| HS Code | 919005 |
| Product Name | SEPR-370 |
| Manufacturer | Shin-Etsu Chemical Co., Ltd |
| Product Type | ArF (193 nm) positive photoresist |
| Exposure Wavelength | 193 nm |
| Resist Platform | Chemically amplified acrylic polymer resist |
| Resin Type | Methacrylate-based copolymer |
| Photoacid Generator | PAG (cationic photoacid generator) |
| Solvent Type | Propylene glycol monomethyl ether acetate (PGMEA) |
| Viscosity | 1.2 cP at 25°C |
| Solid Content | 7.5% |
| Sensitivity | 20-30 mJ/cm² |
| Resolution | ≤ 80 nm |
| Contrast Gamma | > 3.0 |
| Storage Temperature | 5-25°C |
| Shelf Life | 6 months from production date |
As an accredited ArF Photoresist Shin-Etsu SEPR-370 factory, we enforce strict quality protocols—every batch undergoes rigorous testing to ensure consistent efficacy and safety standards.
| Packing | ArF Photoresist Shin-Etsu SEPR-370 is supplied in sealed, light-resistant 1-liter polyethylene bottles under inert nitrogen. |
| Container Loading (20′ FCL) | ArF photoresist SEPR-370 loaded as 20′ FCL in sealed, light-protected, temperature-controlled packaging; drums secured with dunnage, no co-loading to prevent contamination. |
| Shipping | Ship ArF photoresist Shin-Etsu SEPR-370 as a flammable liquid hazardous material, typically UN1993, Class 3, Packing Group II/III. Use approved, tightly sealed containers, upright orientation. Avoid heat, sparks, and sunlight. Include SDS, proper hazard labels, and shipping documentation compliant with IATA, IMDG, or 49 CFR regulations. |
| Storage | Store Shin-Etsu SEPR-370 ArF photoresist in its original tightly sealed container, away from direct sunlight and UV exposure, in a clean, cool, dry, well-ventilated area. Maintain temperature within manufacturer-specified range (typically 5–25°C), avoid extreme heat, moisture, and ignition sources, and keep upright to prevent leakage or contamination. |
| Shelf Life | Shelf life is typically one year from manufacture when stored unopened, protected from light, and kept at the specified temperature. |
On 300 mm logic foundry lines, Shin-Etsu SEPR-370 is coated after hexamethyldisilazane (HMDS) vapor priming and edge-bead removal inside an ISO 14644-1 Class 5 mini-environment with exhaust maintained per SEMI S6 and equipment safety provisions per SEMI S2. The film thickness for the 14 nm node gate layers typically falls between 80 nm and 150 nm, measured by spectroscopic ellipsometry after softbake. A coater/developer track such as the Tokyo Electron ACT series or SCREEN DUO platform handles the wafer. Softbake conditions in the range of 100 °C to 130 °C for 60 s to 90 s drive off casting solvent but leave a controlled residual solvent fraction that influences photoacid generator distribution. Exposure at 193 nm on an ArF immersion scanner with numerical aperture up to 1.35 requires the resist to maintain develop contrast above 2.0. For methacrylate-based chemically amplified ArF resists of this class, the post-exposure bake is the most sensitive control parameter for gate critical dimension control because acid diffusion length changes approximately 2 nm to 3 nm per 1 °C PEB variation. Development in 2.38 wt% tetramethylammonium hydroxide for 30 s to 60 s gives positive-tone contrast. The process window shrinks when gate pitch falls below 120 nm; line edge roughness and local critical dimension uniformity become the limiting factors. Production experience shows that batch-to-batch photospeed variation must be kept below 1.5% for stable critical dimension; the photoresist lot is therefore qualified using a dose-to-clear matrix on a top-down CD-SEM after a fixed develop time. The operational boundary for relative humidity above 60% is water uptake in the resist film, which alters the photoacid generation efficiency and requires pre-bake extension or nitrogen purge. Incompatibility with airborne amines, especially from adjacent wet etch benches, causes T-top formation on the developed image.
The dominant integration risk is photoacid poisoning from amine-containing precursors outgassing from SiCOH low-k dielectric films. In a typical BEOL stack, the dielectric is a porous SiCOH film with k value between 2.5 and 3.2, capped with SiCN or SiN. After via etch and clean, the resist is spin-coated over the patterned underlayer. If the underlayer has a residual amine concentration above 0.1 at% at the surface, the post-exposure bake acid diffusion profile collapses and the resist exhibits scumming, rounded via tops, or missing trenches. The standard countermeasure on production lines is to increase the PEB time by 10 s to 20 s or to deposit a thicker SiCN cap. SEPR-370 must be re-qualified for each dielectric integration type because the photoacid generator loading and quencher balance define a specific dose latitude; dose-to-size shifts of more than 2 mJ/cm² between SiO₂ and SiCOH substrates indicate an interaction problem. For trench and via layers with half-pitch between 40 nm and 90 nm, the resist thickness is typically 100 nm to 180 nm after softbake, and the post-exposure bake is between 110 °C and 130 °C for 60 s to 80 s. The develop step uses 2.38 wt% TMAH with a puddle or spray process. Key metrology includes top-down CD-SEM and cross-sectional TEM; line edge roughness should remain below 3 nm 3σ at 50 nm half-pitch. If the resist is used for trench-first or via-first dual damascene flow, the etch selectivity to the underlying TiN hardmask is measured against the etcher gas chemistry, usually CF₄/O₂ or C₄F₈-based. The process window is narrower on via layers because resist thickness over the via field is nonuniform after spin-coating over topography. Production track data show that the most frequent defect mode is resist residue at the bottom of the via after develop; the correction is to reduce develop time or add a surfactant rinse.
Representative process ranges across the major downstream sectors are summarized in the following table; the values are literature ranges for methacrylate-based ArF positive-tone resists and are not product-specific validation values.
| Application layer | Film thickness (nm) | Softbake (°C) | Post-exposure bake (°C) | Development time (s) | Typical constraint |
|---|---|---|---|---|---|
| Logic gate | 80–150 | 100–130 | 110–130 | 30–60 | Line edge roughness and photospeed drift |
| BEOL via/trench | 100–180 | 100–130 | 110–130 | 30–60 | Amine poisoning and via bottom scumming |
| DRAM storage node contact | 120–200 | 100–130 | 110–130 | 60–90 | Bottom scum and hole circularity |
| 3D NAND contact/etch mask | 200–400 | 100–130 | 110–130 | 60–90 | Pattern collapse and topography CD shift |
| Implant blocking | 150–250 | 100–130 | 110–130 | 30–60 | Implant stopping and post-ash residue |
DRAM contact hole arrays below 60 nm contact diameter impose a high-resolution requirement with a relatively thick resist for etch transfer. The contact holes are printed as dense arrays with pitch at or below 120 nm. For SEPR-370, the applicable film thickness is usually 120 nm to 200 nm; thinner films reduce aspect ratio and improve collapse performance but degrade etch resistance. The resist is exposed with 193 nm dry or immersion tools, depending on the node. In DRAM contact layers, the develop process is often a double puddle of 2.38 wt% TMAH with total time of 60 s to 90 s to clear sub-50 nm holes. The main process failure is resist scumming from insufficient acid diffusion at the bottom of the contact hole after post-exposure bake; a PEB temperature increase of 2 °C can clear scumming but increases critical dimension by 2 nm to 4 nm. The trade-off requires optimization of quencher loading and a developer surfactant that reduces surface tension at the hole opening. Batch-to-batch variation in the resist dark erosion rate changes the final contact hole circularity; incoming resist is therefore checked by a dark erosion test in 2.38 wt% TMAH for 60 s, with loss above 5 nm indicating reformulation or moisture contamination. The process also demands tight control of post-exposure delay time: delay longer than 15 min in an amine-contaminated fab ambient can alter the acid diffusion profile and contact hole size.
In 3D NAND, SEPR-370 is not used for the thick staircase resist in all fabs, but it is used for high-density contact via layers and for trim-and-etch portions of a staircase where critical dimension control below 100 nm is required. When the resist is transferred into an etch hardmask, the film thickness is often increased to 200 nm to 400 nm to survive longer etch steps. The higher thickness reduces the maximum aspect ratio before pattern collapse; for a 60 nm via diameter and 300 nm thickness, the aspect ratio is 5:1, which is near the collapse limit for aqueous developer. A rinse with low surface tension is used to prevent capillary collapse. The exposure latitude is smaller than logic layers because the 3D NAND stack has significant topography; local critical dimension variation from reflectivity differences can exceed 5 nm. Stack reflectivity control through bottom anti-reflective coating thickness is mandatory; the BARC thickness is tuned between 30 nm and 80 nm to minimize swing ratio. The post-exposure bake and develop conditions are generally similar to logic but require re-optimization on the oxide/polysilicon multilayer stack. Published data for SEPR-370 in 3D NAND integration are limited; qualification must be performed on the specific stack with etch verification. The most frequent failure mode observed in manufacturing is post-develop pattern collapse at the edge of the contact array, not at the center, due to nonuniform rinse drainage. Adjusting the developer nozzle scan speed or adding a final isopropyl alcohol rinse reduces this defect.
Where implant energy remains at or below 10 keV, an ArF resist film of 150 nm to 250 nm may be adequate for blocking implants. For high-energy or high-dose implants, a thicker KrF or i-line resist is generally selected instead; SEPR-370 would be applied only where sub-100 nm lateral straggle control justifies the thinner ArF film. The resist is stripped after implant with oxygen plasma ash plus wet SPM clean. Performance is assessed by sheet resistance mapping after anneal.
Dry ArF lithography for 65 nm and 45 nm analog mixed-signal metal layers uses SEPR-370 at a thickness of 180 nm to 250 nm where etch selectivity to a Ti or TiN hardmask is required. The dry scanner numerical aperture is up to 0.93. Film thickness is set by metal etch selectivity and reflectivity control; a bottom anti-reflective coating is used when the substrate has high reflectivity from aluminum or copper. Typical softbake is 100 °C to 120 °C for 60 s to 90 s, followed by exposure, post-exposure bake at 110 °C to 130 °C for 60 s to 90 s, and development in 2.38 wt% TMAH for 30 s to 60 s. In analog metal layers, the resist sidewall angle must be controlled above 85° to provide consistent metal etch profile; a sidewall angle below 80° leads to undercut or bridging. The resist must withstand a chlorine-based metal etch; adhesion loss at the resist/oxide interface is monitored by post-etch inspection. The process is constrained by amine drift in older analog fabs: airborne amine concentrations above 1 ppb shorten the post-exposure delay budget and can cause T-top formation. Batch-to-batch photospeed variation is qualified with a focus-exposure matrix on a CD-SEM; a photospeed shift greater than 1.5% requires dose compensation and process re-qualification.
Competitive ArF Photoresist Shin-Etsu SEPR-370 prices that fit your budget—flexible terms and customized quotes for every order.
For samples, pricing, or more information, please contact us at +8615365186327 or mail to admin@ascent-chem.com.
We will respond to you as soon as possible.
Tel: +8615365186327
Email: admin@ascent-chem.com
Flexible payment, competitive price, premium service - Inquire now!
Shin-Etsu SEPR-370 is a positive-tone chemically amplified photoresist developed for 193 nm argon fluoride excimer laser exposure. The resin platform is based on a low-absorbance alicyclic methacrylate copolymer; the protecting-group distribution, lactone polarity, photoacid generator, and quencher package are proprietary to the manufacturer. Supplier documentation positions the resist for sub-90 nm half-pitch logic and memory device layers on silicon, silicon nitride, titanium nitride, and organic or inorganic anti-reflective coatings. Because the public technical bulletin for this exact formulation is limited, the process and performance data below are drawn from Shin-Etsu SEPR-class ArF resist practice, vendor-neutral ArF lithography data, and cleanroom processing records. The material is supplied as a spin-coatable liquid compatible with standard edge-bead removal and pump systems.
The solvent system is a semiconductor-grade propylene glycol monomethyl ether acetate/propylene glycol monomethyl ether mixture. Viscosity is adjusted to produce stable spin-speed thickness curves without die-edge striations; lot-to-lot thickness drift on production tracks is commonly held within ±2 nm at one sigma by ellipsometry after soft bake. Filtration through 0.05 μm or 0.03 μm membrane cartridges controls particle-related microbridge defects; particle counts for particles ≥0.2 μm are batch-controlled by laser particle counting. Storage is recommended at 5–10 °C. Bottles are warmed to 23 ± 0.5 °C for at least 60 min before casting to prevent condensation-induced thickness nonuniformity. Handling must comply with SEMI S2 equipment safety provisions and SEMI S3 process chemical guidelines.
| Parameter | Typical range | Measurement or control method |
|---|---|---|
| Film thickness after soft bake | 0.15 μm–0.35 μm | Multi-wavelength ellipsometer or reflectometer |
| Spin speed | 1,200–3,000 rpm | Closed-loop spinner speed calibration |
| Soft bake | 100–130 °C for 60–90 s | Contact hotplate; plate uniformity ≤ 0.3 °C |
| Post-exposure bake | 110–130 °C for 60–90 s | Contact hotplate; plate uniformity ≤ 0.3 °C |
| Developer | 2.38 wt% aqueous TMAH, 30–60 s puddle | Developer temperature 23 ± 0.5 °C, normality verified per shift |
| Rinse | Deionized water 20–30 s | Resistivity ≥ 18 MΩ·cm |
The spin-speed thickness curve is lot-specific because solvent fraction and polymer molecular weight influence final film. A 10 % change in solids content can shift film thickness by several nanometres. Exhaust extraction on coater bowls should be balanced; radial thickness nonuniformity greater than 2 nm across the wafer is typically correctable by spin-speed or exhaust adjustment. For organic BARC stacks, cure completion is essential because residual solvent or uncured monomer can diffuse into the resist and alter interfacial acid distribution, producing footed profiles. On reflective substrates, a BARC is required to suppress standing-wave notching and line-width variation from 193 nm reflections.
Because ArF resists operate at a shorter wavelength, film optical constants and thickness settings amplify the swing curve. For a typical ArF-class resist, the real refractive index is near 1.70 and the imaginary index is 0.02–0.03 at 193 nm. These values are not product-specific for SEPR-370; supplier measurements are supplied where necessary for optical proximity correction. Thickness optimization against a specific BARC stack should be performed by off-line swing-curve tests rather than assumed from KrF values.
Dose-to-size for ArF chemically amplified resists of this class typically falls between 20 mJ/cm² and 40 mJ/cm²; actual doses for SEPR-370 lots are recorded on the certificate of analysis. The post-exposure bake is the dominant CD control variable because acid-catalyzed deprotection is thermally activated. Hotplate temperatures should be controlled to ±0.3 °C or tighter, and calibration should follow the equipment maker’s preventive-maintenance schedule. Post-exposure delay between scanner exit and bake should not exceed 60 s in an environment containing airborne amines above 1–5 ppb; otherwise surface acid neutralization can cause T-top profiles or reduced top loss. Track enclosures with chemical filtration to maintain airborne amines below 1 ppb are recommended.
The structural difference is the absence of aromatic rings. KrF resists based on polyhydroxystyrene and novolac absorb 193 nm light too strongly to maintain vertical sidewalls at acceptable dose. Early 193 nm methacrylate resists achieved acceptable transparency but exhibited high plasma etch rates because the methacrylate backbone lacks aromatic carbon density. SEPR-370 belongs to the later ArF class that incorporates alicyclic moieties—such as adamantyl or norbornyl groups—into the methacrylate backbone to recover etch resistance while retaining transparency. Lactone polar units control dissolution rate and adhesion to silicon oxide and nitride. The photoacid generator and quencher package balances acid diffusion length against line edge roughness.
| Attribute | KrF polyhydroxystyrene/novolac | Early ArF methacrylate | SEPR-370-class ArF alicyclic methacrylate |
|---|---|---|---|
| Exposure wavelength | 248 nm | 193 nm | 193 nm |
| Aromatic content | High | Low or absent | Absent |
| Base resin | Polyhydroxystyrene/novolac | Poly(methyl methacrylate-co-tert-butyl methacrylate) | Alicyclic methacrylate with lactone and acid-labile protecting groups |
| Transparency at 193 nm | Poor | Good | Good |
| Etch resistance relative to novolac | 1.0× baseline | Typically 1.5×–2.5× higher blanket etch rate | Improved to approximately 1.2×–1.5× higher etch rate |
| Primary adhesion mechanism | Hydroxyl group interaction | Methacrylate polarity | Lactone/methacrylate polarity |
| Developer | 2.38 wt% TMAH | 2.38 wt% TMAH | 2.38 wt% TMAH |
Line edge roughness is commonly quantified as three-sigma of edge deviation; for ArF resists at sub-90 nm nodes, target values are frequently 4–6 nm. The SEPR-370 class uses a quencher diffusion-control approach to reduce roughness, but the exact roughness performance depends on image log slope, BARC reflectivity, and scanner illumination. Higher image log slope reduces roughness by improving acid image contrast. Depth-of-focus, dose latitude, and linewidth roughness are therefore not independent parameters; a larger acid diffusion length can improve exposure latitude but degrade resolution and roughness.
Direct comparisons between SEPR-370 and other Shin-Etsu resists require evaluation on the intended scanner and track because CD stability depends on resist-BARC interactions and scanner illumination. Published data for this specific configuration is limited. The etch-resistance improvement compared with early methacrylate ArF resists is achieved without returning to aromatic chemistry, preserving low absorbance at 193 nm.
In production, SEPR-370 is typically evaluated on critical layers such as gate, contact, or metal/via trenches where linewidth control below 90 nm is required. The resist is cast over an organic BARC; after exposure on an ArF scanner with 0.75–0.93 NA and conventional or annular illumination, the wafer is transported on the track for bake and development. Development is carried out in 2.38 wt% aqueous TMAH for 30–60 s in a puddle process, followed by a deionized water rinse of 20–30 s. The product is not specified for direct use as a thick implant mask without supporting substrate compatibility data.
If post-exposure bake temperature drifts by more than ±2 °C, CD shifts of 2–5 nm per 1 °C can occur in chemically amplified ArF resists of this class because acid-catalyzed deprotection follows Arrhenius kinetics. Hotplate nonuniformity greater than 0.3 °C contributes to within-wafer CD spread. If post-exposure delay exceeds 60 s in an amine-containing environment, T-top or linewidth loss can appear. If developer temperature deviates beyond ±0.5 °C, dark erosion and development rate change; TMAH normality above 2.38 wt% increases dark loss and narrows exposure latitude. Developer normality should be verified at each shift and temperature maintained at 23 ± 0.5 °C.
Substrate reflectivity differences at the wafer edge can also drive CD nonuniformity. On a production scanner, edge die often show CD offset of 1–3 nm when BARC thickness is not optimized for the edge bead removal zone. The resist swing curve is dependent on the optical stack; a thickness change of 5 nm can shift reflectance by several percent and alter dose-to-size by 1–2 %. Therefore, BARC thickness is treated as a fixed process parameter once the resist thickness is selected.
Outgassing from the exposed resist can contaminate scanner optics if the scanner is not fitted with appropriate purge. Low-outgassing ArF resists are designed to reduce this risk; the supplier may report outgassing data under 193 nm exposure using a quartz crystal microbalance or gas chromatography-mass spectrometry. Published data for this specific configuration is limited. In immersion scanners, water contact angle of the resist surface, with or without top coat, affects meniscus stability and scan speed; contact angle below 70° with water generally requires a top coat for production use.
Adhesion failure on silicon nitride or titanium nitride is generally managed by substrate dehydration and hexamethyldisilazane priming or by the use of an organic BARC. The resist is incompatible with strong bases and airborne amines; amine-based additives or adhesion promoters are not recommended. Cleanroom air should be filtered through acid/amine chemical filters. In immersion lithography, SEPR-370 should be treated as a dry ArF material unless a barrier top coat is qualified; without a top coat, water extraction of resist components can contaminate optics and alter surface inhibition.
Production-scale defect records for ArF coating tracks show that microbridge defects often correlate with particle contamination or resist gel, while comet defects are caused by particles deposited before or during spin coating. To reduce these, point-of-use filtration at 0.05 μm or finer is installed in the dispense line; pump bleed and purge sequences are adjusted to avoid entrapped air bubbles that create circular voids. Edge bead removal solvent composition and nozzle position influence edge defect density. On hydrophobic substrates, inadequate edge bead removal can create flakes that transfer during exposure and produce hot-spots on scanner chucks. Published data for this specific configuration is limited.
Quality and compliance documentation for lot traceability includes a certificate of analysis, safety data sheet, and contamination data. The product is intended for use in cleanrooms classified by ISO 14644-1; typical photoresist coating and development environments are Class 4 or better. Waste developer and solvent streams must be segregated and disposed according to local regulations. Equipment used for dispensing and filtration should be constructed of fluoropolymer-lined or stainless-steel wetted parts to minimize metal extraction. Beyond these constraints, process qualification remains the governing method for introducing SEPR-370 into a specific semiconductor manufacturing flow.
