In Fmoc/tBu solid-phase synthesis of peptide active pharmaceutical ingredients destined for injectable GLP-1 receptor agonist dosage forms, HBTU is used as a pharma-grade uronium-type coupling activator. The reagent is pre-dissolved in anhydrous DMF or NMP at 0.2–0.5 M, the resin-bound carboxyl component is added, and DIPEA is introduced at 1.5–2.0 equivalents relative to the carboxyl component. Preactivation is maintained for 1–3 min at 20–25°C. Coupling to the deprotected amine resin proceeds for 30–45 min. Residual free amine after coupling is checked by Kaiser or chloranil test; a positive response triggers a second coupling with fresh HBTU solution. Synthesis is performed at resin substitution between 0.3 and 0.8 mmol/g. HBTU is selected over carbodiimide activators because the in situ HOBt ester suppresses oxazolone-mediated racemisation at C-terminal cysteine and histidine residues. Prolonged preactivation of HBTU in the absence of the carboxyl component is avoided because uronium reagents can guanidinylate unprotected amino groups in the reaction vessel, generating tetramethylguanidinium-related side products that complicate preparative HPLC purification.

After chain assembly, the peptide is cleaved from the resin with TFA/TIS/H₂O 95:2.5:2.5 v/v/v. HBTU-derived HOBt and tetramethylurea are removed from the crude peptide by preparative reversed-phase HPLC using an octadecylsilyl bonded silica column of 10–15 μm particle size and 100 Å pore size. The mobile phase is an acetonitrile/water gradient containing 0.1% TFA. Fractions are pooled, concentrated below 35°C, and lyophilised. For an injectable peptide API, the release specification includes assay by HPLC-UV of not less than 99.0%, total impurities not more than 1.0%, and unspecified individual impurities controlled per ICH Q3A(R2) at reporting threshold 0.05% w/w for a maximum daily dose ≤2 g/day. A validated LC-MS/MS method with an LOQ of 0.05% w/w is applied to monitor intact HBTU, HOBt, and tetramethylurea in the final API powder. HBTU is not a pharmacopoeial-listed impurity in finished peptide monographs, so the acceptance limit is justified through process validation and toxicological data; published data for this specific configuration is limited, and therefore an in-house limit of not more than 0.05% w/w for each reagent-derived impurity is commonly adopted. Residual DMF from the coupling solution is controlled under ICH Q3C(R8) at 880 ppm for an injectable product when the daily dose is 10 g.

SPPS campaigns using HBTU show batch-to-batch variability when the reagent is stored improperly. HBTU is moisture-sensitive; its hexafluorophosphate salt hydrolyses in humid air to produce HOBt and tetramethylurea. Pharma-grade material must be stored in original sealed containers at 2–8°C under nitrogen. Reagent lot qualification includes assay by HPLC-UV not less than 99.0%, water content below 0.5% by Karl Fischer, and residual DMF by GC below 500 ppm. If the reagent is added from a container that has been repeatedly opened, coupling efficiency drops due to hydrolysis of the uronium centre to inactive HOBt; the resulting incomplete coupling is detected by a positive Kaiser test and corrected with double coupling. The production line therefore uses pre-weighed single-use aliquots to avoid moisture ingress.

What Residual Reagent Controls Apply When the Peptide API Is Compressed into Oral Tablets?

Compression of a synthetic peptide API into an oral tablet introduces an entirely different residual-solvent and residual-reagent matrix than an injectable. Oral peptide tablets using absorption enhancers such as sodium N-[8-(2-hydroxybenzoyl)amino]caprylate require the peptide API to be diluted into a relatively hydrophobic core containing high enhancer mass and only a small amount of API. HBTU-derived HOBt, if not cleared below 0.05% w/w, can function as a weakly acidic contaminant in the tablet core; HOBt has a pKa near 4.6, and the resulting low-pH microclimate can shift the dissolution behaviour of methacrylic acid–ethyl acrylate copolymer coatings intended to release at pH 5.5. In addition, residual tetramethylurea is water-miscible and hygroscopic, and at levels above 0.2% w/w in the API it may increase water uptake in the core during stability storage, altering tablet hardness and disintegration time. The oral tablet process therefore uses API with residual impurities controlled by the same LC-MS/MS method as the injectable grade, but with additional monitoring of brown or amber discoloration during forced degradation because HOBt oxidation products can form in the presence of crospovidone at accelerated storage conditions of 40°C/75% RH.

Compression runs on rotary presses operating at 20–60 rpm with pre-compression force 5–12 kN and main compression force 8–20 kN require peptide API with bulk density above 0.25 g/mL and residual moisture below 3.0% to prevent sticking and picking. The analytical data package must comply with FDA 21 CFR 211.194 and USP <621> for chromatographic reproducibility. Residual solvent compliance is governed by USP <467>. The table below summarises the control matrix for solvents commonly encountered in HBTU-mediated SPPS and downstream purification.

Lyophilised Injection Cakes and Reagent-Derived Impurity Clearance

For lyophilised injectable peptide products, the final dosage form is aseptically filled after sterile filtration, but soluble reagent-derived impurities are not removed by filtration. The API must therefore enter the formulation train with HBTU, HOBt, and tetramethylurea levels already at the release limits. The formulation is prepared in water for injection, pH adjusted to 5.0–6.5 with dilute hydrochloric acid or sodium hydroxide, and tonicity adjusted with mannitol to 280–320 mOsm/kg. Freeze-drying is conducted in a lyophiliser with shelf temperature ramp from -40°C to 25°C over 24–48 h. Primary drying is maintained below the collapse temperature; for peptide-mannitol cakes this is typically -25°C to -15°C, with chamber pressure 50–100 mTorr. Secondary drying at 20–25°C reduces residual moisture below 1.0%. The non-volatile PF₆ counterion from HBTU, if not removed by preparative HPLC, will remain in the final cake and can alter the ionic strength of the reconstituted solution. Therefore the API is screened for total phosphorus by ICP-MS before lyophilisation, with an in-house limit commonly set at 10 ppm. This limit is not a pharmacopoeial requirement; it is derived from the known effect of ionic impurities on reconstitution time and injection-site discomfort. The freeze-dried cake must also pass USP <1> injectable product tests and USP <790> visible particulate matter criteria after reconstitution.

Lyophilisation of synthetic peptide APIs manufactured with HBTU includes a pre-filtration hold time validation. A sterile-filtered peptide solution can be held at 2–8°C for up to 24 h before freeze-drying if solution stability is demonstrated; residual HBTU-derived hydrolysis products may continue to form during this hold if the peptide contains free amino groups. The hold time study includes measurement of HOBt and tetramethylurea at 0 h, 12 h, and 24 h by LC-MS/MS. If the concentration of either species increases by more than 10% relative to time zero, the hold time is reduced. The freeze-drying chamber is cleaned to prevent cross-contamination of reagent-derived compounds; a swab recovery of less than 1 ppm for HBTU-related species is used in the cleaning validation protocol. This is not a pharmacopoeial requirement but arises from the high sensitivity of peptide injections to non-volatile organic impurities.

When PLGA Microspheres Expose Residual Coupling Reagent Reactivity

Long-acting injectable microspheres and granule depots of synthetic peptide APIs are commonly prepared by double emulsion solvent evaporation. The inner aqueous phase contains the peptide, while the organic phase contains PLGA. Residual HBTU-derived species in the peptide API are transferred into the inner aqueous phase. HOBt is weakly acidic and can drive the inner aqueous phase pH below 4.0; PLGA ester hydrolysis is acid-catalysed, and this low pH can accelerate molecular weight loss of PLGA during solvent removal and produce a widened particle size distribution and increased initial burst release. Published data for HBTU residue effects in PLGA microspheres is limited, but the acid-catalysed PLGA degradation mechanism is established in polymer literature. Consequently, peptide API batches intended for PLGA microspheres are released only when HOBt is not more than 0.05% w/w and tetramethylurea is not more than 0.1% w/w by LC-MS/MS. The double emulsion is formed with a high-shear rotor-stator disperser at 8,000–15,000 rpm, and the resulting microspheres are collected with a particle size Dv50 of 30–50 μm. Solvent removal is performed in a stirred tank at 25°C with controlled ramp rate of 1°C/min, and residual dichloromethane is reduced below 600 ppm per ICH Q3C(R8) Class 2. If the peptide API has not been fully desalted, residual HBTU-derived ionic species can also destabilise the primary W/O emulsion, causing coalescence and irregular granule morphology. The final microsphere product is assessed for in vitro release using USP <711> apparatus 4 or a modified USP <711> method.

PLGA microspheres prepared from APIs with residual HBTU-derived acids show an altered in vitro release at the early time point. The initial burst is influenced by the pH of the inner aqueous phase; pH values below 4.0 can increase PLGA hydrolysis during the solvent-evaporation step. In a double emulsion, the inner water phase is often buffered at pH 5.0–6.0; if HOBt is present at 0.2% w/w in the peptide mass, the buffer capacity of a 50 mM acetate or phosphate phase may be insufficient to maintain pH, because the amount of HOBt is approximately 0.002 mmol per mg peptide at that level. The resulting pH shift may be small in absolute terms but occurs at the W/O interface where PLGA contacts the inner water phase. Microsphere hardening in an aqueous quenching bath at 4°C for 4 h is monitored by pH and conductivity. The final product is characterised by scanning electron microscopy for surface pores; residual HBTU-derived ionic impurities can increase surface porosity, leading to fast initial release and possible needle clogging during injection. A pre-encapsulation API specification of not more than 0.05% w/w HOBt and 0.1% w/w tetramethylurea reduces this risk. The PLGA grade and lactide:glycolide ratio are selected with the peptide manufacturer because strongly acidic residues can catalyse the hydrolysis of high-lactide PLGA more rapidly; high-lactide PLGA has slower inherent degradation but is more sensitive to acid catalysis.

For enteric-coated peptide granules, HBTU-mediated synthesis is followed by fluid-bed layering and pH-sensitive coating. The API is dissolved or suspended in an aqueous binder solution and sprayed onto microcrystalline cellulose spheres in a fluid-bed granulator fitted with a Wurster insert. Inlet air temperature is maintained at 30–35°C, product temperature at 25–28°C, and spray rate at 5–10 g/min per kg core. Granules are dried to a water content below 2.0% and then coated with an aqueous dispersion of methacrylic acid–ethyl acrylate copolymer, Type C, often at 15–25% w/w polymer weight gain. HOBt residues in the API can acidify the granule surface microclimate and may destabilise the aqueous latex dispersion during curing at 40°C, leading to incomplete coalescence of the enteric film. The acid resistance of the finished granules is tested by USP <711>: not more than 10% release in 0.1 M HCl after 2 h, then not less than 75% release in pH 6.8 phosphate buffer within 45 min. Granule size is controlled by sieve analysis with Dv50 between 300 μm and 500 μm; oversized material is milled and reintroduced only if the enteric coating fraction remains uniform. This process-specific requirement exists because HBTU-derived residues are not removed by fluid-bed drying, and any pH shift in the coated granule population can alter the release profile of the entire batch.

Enteric-coated peptide granule manufacturing is carried out in a fully perforated fluid-bed coater with bottom spray. The coating dispersion is continuously stirred at 500–1,000 rpm; the inlet air humidity is controlled to a dew point below 5°C to ensure film formation. HBTU-derived HOBt is water-soluble and can migrate to the surface of the peptide layer during drying, where it interacts with the anionic methacrylic acid copolymer dispersion. This interaction can increase the minimum film-forming temperature by 2–5°C, which is relevant because the enteric coating is applied at product temperature 25–28°C. If the product temperature falls below the modified minimum film-forming temperature, the film becomes porous and the granules fail the acid resistance stage. The granule formulation is therefore adjusted with 1–3% w/w talc or glyceryl monostearate to buffer surface acidity and improve coating wetting. During process development, HOBt spiking studies are conducted at 0.05%, 0.1%, and 0.2% w/w in the API to determine the failure threshold. Published data for this specific configuration is limited, but the spiking study is designed according to a 3-factor Doehlert matrix covering inlet air temperature, HOBt level, and polymer weight gain. The resulting granules are packed in high-density polyethylene bottles with desiccant and stored at 25°C/60% RH for stability testing.

Capsule Fill Mass Variability Tracks Peptide API Particle Size and Reagent Residue

Synthetic peptide APIs intended for hard-shell capsule filling are often low-dose actives dispersed in a ternary excipient system. The API must be milled and sieved to a particle size Dv90 below 250 μm before blending with microcrystalline cellulose and lactose monohydrate. Residual HBTU-derived tetramethylurea can increase hygroscopicity of the API powder, and if residual moisture rises above 3.0% w/w, dosator-type capsule machines show erratic fill weight and occasional nozzle blockage. Blend uniformity is assessed by USP <905> on 10 sampling points, with acceptance value not more than 15.0% RSD. Capsule filling is performed at 30–40% RH. Capsule fill weight ranges from 100 mg to 500 mg, with size 0 or 1 capsules; API content ranges from 1 mg to 20 mg. The final capsule product is tested for disintegration per USP <701>; for immediate-release capsules, disintegration must occur within 30 min in 0.1 M HCl at 37°C. If the peptide API is enteric-coated in a multi-particulate capsule, disintegration is assessed after the acid stage and buffer stage. In all cases, HBTU is not present as an excipient in the finished capsule; it is a process reagent from API synthesis, and its residual content is controlled at the API stage. Capsule products containing synthetic peptide APIs manufactured with HBTU are supported by batch records that document the complete clearing of reagent-derived impurities before blending, as required by FDA 21 CFR 211.100 and 21 CFR 211.160.

Capsule products are packaged in blisters or bottles; residual HBTU-derived species are not volatile but can interact with gelatin capsule shells. Gelatin shell crosslinking is not directly caused by HBTU residues, but acidic HOBt can contribute to shell brittleness if the API and excipient blend is stored above 40°C. For that reason, finished capsules are stored below 25°C. Capsule filling with low-dose peptide APIs requires geometric dilution because the API is cohesive and can form agglomerates with residual tetramethylurea. The blend is passed through a 600 μm mesh screen before final lubrication with magnesium stearate at 0.5–1.0% w/w for 3–5 min; longer mixing time can increase the hydrophobic character of the blend and slow dissolution. Dissolution testing of the capsule is performed in USP <711> apparatus 2 at 50 rpm in 900 mL of 0.1 M HCl. For immediate-release capsules, not less than 80% of label claim must be released within 30 min; if the capsule contains enteric-coated granules, the acid stage is run for 2 h, followed by pH 6.8 buffer. The capsule filler, typically a dosator machine with 10 to 30 dosing stations, is calibrated before each lot with placebo powder to ensure fill weight RSD below 2.0%. The API particle size is adjusted by air-jet milling to Dv90 below 100 μm for blend uniformity; milling can generate heat and may promote HOBt oxidation, so the mill is cooled with nitrogen and the product temperature is maintained below 30°C.