HBTU / HATU guanidination of free amine

Free α-amine attacks the uronium/guanidinium carbon of HBTU/HATU faster than the carboxyl partner does — gives a permanent Nα-tetramethylguanidinium adduct that stops chain elongation. +98 Da and your peptide is now truncated.

Why it happens (mechanism)

HBTU/HATU consist of a benzotriazolyl + a tetramethyluronium center. The intended path: carboxyl + uronium → O-acylisouronium → active ester (OBt/OAt). But the same uronium carbon is electrophilic toward amines too. If the free α-amine of the resin-bound peptide encounters HATU before pre-activation has consumed it, the amine attacks the uronium carbon → stable tetramethylguanidinium urea. The peptide can no longer accept the next coupling.

When it strikes (triggers)

Charging HATU + carboxyl + DIEA + amine all together without pre-activation. Excess HATU vs. carboxyl (e.g., from imprecise weighing of poorly soluble Fmoc-AA). Sluggish carboxyl activation (hindered residues like Aib, MeVal). Head-to-tail cyclization (carboxyl + amine in same molecule). Microwave-aggravated.

How to spot it (MS signature)

+98 Da (or +97 depending on exact protonation) on the N-terminal residue. Truncated peptide; chain stops at this residue. Sometimes paired with a +126 Da DIC-derived hydantoin variant.

How to prevent it

• Pre-form the active ester: mix carboxyl + HATU + DIEA, wait 2-5 min, then add to the resin. Don't dump all together.
• Use 0.95 eq HATU vs. carboxyl (slight understoichiometry — sacrifices ~5% carboxyl to prevent guanidination, well worth it).
• For head-to-tail cyclization: switch to PyBOP / PyAOP (phosphonium-type, inert to amines). Don't use HATU for cyclization.
• For very hindered couplings (e.g., -Aib-Aib-): use DIC + Oxyma instead of HATU + DIEA.

If it already happened (salvage)

• Guanidinium urea is stable. The peptide is truncated. Re-synthesize.

Source

Yi Yang, Side Reactions in Peptide Synthesis (Elsevier, 2016), Chapter 5, §5.2.