, developed, rinsed, and analyzed according to Example 2.

[0105] (HD1) The samples were tested at baking temperatures after exposure at 120 ℃, 140 ℃, 160 ℃, and 180 ℃, and then developed and rinsed with the prepared developer composition.

[0106] (HD2) The samples were tested at baking temperatures after exposure at 120 ℃, 160 ℃, and 180 ℃, and then developed and rinsed with the prepared developer composition.

[0107] (HD3)The samples were tested at baking temperatures after exposure at 120 ℃, 140 ℃, 160 ℃, and 180 ℃, and then developed and rinsed with the prepared developer composition.

[0108] Processed at a baking temperature after exposure to 160 ℃ ( HFO A control water wafer that was developed and rinsed in ) was also prepared.

[0109] The microbridge defect analysis methodology described in Example 2 was applied to line-space patterns generated using a 16 nm target CD. The number of fitted defects per 150 images for the 16p32 field of each wafer is E in Fig. 16. size Plot for (at 16p32). The PEB temperature of each wafer is indicated by data point labels. For wafers processed at 160 ℃ PEB, a clear reduction in defects was observed in the developer solution containing diol additives, and the developer solution ( HD3 )(2 volume% 1,2-propanol in 2-heptanone) is ( HD0 It shows the largest decrease (~3x) compared to ). HD3 )-The developed wafer in all cases ( HD1 ) and ( HD2 As it was found that the defect rate was lower than that observed in ), this trend continued at all tested PEB temperatures.

[0110] Example 9. 2-Heptanone and water developer composition

[0111] This example demonstrates improved patterning performance obtained by using water as a developer additive. An organotin hydroxide photoresist pattern is developed with a developer using a mixture of 2-heptanone and water. High-resolution imaging is used to demonstrate that microbridge defects are significantly reduced.

[0112] A developer solution composed of a mixture of 2-heptanone and H2O was prepared by mixing 18 MΩ DI H2O with 2-heptanone (determined to have a trace H2O concentration of 150 ppm) to achieve the final concentrations presented in Table 3. The final concentrations were confirmed by Karl Fischer titration.

[0113]

[0114] An organotin oxide hydroxide photoresist precursor solution and a photoresist film were prepared and treated as described in Example 2 to form a line-space pattern with a pitch of 32 nm. Developer solution HW1-HW4 was utilized in the development stage. All wafers were rinsed with the corresponding developer solution. Each of the four developer solutions was tested on PEB-treated wafers at 140 °C, 160 °C, and 180 °C. The number of fitted defects per 150 images for the 16p32 field of each wafer is E in Fig. 17. size Plot for (16p32). The PEB temperature for each wafer is indicated by data point labels. A tendency for defects to decrease as H2O concentration increases is observed in all tested PEBs, and the solution ( HW4 A wafer developed with )(8000ppm) is in a solution containing only a trace amount (150 ppm) of H2O ( HW1 Compared to wafers developed with ), microbridge density decreases by 2–3 times and < 2 mJ / cm² 2 It represents the general dose shift of.

[0115] Example 10. 2-Heptanone and 1,2-Hexanediol developers under various rinsing conditions

[0116] This example demonstrates improved patterning performance obtained by rinsing with a diol additive solution. Organotin oxide hydroxide photoresist patterns were developed with a developer solution of 2-heptanone and 2 volume% 1,2-hexanediol and rinsed under various rinsing conditions. Using high-resolution imaging, it was shown that rinsing with a developer solution containing a diol additive resulted in a further reduction of microbridge defects compared to rinsing with a developer solution without a diol additive. Before formulating the developer solution, all glassware was rinsed three times with isopropyl alcohol and baked at 100 °C for 24 hours. Preparation was carried out in a fume hood. A known volume of 2-heptanone (EUVR developer, TOK America, 100-300 ppm H2O) was dispensed into glass vials. Next, add the volume required to achieve 2 volume% of 1,2-hexanediol additive to the bottle using a graduated cylinder or automatic pipette to obtain the developer solution ( HD2 Prepared ). Closed the stopper on the solution and placed it on a shaker table to mix thoroughly.

[0117] The wafer was coated, exposed, developed, rinsed, and analyzed according to Example 2.

[0118] 12 wafer samples were processed at a baking temperature after exposure to 160°C, and then a developer ( HD2 It was developed using ). After development, one of four rinsing conditions was used.

[0119] The microbridge defect analysis methodology described in Example 2 was applied to line space patterns generated using a 16 nm objective CD. The average number of fitted defects per 150 images for the 16p32 field is plotted for three wafers under each washing condition in Fig. 18. Error bars in the figure represent ±1 standard deviation. Rinsing condition (A) was no rinsing. Rinsing condition (B) was solution ( HFO It was a 10-second rinse using ). Rinsing condition (C) was solution ( HD2It was a 10-second rinse using ). Rinsing condition (C) was solution ( HFO Rinse with ) for 10 seconds, then the solution ( HD2 It was to rinse for 10 seconds with ).

[0120] As shown in FIG. 18, rinsing condition (B) exhibited the highest defect rate and variability. Rinsing condition (D), ( HF2 After using ) ( HD2 It was found that a two-step rinse using ) reduced defects compared to rinse condition (B). Rinse condition (A), which eliminated the rinse step, reduced defects by more than three times compared to rinse condition (B), with a relative standard deviation of ~14%. Rinse condition (C) showed the lowest number of defects per 150 images as well as a reduction in wafer-to-wafer variability.

[0121] This example is a diol-containing solution ( HD2 It shows that a single rinse using ) reduces microbridging defects compared to rinsing with a solvent without diol additives or not rinsing. Surprisingly, a single rinse using a diol-containing solution rinse, the diol-containing solution ( HD2 ) had much fewer defects than the 2-step rinse used as the second step.

[0122] Example 11. PGMEA and acetic acid developer composition having a hexagonal columnar pattern

[0123] This example demonstrates improved performance of hexagonal columnar patterning obtained using acetic acid as a developer additive. An organotin oxide hydroxide photoresist hexagonal columnar pattern is developed with a developer solution using a blend of PGMEA and acetic acid and compared with a 2-heptanone developer solution. High-resolution imaging is used to demonstrate that microbridge defects are significantly reduced.

[0124] The organotin oxide hydroxide photoresist precursor solution and photoresist film were prepared and processed as described in Example 2, except that the following process was used to expose and pattern the resist-coated substrate following the PAB. A hexagonal array of 23 nm diameter pillars with a 40 nm vertical and 70 nm horizontal pitch was projected onto each coated wafer with a fixed focus using an ASML NXE:3300B EUV scanner with a numerical aperture of 0.33 and hexapole illumination. After each exposure, a field of an approximately annular array was generated around the center of the wafer in steps at various exposure doses.

[0125] Developer solution ( PA2 ) and ( PA3 ) was prepared by mixing acetic acid and PGMEA as described in Example 5. All wafers were rinsed with the corresponding developer solution.

[0126] The developer solution was tested on wafers treated with PEB at 170 ℃ and 190 ℃, and ( HF0 It was compared with ). Fig. 19 shows (a) baking temperature after exposure at 170 ℃ and (HFO) Developer, (b) baking temperature after exposure at 190°C and (PA2) Developer, (c) baking temperature after exposure at 190°C and (PA3) E for wafers treated with a developer size Representative SEM images are shown near (e.g., ~23 nm CD). Fig. 19 also shows E sizeThe values ​​and image pillar 3-sigma LCDU (three times the standard deviation of the measured pillar CD per image) are presented. Microbridges between pillars were detected using the image analysis software KOLONA (stylized KOLONA) by adjusting the microbridge defect analysis methodology described in Example 2. The average number of fitted defects per 150 images for 27 nm CD pillar patterns printed on a 70x40y hexagonal grid is listed for each condition in Fig. 19. Treated with PEB at 190 °C and solution (PA2) and (PA3) The wafer developed in is processed at 170 ℃ and the solution (HF0) It was found to exhibit substantially lower microbridge defects and image LCDU compared to wafers developed by. (HFO) Compared to that, the patterning performance improved by acetic acid containing a developer was obtained even at lower doses for smaller sizes.

[0127] Example 12. Primary and secondary alcohol developer composition

[0128] This example demonstrates improved patterning performance obtained using primary and secondary alcohols as developer solutions. Organotin oxide hydroxide photoresist patterns are developed with the primary alcohol n-propanol and the secondary alcohol propylene glycol methyl ether and compared with 2-heptanone. High-resolution imaging is used to demonstrate that microbridge defects are significantly reduced.

[0129] The wafer was coated, exposed, developed, rinsed, and analyzed according to Example 2.

[0130] pure solvent (HFO) (2-Heptanone, TOK America), (PGO) (propylene glycol methyl ether, PGME, Sigma Aldrich, 99.5%) and (NP0)A developer solution containing n-propanol (Sigma Aldrich 99.5%) was used exactly as received from the manufacturer. Each of the three developer solutions was tested on wafers treated with PEB at 160 °C and 180 °C. All wafers were rinsed with the corresponding developer solution. The microbridge defect analysis methodology described in Example 2 was applied to a line-space pattern generated using an 18 nm objective CD printed by overexposure of a projected 16p32 image.

[0131] Representative SEM images captured in a 32 nm pitch pattern field with an average line CD closest to 16 nm are presented for each wafer in Fig. 20. Projected dose (mJ cm⁻¹) -2 ), line CD (nm) and LWR (nm) are presented for each image. Generally (PG0) and (NP0) In all (HF0) Compared to, improvements in line quality are observed in terms of increased dose, reduced LWR, and visible microbridge defects in an equivalent PEB. (NP0) While the developer solution shows a greater increase in dose, (PG0) It was shown that the line width roughness was improved more significantly. Figure 21 shows a log-linear plot of the number of defects counted per field for 18 nm lines with a 32 nm pitch as a function of the dose required to print a 16p32 pattern in the same reticle field. (PG0) and (NP0) Both of them (HF0) A decrease in the number of microbridges is observed at both 160 °C and 180 °C compared to. Under approximately the same dose conditions (e.g., 180 °C, NP0 vs. 160 ℃, HF0 )at (NP0)It was found that the microbridge density decreased by more than 5 times. These results indicate that patterning performance is improved when using primary and secondary alcohols compared to 2-heptanone, and primary alcohols as developers substantially reduced microbridging defects.

[0132] The above embodiments are exemplary and not limiting. Additional embodiments are also included within the scope of the claims and the concept of the invention. Furthermore, while the invention has been described by reference to specific embodiments, those skilled in the art will recognize that changes in form and detail may be made without departing from the spirit and scope of the invention. The references cited above are limited to not include any subject matter expressly contrary to the content of this application. To the extent that a specific structure, composition, and / or process is described by a component, element, component, or other part in this specification, the disclosure of this application is understood to encompass, unless specifically stated otherwise, embodiments comprising a specific embodiment, a specific component, element, component, other part, or combination thereof, as well as embodiments essentially composed of such specific embodiment, specific component, element, component, other part, or combination thereof, which may include additional features that do not alter the fundamental nature of the subject matter as presented in this discussion. The use of the term "about" in this specification refers to a measurement error for a specific parameter unless otherwise explicitly indicated.

Claims

Claim 1 A method for forming a developed patterned layer comprising an organometallic oxide / hydroxide network by developing a radiation-exposed organometallic patterning layer, wherein the method comprises contacting the radiation-exposed organometallic patterning layer with a developer composition comprising a solvent blend containing an ester and a carboxylic acid. Claim 2 A method according to claim 1, wherein the radiation-exposed organometallic patterning layer has an initial dry thickness before contacting the developer composition, and the initial dry thickness is 1 nm to 50 nm. Claim 3 In claim 1, the organometallic oxide / hydroxide network is formula R z SnO (2-(z / 2)-(x / 2)) (OH) x It is represented as (where 0 < z ≤ 2 and 0 < (z + x) ≤ 4, and R is a hydrocarbyl group forming a carbon bond with a tin atom), or hydrocarbyl ligands R N A method comprising a combination of compositions having, wherein N is a tin composition incorporated into an oxo / hydroxo network. Claim 4 A method according to claim 1, wherein the carboxylic acid comprises acetic acid, formic acid, oxalic acid, 2-ethylhexanoic acid, or a combination thereof. Claim 5 A method according to claim 1, wherein the carboxylic acid comprises acetic acid. Claim 6 The method according to claim 1, wherein the ester is polypropylene glycol methyl ether acetate (PGMEA), ethyl acetate, ethyl lactate, n-butyl acetate, butyrolactone, or a combination thereof. Claim 7 A method according to claim 1, wherein the ester comprises PGMEA. Claim 8 A method according to claim 1, wherein the solvent blend comprises PGMEA and acetic acid. Claim 9 A method according to claim 1, wherein the solvent blend further comprises an aromatic compound, a ketone, an ether, a monohydroxyl alcohol, a polyhydroxyl compound, a pyrrolidone, a glycol ether, a diol, or a mixture thereof. Claim 10 The method of claim 1, wherein the solvent blend comprises 2-butanol, 2-heptanol, propylene glycol methyl ether, methanol, ethanol, propanol, isobutanol, pentanol, ethylene glycol, propylene glycol, glycerol, 2-pyrrolidone, 1-ethyl-2-pyrrolidone, N-methyl-2-pyrrolidone, ethylene glycol monomethyl ether, formic acid, acetic acid, oxalic acid, 2-ethylhexanoic acid, 1,2-hexanediol, 1,2-propanediol, 1,3-propanediol, or a combination thereof. Claim 11 The method of claim 1, wherein the solvent blend comprises 4-methyl-2-pentanol, cyclohexyl acetate, dibutyl oxalate, benzene, xylene, toluene, propylene glycol monomethyl ether acetate, ethyl acetate, ethyl lactate, n-butyl acetate, butyrolactone, methyl ethyl ketone, acetone, 2-butanone, cyclohexanone, 2-heptanone, 2-octanone, tetrahydrofuran, dioxane, anisole, or a combination thereof. Claim 12 A method according to claim 1, wherein the solvent blend further comprises 2-heptanone. Claim 13 A method according to any one of claims 1 to 12, wherein the solvent blend comprises 0.25 volume% to 45 volume% of a carboxylic acid and 55 volume% or more of an ester. Claim 14 A method according to any one of claims 1 to 12, wherein the solvent blend comprises 2-heptanone, PGMEA, or both, and 2 to 20 volume% of acetic acid. Claim 15 A method according to any one of claims 1 to 12, wherein the solvent blend comprises a water additive of up to 10 volume%. Claim 16 A method according to any one of claims 1 to 12, wherein the contact comprises forming a negative tone image. Claim 17 A method according to any one of claims 1 to 12, wherein the contact comprises carrying the developer composition as a reaction gas. Claim 18 A method according to any one of claims 1 to 12, wherein the contact step is performed using a puddle method, a dip method, a spin coating method, or a spray method, and the developer composition is applied to the surface of a radiation-exposed organometallic patterning layer and dried by spinning, blowing, or both for a selected time period of 2 seconds to 30 minutes. Claim 19 A method according to any one of claims 1 to 12, further comprising the step of rinsing the developed pattern with a rinsing solution, wherein the rinsing solution comprises the developer composition. Claim 20 A method according to any one of claims 1 to 12, further comprising the step of baking the radiation-exposed organometallic patterning layer at a temperature of 120°C to 190°C (PEB temperature) before contacting the radiation-exposed organometallic patterning layer with the developer composition. Claim 21 A method according to claim 20, wherein the contact duration, PEB temperature, and developer composition are selected such that the developed patterned structure has a desired feature size or a desired defect rate or both. Claim 22 A method according to claim 21, wherein the contact duration is 2 seconds to 30 minutes. Claim 23 A method according to any one of claims 1 to 12, wherein the developed patterned layer has an average linewidth roughness of 5 nm or less. Claim 24 A method according to any one of claims 1 to 12, wherein the developer composition further comprises one or more characteristic-modifying additives, and the developer composition comprises 5% by weight or less of an additive. Claim 25 A method according to any one of claims 1 to 12, further comprising the step of forming a radiation-exposed organometallic patterning layer, wherein the step of coating a substrate with an organometallic radiation-sensitive organometallic resist material to form a radiation-sensitive organometallic resist layer on the surface of the substrate, and the step of exposing the radiation-sensitive organometallic resist layer to extreme ultraviolet light using a patterned mask to form a radiation-exposed organometallic patterning layer comprising an exposed portion and an unexposed portion. Claim 26 In paragraph 25, the step of exposing the radiation-sensitive organometallic resist layer is 100 mJ / cm² 2 A method comprising irradiating extreme ultraviolet rays at a dose lower than or equal to the following. Claim 27 A method according to any one of claims 1 to 12, further comprising: a step of depositing a material based on the developed patterned layer or a step of etching a substrate based on the developed patterned layer; and a step of removing the developed patterned layer to form a treated substrate.

Citation Information

Patent Citations

  • Developer for resist

    JP2000029228A

  • Developing or rinsing solution for lithography and pattern forming method using the same

    JP2013257379A

  • Organic solvent developer

    JP2015031923A

  • Organotin oxide hydroxide patterning compositions, precursors, and patterning

    KR1020180054917A

  • Developer compositions and methods of forming photolithographic patterns

    US20120219755A1