Optimization of Lipid Nanoparticle Analysis Across Universal Detectors

Significance of the topic

Lipid nanoparticles (LNPs) are central to modern nucleic acid delivery, including mRNA vaccines. Accurate, precise quantitation of the four core lipid components (ionizable lipid, phospholipid, PEG-lipid, and cholesterol) is essential to control potency, stability and manufacturability. Because these lipids lack chromophores, universal detectors such as charged aerosol detection (CAD) and evaporative light scattering detection (ELSD) are commonly used; comparative performance data are needed to guide analytical method selection for formulation development and quality control.

Objectives and overview of the study

This work evaluated and optimized a reversed‑phase UPLC method for a representative LNP formulation (SM‑102 ionizable lipid, DSPC phospholipid, DSPE‑PEG 2000, and cholesterol) to compare CAD and ELSD performance. Key aims were to determine limits of quantitation (LOQ), linearity, repeatability, recovery (accuracy), and chromatographic resolution for a QC‑relevant concentration range (standards from 5 to 500 µg/g).

Methodology and instrumentation

Method summary:

• Column: CORTECS Phenyl, 90 Å, 1.6 µm, 2.1 x 100 mm.
• Mobile phases: 10 mM ammonium acetate in (A) 90/10 methanol/water and (B) 90/10 acetonitrile/water; gradient elution used to separate the four lipids.
• Chromatographic conditions: flow 0.40 mL/min, injection 5 µL, column 30 °C, sample 12 °C; representative gradient moved from high %A to high %B and back.
• Calibration: nine concentration levels spanning 5–500 µg/g; standards contained the four lipids at equal concentrations while representative samples were prepared at a typical formulation ratio (SM‑102 : cholesterol : DSPC : DSPE‑PEG 2000 ≈ 50 : 38.5 : 10 : 1.5).
• LOQ criteria: signal‑to‑noise ≥ 10 and area %RSD ≤ 15% across six injections.

Instrumentation used:

• ACQUITY Premier UPLC system (Waters).
• Detector A: Charged aerosol detector (CAD) — evaporator temperature ~40 °C, power function value used, data rate 2 Hz.
• Detector B: Evaporative light scattering detector (ELSD) — drift tube ~50 °C, nebulizer/gas pressure ~40 psi, gain adjusted, data rate 2 Hz.

Sample concentrations used to demonstrate detector performance reflected LOQ differences: for CAD a representative sample contained ~167 µg/g SM‑102, 128 µg/g cholesterol, 33 µg/g DSPC, 5 µg/g DSPE‑PEG 2000; for ELSD a higher concentration sample was used (~500, 385, 100, 15 µg/g respectively) because of ELSD’s higher LOQs.

Main results and discussion

LOQ and sensitivity:

• CAD achieved lower LOQs across all four lipid components, with an operational LOQ of 5 µg/g for each component in this study.
• ELSD required higher LOQs for some analytes (10–15 µg/g for DSPC and DSPE‑PEG 2000; an overall illustrative ELSD LOQ standard was shown at 15 µg/g).

Linearity and dynamic range:

• Both detectors produced strong log‑log calibration linearity over the LOQ to 500 µg/g range (R² ≥ 0.994). CAD consistently gave higher R² values and maintained linearity across a wider concentration range for all components.

Repeatability (area %RSD at LOQ):

• CAD demonstrated substantially better repeatability at low concentrations. For cholesterol, SM‑102 and DSPE‑PEG 2000 the area %RSD values on CAD were roughly half those observed with ELSD.
• For DSPC the repeatability was similar between detectors, but CAD still showed superior performance at the low end.

Accuracy (sample recovery):

• Using detector‑specific calibration curves, CAD recoveries clustered tightly near 100% for all components (examples: cholesterol ≈ 106%, DSPC ≈ 94%, SM‑102 ≈ 104%, DSPE‑PEG 2000 ≈ 103%).
• ELSD recoveries were more variable (examples: cholesterol ≈ 84.5%, DSPC ≈ 79.8%, SM‑102 ≈ 137.2%, DSPE‑PEG 2000 ≈ 98.6%), indicating potential bias and larger uncertainty for some analytes at the studied levels.

Chromatographic resolution:

• CAD provided improved resolution on the critical DSPC/SM‑102 pair compared to ELSD. Improved separation is attributable in part to the ability to inject and detect at lower mass loadings when using the more sensitive CAD, reducing peak broadening and coelution effects.

Overall discussion:

• The detector physical principles explain much of the observed difference: CAD measures nonvolatile aerosolized analyte mass and tends to give more uniform, mass‑related response with high sensitivity; ELSD signal depends on analyte scattering after solvent evaporation and can show poorer sensitivity and greater variability for low‑response species such as PEG‑lipids.
• For LNP lipid panels where species span a wide polarity and concentration range, CAD offers better low‑end quantitation, repeatability and accuracy, improving confidence in QC and formulation analytics.

Benefits and practical applications of the method

Practical advantages when using CAD with this UPLC method:

• Lower and more consistent LOQs enable measurement of low‑abundance lipid components (e.g., PEG‑lipids) without excessively concentrating samples.
• Better repeatability at LOQ enhances assay precision for release and stability testing.
• Improved resolution on critical pairs reduces the need for extensive method reoptimization or additional separation steps.
• Robust recoveries near 100% support accurate assay reporting for formulation development, in‑process control and final product QC.

These attributes make the CAD‑based LC method well suited to LNP formulation characterization, lot release assays, comparability studies and stability monitoring.

Future trends and potential applications

Key directions and opportunities building on this work include:

• Broader adoption of CAD for routine LNP QC and regulatory submissions, supported by inter‑laboratory method harmonization and standardized calibration practices.
• Integration with orthogonal detectors (e.g., LC‑MS) for structural confirmation of novel ionizable lipids and degradants while relying on CAD for robust quantitation.
• Method miniaturization and automation to increase throughput for formulation screening and manufacturing support.
• Continued improvement of stationary phases and gradients to further resolve complex lipid mixtures and related impurities.
• Development of certified reference materials and standardized reporting workflows to reduce variability across sites and platforms.

Conclusions

The comparative study demonstrates that CAD outperforms ELSD for quantitative LC analysis of a representative LNP formulation: CAD delivered uniformly lower LOQs (5 µg/g), superior repeatability and tighter recoveries, and better chromatographic resolution for a critical lipid pair. Although both detectors can produce acceptable log‑log calibration linearity, CAD’s greater sensitivity and precision make it the recommended universal detector for LNP lipid quantitation in formulation development and QC contexts. ELSD may remain useful where CAD is unavailable or for complementary analyses, but caution is warranted for PEG‑lipids and low‑level components.