Jingrui Electric Material supplies an i-line positive-tone photoresist based on a cresol novolac resin, a diazonaphthoquinone-4-sulfonate photoactive compound, and propylene glycol monomethyl ether acetate as the primary casting solvent. Model designations correspond to viscosity and film thickness class; model-specific datasheets should be consulted before lot acceptance because publicly released data for the Jingrui Electric Material i-line configuration are limited. The material is intended for 365 nm exposure on silicon, silicon dioxide, silicon nitride, aluminium, and selected barrier layers in semiconductor and MEMS lithography. Spin-coated film thickness typically falls between 1.0 µm and 3.5 µm at 2000–4000 rpm, depending on viscosity grade and ambient temperature. Bulk metal specifications for commercial i-line resists of this class commonly limit total trace metals to less than 50 µg/L by ICP-MS, with individual metal limits below 10 µg/L for sodium, potassium, iron, chromium, and copper. The product is filtered during manufacture and is dispensed through 0.1 µm absolute filtration on automated coat tracks.

The material is a clear, amber to reddish liquid with a closed-cup flash point governed primarily by PGMEA; a typical value is 42 °C. Storage at 5–25 °C in sealed, light-tight containers is required, and shelf life is commonly 12 months from the date of manufacture when stored under these conditions. Dispense systems should use an all-fluoropolymer fluid path with a 0.1 µm polytetrafluoroethylene membrane filter. Dispense pressure is maintained between 0.1 MPa and 0.3 MPa; pressure spikes above 0.5 MPa can shear the novolac polymer and alter viscosity by more than 3%. Particle counts measured after filtration are typically below 50 particles/mL at a 0.5 µm detection threshold using a laser particle counter calibrated to ISO 21501-4.

What Distinguishes a 365 nm DNQ/Novolac Resist from g-Line and KrF Platforms?

The primary functional difference is the exposure wavelength and amplification mechanism. A DNQ/novolac resist relies on photochemical conversion of diazonaphthoquinone to indene carboxylic acid at 365 nm, creating base-soluble regions without catalytic amplification. In contrast, KrF resists at 248 nm are typically chemically amplified systems using photoacid generators and protected polyhydroxystyrene. g-Line resists at 436 nm share the DNQ/novolac chemistry but exhibit lower resolution because of longer wavelength. Under contact or proximity exposure, i-line resists achieve 0.35–0.50 µm line/space resolution, while g-Line products are generally limited to 0.50–0.80 µm on equivalent tooling. The i-line resist also has a higher DNQ extinction coefficient at the exposure wavelength, requiring adjusted solids content and novolac molecular weight to control sidewall angle and scum formation. Unlike chemically amplified resists, the Jingrui Electric Material i-line DNQ/novolac product does not require immediate post-exposure bake to activate acid catalysis; however, post-exposure bake is used to reduce standing wave artifacts. Published data for the Jingrui Electric Material i-line grade are limited; these differences are representative of the product class and may vary by formulation.

Low-viscosity grades are designed for films in the 0.8–1.5 µm range, while high-viscosity grades support single-coat thicknesses from 5 µm to 15 µm. The main formulation differences are cresol novolac molecular weight, solids content, and solvent composition. Low-viscosity grades often contain ethyl lactate or butyl acetate as a co-solvent to improve planarization, while thick-film grades use slower-evaporating solvents to reduce edge bead. These formulation changes affect film shrinkage, contrast, and thermal flow temperature. Compared with polyimide resists or epoxy-based SU-8, the DNQ/novolac i-line product is selected when stripping is required after pattern transfer because novolac-based masks are removed more readily in heated NMP or oxygen plasma. This difference is operationally significant; SU-8 crosslinks and cannot be stripped once hardbaked, whereas novolac-based i-line resists retain solubility in organic strippers before hardbake and can be removed by plasma ashing after hardbake.

Coating uniformity on 200 mm wafers is evaluated by spectral reflectometry or ellipsometry using a 49-point polar map; total thickness variation after softbake typically remains below 30 nm for films in the 1.2–1.8 µm range. Softbake on a calibrated hotplate at 100 °C for 60 s removes 70–80% of residual PGMEA, leaving residual solvent levels below 2 wt%. If softbake temperature exceeds 120 °C, partial thermal decomposition of the DNQ component can reduce photospeed and broaden the development window. In production, hotplate temperature uniformity of ±1 °C across the contact surface is used to maintain critical dimension uniformity below 0.02 µm.

Adhesion promotion on oxide and nitride substrates is normally carried out with hexamethyldisilazane vapour at 110–120 °C for 30–60 s in a sealed chamber. The contact angle of deionized water after priming typically increases to 55–70°, indicating surface silylation. Insufficient HMDS coverage produces undercut and lifted lines during development; excessive coverage can increase surface hydrophobicity beyond 75° and cause resist dewetting. Liquid amine primers are avoided because they can alter dissolution selectivity and produce residue at the substrate interface.

Exposure Dose, Focus Latitude, and Standing Wave Control

Process windows are established by measuring linewidth linearity across a dose matrix from 60 mJ/cm² to 180 mJ/cm² and a focus offset range of ±5 µm on a contact aligner. A mask aligner with a 365 nm bandpass filter and intensity uniformity of ±3% across the exposure field is used to reduce linewidth non-uniformity. Without a bandpass filter, longer-wavelength g-line and h-line content can contribute to exposure, degrading resolution and process latitude. For a 1.0 µm film, the optimum dose typically falls between 100 mJ/cm² and 140 mJ/cm². Post-exposure bake at 110–120 °C for 60–90 s reduces standing wave ridge formation by thermally diffusing partially converted DNQ species. Where post-exposure bake is omitted, sidewall roughness can increase by 20–40% and the usable depth of focus narrows by approximately 0.5 µm. The Jingrui Electric Material i-line product should be qualified using a focus-exposure matrix because published data for this specific configuration is limited.

Standing wave ridge amplitude can be measured by cross-sectional scanning electron microscopy or atomic force microscopy. For a 1.0 µm film on silicon, post-exposure bake typically reduces ridge amplitude from 50–80 nm to 10–20 nm. The contrast curve is derived from normalized film thickness after development versus log dose; a high-contrast resist shows a slope above 3.0, abrupt clearing at E0, and dark erosion below 20 nm/min. Typical E0 values for 1.2 µm films are 30–50 mJ/cm², and dose to size for 0.5 µm dense lines is commonly 2–3 times E0.

Development is performed with 2.38 wt% tetramethylammonium hydroxide at 21–23 °C. Immersion development in a tank with ultrasonic agitation or puddle development on a single-wafer developer can be used; puddle processes provide better lot-to-lot reproducibility because developer temperature and contact time are controlled to ±0.5 °C and ±1 s. Typical development time for a 1.5 µm film is 60–90 s. Dissolution rate monitors using interferometry at 410 nm record thickness loss in developer and are used to detect batch-to-batch shifts in dark erosion. Developer normality drift above 0.01 N from the bulk bath value is a known source of critical dimension shift.

When Film Thickness Exceeds 3 µm for Bump and Deep Cavity Applications

For thick-resist applications such as solder bump plating and deep silicon etching, the viscosity grade is adjusted to deliver 6–15 µm of film thickness after a single spin coat at 1000–2000 rpm. Solvent release from these films requires a low-temperature softbake ramp starting at 60 °C for 90 s, followed by 100 °C for 120–180 s, because rapid solvent evaporation at full setpoint produces surface skinning and bubble formation. The exposure dose for thick DNQ/novolac films must be increased nonlinearly; at 10 µm, doses of 400–800 mJ/cm² and multiple exposure passes may be required to maintain sidewall angles above 80°.

In this thickness regime, the primary failure mode is incomplete development at the resist-substrate interface. Multiple puddle development cycles with fresh developer every 45–60 s are used to reduce scum and foot formation. After hardbake, the resist mask is exposed to acidic copper sulfate electrolytes in bump plating; dimensional stability is maintained only when the hardbake temperature remains below 150 °C for films thicker than 12 µm. Published data for the Jingrui Electric Material thick-film grade under these conditions are limited; the above values are representative of commercial DNQ/novolac thick resists and should not replace process qualification.

Low-viscosity grades are not suitable for thicknesses above 4 µm because edge bead and planarity defects exceed typical total thickness variation limits. High-viscosity grades, in contrast, may exhibit higher particle retention and require slower dispense speeds below 0.5 mL/s to prevent microbubble entrapment on 200 mm substrates.

Reference quality control values for this product class are summarized in the table below. These values are not lot-release specifications for the Jingrui Electric Material grade unless confirmed by the supplier datasheet.

Thermal Stability of Unexposed Resist Film Is Controlled by Residual Solvent Content

After softbake, residual PGMEA acts as a plasticizer and lowers the glass transition temperature of the novolac film. Thermal flow during hardbake is therefore controlled by the softbake profile. For hardbake temperatures between 140 °C and 180 °C, dimensional stability is maintained only when residual solvent is below 2 wt%. In plating applications, hardbake at 120 °C for 30 min is used to stabilize the resist mask before exposure to acidic copper sulfate electrolytes. Resist cracking during hardbake is observed on films above 15 µm if the temperature ramp exceeds 5 °C/min. Thermal degradation of unexposed DNQ/novolac films begins at sustained temperatures above 200 °C; therefore, hardbake beyond this point is not recommended for process stability.

Regulatory compliance is typically documented by the following matrix for commercial i-line photoresist formulations; supplier lot certificates should be requested for the Jingrui Electric Material product.

The material is stripped after pattern transfer using oxygen plasma ashing followed by sulfuric acid/hydrogen peroxide mixture. In dry etching, the etch selectivity of the resist relative to silicon dioxide depends on gas chemistry; under CHF3/CF4 plasma, selectivity to oxide is typically 0.5:1 to 1:1. The specific Jingrui Electric Material i-line resist selectivity must be verified on the production etcher because published data for this configuration is limited. For lift-off processes, the resist is applied as a bilayer or image-reversal-capable grade, where the undercut profile is controlled by flood exposure and reversal bake. Those grades require a reversal bake at 110–120 °C for 60–90 s, followed by flood exposure of 200–300 mJ/cm² and development in 2.38 wt% TMAH.

The material is incompatible with strong alkaline cleaners that exceed 3 wt% NaOH at sustained temperatures above 50 °C. Open reservoirs in humidity above 60% RH should be avoided because moisture uptake can hydrolyze the DNQ component and reduce shelf life. When pre-drying is required at relative humidity above 60%, the wafer is dehydrated at 150 °C for 60 s before HMDS priming. Batch-to-batch viscosity variation of ±5% from the target value can shift film thickness by 0.1–0.2 µm on a 200 mm spin coater; automated viscosity adjustment by solvent addition is performed before release to coat tracks.