Atomic Absorption, ICP-MS, or HPLC-DAD — Which Method for Copper Quantification in Cu-Peptide Lots

Three analytical methods are commonly listed on Cu-peptide release reports: atomic absorption spectroscopy (AAS), inductively coupled plasma mass spectrometry (ICP-MS), and reversed-phase HPLC with diode-array detection (HPLC-DAD). They all measure copper content, but they answer different questions, have different sensitivities, and have different blind spots. Treating the three as interchangeable means missing the failure modes only one of them would catch.

This Note unpacks what each method measures, what its release-relevant strengths and blind spots are, and how a chemistry-led atelier combines them rather than picking one.

What each method actually measures

**Atomic absorption spectroscopy (AAS, including flame and graphite-furnace variants)** atomizes the sample (typically in a flame or a heated graphite tube), then measures absorbance of light at copper's characteristic atomic line (324.8 nm primary). The reading is proportional to the total copper concentration in the sample — bound or unbound, in any oxidation state, in any complex. Typical detection limit is ppb (μg/L); modern graphite-furnace variants reach ppt.

**Inductively coupled plasma mass spectrometry (ICP-MS)** ionizes the sample in an argon plasma at ~6,000 K (full atomization) and mass-analyzes the resulting ions. Copper is quantified at m/z 63 and 65 (the two natural isotopes). Like AAS, this measures total copper regardless of speciation; unlike AAS, ICP-MS measures essentially every element simultaneously, with detection limits at parts-per-trillion (pg/g) for most metals.

**HPLC with diode-array detection (HPLC-DAD)** separates the peptide-Cu complex from any free copper, free peptide, or other UV-absorbing species, then quantifies the complex peak by its UV absorbance. The DAD records absorbance across a wavelength range (typically 200-800 nm) per peak, so the d-d band can be tracked alongside the peptide-bond UV signal. The reading is the integrated peak area for the peptide-Cu complex — not total copper.

What each method catches that the others miss

• **AAS catches**: total Cu content, with high reliability + low instrument cost + minimal sample prep. Misses: speciation (free vs bound Cu look the same), trace metal contaminants (only Cu is measured).
• **ICP-MS catches**: total Cu at ppt sensitivity + every other trace metal in the sample at the same time + isotope ratios for forensic / provenance work. Misses: speciation, requires more sample prep and skilled operators.
• **HPLC-DAD catches**: peptide-Cu complex specifically (separates from free Cu, from uncoordinated peptide, from coordination variants); also shows the d-d band for coordination verification. Misses: total Cu (anything not in the labeled complex peak is invisible), and ppb-or-below trace contaminants.

The release question each method answers

**'Is the labeled copper content actually in the vial?'** — AAS or ICP-MS. Both quantify total Cu. A 1% GHK-Cu sample with labeled ~12% Cu by mass should show ~12% by AAS or ICP-MS within analytical tolerance. Disagreement of 10%+ from theoretical is a real composition issue.

**'Is the copper actually bound to the peptide, or is some of it free?'** — HPLC-DAD. The peptide-Cu complex peak area gives bound-Cu content; subtracting that from the AAS/ICP total Cu gives free Cu content. Excess free Cu (above ~5% of total Cu) is a real quality concern for cosmetic actives because of pro-oxidant effects.

**'Are there trace metal contaminants I should worry about?'** — ICP-MS, which simultaneously screens for the other ~30 metals of cosmetic-regulatory interest (Pb, As, Hg, Cd, Cr, Ni, etc.) at concentrations below cosmetic-product limits. AAS could be configured for any of these but is more typically run for one or two specific elements.

**'Is the copper coordination geometry intact?'** — HPLC-DAD with diode-array gives the d-d band along with the complex peak; UV-Vis on the bulk solution gives the same band at higher concentration. Both should agree.

Standard Cupratec release stack

Cupratec's standard release packet for GHK-Cu and AHK-Cu uses:

• **AAS for primary Cu quantification** — fast, robust, well-suited for the ppb-ppm range of routine release samples. Reported on every COA.
• **HPLC-DAD for peptide-bound Cu** — quantifies the complex peak specifically, computes the Cu:peptide molar ratio (combined with HPLC peptide quantification), shows the d-d band per peak.
• **UV-Vis bulk on reconstituted sample** — independent check of the d-d band at higher concentration, with the ΔE colour reading on the lyophilate as a separate orthogonal signal.
• **ICP-MS on request for trace-metal screen** — typical use case is when a finished-product launch needs the heavy-metal panel for regulatory notification (CPNP for EU, equivalent for other jurisdictions).

How a formulator should read the release packet

The most useful comparison: AAS-derived total Cu vs HPLC-DAD-derived bound Cu. The bound fraction (bound Cu / total Cu) should be ≥ 0.95 for a clean Cu-peptide lot. If the ratio is materially lower (say 0.85), part of the copper in the vial is not bound to the peptide — either uncoordinated free Cu or coordinated to something other than the labeled peptide. Either way, the active fraction is less than the total Cu suggests.

If the release packet shows only one of the three methods, request the other two on the next lot. The cost of one ICP-MS or one HPLC-DAD run is modest compared to the cost of a finished-product reformulation triggered by an undetected free-Cu issue.

Total Cu and bound Cu are different numbers. A release packet that shows only one is answering only one of the questions a formulator actually needs answered.