| HS Code | 727034 |
| Product Name | High Selectivity Silicon Dioxide/Silicon Nitride Etchant, Electronic/EL Grade |
| Chemical Family | Buffered hydrofluoric acid / ammonium fluoride aqueous mixture |
| Grade | Electronic / EL grade |
| Appearance | Clear, colorless liquid |
| Physical State | Liquid |
| Boiling Point | Approximately 100 degrees Celsius at 1 atm |
| Flash Point | None; non-flammable aqueous solution |
| Specific Gravity | Approximately 1.15 to 1.30 at 20 degrees Celsius |
| Ph | Acidic; approximately 4.5 to 6.0 |
| Solubility In Water | Fully miscible with water |
| Selectivity Sio2 To Si3n4 | Greater than 100 to 1 in favor of silicon dioxide |
| Etch Rate On Thermal Sio2 | Approximately 700 to 1000 angstroms per minute at 25 degrees Celsius |
| Etch Rate On Si3n4 | Less than 10 angstroms per minute at 25 degrees Celsius |
| Metal Impurity Level | Individual trace metals typically less than or equal to 10 ppb |
| Storage Temperature | 15 to 30 degrees Celsius |
| Shelf Life | At least 6 months when unopened in original container |
As an accredited High selectivity silicon dioxide/silicon nitride Etchant Electronic/EL Grade factory, we enforce strict quality protocols—every batch undergoes rigorous testing to ensure consistent efficacy and safety standards.
| Packing | Supplied in a 1-gallon (3.78 L) HDPE container with tamper-evident seal, ensuring ultrapure EL-grade etchant integrity. |
| Container Loading (20′ FCL) | 20′ FCL: sealed chemical drums palletized, strapped, and secured; labeled, ventilated, and loaded per hazardous material regulations for safe transport. |
| Shipping | Ship as Corrosive Liquid, Acidic, Inorganic, n.o.s., UN 3264, Hazard Class 8. Use UN-certified HDPE drums or compatible containers with secure, vented closures. Protect from damage, keep upright, and label with the proper shipping name, UN number, and Class 8 corrosion hazard. Include emergency response information for spills or leaks. |
| Storage | Store in a tightly sealed original container in a cool, dry, well-ventilated area, away from direct sunlight and incompatible materials. Keep container upright and protected from physical damage. Ensure secondary containment to prevent spills. Avoid moisture and extreme temperatures; inspect regularly for leakage or degradation. Always follow the manufacturer’s SDS guidelines. |
| Shelf Life | Store tightly sealed at room temperature. Shelf life is typically 12 months from manufacture date. |
In gate-first high-k/metal gate front-end integration, post-spacer etch cleaning removes sacrificial silicon dioxide while minimizing attack on silicon nitride spacers. On a single-wafer spray processor with a 200 mm rotating chuck, the etchant is metered at 1.2 L/min through a 0.22 µm PTFE point-of-use filter and blended with 18.2 MΩ·cm deionized water conforming to ASTM D1193-06 Type E-1. The working mixture is a 200:1 volumetric dilution of electronic-grade 49 wt% HF, yielding thermal silicon dioxide etch rates of 1.5–2.5 nm/min at room temperature on 25 nm thermal oxide monitor wafers. Silicon nitride spacer loss remains below 0.3 nm per 60 s dispense cycle as measured by spectroscopic ellipsometry on 50 nm blanket plasma-enhanced chemical vapour deposition nitride, yielding an SiO2:Si3N4 selectivity of at least 5:1. Supplier Certificates of Analysis reference SEMI C7-2021 trace metal tables and report particle counts below 50 particles/mL at 0.2 µm. The etching sequence is not followed by a sulfuric acid-hydrogen peroxide over-strip because that sequence raises interface roughness on the high-k dielectric. The resulting dielectric stacks are integrated into advanced logic and 3D NAND gate stacks where wet-clean selectivity directly affects threshold-voltage stability and capacitor retention margin.
For oxide passivation opening over a-IGZO or low-temperature polycrystalline silicon thin-film transistors, the wet etch must clear PECVD silicon dioxide without consuming the underlying silicon nitride etch-stop layer. A 7:1 volumetric mixture of 40 wt% ammonium fluoride and 49 wt% hydrofluoric acid is dispensed in a Gen 8.5 in-line wet station at 25.0 °C ± 0.5 °C; recirculation passes through 0.1 µm fluoropolymer filtration. The ammonium fluoride shifts the equilibrium toward bifluoride, which attacks silicon dioxide while reducing free fluoride concentration and preserving silicon nitride. Static immersion tests on 100 nm PECVD silicon nitride monitor wafers show a SiO2:Si3N4 etch-rate ratio of not less than 8:1 under bath conditions. A 300–500 nm oxide layer requires 90–120 s immersion; cross-section scanning electron microscopy monitors sidewall undercut below 0.30 µm per side and taper angle between 30° and 45°. Post-etch optical reflectance confirms remaining silicon nitride thickness is not less than 50 nm. The bath is pH-stabilized at 4.8–5.2 to prevent photoresist lifting; isopropyl alcohol addition is avoided because it accelerates photoresist adhesion loss despite improving wetting. The etchant, filtration media, and rinse water comply with SEMI C8-2021 for ammonium fluoride and ISO 14644-1:2015 Table 1 for ISO Class 4 environment. Resulting Gen 8.5 array substrates feed OLED backplane and LCD TFT module assembly, where contact resistance and pixel-charging uniformity are set by the etched contact edge profile.
Sacrificial oxide release in surface-micromachined inertial sensors uses vapour-phase hydrofluoric acid to prevent stiction following release. The chamber is held at 35.0 °C ± 1.0 °C with anhydrous hydrogen fluoride delivered at 0.3–0.8 L/min and ethanol co-flow at 0.2 L/min to suppress capillary condensation inside 1.5 µm-thick sacrificial silicon dioxide cavities. The etchant’s selectivity against a 0.35 µm low-stress silicon nitride etch stop is specified above 100:1 in supplier qualification data. Uncontrolled over-etch would reduce suspension thickness and shift comb-drive resonance frequency. On-line optical interferometry tracks oxide release progress and terminates the cycle when the oxide is cleared. When chamber exhaust hydrogen fluoride approaches 0.5 ppm, cartridge abatement is triggered under SEMI S2 equipment safety protocols. The etchant is supplied in electronic grade with trace metal certificates indicating individual cations below 10 ppb, and the process bay meets ISO 14644-1:2015 ISO Class 4 particle limits. Published selectivity data under production vapour release on thin nitride stops remain limited; qualification therefore uses wafer-level electrical resonance testing to bound final device performance. Released MEMS accelerometers and gyroscopes enter automotive yaw-rate sensors and mobile inertial measurement units after wafer-level electrical resonance screening.
Passivation opening over aluminum bond pads in fan-in wafer-level chip-scale packaging uses a low-pH fluoride etchant to remove a thin bilayer of silicon dioxide and silicon nitride without corroding the aluminum-copper pad underneath. The process uses a 30:1 volumetric NH4F/HF mixture diluted with deionized water to a working temperature of 21.0 °C ± 0.5 °C. The etch requires an SiO2:Si3N4 selectivity of at least 10:1 to clear 0.8 µm of silicon oxide and 0.3 µm of plasma silicon nitride defined by a spin-on polyimide mask with openings of 30–40 µm diameter on a 300 mm wafer. Endpoint detection by optical emission stops the process before pad surface roughening exceeds 1.5 nm root-mean-square. The fabrication lot is sampled for residual fluoride contamination by ion chromatography; acceptance is below 1.0 × 10¹⁰ atoms/cm² before polyimide curing. The etchant must comply with SEMI C7-2021 and SEMI C8-2021, while the final packaging line meets RoHS 2011/65/EU and REACH EC 1907/2006 restrictions. Published data for this specific bilayer configuration is limited, so engineering runs compare focused ion beam cross-sections against the previous qualified process of record. Qualified fan-in WLCSP devices are assembled into smartphone power management integrated circuits and radio-frequency front-end modules.
For self-aligned spacer removal before epitaxial raised source/drain formation, an 85 wt% electronic-grade phosphoric acid bath is maintained at 160.0 °C ± 1.0 °C with a water vapour sparger to control boiling point and prevent polymerised phosphorus species from raising viscosity. A quartz recirculating wet bench processes 200 mm wafers in batches of 50; low-stress silicon nitride etch rate is 4.5–5.5 nm/min while tetraethyl orthosilicate silicon dioxide loss is held below 0.5 nm per 5 min cycle, yielding a Si3N4:SiO2 selectivity of not less than 10:1 after 1 h of bath aging. The bath temperature bump is limited to 0.3 °C/min during wafer entry to avoid thermal shock cracking. Bath lifetime is extended by monitoring phosphoric acid density and water content with an in-line refractometer, discarding the bath when dissolved silicon exceeds 80 ppm to prevent particle formation. The etch tool is equipped with an exhaust scrubber rated under SEMI S8 safety guidelines, and the cleanroom is classified to ISO 14644-1:2015 ISO Class 4. Stripped wafers proceed to RF silicon-on-insulator switches and low-power embedded non-volatile memory fabrication, where junction leakage is a gate-level yield parameter.
During TOPCon cell fabrication, front-side silicon nitride antireflection coating removal uses a hot phosphoric acid spray on a horizontal conveyor tool after emitter formation and before tunnel oxide passivated contact annealing. The etchant is blended to 75–80 wt% H3PO4 and held at 150–160 °C; the spray nozzle bank operates at 2.0–3.0 bar and the belt speed is 0.4–0.8 m/min. The removal must clear the PECVD silicon nitride mask layer without reducing the underlying silicon dioxide tunnel layer by more than 0.2 nm; inline ellipsometry at 5-point wafer mapping rejects cells with center-to-edge silicon dioxide thickness variation greater than 0.5 nm. The Si3N4:SiO2 selectivity is maintained above 10:1 by monitoring water content with a density-controlled sparging skid upstream of the spray bar. The chemical handling system must satisfy SEMI S2, and waste discharge is treated by fluoride and phosphate abatement to meet local semiconductor effluent limits. The incoming electronic/EL-grade etchant is filtered to 0.1 µm and trace metal data for copper and iron are below 10 ppb to avoid minority carrier lifetime degradation. Processed n-type TOPCon cells are later laminated into IEC 61215-certified modules for utility-scale solar installations.
Competitive High selectivity silicon dioxide/silicon nitride 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!
High selectivity silicon dioxide/silicon nitride etchant supplied as Electronic/EL Grade is a buffered hydrogen fluoride solution formulated for wet removal of silicon dioxide with minimum silicon nitride loss. Product model designations typically encode the NH4F:HF weight ratio, purity class, and container volume. Production-grade ratios include 4:1, 6:1, 7:1, and 10:1; the 7:1 composition is widely used for high-selectivity oxide release. Electronic/EL grade means the etchant is filtered to submicrometer particle limits and filled in cleanroom conditions. It is delivered with lot-specific certificates of analysis, and the wetted packaging is normally PFA or FEP.
In a 7:1 product, the hydrofluoric acid assay is controlled near 6.0 wt% to 7.0 wt%, and ammonium fluoride provides buffering capacity. That compositional window shifts the etch-active fluoride speciation toward HF2− and lowers free-HF activity. The result is a process chemistry that removes thermal oxide at a practical rate while suppressing silicon nitride hydrolysis, occupying a different process space from technical-grade BOE, unbuffered dilute HF, or low-purity oxide removers. The model-specific assay, cation budget, and particle burden are the main specifications used by process engineers to compare sources.
Fluoride speciation in NH4F/HF solutions is governed by the equilibria HF ⇌ H+ + F− and HF + F− ⇌ HF2−. At high fluoride-to-HF ratios, bifluoride is the dominant etch-active species. In thermal silicon dioxide, fluoride attack at siloxane bridges produces SiF62−, and the dissolution rate exhibits first-order dependence on HF2− and proton activity in the acidic pH range. Silicon nitride, by contrast, requires protonation and hydrolysis of Si–N bonds before fluoride attack. The kinetic barrier for that hydrolysis is higher for densified LPCVD silicon nitride, which is why selectivity can exceed 100:1.
Representative supplier data for 7:1 formulations report thermal silicon dioxide etch rates of 100 nm/min to 140 nm/min at 25 °C for undensified thermal oxide, while as-deposited LPCVD silicon nitride remains below 1 nm/min. When the bath temperature is raised to 35 °C, oxide etch rates increase to approximately 160 nm/min to 200 nm/min, and nitride etch rates remain below 3 nm/min for optimized buffering. The selectivity value is therefore not a fixed constant; it is a temperature- and film-dependent process variable.
The process margin narrows when PECVD nitride replaces furnace LPCVD nitride. PECVD silicon nitride can contain 10 at% to 20 at% hydrogen and may have lower density and higher oxygen content. Those factors increase wet etch rates by two to five times compared with stoichiometric LPCVD nitride. Silicon-rich nitride can also etch faster, especially after plasma exposure forms a surface oxide. Production lines using fixed etch times must re-qualify selectivity when the nitride deposition tool, silane-to-ammonia ratio, or annealing condition is changed.
Bath aging introduces a separate threshold. On a production-scale recirculating wet bench with a 120 L PFA bath, point-of-use filtration at 0.02 µm, and temperature control at 25.0 ± 0.5 °C, dissolved silicon and organic residues accumulate as patterned wafers are processed. Etch rate drift of ±5% is commonly managed by replenishment based on cumulative wafer area treated. When pH rises above 4.0 or dissolved silicon exceeds the supplier-defined limit, nitride selectivity can degrade before bulk oxide etch rate falls. At that point, partial bath exchange is more effective than extending immersion time.
In front-end semiconductor manufacturing, the high-selectivity etchant is used in recirculating wet etch stations with PFA and PTFE wetted materials. The bath is blanketed with filtered nitrogen, filtered through 0.02 µm or 0.05 µm fluoropolymer cartridges, and held at 25 °C or 30 °C, depending on the oxide budget. Immersion time is calculated from monitor wafer oxide thickness. For a 50 nm thermal oxide film in a 7:1 bath at 25 °C, nominal immersion is approximately 25 s based on the mid-range oxide etch rate; overetch is then added according to the minimum nitride thickness that can be sacrificed. Wafer-to-wafer repeatability of ±1.5% in oxide removal is achievable when temperature control remains within ±0.5 °C and replenishment is automated by fluoride-selective electrode or cumulative wafer-area count.
Dopant-dependent oxide etch rate is integrated into process control. Boron- and phosphorus-doped thermal oxides generally etch faster than undoped oxide, so the apparent selectivity against nitride can shift by 10% to 20% depending on dopant concentration and activation anneal. Published data for a specific doped-oxide configuration is limited; each process must measure monitors on the user's furnace profile. In nitride-spacer release applications, endpoint is often inferred from monitor wafers because direct optical endpoint is difficult in wet benches. In-line ellipsometry and X-ray photoelectron spectroscopy are used to verify remaining nitride thickness.
MEMS release applications use this chemistry when a sacrificial oxide must be removed from beneath a nitride or nitride-shielded structure. The wet etch route offers higher throughput than vapor HF release, but it introduces stiction risk in high-aspect-ratio devices. After the oxide is cleared, the wafer is transferred through a rinse and drying sequence; for fragile microstructures, IPA vapor drying or supercritical CO2 drying is used. The etchant is not suitable for release of structures with exposed aluminum, titanium, or titanium nitride, because those materials are attacked by fluoride-containing media.
The electronic/EL-grade product has tighter trace-metal and particulate controls than semiconductor-grade or technical-grade BOE. The certificate of analysis is prepared by an ISO 17025-accredited laboratory and reports assay, trace metals, anions, and particle burden. Incoming quality control on the wafer fab side often repeats critical cation analysis by ICP-MS after package opening. The limits in Table 1 are representative supplier specifications for high-selectivity mixtures; individual models may differ. Model nomenclature varies by manufacturer, but it usually includes a prefix, ratio code, grade code, and packaging code. A product code containing “BOE-7:1-EL” identifies a 7:1 buffered oxide etchant with electronic/EL purity. Purchasing specifications should state whether the etchant is pre-blended or mixed from separate HF and NH4F sources. Pre-blended electronic/EL grade reduces point-of-use blending errors and is supplied with certification; on-site blending from technical-grade raw materials can introduce trace metal contamination and batch inconsistency.
| Property | Representative electronic/EL-grade specification | Measurement method |
|---|---|---|
| NH4F concentration | 32.0 wt% to 36.0 wt% | acid-base titration |
| HF concentration | 5.8 wt% to 7.2 wt% | acid-base titration |
| Trace cations per element | ≤100 ppb | ICP-MS |
| Particles ≥0.2 µm | ≤100 particles mL⁻¹ | laser particle counter |
| Chloride | ≤500 ppb | ion chromatography |
| Sulfate | ≤500 ppb | ion chromatography |
| Specific gravity at 25 °C | 1.10 to 1.15 | digital density meter |
Mobile-ion control is critical for gate oxide reliability. Sodium and potassium are controlled in electronic/EL grade to prevent drift in threshold voltage and oxide breakdown fields. Metal impurities such as iron, copper, and aluminum can increase surface microroughness and degrade carrier lifetime. The electronic/EL-grade limits reflect compatibility with front-end processing before metallization, not only with back-end cleanrooms. The limits align with SEMI C7 guidance for HF-based etchants. Packaging in PFA or FEP with sealed foil overwrap limits recontamination. Storage at 5 °C to 25 °C is recommended; repeated opening accelerates composition drift because HF vapor loss and water uptake alter the ratio. Final rinse water is specified to meet ASTM D5127-13 electronic-grade water quality.
In batch oxide release over nitride spacers, the product is typically introduced after dry etch residue removal and before final deionized-water rinsing. The etch rate reference is established by measuring monitor wafers with the same film stack. A typical process uses a 7:1 bath at 25 °C, with endpoint set by oxide monitor thickness and a fixed overetch of 10% to 20% of the oxide removal time, not of nitride thickness. However, the overetch fraction must be constrained by the nitride thickness budget and the film's wet etch rate. If the nitride stop layer is 10 nm and the product etches LPCVD nitride at <1 nm/min, an overetch of 10 s removes less than 0.2 nm of nitride under ideal conditions; that margin disappears if the film is PECVD or plasma-damaged.
In single-wafer tools, uniformity depends on rotation speed, nozzle geometry, and etchant dispense rate. Because single-wafer exposure times are short, oxide etch rate differences of 2 °C in bath temperature can produce measurable center-to-edge variation. Recirculation and in-line concentration monitoring are therefore more important for single-wafer use than for immersion batch use. Field data from production tools show that lot-to-lot etch variation often correlates with nitride film density drift from the deposition furnace and with photoresist residue carryover, not with etchant lot variance.
The etchant is incompatible with aluminum, titanium, and titanium nitride. If these materials are exposed at the wafer edge or through pinholes in an overlying oxide, localized metal loss occurs during the etch and following rinse. Wet benches dedicated to fluoride chemistry prevent cross-contamination with ammonia-based SC-1 solutions or amine-containing solvents; mixing such residues with HF-bearing baths creates localized reactions, salt precipitation, and airborne fluoride hazards. Equipment used with this product must comply with SEMI S2 and applicable local ventilation requirements.
High-selectivity product differs from standard BOE mainly in buffering strength and selectivity persistence. Standard BOE at 6:1 removes thermal oxide at similar rates but has lower tolerance to bath loading; as the bath ages, nitride attack increases and the selectivity window narrows. Dilute HF at 0.5 wt% produces smooth surfaces and is used in single-wafer cleaning, but its oxide etch rate is too low for thick sacrificial oxide removal. Vapor HF is an alternative for MEMS release because it avoids liquid stiction, but its throughput is lower and moisture control is critical. Hot phosphoric acid at 160 °C removes silicon nitride selectively to oxide, inverting the etch direction and making it unsuitable for oxide release over nitride stop layers.
| Etchant system | Thermal SiO₂ removal rate | LPCVD Si₃N₄ removal rate | Selectivity direction | Primary process role |
|---|---|---|---|---|
| High-selectivity BOE 7:1 electronic/EL | 100–140 nm/min at 25 °C | <1 nm/min | SiO₂ > Si₃N₄ | oxide release over nitride |
| Standard BOE 6:1 | 80–120 nm/min at 25 °C | 1–2 nm/min as-aged | SiO₂ > Si₃N₄ but lower margin | general oxide strip |
| Dilute HF 0.5 wt% | 5–15 nm/min at 25 °C | <0.2 nm/min reported | SiO₂ > Si₃N₄ | single-wafer surface cleaning |
| Hot phosphoric acid 85% | <0.5 nm/min at 160 °C | 4–8 nm/min on LPCVD nitride | Si₃N₄ > SiO₂ | nitride strip |
Selection between these chemistries depends on the oxide removal amount, the nitride stop-layer thickness, the exposed materials, and the available process margin. A high-selectivity wet fluoride etchant is normally selected when oxide removal is sufficiently large that dHF would require excessive process time, when nitride loss must remain below a few angstroms, and when the wafer has no exposed HF-sensitive metal. It is not a direct replacement for hot phosphoric acid, nor for vapor HF release tools, and process qualification must compare the full film stack rather than isolated etch rates.
High-selectivity wet etching is not appropriate for all silicon-containing dielectrics. Silicon oxynitride etch rates lie between those of oxide and nitride; if the film is oxygen-rich, the apparent selectivity collapses. Pad oxide under a nitride mask can be attacked at pinholes or stress-induced cracks, producing lateral undercut. The product is not designed for aluminum, copper, germanium, or III-V lift-off applications. Published data for those material systems is limited, and users must qualify any use outside silicon dioxide and silicon nitride structures.
The process window is bounded by temperature, buffer ratio, dissolved silicon, and nitride film density. Bath operation above 40 °C accelerates nitride hydrolysis and can reduce selectivity. Operation below 15 °C may require etch-time correction and longer immersion. The etchant should not be mixed with strong bases, amines, or incompatible solvents. Waste neutralization must account for fluoride content and local discharge limits. The final process specification should fix the model, the bath replenishment procedure, and the allowable overetch budget, rather than relying only on the supplier's selectivity claim.
