A Novel Capillary Electrophoresis Capacitively Coupled Contactless Conductivity Detection (CE‑C⁴D) Method for Quantification of Ionic Excipients in Doxorubicin HCl Liposome Injection

Abstract

Pegylated liposomal doxorubicin (DOX) was recognized for its efficient drug delivery, reduced toxicity, and prolonged circulation. Ammonium sulfate ((NH₄)₂SO₄) plays a critical role in DOX loading and retention within liposomes, and intraliposomal (NH₄)₂SO₄ concentrations above 200 mM are essential for stable DOX sulfate nanocrystal generation and liposomal stability. Currently available methods, such as ion chromatography, for the selective quantitation of liposomal encapsulated ions require time-consuming separation steps. In this study, a capillary electrophoresis-based separation method coupled with capacitively coupled contactless conductivity detection (CE-C⁴D) was developed and validated for the direct quantitation of total (internal + external) and external ionic excipients in DOX liposomal formulations. Cation and anion analyses, conducted under positive and negative polarities, respectively, demonstrated distinct migration orders (NH₄ ⁺, K⁺, Na⁺, Li⁺ for cations; Cl^–, SO₄ ^2–, F^– for anions) with excellent baseline separation and peak resolution. The method demonstrates linearity with correlation coefficients consistently above 0.999 and low limits of quantitation (below 8.22 μM) for all analyzed ions. The optimized CE-C⁴D method allowed accurate quantitation of total and external ion concentrations including NH₄ ⁺, K⁺, Na⁺, Cl^–, and SO₄ ^2– ions in liposomal DOX formulations. All tested liposomal DOX formulations contained an excess of 200 mM (NH₄)₂SO₄ within liposomes. This method provides an alternative for accurate quantitation of ion species present in liposomal formulations, supporting both brand and generic liposome formulation development and regulatory assessment.

Introduction

Over the past six decades, extensive research has propelled liposomal drug delivery systems to the forefront of medicine and pharmaceutical sciences, establishing liposomes as versatile carriers for various therapeutic agents, including chemotherapeutics, antibiotics, vaccine antigens, and gene delivery. − These lipid-based vesicles offer a unique platform for delivering both hydrophilic and hydrophobic active pharmaceutical ingredients with selectivity toward target tissues, thereby reducing toxicity through lower doses. − The approval of the first liposomal doxorubicin (DOX) HCl formulation, Doxil, by the U.S. Food and Drug Administration (US FDA) in 1995 marked a significant milestone. Subsequently, five generic versions of Doxil were approved by U.S. FDA. DOX-loaded liposomes, composed of hydrogenated soy phosphatidylcholine, cholesterol, and polyethylene glycol conjugated distearoyl-phosphatidylethanomine (mPEG-2000-DSPE), have become a pivotal treatment option of ovarian cancer, multiple myeloma, and AIDS-related Kaposi’s sarcoma.

One of the key factors contributing to the performance of liposomal DOX formulations is the use of ammonium sulfate ((NH₄)₂SO₄), employed for active loading and crystallization of DOX within the liposomes. Figure shows the mechanism of (NH₄)₂SO₄ gradient-assisting active DOX loading. The methodology involves introducing DOX·HCl solution into (NH₄)₂SO₄-filled liposomes, resulting in the charge-neutral DOX transversion through the liposomal membrane, followed by protonation to its ionic form inside the liposome, culminating as crystal ((DOX-NH₃)₂SO₄) with sulfate ions. ,

1.

Schematic illustration of ammonium-induced drug loading of liposomal doxorubicin.

The transmembrane (NH₄)₂SO₄ gradient and total (NH₄)₂SO₄ concentrations are well documented to play a pivotal role in achieving sufficient encapsulation efficiency, retention, release rates, and stability of DOX within liposomes. Wei et al. discussed the importance of having a critical intraliposomal (NH₄)₂SO₄ concentration over 200 mM to generate DOX sulfate nanorod crystals and resultant formulations to have a monophasic slow DOX release profile compared to a biphasic release profile with an early rapid release for liposomal DOX formulations formulated with less than 200 mM (NH₄)₂SO₄.

Consequently, several analytical techniques have been employed to selectively quantify intra- and extra-liposomal sulfate and other inorganic ions in liposomal DOX formulations, including ion chromatography (IC), , RP-HPLC with charged aerosol detection, and capillary electrophoresis (CE) coupled with inductively coupled plasma mass spectrometry (ICP-MS). While IC has traditionally been employed predominantly, it necessitates the separation of suspension media from liposomes by filtration for selective quantitation of external ions. The separation process is challenging, as it may overestimate ion concentration or have variable results due to potential liposome rupture, evaporation of ammonia, or adsorption of ions onto the filter membrane. CE-ICP-MS was later developed as an alternative, but its application is restricted to sulfate because of technical constraints for other ions, and it remains costly and operationally complex.

In this investigation, we utilized CE coupled with capacitively coupled contactless conductivity detection (C⁴D) for the selective quantitation of inorganic ions in liposomal DOX formulations, as illustrated in Scheme . CE provides high-resolution separation of inorganic ions, and the C⁴D enables sensitive, noninvasive conductivity-based detection. By exploiting the differences in electrophoretic mobility between analyte ions and the background electrolyte, CE-C⁴D offers a robust approach for direct injection of liposomal DOX formulations without requiring filtration or separation steps.

1. Illustration of Free and Intraliposomal Ion Quantitation Using CE-C⁴D Method.

In this study, in addition to optimizing key analytical conditions, including background electrolyte (BGE) composition, crown ether concentration, capillary internal diameter, injection volume, and C⁴D detection parameters, we evaluated filtration versus direct injection approaches and compared the CE-C⁴D method with IC and ICP-MS. This CE-C⁴D method ensured a high sensitivity, reproducibility, and compatibility with intact liposomes.

Experimental Method

Materials

Liposomal Formulations

Doxil 2 mg/mL (Lot No MHzTC00, liposomal DOX formulation; Baxter Healthcare Corporation, Deerfield, IL, USA) and its generic formulations (Sun Pharmaceutical Industries Ltd., Lot HAD4043A; Dr. Reddy’s Laboratories Ltd., Lot 501887; Ayanna Pharma Ltd., Lot 050122; Zydus Pharmaceuticals, Lot P200135) were obtained from WEP Clinical (Morrisville, NC, USA).

Chemicals and Standards

Doxorubicin hydrochloride (secondary pharmaceutical grade, CAS: 25316-40-9), sodium dodecyl sulfate (SDS, Reagent Plus, CAS: 151-21-3), lithium fluoride (LiF, CAS:7789-24-4), and sulfate standard for IC (Na₂SO₄, TraceCERT) were purchased from Sigma-Aldrich (St. Louis, MO, USA). l-Histidine (Reagent Plus, CAS: 71-00-1), sodium hydroxide (certified ACS grade, CAS: 497-19-8), methanol (MeOH, Optima LC/MS grade, CAS: 67-56-1), and ethylenediaminetetraacetic acid (EDTA, CAS: 6381-92-6) were purchased from Fisher Scientific (Fair Lawn, NJ, USA). 2-Hydroxy-4-morpholinepropanesulfonic acid (MOPSO, 98%, CAS: 68399-77-9), nitric acid (Optima, CAS: 7697-37-2), and N,N-dimethyldodecylamine N-oxide (CAS: 1643-20-5) were purchased from Thermo Fisher Scientific (Waltham, MA, USA).

Buffer Components

2-(N-Morpholino)­ethanesulfonic acid monohydrate (MES monohydrate, 99+%, CAS: 145224-94-8) was purchased from Acros Organics (New Jersey, USA). Sucrose (ultrapure, CAS: 57-50-1) and 18-crown-6 ether (99%, CAS: 17455-13-9) were purchased from Thermo Scientific (Ward Hill, MA, USA). Ammonium standard (NH₄Cl, CAS: 12125-02-9) was purchased from Agilent Technologies (Santa Clara, CA, USA).

Instrumentation and Supplies

250 μL CE vial inserts with polymer feet (Part no: 5181–1270) and CE vials (Part no: 5182–0562) were purchased from Agilent Technologies (Santa Clara, CA, USA). CE fused silica capillary tubing (I.d. 25 μm, Part no: 1068150011; 50 μm, Part no: 1068150017; and 75 μm, Part no: 1068150019) was purchased from Molex, LLC (Lisle, IL, USA).

Other Materials

Arsenic acid sodium salt heptahydrate (CAS: 10048–95–0) was obtained from MP Biomedicals, LLC (Solon, OH, USA). Deionized water (>18.2 MΩcm^–1) was obtained from a Millipore Milli-Q Plus unit (Millipore, Billerica, MA, USA). All solvents were filtered through 0.2 μm filters (Fisher Scientific, Pittsburgh, PA, USA).

CE-C⁴D

The capillary electrophoresis experiments were performed using an Agilent CE 7100 (Agilent Technologies, Santa Clara, CA, USA) instrument that was equipped with both a diode array optical absorbance detector and a C⁴D detector (eDAQ, New South Wales, Australia). The C⁴D detector system contains two parts: the C⁴D head stage (Part No: ET120, eDAQ, New South Wales, Australia) and control unit (Part No: ER815, eDAQ, New South Wales, Australia), and C⁴D Profiler V2 software for data acquisition. The analog output from the C⁴D detector was converted to a digital signal through the CE instrument at a data collection frequency of 16.67 Hz.

The C⁴D detector features two circular electrodes around the capillary with a 2.5 mm gap. One electrode acts as the transmitter, applying a high-frequency AC voltage, while the other functions as the receiver, capturing the transmitted signal. , The received signal correlates to the conductivity of the solution inside the capillary. This design enables quantitation of analytes by exploiting the conductivity difference between migrating ions and the BGE, requiring analyte ions to exhibit an electrophoretic mobility higher than that of the BGE to yield a positive signal.

Experiments used fused silica capillaries (25 μm i.d., 52 cm total length, 37.5 cm to the detector). The optimized BGE contained 146.1 mM sucrose, 100 mM MES, 3.5 mM 18-crown-6, and 78 mM l-histidine, adjusted to pH 6.00 ± 0.05. This composition was selected to maintain liposome integrity while providing optimal ion separation. The BGE was heated to 40 °C and vacuum degassed using an oil-free vacuum pump (Welch, Model 2154B-01, Niels, IL USA). A polyamide-coated fused silica capillary was cleaned with a solution of 25.0 mM EDTA, 12.5 mM SDS, and 12.5 mM N,N-dimethyldodecylamine N-oxide at pH ∼ 12.5.

The capillary used for CE separation was conditioned for 15 min with the cleaning solution and BGE before preceding each experiment. The C⁴D Profiler V2 software was employed to systematically modify the frequency, amplitude, and head-stage gain. This adjustment process was employed for diverse conditions, encompassing variations in the BGE content, BGE concentration, temperature, voltage, and capillary internal diameter. The software settings were configured by using 90% of the BGE as the reference base.

The capillary was equilibrated with freshly replenished BGE before each run for 4 min and flushed with cleaning solution for 2 min afterward. Prior to injection, 146.1 mM sucrose in 10 mM histidine was hydrodynamically injected at 50 mbar for 10 s. Samples were injected at 50 mbar for 50 s with the capillary temperature maintained at 15 °C. Cation analysis was performed at +30 kV for 5 min, while anion analysis at −30 kV for 5 min. Conductivity was measured using an eDAQ C⁴D detector at 316.9790 kHz, 100% amplitude, and head gain on, with a data collection frequency of 16.67 Hz to ensure sufficient points for peak integration.

Calibrations and Quality Control

Standard ion solutions were prepared in 146.1 mM sucrose and 10 mM histidine buffer, covering concentration ranges of 0.50–8.00 mM for NH₄ ⁺, 0.50–5.00 mM for Cl^– and SO₄ ^2–, and 0.10–1.00 mM for Na⁺, and K⁺. Each included 1.00 mM LiF as the internal standard. Calibration samples were freshly prepared and analyzed in triplicate. The limits of detection (LOD) and quantification (LOQ) were determined according to ICH guidelines, calculated as 3.3 × (standard deviation/slope) and 10 × (standard deviation/slope), respectively.

The precision of the CE-C⁴D method was assessed for both intraday and interday analyses using quality control (QC) samples at three concentration levels: LQC (1.00 mM NH₄ ⁺, 0.25 mM Na⁺ and K⁺, and 0.75 mM Cl^– and SO₄ ^2–), MQC (5.00 mM NH₄ ⁺, 0.50 mM Na⁺ and K⁺, and 2.50 mM Cl^– and SO₄ ^2–), and HQC (7.50 mM NH₄ ⁺, 0.75 mM Na⁺ and K⁺, and 4.50 mM Cl^– and SO₄ ^2–). For intraday precision, six replicate QC samples (n = 6) at each concentration level were prepared and analyzed within a single day. For interday precision, the analysis was repeated over three independent days (n = 3).

Quantification of External Inorganic Ions by CE-C⁴D – Direct Injection Method

The DOX formulation (50.0 μL) was spiked with 50.0 μL of 2.00 mM LiF to reach LiF final concentration of 1.00 mM. The sample was placed in a vial insert inside a CE vial and directly injected with minimal agitation to prevent liposome rupture.

Quantification of External Inorganic Ions by CE-C⁴D – Filtration Method

The DOX formulation was filtered through 0.02 μm filters (Cytiva 68091102; Fisher Scientific, Pittsburgh, PA, USA). A 50.0 μL aliquot of filtrate was spiked with 50.0 μL of 2.00 mM LiF to reach 1.00 mM LiF final concentration before injection.

Quantification of Total Inorganic Ions by CE-C⁴D

A 100.0 μL aliquot of DOX sample was mixed with 200.0 μL methanol and agitated at 50 °C for 10 min in a Thermomixer R (Eppendorf, Enfield, CT, USA) to disrupt the liposome. An additional 200.0 μL solution containing the LiF, histidine, and sucrose was then added (final concentrations: 1.00 mM LiF, 10 mM histidine, 146.1 mM sucrose). A 100.0 μL aliquot was transferred to a vial insert within a CE vial for injection in CE-C⁴D.

Intraliposomal inorganic ion concentrations were calculated by subtracting external from total ion values, a widely accepted standard approach for liposome analysis.

Quantification of Total Sulfate and Sodium Ions by ICP-MS

DOX formulations (50.0 μL) were digested with 500.0 μL of concentrated HNO₃ overnight at room temperature. Subsequently, the digested samples were diluted with internal standards: sodium arsenate (100 ng/mL) for sulfate and scandium (100 ng/mL) for sodium. Sulfate was measured using Agilent 8800 ICP-Triple Quad in O₂ reaction cell MS/MS operational mode (³²S, 32 amuàS¹⁶O⁺, 48 amu; ⁷⁵As, 75 amuà⁷⁵AsO⁺, 91 amu). Sodium was analyzed in no-gas mode (²³Na⁺, 23 amu; ⁴⁵Sc⁺, 45 amu). Calibration curves covered 100.0–3000.0 ng/mL, for sulfate and 10.0 – 500.0 ng/mL for sodium, with a constant 100 ng/mL internal standard.

Quantification of Total and External Ions by Ion Chromatography

External ions were first separated using the filtration method above and analyzed with a Metrohm 930 Compact IC Flex (Metrohm USA Inc., TX, USA) equipped with a conductivity detector. Cations were analyzed using a Metrosep C6 – 250/4.0 column (30 °C, 0.7 mL/min, 5 mM HNO₃ eluent). Anions were analyzed using a Metrosep A Supp 5–100/4.0 column (30 °C, 0.6 mL/min, 2.81/0.88 mM CO₃ ^2–/HCO₃ ^– eluent, 0.5 M H₂SO₄ suppressor). Total ions were determined after liposome disruption with 40% methanol and measured with the same system.

All the measurements across CE-C⁴D, ICP-MS, and IC methods were performed in triplicate, and results are reported as mean ± standard deviation (SD).

Results and Discussion

CE Separation Principle

Anions and cations in a solution migrate in opposite directions due to their charge-dependent electrophoretic mobilities when placed in an electric field. In CE, the electric field is applied to the BGE across the inlet (injection end) and outlet (detector end) with a high-voltage power supply. The negatively charged silica capillary surface generates an electroosmotic flow (EOF), a bulk liquid movement toward the cathode caused by cation accumulation near the capillary inner wall. As EOF has a plate flow profile, it does not generate shear force on liposomes, unlike other separation techniques. It is possible to detect both cations and anions in a single run if the EOF is high enough to overcome the electrophoretic mobility. However, most of the inorganic anions have high electrophoretic mobilities, and they migrate in opposite directions based on charge and BGE solution parameters. Thus, we opted to analyze cations and anions separately, resulting in better peak resolution since it eliminated the need of maintaining electrophoretic mobilities of anions below EOF to elute all ions in a single run.

Positive polarity was applied for cation analysis, resulting in both electrophoretic mobility and EOF to assist the movement of the cations to the opposite electrode where the cations are detected. Under these conditions, cations having the highest electrophoretic mobility reach the detector first (e.g., NH₄ ⁺ ion), and the overall analysis time is very short (about 4 min). Figure A shows the cations identified in liposomal DOX formulation in the cation analysis using +30 kV across the capillaries. Given the nanoscopic size of liposomes, the potential drop across a single vesicle remains negligible, even at an applied voltage of 30 kV across the capillary, thus maintaining the membrane integrity. The NH₄ ⁺ ion has the highest electrophoretic mobility and, thus, arrives first at the detector. While the NH₄ ⁺ and K⁺ ions have similar electrophoretic mobilities, selective complex formation of potassium with 18-crown-6 alters the potassium migration, resulting in their separation and detection. The Na⁺ and Li⁺ (internal standard) ions arrive at the detector later than the NH₄ ⁺ and K⁺ ions.

2.

Electropherograms of ion separation using CE-C⁴D method. (A) Cation analysis under positive polarity (+30 kV) showing separation of NH₄ ⁺ (1.97 min), K⁺ (2.35 min), Na⁺ (2.70 min), and Li⁺ (3.29 min) internal standard. (B) Anion analysis under negative polarity (−30 kV) showing separation of Cl^– (2.60 min), SO₄ ^2– (3.89 min), and F^– (4.30 min) internal standard. Experiment conditions: 25 μm i.d. fused silica capillary (52 cm total length, 37.5 cm to detector), 100 mM MES BGE with 146.1 mM sucrose, 78 mM l-histidine, and 3.5 mM 18-crown-6 at pH 6.0, 15 °C capillary temperature, hydrodynamic injection at 50 mbar for 50 s.

Anion analysis was carried out using negative polarity, and anions injected from the inlet end migrated to the outlet end due to their electrophoretic mobility. Although in this case, the EOF is in the opposite direction (toward the inlet), the electrophoretic mobilities of inorganic anions were large enough to overcome EOF. The sequence of migration is dictated by the electrophoretic mobility, with the chloride ions exhibiting the greatest electrophoretic mobility experiencing a shorter migration time and arriving at the detector first, as illustrated in Figure B. In contrast, the internal standard, fluoride (F^–) ions, characterized by lower electrophoretic mobility, were the last to reach the detector.

The peaks were assigned by running individual analytes under identical conditions to compare retention times. Figure S1 shows all of the analytes identified in liposomal DOX formulations. Figure S1 also confirms the specificity of the optimized method, and the zero calibrator (blank) was free of interferences at the retention times of the analytes and the internal standards.

The resolution between peaks was calculated according to USP guideline and found to be greater than 1.5 (specific values are tabulated in Table S1), indicating baseline separation for all the ions. CE has an EOF-driven flat flow profile and single flow path, as it is an open tubular channel. Thus, CE is less susceptible to peak broadening compared to pressure-driven packed column separation techniques, which have parabolic flow profiles and multiple flow paths. This remarkable property allows CE to achieve baseline separation even at lower calculated peak resolutions.

Method Optimization

Background Electrolyte Optimization

The CE BGE was optimized for improved resolution of anions and cations while maintaining liposome integrity. Various buffer systems, including histidine, citric acid/histidine, acetic acid/histidine, MOPSO/histidine, and MES/histidine, were explored. We chose a zwitterionic organic MES buffer for its low conductivity and minimal susceptibility to electrolysis at the high voltage electrodes. Also, the low conductivity of the MES buffer gives better resolution and generates minimum heat (Joule heating) inside the capillary. We incorporated histidine and sucrose in the BGE system to ensure the integrity and stability of liposomes throughout the separation process, maintaining the buffer pH at 6.00 ± 0.05. This pH closely aligns with that of the formulation (6.5); histidine serves as the buffering agent, while sucrose maintains isotonicity and provides stability for the liposomes.

The optimization of buffer concentrations involved adjusting MES concentrations from 50.0 to 200.0 mM while maintaining constant histidine and sucrose concentrations. In both cation and anion analyses, electropherograms for each BGE with different MES concentrations were generated (Figure S2), and resolutions were calculated (Table S2) using eq . In cation analysis with a 25 μm internal diameter capillary, peak resolutions and symmetry improved with increasing MES concentration. In anion analysis, peak resolutions initially increased with buffer concentration, however, at higher MES concentrations, the resolution between the SO₄ ^2– and F^– ions decreases. Peaks at low buffer concentrations exhibited significant broadening and skewing due to electrodispersion, while higher concentrations resulted in more symmetrical peaks. Based on resolution calculations and peak symmetry analysis, 100.0 mM MES was selected as the optimal BGE concentration, providing the best balance of analysis time, sensitivity, and resolution. This experimental result is consistent with a previous report that relatively high concentrations of BGE are required for best sensitivity and resolution in narrow capillaries.

$$R_{\mathrm{s}}=\frac{1.18(t_{\mathrm{R}2}-t_{\mathrm{R}1})}{(W_{\mathrm{h}1}+W_{\mathrm{h}2})}$$ 1

where R _s: resolution; t _R1, t _R2: retention times of the peaks; W _h1, W _h2: peak widths at half height and t _R2 > t _R1.

Optimization of the C⁴D

The C⁴D detector signal was optimized for the frequency, amplitude, and head-stage gain for each condition (e.g., BGE content, BGE concentration, voltage, temperature, and capillary internal diameter) prior to experiments. Figure S3 shows the C⁴D Profiler V2 software response automatically generated for every combination of frequency, amplitude, and head-gain stage for optimized BGE (90% of optimized BGE was used as the base response), 25 μm capillary at 15.0 °C and +30 kV voltage. The optimum conditions with its response for all the different conditions used in this study are reported in Table S3.

Capillary Diameter Selection

The utilization of narrow internal diameter (i.d.) capillaries poses a challenge with standard optical detection methods due to the reduction in optical path length, as it diminishes sensitivity and introduces noise. However, this limitation can be effectively overcome by employing a C⁴D detector, which does not experience the sensitivity reduction associated with optical detection in small capillaries. In fact, the use of narrower capillaries in conjunction with a C⁴D detector offers enhanced resolution, benefiting from a greater number of theoretical plates (N). Direct comparison between capillaries of different diameters is impractical due to distinct optimized buffers for each diameter; higher diameter capillaries necessitate lower buffer concentrations, while lower diameter capillaries demand higher buffer concentrations. , We conducted experiments using capillaries with 25, 50, and 75 μm i.d., employing both low (50 mM MES) and high (150 mM MES) buffer concentrations. Capillary sensitivity in CE-C⁴D was determined by dividing the NH₄ ⁺ peak area by the corresponding injection volume, where sensitivity was calculated using eq , while the injection volume is determined through the CEToolbox application, eq .

$$\mathrm{Sensitivity}=\raisebox{1ex}{${\mathrm{NH}}_4^{+}\mathrm{peak}\mathrm{area}$}\!\left/ \!\raisebox{-1ex}{$\mathrm{injection}\mathrm{volume}(\mathrm{V})$}\right.$$ 2

$$V=\frac{\Delta P{d}^4\pi t}{128\eta L}$$ 3

where V: hydrodynamic injection volume; ΔP: applied pressure; d: capillary internal diameter; t: time of injection; η: viscosity; and L: length of the capillary.

Electropherograms obtained for different injection volumes and capillary internal diameters for a cation mix of NH₄ ⁺, K⁺, Na⁺ , and Li⁺ in 50 mM MES buffer and 150 mM MES buffer are presented in Figures S4 and S5, respectively, and the corresponding sensitivity calculation data are tabulated in Table S4. The sensitivity ratio of 25 μm to 50 μm to 75 μm capillaries in low concentration buffer was found to be 8:4:1. In higher buffer concentrations, a substantial two-fold decrease in sensitivity was observed for 50 μm capillary. However, the sensitivity of 75 μm capillary at higher buffer concentrations could not be calculated due to highly distorted peaks observed for the larger capillary diameter. Consequently, the 25 μm internal diameter capillary, offering the best sensitivity, was chosen as the optimal capillary diameter.

Injection Volume Optimization

In this study, a hydrodynamic injection mode available in the Agilent CE7100 instrument was employed, allowing manipulation of pressure and time parameters to produce varying injection volumes. Different concentrations of a standard cation mixture (comprising NH₄ ⁺, K⁺, Na⁺ , and Li⁺) were investigated using diverse hydrodynamic injections in a 25 μm capillary with 100 mM MES BGE. The resulting electropherograms for each injection pressure and varying injection times are depicted in Figure S6, with corresponding resolution data tabulated in Table S5. As anticipated, resolution exhibited a diminishing trend with increasing injection volume. Consequently, an injection volume produced with 50 mbar and 50 s was selected for subsequent experiments.

18-Crown-6 Concentration Optimization

The NH₄ ⁺ and K⁺ ions exhibit similar electrophoretic mobilities, thus requiring manipulation for effective separation of the two ions. The addition of 18-crown-6 complexing agent alters their electrophoretic mobilities differently, facilitating distinct separation and enhancing peak shape and resolution. In this study, varying concentrations of 18-crown-6 (0.0–10.0 mM) in the BGE were explored to identify an optimal concentration for separation. Electropherograms in Figure illustrate the separation of NH₄ ⁺, K⁺, Na⁺, and Li⁺ ions in 100 mM MES BGE with different 18-crown-6 concentrations, with resolution data presented in Table S6. Both 2.5 and 3.5 mM 18-crown-6 in the BGE effectively separated NH₄ ⁺ and K⁺, but for liposomal samples with higher NH₄ ⁺ concentrations, 3.5 mM 18-crown-6 in the BGE was deemed the most suitable concentration.

3.

Electropherograms obtained for the separation of ion standards (1.0 mM NH₄ ⁺, K⁺, Na⁺, and Li⁺ each) using 100 mM MES BGE containing different 18-crown-6 concentrations (0.0–10.0 mM) to evaluate the effect of 18-crown-6 concentration on separation of NH₄ ⁺ and K⁺ ions.

Cleaning Solution Optimization

Run-to-run variation in cation mode during formulation analysis posed a significant challenge attributable to molecule adsorption to the capillary wall, altering the wall charge, and, consequently, the EOF. Commercially available cleaning solutions failed to mitigate the migration time inconsistencies. Therefore, an in-house cleaning solution comprising an ion complexing agent (EDTA), an anionic detergent (SDS), and a nonionic detergent (N,N-dimethyldodecylamine N-oxide) was prepared. Flushing the capillary for 2 min with the in-house cleaning solution, followed by 4 min with the BGE, proved effective in ensuring consistent retention times across all analyses, enhancing reproducibility. Figure S7 displays electropherograms for consecutive runs of the same DOXIL sample, illustrating the impact of flushing between runs on migration time consistency. Crucially, electropherograms from successive trials with the cleaning solution yielded more robust data with consistent migration times.

Other Optimizations/Considerations

The BGE and stacking buffer were filtered and vacuum degassed to eliminate particles and dissolved gases, preventing potential CE capillary clogging and noisy background. A stacking buffer composed of 146.1 mM sucrose solution with 10 mM histidine was employed to enhance resolution and was injected before the sample. This plug is characterized by low conductivity and a high voltage gradient, which facilitated faster ion migration until encountering the boundary of the running buffer with a high electrolyte concentration, resulting in sharper peaks, a phenomenon well-documented in the literature. Electro-stacking not only improved resolution but also heightened CE method sensitivity by elevating signal-to-noise ratios. Maintaining the capillary at 15.0 °C minimized the risk of liposomal rupture, averting erroneous results and signal peak broadening due to Joule heating. An optimum voltage of +30 kV for cation mode and −30 kV for anion mode was selected for effective separation. The addition of LiF as an internal standard had no impact on the liposomal DOX formulation, while the Li⁺ ion in the cation mode and the F^– ion in the anion mode exhibited peaks with distinct retention times, thus preventing overlap with other analytes.

Calibration and Quality Control

To emulate the viscosity conditions of the DOX formulation, sucrose and histidine were incorporated into the standard samples. The optimized CE-C⁴D method, validated in accordance with the ICH guideline Q2­(R2) for bioanalytical method validation, exhibited robust performance. The method demonstrated excellent linearity, with a correlation coefficient (R ²) consistently above 0.999 (Figure S8) and high sensitivity with LOD values below 3.0 μM for all ions (Table ). This increased sensitivity can be attributed to the enhanced ability of C⁴D to detect minute conductivity differences within the analyte plug and the background.

1. LOD and LOQ for Identified Cations and Anions (as Micromoles per Liter).

	inorganic ion	retention time (min)	LOD (μM)	LOQ (μM)
cation	NH4 +	1.97	2.76	8.22
	K+	2.35	1.25	3.72
	Na+	2.70	1.88	5.58
anion	Cl–	2.60	1.81	5.38
	SO4 2–	3.89	1.00	2.86

The precision and accuracy were assessed in terms of relative standard deviation (RSD) and relative error (Error%), respectively. Results are presented in Table , which shows intra- and interday variability across three concentration levels (LQC, MQC, HQC) for all ions. Both intraday and interday evaluations gave consistently low values for RSD (<3%) and Error% (<1%) across most ions, affirming the method to be highly reproducible and reliable for quantification. Slightly higher variability was observed for Na⁺ at the lowest concentration level (LQC), where the RSD exceeded 7% and Error% reached ∼7%. This behavior is likely influenced by background Na⁺ contributions from reagents, water, or the laboratory environment, as well as minor baseline fluctuations in conductivity detection, which can have a greater impact at lower concentration levels.

2. Intra- and Inter-Day Precision and Accuracy (RSD, Error%) of the CE-C⁴D Method for Quantification of Inorganic Ions in Liposomal DOX Formulations.

		intraday precision (n = 6)			interday precision ( n = 3)
ion (conc., mM)		mean ± SD	RSD	error %	mean ± SD	RSD	error %
NH4 +	1.00 (LQC)	1.00 ± 0.02	2.47	0.43	1.01 ± 0.02	1.79	0.6
	5.00 (MQC)	4.99 ± 0.03	0.51	0.14	5.01 ± 0.06	1.20	0.28
	7.50 (HQC)	7.48 ± 0.08	1.05	0.22	7.55 ± 0.14	1.87	0.62
Na+	0.25 (LQC)	0.23 ± 0.02	7.28	7.22	0.24 ± 0.02	9.82	2.86
	0.50 (MQC)	0.48 ± 0.02	3.36	4.38	0.50 ± 0.03	5.60	0.37
	0.75 (HQC)	0.73 ± 0.01	1.60	2.16	0.75 ± 0.03	3.68	0.19
K+	0.25 (LQC)	0.25 ± 0.00	1.77	1.77	0.25 ± 0.01	2.74	1.14
	0.50 (MQC)	0.50 ± 0.01	2.73	0.60	0.50 ± 0.01	2.74	0.67
	0.75 (HQC)	0.75 ± 0.01	1.77	0.31	0.76 ± 0.02	2.63	1.07
Cl–	0.75 (LQC)	0.75 ± 0.00	0.34	0.00	0.75 ± 0.00	0.42	0.06
	2.50 (MQC)	2.50 ± 0.01	0.27	0.12	2.50 ± 0.01	0.26	0.07
	4.50 (HQC)	4.50 ± 0.01	0.13	0.07	4.50 ± 0.01	0.20	0.05
SO4 2–	0.75 (LQC)	0.75 ± 0.00	0.39	0.20	0.75 ± 0.00	0.48	0.04
	2.50 (MQC)	2.50 ± 0.00	0.17	0.15	2.50 ± 0.01	0.29	0.18
	4.50 (HQC)	4.50 ± 0.01	0.26	0.03	4.50 ± 0.01	0.26	0.15

^aRSD and error % were calculated using unrounded raw data.

Comparison of Total, External, and Intraliposomal Inorganic Ions in DOX Formulations

The CE-C⁴D method was effectively employed for quantifying total and external ion concentrations in liposomal DOX formulations. Figure shows the direct injection and analysis of DOXIL and its generic versions. Table presents the calculated external ion concentrations for each formulation, revealing an external NH₄ ⁺ concentration ranging from 3.5 to 5.2 mM and an external SO₄ ^2– concentration ranging from 0.3 to 1.2 mM based on the total liposomal sample volume. The SO₄ ^2– concentrations are in similar ranges previously reported value by Ansar et al and Wu et al. Notably, formulation 1 exhibited the highest external values for both ions. In addition, 270 μM K⁺ ions were detected in the formulation 2, whereas trace amounts were detected in the formulation 5.

4.

External ions comparison in liposomal DOX formulations 1–5. (A) Cation analysis and (B) anion analysis (no internal standard added).

3. External Ions in DOXIL and Its Generic Versions.

	external ion concentration (average ± SD) mM
formulation	NH 4 +	Na +	K +	Cl –	SO 4 2–
1	5.25 ± 0.53	0.40 ± 0.04		6.13 ± 0.55	1.25 ± 0.18
2	3.51 ± 0.22	0.19 ± 0.03	0.27 ± 0.01	5.68 ± 0.51	0.21 ± 0.06
3	3.71 ± 0.08	0.20 ± 0.02		5.97 ± 0.15	0.53 ± 0.07
4	3.66 ± 0.13	0.28 ± 0.03		5.80 ± 0.11	0.34 ± 0.04
5	3.81 ± 0.07	0.32 ± 0.02	trace amount	6.02 ± 0.38	0.44 ± 0.03
Ansar et al. (CE-ICP-MS)					0.16 ± 0.01–1.28 ± 0.15
Wu et at. (IC)	3.5 ± 0.2				0.8 ± 0.1

Table further illustrates that the total NH₄ ⁺ ion concentration ranges from 9.3 to 12.4 mM, and the total SO₄ ^2– ion concentration ranges from 4.1 to 5.7 mM based on the total liposomal DOX formulation. In this study, the highest analyte concentrations were observed for formulation 1. The total sulfate ion concentrations were validated by the ICP-MS/MS, with results shown in Table S7. These values align with the total sulfate ion data obtained through the CE-C⁴D method, and their variations are limited to 1%. Confirmation of the total ammonium ion was conducted through the ion chromatography method, and the resulting values are presented in Table S8 with other minority ions. The observed variations in the values fall within an acceptable range.

4. Total Ions in DOXIL and Its Generic Versions.

	total ion concentration (average ± SD) mM
formulation	NH 4 +	Na +	K +	Cl –	SO 4 2–
1	12.36 ± 0.48	0.42 ± 0.10		6.33 ± 0.55	5.67 ± 0.14
2	9.42 ± 0.16	0.21 ± 0.06	0.39 ± 0.03	5.86 ± 0.05	4.44 ± 0.29
3	11.00 ± 0.67	0.22 ± 0.06		6.03 ± 0.26	4.44 ± 0.09
4	10.15 ± 0.47	0.34 ± 0.07		5.81 ± 0.26	4.38 ± 0.20
5	9.34 ± 0.61	0.51 ± 0.09	trace amount	6.06 ± 0.14	4.14 ± 0.19
Ansar et al. (CE-ICP-MS)					4.22 ± 0.38–5.78 ± 0.20
Wu et at. (IC)	8.9 ± 0.9				4.2 ± 0.2

The NH₄ ⁺ and SO₄ ^{2 –} concentrations obtained for formulation 1 slightly higher than those claimed in the label (each ml also contains ammonium sulfate, approximately 0.6 mg, equivalent to ∼9 mM of NH₄ ⁺ ion and ∼4.5 mM SO₄ ^2–). The intraliposomal ion concentrations were determined by subtracting external ion concentrations from total ion concentrations. , The values presented for intraliposomal ion concentrations in Table are based on the total liposomal internal volume, 17 μL per 1 mL of liposomal DOX formulation. From the data, it is evident that all formulations contain an excess of (NH₄)₂SO₄ (>200 mM), sufficient to generate rod-like DOX crystals and maintain formulation stability. This finding emphasizes the robustness of the formulations for their intended purposes.

5. Intraliposomal Ions in DOXIL and Its Generic Versions.

	intraliposomal ion concentration (average ± SD) mM		intraliposomal ion concentration (average ± SD) mM
formulation	NH 4 +	SO 4 2–	NH 4 +	SO 4 2–
1	7.56 ± 0.65	4.42 ± 0.23	444.9 ± 38.0	260.2 ± 13.4
2	6.11 ± 0.25	4.23 ± 0.29	359.6 ± 14.7	248.9 ± 17.3
3	7.31 ± 0.68	3.92 ± 0.11	430.2 ± 39.9	230.5 ± 6.7
4	6.68 ± 0.49	4.04 ± 0.21	392.7 ± 28.9	237.7 ± 12.2
5	5.71 ± 0.64	3.70 ± 0.20	336.0 ± 37.5	217.7 ± 11.6
Ansar et al. (CE-ICP-MS)				238.8 ± 22.4–313.5 ± 29.4

^aThe concentrations were calculated based on the total sample volume.

^bThe concentration was calculated based on the total liposome internal volume that is 0.017 mL per mL of DOXIL.

Comparison of External Ion Separation Methods

Accurate analysis of external ions requires their separation from intact liposomes. Here, we compared traditional filtration followed by IC with direct online separation by CE. Ion concentration values obtained from filtration, followed by IC, were slightly higher than those obtained by CE-C⁴D (Table ). This consistent elevation suggests that filtration may have caused partial liposome disruption and leakage of intraliposomal ions, leading to a slight overestimation of external ion concentrations.

6. External Ion Concentrations in DOXIL and Its Generic Versions Measured by CE-C⁴D (Direct Injection) and IC (Filtration).

	sulfate ion concentration, (average ± SD) mM		chloride ion concentration, (average ± SD) mM		ammonium ion concentration, (average ± SD) mM
formulation	CE-C 4 D data	IC data on filtrate	CE-C 4 D data	IC data on filtrate	CE-C 4 D data	IC data on filtrate
1	1.25 ± 0.18	1.40 ± 0.42	6.13 ± 0.55	6.88 ± 0.20	5.25 ± 0.53	6.03 ± 0.28
2	0.21 ± 0.06	0.24 ± 0.03	5.68 ± 0.51	6.84 ± 0.00	3.51 ± 0.22	3.83 ± 0.01
3	0.53 ± 0.07	0.61 ± 0.04	5.97 ± 0.15	7.11 ± 0.04	3.71 ± 0.08	4.71 ± 0.20

Comparison of Methods Available for Inorganic Ion Separation and Detection in Liposomal DOX Formulations

Reliable quantitative methods for inorganic ions are necessary for evaluating the liposomal DOX formulations. We compared three methods including filtration + IC, CE-ICP-MS, and CE-C⁴D (Table ) regarding their strengths and limitations. While traditional IC requires separation by filtration, which can introduce variability due to liposome rupture or ion adsorption, CE-based methods allow direct analysis without a separate separation step prior to injection. In addition, CE provides rapid separation with high resolution while requiring only a few nanoliters of sample. CE-ICP-MS was previously introduced as an alternative, but its application is limited by the inability to analyze all nonmetals, as well as its high cost and operational complexity. Among available approaches, CE-C⁴D offers distinct advantages, including noninvasive detection, compatibility with complex matrices through open channel separation without stationary phases, and eco-friendliness due to minimal solvent and sample requirements. Additionally, C⁴D allows further coupling of CE with another detector, such as UV–vis, fluorescence, or MS to expand analytical capabilities. Overall, CE-C⁴D is a robust, efficient, economical, and green analytical method for inorganic ion analysis of liposomal DOX formulations. Capillary electrophoresis has limited applicability in high-concentration matrices due to poor resolution, migration time shifts, and peak distortion caused by high ionic strengths, Joule heating, and matrix-induced changes in electroosmotic flow. While sample dilution can reduce matrix effects and improve separation, and higher buffer concentrations may enhance buffering capacity and selectivity, these approaches only partially mitigate the problem and may introduce trade-offs such as reduced sensitivity or increased thermal effects.

7. Comparison of Analytical Methods Available for Inorganic Ion Analysis in Liposomal DOX Formulations.

feature	filtration + ion chromatography (IC)	CE-ICP-MS	CE-C 4 D (current method)
sample preparation	require filtration and nano filter	minimal	none, directly injection possible
sample volume required	micro to milliliters	nanoliters	nanoliters
resolution/selectivity	moderate	very high	very high
detection range	wide range of inorganic ions	metals and some nonmetals	wide range of inorganic ions
reliability/reproducibility	results can be influenced by filtration conditions (e.g., liposome rupture or ion adsorption)	high for targeted ions	good reproducibility with careful optimization of conditions
analysis time	30–60 min (including preparation)	∼10 min per specific ion	∼10 min for multiple ions
instrument cost and complexity	moderate to high	high and complex	moderate
eco-friendliness	requires large volumes of solvents/filters	requires high purity gases and solvents	minimal waste, no acidic solvents, environmentally friendly
potential for coupling detectors	limited (UV or conductivity)	ICP-MS only	multiple detectors (UV, LIF, MS) can be integrated
applicability to liposomal formulations	matrix effect, require filtration	moderate matrix tolerance	high matrix tolerance
specific separation column	requires anion/cation columns + guard columns	fused silica capillary	fused silica capillary

This innovative CE-C⁴D methodology not only simplifies the quantitation process but also highlights the synergy between CE separation and C⁴D detection, capitalizing on the inherent advantages of differential ion migration and noninvasive conductivity-based measurement. The combined actions of CE and C⁴D make this method particularly well-suited for the detailed analysis of liposomal DOX formulations. Unlike filtration-based IC, CE-C⁴D avoids potential liposome disruption and provides a direct, accurate assessment of inorganic ions. The CE-C⁴D method is flexible, and it can be coupled with detectors such as UV–vis, fluorescence, or mass spectrometry to expand analytical capabilities. Nevertheless, the method does require separate runs for cation and anion analysis, and optimal performance depends on careful adjustment of the BGE composition and detector parameters for different sample matrices.

Conclusions

A CE-C⁴D-based method was developed and validated for the selective quantitation of total and external ions in liposomal DOX formulations. This method eliminates the need for additional separation steps prior to analysis and demonstrates excellent analytical performance with detection limits below 3.0 μM, high precision (RSD < 5%), and rapid analysis time (∼10 min). Analysis of five FDA-approved liposomal DOX formulations using the CE-C⁴D confirmed that all formulations contain adequate internal sulfate ions (>200 mM) to retain DOX as insoluble sulfate crystals. The simplicity, robustness, and minimal sample requirements of the method make it suitable for routine use in formulation development, quality control, brand-generic comparisons, and regulatory assessments of liposomal drug formulations.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c03075.

• Figure S1, CE-C4D ion separation cations (NH₄ ⁺, K⁺, Na⁺ and Li⁺) and anions (Cl^–, SO₄ ^2– and F^–), Table S1, resolution calculated for separation of cations and anions as indicated in Figure , Figure S2, electropherograms’ obtained for ions using background electrolyte with different MES concentrations for comparison of change in separation for cations and anions, Table S2, background electrolyte concentration dependent resolution calculation for electropherograms shown in Figure S2, Figure S3, C4D detector optimization using Profiler V2 software for optimized conditions, Table S3, optimum condition values for C4D detector obtained using C4D Profiler V2 software for buffer, capillary, and voltage system, Figure S4, electropherograms’ obtained for cation mix separated in 50 mM MES concentrations BGE for comparison of effect of different capillary internal diameter, Figure S5, electropherograms’ obtained for Cation mix separated in 150 mM MES concentrations BGE for comparison of effect of different capillary internal diameter, Table S4, relative sensitivity calculated for different capillary internal diameter and injection volumes, Figure S6, ion injection volume effect on peak resolution and separation comparison, Table S5, resolution calculated for Figure S6, Table S6, resolution calculated for the evaluation of effect of crown ether 18-C-6 on analyte separation as indicated in Figure , Figure S7, CE electropherogram for two consecutive runs of the DOXIL with and without cleaning solution, Figure S8, calibration curves for all quantified ions, Table S8, the total sulfate and sodium ions in DOXIL and its generic versions measured using ICPMS-MS, Table S9, and the total ions in DOXIL and its generic versions measured using Ion chromatography (PDF)

The views expressed in this document are those of the authors and should not be interpreted as the official opinion or policy of the U.S. Food and Drug Administration, Department of Health and Human Services, or any other agency or component of the U.S. government. The mention of trades names, commercial products, manufacturers, or organizations is for clarification of the methods used and should not be interpreted as an endorsement of a product or manufacturer.

The authors declare no competing financial interest.