An aqueous bismuth nitrate feed is prepared by dissolving bismuth nitrate pentahydrate in purified water acidified with nitric acid; the free-acid concentration is maintained at or above 0.5 mol/L because bismuth oxynitrate precipitation accelerates rapidly when the pH of a purely nitrate system rises above approximately 1.5–2.0. The dissolution is exothermic, and the vessel is cooled to keep the solution below 35 °C during preparation. For the formula unit BiO(C₇H₅O₃), the reaction with salicylic acid and sodium hydroxide proceeds as Bi(NO₃)₃ + C₇H₆O₃ + 3 NaOH → BiO(C₇H₅O₃) + 3 NaNO₃ + 2 H₂O. When sodium salicylate is used instead of salicylic acid, the sodium hydroxide charge is reduced accordingly; the mass of sodium salicylate required per kilogram of bismuth nitrate pentahydrate is 330.0 g, and the equivalent salicylic acid requirement is 284.7 g, based on the molecular weights of bismuth nitrate pentahydrate (485.07 g/mol), salicylic acid (138.12 g/mol), sodium salicylate (160.10 g/mol), and bismuth subsalicylate (362.09 g/mol). The theoretical bismuth content of bismuth subsalicylate is 57.7% w/w, which provides a mass-balance check for precipitation completeness.

What Limits the Operational pH Window During Direct Precipitation?

The limiting boundaries of the operational pH window are fixed by two competing equilibria. The carboxyl pK_a1 of salicylic acid is 2.97 at 25 °C; the phenolic pK_a2 is approximately 13.7 and does not contribute under acidic processing conditions. Using the Henderson-Hasselbalch equation, the fraction of salicylate anion at pH 2.0 is 9.7%, at pH 2.8 is 40.3%, and at pH 3.4 is 72.9%. Below pH 2.8, the dissolved salicylate concentration is too low to maintain a reproducible precipitation rate, and unreacted bismuth remains in the mother liquor; above pH 3.4, bismuth oxynitrate and hydroxide precipitation compete with salicylate complexation, producing foreign phases that are observed as crystalline impurities by powder X-ray diffraction. The practical control set point is therefore pH 3.0 with a control dead band of ±0.2 units. The pH electrode is a flat-glass autoclaveable device with a PTFE diaphragm, mounted in a retractable fitting; fouling by precipitated product raises the apparent pH by 0.2–0.4 units if the electrode is not cleaned after each batch with 0.1 mol/L hydrochloric acid.

In production-scale equipment, the nitric acid-containing bismuth nitrate feed and the alkaline salicylate solution are delivered through independent PTFE-lined dosing lines that converge in a submerged dip tube below the liquid surface of the glass-lined reactor. The reactor is equipped with a retreat-curve impeller having a diameter of 0.45–0.50 tank diameter, a two-baffle configuration with baffle width T/12, and a jacket capable of maintaining the batch at 60–65 °C with a deviation of ±1 °C. The bismuth nitrate feed vessel is fabricated from 316L stainless steel with a 2B surface finish and is equipped with external cooling; at nitric acid concentrations above 10% w/w and temperatures above 60 °C, PTFE-lined carbon steel is preferred because 316L exhibits pitting and intergranular corrosion under those conditions. Addition time for a production batch is typically 60–120 minutes; addition rate is controlled by peristaltic or PTFE-diaphragm pumps with a flow accuracy better than ±1% of setpoint. Agitation is held at 60–90 rpm in a 3000 L reactor, giving an impeller Reynolds number well above 10,000 and a tip speed of approximately 2.0–3.0 m/s. Scale-up is performed by constant tip speed rather than constant power per unit volume because the precipitation is mesomixing-sensitive; local supersaturation near the addition dip tube controls primary nucleation density. The bismuth nitrate feed is clarified through a 0.45 µm PTFE membrane filter to remove undissolved bismuth nitrate fines and silica particles that would otherwise seed the batch. The salicylate solution is clarified through the same type of filter. A pH controller with a proportional-integral loop modulates sodium hydroxide addition at 1.0 mol/L; stronger base solutions are not used because local pH spikes above 3.5 create bismuth oxynitrate inclusions.

Filterability improves when nucleation is separated from crystal growth.

The separation of nucleation and crystal growth is accomplished through controlled addition, seeding, and a post-addition equilibration hold. Rapid addition of the bismuth nitrate stream produces a steep supersaturation spike at the dip tube, generating a high primary nucleation density and a high specific surface area cake with poor filtration. The same mass of product can be obtained with a filtration time difference of 2–3 hours versus 8–12 hours when the addition time is extended and the pH is maintained within the target window. Seed crystals of bismuth subsalicylate may be introduced at 0.1–1.0% w/w of the expected dry product mass after the first 10–15 minutes of addition; seeding suppresses secondary nucleation and shifts the particle size distribution toward larger, more uniformly agglomerated particles. The post-addition hold is maintained at 60–65 °C for 60–90 minutes to allow Ostwald ripening and desupersaturation before cooling. Agitation during the hold is reduced to 45–60 rpm to avoid crystal breakage while still maintaining complete solids suspension. The particle size distribution of the dried product is measured by laser diffraction according to ISO 13320:2020; particle size is used as process monitoring rather than as a release test unless process validation demonstrates a direct correlation to dissolution or content uniformity. Filtration is carried out on a pressure nutsche filter with a 20–30 µm polypropylene or PTFE-coated filter cloth, with nitrogen pressure of 0.2–0.4 MPa and a cake thickness limit of 10–15 cm to prevent channels and incomplete washing.

Residual Nitrate, Washing, and Drying End-Point Control

Washing of the filtered cake removes sodium nitrate, excess nitric acid, and unconverted salicylate species. The cake is washed with purified water at 55–65 °C in 2–3 displacement volumes. The wash filtrate conductivity is monitored with a conductivity probe that is calibrated against USP <645>; washing continues until the conductivity falls below 100 µS/cm or until the nitrate concentration in the filtrate falls below the limit established during process validation. Residual nitrate in the final product is determined by ion chromatography with conductivity detection using a method validated under ICH Q2(R1). If ethanol or methanol is used to improve drying, the residual solvent content is controlled according to ICH Q3C and measured by headspace gas chromatography according to USP <467>. The wet cake is not exposed to chloride-bearing materials; even trace chloride can produce bismuth oxychloride, which is difficult to detect at low levels and alters the infrared spectrum. The wet cake is dried in a vacuum tray dryer at 70–75 °C and 150–180 mbar absolute for 8–12 hours, with tray loading limited to 2–3 kg/m² to prevent non-uniform heat transfer. Loss on drying is measured by USP <731>; a typical endpoint is below 0.5%, but the final limit is established by the compendial monograph and process validation. Water content may also be determined by Karl Fischer titration according to USP <921> if the hydration state of the product must be distinguished from surface moisture. The dried material is passed through a 0.5 mm conical mill to break soft agglomerates without generating fines; the mill is nitrogen-inerted if dust deflagration hazards exist.

When cooling is imposed before desupersaturation is complete

Cooling is used to reduce residual solubility and increase yield, but the thermal ramp must follow the pH equilibration hold. If the batch is cooled from 60–65 °C to 20–25 °C before the salicylate complexation equilibrium has reached its endpoint, the drop in temperature increases supersaturation and triggers secondary nucleation on existing crystals, producing a bimodal particle size distribution and a less filterable slurry. The cooling ramp is therefore specified as 0.2–0.5 °C/min over 90–180 minutes, with continuing pH control at pH 3.0 ± 0.2. A rapid cooling rate above 0.5 °C/min can also trap mother liquor in the crystal lattice and increase residual nitrate after washing, because fast crystal growth forms liquid inclusions. Field observations from production vessels indicate that wall encrustation occurs when the jacket-batch temperature difference exceeds 10 °C; this encrustation reduces heat-transfer coefficients and must be removed by acid cleaning. After the slurry reaches 20–25 °C, it is held for at least 30 minutes before filtration to allow full particle equilibration. The cooling jacket is operated with a temperature differential of no more than 10 °C between jacket and batch to prevent wall nucleation and encrustation on glass-lined surfaces.

Analytical control of the precipitated material is organized around three production stages: raw material qualification, in-process monitoring, and release testing. Bismuth nitrate pentahydrate, salicylic acid, sodium hydroxide, nitric acid, and purified water are tested according to 21 CFR 211.84; bismuth nitrate assay is corrected for the water of hydration before weighing. In-process sampling and testing during precipitation and drying are governed by 21 CFR 211.110 and ICH Q7 Section 8.1. The pH meter is calibrated before each batch with two buffer solutions and checked after the batch; the measurement is performed according to USP <791>. Powder X-ray diffraction according to USP <941> is used to detect bismuth oxynitrate or bismuth oxychloride contamination; a sharp foreign reflection in the low-angle region can indicate incomplete washing or chloride contact. Particle size distribution is measured by laser diffraction according to ISO 13320:2020; the sample is dispersed in purified water with ultrasonication and the obscuration is held within the instrument manufacturer’s recommended range. The residual nitrate method uses ion chromatography with a column and suppressor configuration that separates nitrate from sulfate and phosphate; method validation is performed according to ICH Q2(R1). The final product is tested for bismuth content by complexometric titration or atomic absorption spectrophotometry; the bismuth content is reported on the dried basis.

Published data for this specific configuration is limited where the compendial monograph does not specify particle size acceptance criteria; process capability analysis is used to set alert and action limits for particle size and residual nitrate.