Skip to content

Mass Spectrometry-Based Characterization of Electrolytic Decomposition Products of Carbamazepine and Aripiprazole Under Different Electrode Conditions

FieldValue
JournalApplied Sciences
Published2026-08-12
AuthorsMasamitsu Maekawa, Hayahito Ishii, Kenji Miyata, Shunsuke Yokomi, Ryosuke Segawa
DOI10.3390/app16168029

Metal electrodes removed carbamazepine faster than boron-doped diamond electrodes in 1% NaCl. For carbamazepine at 50 mg/L, apparent pseudo-first-order rate constants were 4.09 to 5.15 min⁻¹ for metal and 0.26 to 0.72 min⁻¹ for diamond. Reported half-lives were 0.10 to 0.16 min for metal and 0.96 to 2.63 min for diamond. Metal gave aromatic intermediates consistent with partial oxidation, while diamond gave more polar products consistent with deeper fragmentation.

Both parent compounds declined under every tested condition. The metal electrode generally lowered residual drug concentrations faster under comparable conditions. For aripiprazole at the higher starting concentration, unchanged drug did not fully disappear with diamond, while metal showed a trend toward complete disappearance.

Metal conditions yielded relatively hydrophobic intermediates that retained aromatic chromophores. Diamond conditions yielded more polar, low-molecular-weight products and a more diverse distribution of detected ions. The authors interpret these patterns as partial oxidation and deeper skeletal fragmentation, respectively. Prolonged electrolysis removed most detectable intermediates with both electrode types.

The representative carbamazepine rate data show this difference.

Compound and starting levelElectrodeApparent rate constantApparent half-lifeReported scope
Carbamazepine, 50 mg/LMetal4.09 to 5.15 min⁻¹0.10 to 0.16 minRepresentative runs with variable current
Carbamazepine, 50 mg/LBDD diamond0.26 to 0.72 min⁻¹0.96 to 2.63 minRepresentative runs with variable current

Survival rate was calculated as residual concentration divided by concentration at 0 s. Direct comparison across batches was limited by differences in applied current and initial drug concentration.

Electrochemical platform and mass spectrometry

Section titled “Electrochemical platform and mass spectrometry”

The Eleca system continuously recirculated 2 L of solution between the electrolytic cell and a 2 L recirculation tank. It compared a BDD electrode with a proprietary metal oxide-coated electrode under constant-current operation. The electrolyte was 1% NaCl with conductivity of approximately 17 mS/cm. Current ranged from 1 to 5 A, corresponding to 20 to 100 mA/cm² for an estimated surface area of approximately 50 cm². Applied voltage ranged from 5 to 15 V. Initial pH was approximately 6.0 to 7.0 and rose during electrolysis.

Carbamazepine was prepared by dissolving one 100 mg tablet overnight, giving an estimated 50 mg/L. Aripiprazole came from one 3 mg tablet, giving an estimated 1.5 mg/L. Residual parent drugs were measured by positive-mode LC-MS/MS on a QTRAP 6500 with selected reaction monitoring. Transformation products were examined in positive mode by LC/PDA/HRMS/MS over m/z 50 to 500. Assignments used accurate mass, retention time, UV absorption at 254 and 280 nm, and MS/MS where signal quality allowed.

Scavenger tests added 10% methanol or 10% tert-butanol before electrolysis. Methanol scavenges hydroxyl radicals and chlorine radicals, while tert-butanol was used as the hydroxyl-radical scavenger.

After 40 s with the metal electrode, the parent peak fell and new UV peaks appeared from 5 to 9 min. LC/HRMS detected m/z 253.0972, consistent with mono-oxygenated carbamazepine, at several retention times. The authors tentatively favored monohydroxy isomers over the active 10,11-epoxide metabolite. Published retention data placed the epoxide at less than half the carbamazepine retention time.

Tentative candidates included m/z 208.0755, consistent with deamidated 9-acridinecarboxaldehyde. Other candidates were m/z 196.0754, a ring-contracted deamidated product, and m/z 230.0366, a chlorinated product. The latter carried a chlorine isotope pattern.

At 1800 s with metal, most ions seen at 40 s had largely disappeared. A residual UV-absorbing peak remained near 3.5 min. With diamond after 3000 s, that 3.5 min peak was absent. Multiple low-intensity peaks occupied 3 to 6 min, and HRMS showed greater ion diversity.

The proposed deamidation removes the polar carbamoyl group used for voltage-gated Na+ channel binding. The authors infer that loss of this group could disrupt the pharmacophore. No biological assay tested residual activity.

After 40 s with the metal electrode, aripiprazole declined and UV-active products appeared at 2 to 6 min. Ion m/z 303.1017 matched C14H21N2OCl2+ and had a -3.3 ppm mass error. MS/MS fragments at m/z 231.044 and 91.054 supported its tentative assignment as 1-(2,3-dichlorophenyl)-4-(4-hydroxybutyl)piperazine.

Ions at m/z 337.0626 and 339.0419 carried three chlorine atoms. Their M+2/M ratios were 0.943 and 0.956, respectively. The 33.961 Da difference between m/z 303.1017 and m/z 337.0626 was interpreted as chlorination.

After 1800 s with metal, the aripiprazole peak disappeared, but UV-absorbing peaks remained at 2 to 5 min. The diamond profile at 3000 s contained almost no UV-detectable or ionizable products.

The pre-electrolysis sample already contained m/z 273.1666, 365.1046, and 245.0625 without corresponding UV peaks. The authors treated these as possible background or tablet excipient components rather than confirmed aripiprazole degradation products.

The expected dehydroaripiprazole ion at m/z 446.14 was not detected with either electrode. The expected photo- or thermo-oxidation monoxide ion at m/z 464.15 was also absent. These findings differ from the cited earlier studies.

The candidate at m/z 303.1017 retained the dichlorophenylpiperazine pharmacophore but lacked the quinolinone group and intact butoxy chain. The authors predict reduced D2 receptor affinity, but they did not test this inference.

Total chlorine concentration rose faster with the metal electrode, while pH increased under both electrode conditions. The authors suggest that metal conditions favored active chlorine species such as HOCl and OCl-. Diamond conditions appeared to favor a greater relative contribution from hydroxyl radicals.

Scavenger tests supported this interpretation. The 10% methanol treatment inhibited degradation under both electrode conditions. The 10% tert-butanol treatment produced stronger inhibition under diamond conditions. This agrees with earlier work reporting active chlorine dominance during carbamazepine degradation in NaCl electrolyte.

Direct verification of the dominant reactive species was not performed. The relative contributions of hydroxyl radicals, HOCl, OCl-, chlorine radicals, chlorine oxide radicals, and direct electron transfer remained unresolved.

For treatment design, metal appears better when rapid parent-compound removal is the target. Diamond appears more effective at reducing detectable aromatic and ionizable intermediates during prolonged treatment. Electrode selection therefore depends on the required treatment endpoint and the fate of chlorinated intermediates.

The 1% NaCl matrix did not represent dissolved organic matter, surfactants, co-existing pharmaceuticals, or variable pH and conductivity in medical facility wastewater. The proprietary metal oxide composition and limited electrode characterization also restrict exact replication.

Low product concentrations and the saline matrix reduced electrospray signal. Most product spectra were weak, and no tentative assignment was confirmed with authentic reference standards. Targeted MS/MS, synthesized standards, or orthogonal analysis such as NMR would be needed for stronger confirmation.

The study did not measure TOC, COD, receptor binding, or acute and chronic toxicity. It also did not calculate electrical energy per order, kWh/g, or kWh/mol. Chlorinated CBZ products may have toxicity or environmental persistence independent of pharmacological activity.

Parent-compound disappearance therefore does not demonstrate complete mineralization, detoxification, or environmental safety.

Boron-doped diamond is the working-electrode material for the paper’s pharmaceutical wastewater electrolysis in 1% NaCl. 6CCVD’s diamond materials catalog lists boron-doped diamond and metallization with Au, Pt, Pd, Ti, W, or Cu, while engineering support can be used to define a custom electrode build for laboratory comparison with the paper’s cell.

View Original Abstract

Background: Carbamazepine (CBZ) and aripiprazole (ARI), frequently detected in medical facility effluents, are resistant to conventional wastewater treatment. However, a systematic comparison of their electrolytic degradation product profiles under different electrode conditions using an identical operating platform has not been previously reported. Methods: CBZ and ARI in NaCl solutions were subjected to electrolytic oxidation using the Eleca® system with either a boron-doped diamond (BDD) or a metal electrode. Residual drugs were quantified by LC-MS/MS and pseudo-first-order degradation kinetics were calculated. Transformation products were characterized by LC/PDA/HRMS/MS based on accurate mass measurements, MS/MS fragmentation, chromatographic behavior, and UV absorption. Radical scavenger experiments using methanol and tert-butanol were conducted to investigate reactive species’ contributions to degradation. Results: Both drugs were rapidly degraded under all conditions, with the metal electrode consistently exhibiting higher apparent pseudo-first-order rate constants than the BDD electrode. The metal electrode generated hydrophobic intermediates apparently retaining aromatic skeletons (consistent with partial oxidation), whereas the BDD electrode yielded more polar, low-molecular-weight products consistent with more extensive skeletal fragmentation (deep oxidation). Scavenger experiments suggested a greater relative contribution of hydroxyl radicals (•OH) under BDD electrode conditions. Transformation products were transiently detected but eliminated by prolonged electrolysis. Conclusions: Electrode material critically determines the degradation pathway, reactive species distribution, and product profile. Structural analysis of the tentatively identified transformation products suggests loss of pharmacophore integrity in several intermediates; however, residual pharmacological risk has not been experimentally validated and requires biological sassays for confirmation. Furthermore, the observed disappearance of the parent compounds does not demonstrate complete mineralization, detoxification, or environmental safety, as total organic carbon measurements and ecotoxicological assays were not performed. This study provides the first comparative characterization of transformation products of recalcitrant psychotropic drugs under metal vs. BDD electrode electrolysis, offering mechanistic insights for pharmaceutical wastewater treatment optimization.