Sensors Magazine 2011-01-01

Ionophore-based potentiometric sensors for the flow-injection determination of promethazine hydrochloride in pharmaceutical formulations and human urine.

Ahmed Khudhair Hassan, Bahruddin Saad, Sulaiman Ab Ghani, Rohana Adnan, Afidah Abdul Rahim, Norariza Ahmad, Marina Mokhtar, Suham Tawfiq Ameen, Suad Mustafa Al-Araji

Index: Sensors (Basel.) 11(1) , 1028-42, (2011)

Full Text: HTML

Abstract

Plasticised poly(vinyl chloride)-based membranes containing the ionophores (α-, β- and γ-cyclodextrins (CD), dibenzo-18-crown-6 (DB18C6) and dibenzo-30-crown-10 (DB30C10) were evaluated for their potentiometric response towards promethazine (PM) in a flow injection analysis (FIA) set-up. Good responses were obtained when β- and γ-CDs, and DB30C10 were used. The performance characteristics were further improved when tetrakis(4-chlorophenyl) borate (KTPB) was added to the membrane. The sensor based on β-CD, bis(2-ethylhexyl) adipate (BEHA) and KTPB exhibited the best performance among the eighteen sensor compositions that were tested. The response was linear from 1 × 10(-5) to 1 × 10(-2) M, slope was 61.3 mV decade(-1), the pH independent region ranged from 4.5 to 7.0, a limit of detection of 5.3 × 10(-6) M was possible and a lifetime of more than a month was observed when used in the FIA system. Other plasticisers such as dioctyl phenylphosphonate and tributyl phosphate do not show significant improvements in the quality of the sensors. The promising sensors were further tested for the effects of foreign ions (Li(+), Na(+), K(+), Mg(2+), Ca(2+), Co(2+), Cu(2+), Cr(3+), Fe(3+), glucose, fructose). FIA conditions (e.g., effects of flow rate, injection volume, pH of the carrier stream) were also studied when the best sensor was used (based on β-CD). The sensor was applied to the determination of PM in four pharmaceutical preparations and human urine that were spiked with different levels of PM. Good agreement between the sensor and the manufacturer's claimed values (for pharmaceutical preparations) was obtained, while mean recoveries of 98.6% were obtained for spiked urine samples. The molecular recognition features of the sensors as revealed by molecular modelling were rationalised by the nature of the interactions and complexation energies between the host and guest molecules.