Blog

Determination of Trace Impurities in Nanomedicines: What Three Case Studies Teach Us

Dr. Maria Marioli November 12, 2026 in Nanomedicines

Impurities in nanomedicines are critical quality attributes (CQAs) as they can have a direct impact in product safety. Detection and quantification of low abundance impurities in these products is one of the more demanding tasks in pharmaceutical analysis because they are present at ppm levels in a complex product matrix, where conventional analytical approaches may fall short.

At iMNC26, the International MicroNanoConference, held on November 2 and 3, 2026, at the University of Twente in Enschede, Netherlands, Ardena shared its experience with this challenge. The annual event, organized by MinacNed together with NanoLabNL, connects research and industry around micro-, nano-, and quantum technology.

Dr. Maria Marioli, Principal Scientist and Group Lead at Ardena, gave a talk titled “Detection and quantification of trace impurities in nanomedicines. Are current analytical tools sufficient or are new technologies needed?”

“Determining impurities in nanomedicine products can be challenging, and it is a great
showcase of our expertise,” said Dr. Marioli.

Impurities present in nanomedicines

The impurities in nanomedicines can be categorized into two main categories:

The impurities related to main components are either present in the starting materials or they are degradation products. Those related to the active pharmaceutical ingredient (API) require strict monitoring and control using phase-appropriate developed and validated analytical methods, and their limits must comply to ICH Q3B. Impurities related to the excipients used as structural components (such as lipids or polymers) also require quantification, and similar limits may apply, particularly for novel excipients. Aggregates such as dimers, oligomers or higher-order aggregates are also considered impurities, and are typically controlled with particle size measurements.

Organic solvents or other compounds used as manufacturing aids (e.g., coupling agents, catalysts, stabilizers) as well as unreacted components, are typically removed during manufacturing by a diafiltration step, such as tangential flow filtration (TFF). Phase-appropriate analytical methods should be applied to ensure that any residual amount remains below acceptance limits. It is also necessary to confirm the absence of leachable from contact materials or contaminants including microbiological contamination. Limits for residual solvents and elemental impurities should comply with ICH Q3C and Q3D, respectively.

Why trace impurities are hard to measure

The impurities must be extracted from a complex matrix because nanomedicines are inherently multi-component systems. Sample preparation often requires multiple steps, such as nanoparticle disruption and the removal of interfering components. In addition, every nanomedicine is unique, and therefore the impurities and the corresponding methods are product-specific. Particularly challenging analytes are those that are UV-inactive and nonvolatile, covalently bound to the particles, prone to degradation, or unspecified that require untargeted analysis. Because impurities are often present at ppm levels, extensive sample preparation is prone to cause significant underestimation (due to impurity degradation or adsorption to glassware), or overestimation (due to degradation of the main components or the introduction of contaminants).

A structured workflow for method development

The technology selection depends on analyte properties, matrix and the desired technology performance characteristics.

Dr. Marioli showed how analyte properties drive the choice of technique

Analyte properties Analytical technique 
Volatile or volatile after derivatization (Headspace) Gas Chromatography (HS-GC) with FID/MS 
Semi-volatile or semi-volatile after derivatization Gas Chromatography (GC) with FID/MS 
Non-volatile – Liquid Chromatography (LC) with UV detection or CAD/MS detection (for UV-inactive compounds) – UV/Fluorescence spectrometry (typically after derivatization) 

She also proposed a workflow including the sample preparation steps requiring consideration during method development.

1. Particle disruption

Her decision workflow starts by determining whether the compound is encapsulated. If so, the nanoparticles must be disrupted to fully release the compound into solution. When the nanoparticle components are non-covalently bound, dilution in a suitable organic solvent is often sufficient to release the compound. If the nanoparticle components are covalently bound, chemical or enzymatic treatment is necessary.

2. Remove interference from other matrix components

Due to the multi-component matrix, other components may interfere with the measurement. For instance, the target impurities may co-elute with nanoparticle components in LC or interact with them, resulting in low recovery or inaccurate quantification. In such cases, the analyte must be separated from interfering matrix components using techniques such as centrifugal ultrafiltration, centrifugation, liquid-liquid extraction, or solid-phase extraction. Alternatively, derivatization can be used to improve selectivity and eliminate matrix interference, provided that the derivatization reaction is specific to the target analyte.

3. Compatibility of the prepared solutions with the analytical technique

Before measurements, it is important to ensure that the prepared solutions are compatible with the analytical technique. For example, in LC if the sample solvent has higher elution strength than the mobile phase, the analyte may elute in the void volume. Similarly, highly acidic solutions may need to be neutralized before analysis to prevent incompatibility with the chromatographic column or adverse effects on analyte retention.

Lessons from three case studies

Three case studies presented by Dr. Marioli show that the approach and the complexity differ depending on the analyte and the product matrix.

Lipid nanoparticles (Case 1)
The first case study involved a product containing lipid nanoparticles encapsulating a small-molecule API. The nanoparticle components were non-covalently bound; therefore, the API-related impurities were released upon dilution in an organic solvent. In this case, the API and its related impurities were UV-active, whereas the lipid excipients were not. The diluent was compatible with the LC method, and the lipids eluted during the run without causing interference because they were UV-inactive.

Polymeric micelles (Case 2)
The second case study involved a product based on polymeric micelles consisting of a block copolymer encapsulating a small-molecule API. Similar to the first case, the non-covalently bound nanoparticles were readily disrupted by dilution in an organic solvent. Because the API and its related impurities were UV-inactive, LC-CAD was used. However, the block copolymer was also detected and co-eluted with the impurities, preventing accurate quantification. A centrifugal ultrafiltration filter with a 3 kDa molecular-weight cut-off removed the polymer but reduced recovery of both the API and its related impurities. A mixedmode SPE method (reversed-phase/anion-exchange) was then used instead. The uncharged copolymer was removed, while the charged impurities were retained. Although this approach effectively removed the polymer, the acidic elution solution used in the final SPE step caused API degradation, requiring optimization of its concentration.

Metal nanoparticles (Case 3)
The last case presented was a product contained metal nanoparticles with a biodegradable organic coating. The target analyte was a residual fatty acid derivative, used as a stabilizer for an intermediate product during manufacturing and later removed by TFF. Its release and isolation for chromatographic quantification required chemical treatment, liquid-liquid extraction, concentration, and derivatization. Although the analyte was present at very low levels, several additional peaks were observed in the chromatogram. These were initially considered related impurities but were later suspected to be leachable originating from consumables and not present in the drug product. Analysis of blanks collected that were including all sample preparation steps confirmed this hypothesis.

Controls that reduce the risk

The following controls were proposed to ensure reduction of associated risks:

Risk  Control 
Increase in degradation products due to sample preparation Check stability of the main components 
Risk of contamination Investigate with blanks (including all sample preparation steps) and understand the manufacturing process 
Low recovery due to interaction with matrix or extensive sample preparation Check accuracy by spiking 
The analyte is not fully released from the nanoparticles Ensure nanoparticles are fully disrupted (check visually if samples change from turbid to transparent after disruption, or with size measurements) 
Method is not reproducible  Measure in different days with different columns/equipment (S/N can be different) 

Are new analytical technologies needed?

Current analytical tools can deliver reliable determination of impurities when supported by a careful workflow and the right controls. However, even with carefully designed methods and controls, there remains a risk of irreproducible results, particularly when methods are transferred between laboratories. Dr. Marioli concluded that the field needs new technologies capable of measuring impurities at ppm levels in complex matrices, including encapsulated impurities without the need of releasing them from the particles.

Want to Learn More? 

Our nanomedicine scientists at Ardena Oss tackle challenges like these as part of a full CDMO offering.  

Explore Ardena’s nanomedicine services, or reach out to the team to discuss your program. 

Want to Learn More?

Explore our full range of services and discover how Ardena can support your drug development journey.

View All Blog