| HS Code | 827889 |
| Product Name | Silicon Oxide Etchant Electronic/EL Grade |
| Chemical Composition | Buffered solution of ammonium fluoride and hydrofluoric acid in deionized water |
| Grade | Electronic/EL (high-purity electronic grade) |
| Appearance | Clear, colorless liquid |
| Physical State | Liquid |
| Odor | Sharp, pungent odor |
| Ph | Acidic, approximately 4.5 to 6.0 |
| Specific Gravity | Approximately 1.10 to 1.25 at 20°C |
| Density | Approximately 1.1 to 1.25 g/cm³ at 20°C |
| Boiling Point | Approximately 100°C (aqueous solution) |
| Melting Point | Approximately 0°C (aqueous solution) |
| Vapor Pressure | Water-like vapor pressure at room temperature |
| Solubility | Fully miscible with water |
| Etch Rate | Controlled etch rate for thermally grown silicon dioxide, typically 500 to 2000 Å/min depending on temperature and composition |
| Etch Selectivity | High selectivity of silicon dioxide over silicon and silicon nitride |
| Trace Metal Impurity | Very low, often parts-per-billion (ppb) level |
| Shelf Life | Stable when stored in tightly closed original containers under recommended conditions |
| Storage Temperature | Recommended storage at 15°C to 30°C |
As an accredited Silicon Oxide Etchant Electronic/EL Grade factory, we enforce strict quality protocols—every batch undergoes rigorous testing to ensure consistent efficacy and safety standards.
| Packing | Packaged in a 1-gallon HDPE jerrican with tamper-evident seal, labeled for Electronic/EL Grade silicon oxide etchant. Store upright. |
| Container Loading (20′ FCL) | 20′ FCL container loading of Silicon Oxide Etchant Electronic/EL Grade in UN-approved drums, secured and segregated per chemical safety regulations. |
| Shipping | Silicon Oxide Etchant (Electronic/EL Grade) ships in sealed, corrosion-resistant containers, clearly labeled with hazard warnings. Transport follows strict regulations for corrosive chemicals—ground freight only, upright, protected from moisture and extreme temperatures. Ensure proper documentation and spill-containment measures are in place before dispatch. |
| Storage | Store in tightly sealed original container in a cool, dry, well-ventilated area, ideally between 15–25°C. Keep away from direct sunlight, heat, moisture, and incompatible materials such as alkalis, metals, glass, or strong oxidizers. Use corrosion-resistant secondary containment. Ensure proper labeling and segregation from other chemicals. Maintain adequate ventilation and comply with local regulations. |
| Shelf Life | Shelf life is typically 12 months when stored unopened in original containers at controlled room temperature, away from light and moisture. |
Following plasma stripping of photoresist and dry etch residue removal, 300 mm wafers are immersed in a 7:1 buffered oxide etchant to remove sacrificial SiO₂ before gate oxidation or diffusion. The bath is operated at 25.0 °C ± 0.5 °C in a recirculating PFA or PVDF vessel with 0.05 µm PTFE filtration. At the set point, thermally grown SiO₂ is removed at 100–120 nm/min, while monocrystalline silicon removal remains below 0.1 nm/min under dark conditions due to surface hydrogen termination. A 1 °C deviation from set point alters the removal rate by approximately 7–10%, so the temperature control loop uses a PFA-encapsulated platinum RTD. Bath replenishment is based on free-HF titration and specific gravity; drift in the NH₄F:HF ratio outside 7:1 ± 0.2 changes selectivity to silicon nitride and polysilicon. In production wet benches, 50-wafer cassettes are processed with megasonic agitation at 850 kHz to reduce micro-roughness from local depletion. Particle counts in the bath are maintained below 100 particles/mL at ≥0.2 µm, and trace metals are controlled by ICP-MS to less than 10 µg/L for Na, K, Fe, Cu, Ni, and Zn. The etch is timed rather than endpointed because the oxide layer thickness after plasma strip is known from ellipsometry. After etching, the cassette is transferred through a cascading deionized water rinse with overflow rate above 5 L/min per tank. Bath quality follows SEMI C36 acceptance limits for buffered oxide etchants.
| NH₄F:HF ratio | Thermal SiO₂ etch rate at 25 °C | Relative removal rate | Typical front-end use |
|---|---|---|---|
| 4:1 | 140–160 nm/min | 1.3–1.5× | Bonded oxide etch-back after wafer thinning |
| 7:1 | 100–120 nm/min | 1.0× | Sacrificial oxide strip before furnace processing |
| 10:1 | 65–80 nm/min | 0.6–0.7× | Native oxide touch-up before metallization |
| 20:1 | 30–40 nm/min | 0.3–0.4× | Pre-epitaxy clean where minimal oxide removal is required |
Liquid-phase 6:1 or 7:1 buffered oxide etchant is used to remove sacrificial SiO₂ beneath polysilicon, single-crystal silicon, or SiGe microstructures after deep reactive-ion etching. The release vessel is maintained at 23.0 °C ± 0.3 °C because lateral etch fronts inside high-aspect-ratio gaps are diffusion-limited. In open-bath conditions, 7:1 chemistry removes thermal oxide at 100–120 nm/min at 25 °C; at 23 °C the rate is lower by approximately 15–25%. Structures with sacrificial oxide gaps of 1.5–2.0 µm show an additional apparent rate reduction inside the confined gap. The reduction is geometry-dependent and cannot be transferred from one die layout to another without calibration. Published data for this specific polysilicon/SiO₂/nitride stack are limited; release time should be established with test structures that replicate gap height and vent spacing. Rinsing follows the sequence deionized water, isopropyl alcohol, then low-surface-tension fluid; direct air drying after water is excluded to avoid capillary stiction in comb-drive arrays. The structural silicon nitride etch-stop is verified on every bath startup by ellipsometry: if the nitride removal rate exceeds 2 nm/min, the bath is adjusted or discarded because the etch-stop margin in many MEMS stacks is below 20 nm. Hydrogen bubble accumulation at the undercut front is suppressed by low-power megasonic agitation at 300–500 kHz and by rotating the wafer carrier at 0.5–1.0 rpm. The process uses semiconductor-grade PFA fixtures because stainless steel guide rods release iron and chromium into the bath and increase pitting at aluminum bond pads.
Inline POCl₃ diffusion lines serving TOPCon and PERC cells use EL-grade buffered oxide etchant to lift phosphosilicate glass from the emitter surface. The etch bath is held at 20–22 °C because PSG removal is faster than thermal SiO₂ by a factor of 2–3 at the same temperature and formulation. Time-based endpoint control is used after a daily calibration coupon; over-etch roughens the emitter and reduces sheet-resistance uniformity. Process exhaust extraction is mandatory at the bath surface because volatile HF concentration can exceed 0.5 ppm in the operator breathing zone without scavenging. Iron and copper in the etchant are controlled to less than 10 µg/L by ICP-MS to limit carrier recombination. Particle filtration at 0.1 µm reduces light-induced shunt paths on 182 mm and 210 mm wafers.
Wet etching of PECVD SiO₂ gate-insulator or passivation layers in flat-panel display production must tolerate local area density variations across a 1,500 mm × 1,850 mm glass substrate. Micro-loading appears when the exposed oxide area in high-resolution pixel regions is more than 20:1 relative to open test areas. In linear spray or meniscus-type etching modules, a 7:1 buffered oxide etchant removes PECVD oxide at 80–100 nm/min at 25 °C, while PECVD silicon nitride loss remains below 5 nm/min only when the NH₄F:HF ratio is maintained within ±0.5% by specific gravity and free-HF titration. Local fluoride depletion at high-loading sites produces residual oxide islands visible after backlight inspection; this failure mode is corrected by nozzle pressure adjustment and by reducing cassette crowding. Temperature is controlled to ±0.3 °C because the etch rate of PECVD oxide varies by 10–15% per degree Celsius. The etchant is reclaimed through 0.05 µm filtration until particle counts exceed 100 particles/mL at ≥0.2 µm. Etchant replacement intervals are set by cation contamination rather than usable free fluoride; aluminum and chromium from substrate scribing lines can accumulate above 10 µg/L and alter flat-band voltage in completed TFTs.
For wafer-level fault isolation and physical failure analysis, a 10:1 or 20:1 buffered oxide etchant is dispensed in a single-cassette PFA vessel rather than a production immersion bench. The dielectric film thickness is measured by spectroscopic ellipsometry before removal; etch time is then calculated from the open-bath calibration rate and reduced by an empirically determined margin for film densification. Endpoint is confirmed by surface hydrophobicity after the final rinse, with a water contact angle above 70°. Only EL-grade chemistry with individual trace metal concentrations below 10 µg/L is used, because residual copper or iron can generate secondary electron contrast artifacts in SEM review. In wafer reclaim, the same chemistry strips sacrificial oxide before repolish; particle counts at ≥0.2 µm are limited to fewer than 100 particles/mL to prevent re-entrained defect clusters on reclaimed substrates.
| Parameter | Electronic/EL-grade control target | Reference method |
|---|---|---|
| Individual trace metal, Na, K, Fe, Cu, Ni, Zn | ≤ 10 µg/L per element | ICP-MS, SEMI C1 |
| Total trace metals | ≤ 50 µg/L | ICP-MS |
| Particle count at ≥0.2 µm | ≤ 100 particles/mL | Laser particle counter |
| Free HF content | ± 0.5% of nominal | Titration |
| Chloride and sulfate | ≤ 500 µg/L each | Ion chromatography |
| Cleanroom condition | ISO 14644-1 Class 5 or better | Airborne particle counting |
Fused silica and borosilicate glass substrates are patterned with 7:1 buffered oxide etchant through a 20 nm chromium / 200 nm gold hard mask. The etch is isotropic; therefore the sidewall undercut width is approximately equal to the vertical etch depth. Etch rates are composition-dependent: published data for fused silica in 7:1 chemistry at 25 °C range from 50 to 150 nm/min, while borosilicate glass may etch faster because boron oxide leaches into the fluoride bath. Chromium adhesion failure occurs when pinholes or scribe cracks expose the chromium-glass interface to HF; the resulting undercut appears as ragged line edges under dark-field inspection. Post-etch rinsing uses dilute hydrochloric acid to remove insoluble fluoride salts that form on lead-containing or barium-containing glass surfaces. Process vessels are opaque to short-wavelength light because photochemical effects accelerate mask degradation; this operational boundary is not relevant for production silicon etching but becomes significant in glass microstructuring runs longer than 60 min.
Competitive Silicon Oxide Etchant Electronic/EL Grade 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!
Silicon Oxide Etchant Electronic/EL Grade is a premixed buffered oxide etch formulation based on ammonium fluoride and hydrofluoric acid, supplied for selective removal of thermally grown, CVD, PECVD, and sputtered silicon dioxide films. Representative product designations include EL-6:1, EL-7:1, and EL-10:1, where the ratio denotes the volumetric proportion of 40 % NH₄F solution to 49 % HF solution. The designation Electronic/EL Grade is not an industry-standard grade definition; release limits are supplier-specific and should be verified against the lot certificate of analysis. The solution is commonly packaged in fluoropolymer-lined polyethylene containers and dispensed through point-of-use filtration. Typical pH at 25 °C is 3.2–4.0, and density ranges from 1.11 to 1.16 g/mL depending on ratio. The formulation attacks SiO₂ while maintaining lower etch rates on monocrystalline silicon and typical photoresists than concentrated HF, but undercut and resist adhesion loss must be characterized for each resist chemistry and bake condition.
Documented application areas include pre-diffusion oxide strip, pad oxide removal before gate oxidation, sacrificial oxide release in MEMS, contact surface preparation, and passivation-layer patterning in thin-film electroluminescent display backplanes. It is also employed as a selective oxide etchant over silicon nitride etch-stop layers in shallow trench isolation and gate spacer modules. The product is delivered as a ready-to-use liquid for immersion, spray, or single-wafer wet processing equipment. Use in production requires pre-qualification of etch rate, selectivity, defect density, and trace-metal contamination on the specific film stack and wafer diameter in use.
Electronic/EL Grade material is controlled primarily through feedstock purification, filtration, and packaging discipline. For a representative EL-10:1 formulation, the ammonium fluoride and hydrofluoric acid concentrations are held within narrow ranges to limit etch-rate variation. Acidimetric titration with NIST-traceable sodium hydroxide is used for assay confirmation; density and pH are measured at 25 °C as release checks. The following table gives representative release limits at point of packaging for a 10:1 ratio EL-grade product. Because no single international specification defines Electronic/EL Grade, these values are typical of commercially supplied semiconductor and display-grade buffered oxide etchants, not an absolute standard.
| Parameter | Representative release limit for EL-10:1 | Test method |
|---|---|---|
| HF assay | 4.5–5.0 wt% | Acidimetric titration with NIST-traceable NaOH |
| NH₄F assay | 36.0–37.0 wt% | Acidimetric titration with NIST-traceable NaOH |
| Density at 25 °C | 1.11–1.13 g/mL | Oscillating U-tube density meter |
| pH at 25 °C | 3.4–3.8 | Calibrated pH electrode |
| Fe, Cu, Cr, Ni | ≤5 ppb each | ICP-MS after matrix dilution |
| Na, K, Ca, Mg | ≤5 ppb each | ICP-MS after matrix dilution |
| Li, Mn, Pb, Zn, Al | ≤2–≤5 ppb each | ICP-MS after matrix dilution |
| Total trace metals | ≤50 ppb | ICP-MS summation |
| Particles ≥0.2 µm | ≤25 counts/mL | Liquid particle counter per ISO 21501-2:2019 |
| Particles ≥0.5 µm | ≤5 counts/mL | Liquid particle counter per ISO 21501-2:2019 |
| Non-volatile residue | ≤5 ppm | Controlled evaporation at 105 °C |
Release testing is performed in an ISO 14644-1:2015 Class 5 or better cleanroom. Water used for component dilution and final preparation conforms to ASTM D5127-13 Type E-1.2 for resistivity, total organic carbon, and particle limits. Trace-metal calibration standards are matrix-matched and traceable to NIST SRM 1643f. Point-of-use filtration through 0.05 µm PTFE or PFA membranes is standard for EL-grade products. Sterile or autoclaved filtration hardware is not required because the product is not an aqueous biological preparation, but filter flush volume and membrane extractables must be qualified before first installation.
The primary difference between Electronic/EL Grade and technical-grade buffered oxide etch is the trace-metal, anion, and particle burden of the feedstock and final blend. Technical-grade BOE can contain transition-metal totals in the 100 ppb to 1000 ppb range and may carry chloride, sulfate, nitrate, or phosphate residuals above 100 ppb. EL-grade release limits typically reduce each transition metal to 5 ppb or lower and total trace metals to 50 ppb or lower. Sodium and potassium are controlled below 5 ppb to limit mobile-ion contamination in gate dielectric and electroluminescent device stacks. In non-volatile memory and display backplane processing, mobile alkali ions can shift threshold voltage or cause pixel-level luminance drift; EL-grade chemistry is specified to reduce this risk.
Another difference is particle control. Technical-grade material often lacks any particle specification, whereas EL-grade product is filtered at 0.05 µm and tested by laser light-scattering per ISO 21501-2:2019. Packaging is performed on fluoropolymer filling lines inside ISO 14644-1:2015 Class 5 or better environments. Container materials are high-purity polyethylene or fluoropolymer; borosilicate glass is never used because HF attacks glass and releases calcium, sodium, and silicon. The table below summarizes typical differences between technical-grade BOE and Electronic/EL Grade silicon oxide etchant at point of packaging.
| Parameter | Technical-grade BOE | Electronic/EL Grade silicon oxide etchant |
|---|---|---|
| Total transition metals | 100–1000 ppb typical | ≤50 ppb total; ≤5 ppb per element typical |
| Na + K | Often unspecified | ≤5 ppb each typical |
| Particles ≥0.2 µm | Often unspecified | ≤25 counts/mL typical |
| Point-of-use filtration | 0.2–1.0 µm or none | 0.05 µm PTFE/PFA |
| Assay tolerance | ±10 % of nominal or broader | ±3 % of nominal or tighter |
| Packaging environment | General chemical handling | ISO 14644-1:2015 Class 5 or better |
| Suitability for gate oxide or EL backplane use | Not recommended without purification | Qualified after lot-specific verification |
Technical-grade BOE may also exhibit wider lot-to-lot variation in free HF activity, which shifts SiO₂ etch rate and complicates statistical process control. EL-grade suppliers typically control free HF and NH₄F content within narrow bands, but the etch rate must still be re-qualified on the manufacturing line after any supplier or batch change. In recirculating etch baths, lower sulfate and phosphate residuals reduce the probability of insoluble calcium or magnesium precipitate formation when the bath is exposed to airborne particulate or lower-quality rinse water. This is an operational advantage observed in high-volume wet benches, though published data for specific defectivity improvement are limited and must be generated for the individual process.
For thermally grown SiO₂ removal in a single-wafer spray processor, EL-10:1 at 25 °C typically shows an etch rate of 60–80 nm/min, while EL-6:1 typically falls between 100 and 120 nm/min. These values should be confirmed by spectroscopic ellipsometry on a monitor wafer having the same oxide type because PECVD and sputtered oxides etch faster than thermal oxide. In a recirculating immersion bath, temperature control within ±0.5 °C is common. Process drift of 1 °C can change etch rate by approximately 5–8 %, so closed-loop heating with PTFE heat exchangers and resistance temperature detectors is required. For MEMS sacrificial oxide release in a 100 L PFA bath, batch-to-batch etch rate variation can be held within ±5 % when fresh-solution spiking and titration protocols are followed. Agitation by megasonic nozzles at 0.8–1.2 MHz improves oxide removal uniformity but may lift photoresist at feature edges; compatibility must be evaluated for the specified resist system.
Compared with concentrated HF, the buffered EL-grade formulation reduces silicon attack and photoresist lift, but its SiO₂ etch rate is lower. Compared with vapor HF, it offers higher throughput for sacrificial oxide release and display backplane passivation layers, but it produces isotropic undercut. The selectivity to LPCVD silicon nitride depends on ratio and temperature, with BOE generally etching thermal SiO₂ at roughly 10–30 times the nitride etch rate. This allows oxide removal over nitride stops in shallow trench isolation and gate spacer modules. Silicon-rich nitride and oxynitride films may etch faster than stoichiometric Si₃N₄, so endpoint and selectivity must be measured on the actual film composition.
Process integration is sensitive to bath life and dissolved silicon accumulation. During oxide etching, SiO₂ is converted to hexafluorosilicic acid and water. As dissolved silicon increases in a recirculating bath, etch rate can fall by 10–20 % over repeated wafer batches if no spiking or bath exchange is performed. Precipitation of fluorosilicate salts becomes possible when the bath approaches saturation; continuous 0.05 µm filtration removes pre-existing particulates but does not prevent precipitation caused by chemical saturation. Production baths therefore require periodic titration for free HF, density or refractive index trending, and particle-count monitoring. For single-use or single-pass spray tools, these bath-life effects are reduced, but nozzle clogging and crystallized residues on chamber surfaces remain practical failure modes if the etchant is allowed to dry.
Storage and operational boundaries are as follows. The etchant must remain in fluoropolymer or high-density polyethylene containers at 5–25 °C. Exposure to temperatures below 5 °C can cause crystallization; repeated freeze-thaw cycles can alter ratio by phase segregation. Do not mix with ammonia or amine-containing strippers, because exothermic ammonium fluoride formation shifts pH and may accelerate resist undercut. When oxide etching is performed on aluminum-metallized wafers, the HF content attacks the metallization; a non-HF alternative is required unless the metal is fully masked. Contact with borosilicate glass, aluminum, titanium, zinc oxide, and calcium-containing ceramics must be avoided. Published data for this specific configuration is limited, and process qualification should be repeated after any change in supplier, ratio, rinse water quality, or wafer surface preparation.
