Displaying publications 1 - 20 of 83 in total

Abstract:
Sort:
  1. Shalash M, Makahleh A, Salhimi SM, Saad B
    Talanta, 2017 Nov 01;174:428-435.
    PMID: 28738603 DOI: 10.1016/j.talanta.2017.06.039
    A vortex-assisted liquid-liquid-liquid microextraction method followed by high performance liquid chromatography-diode array detection for the determination of fourteen phenolic acids (cinnamic, m-coumaric, chlorogenic, syringic, ferulic, o-coumaric, p-coumaric, vanillic, p-hydroxybenzoic, caffeic, 2, 4-dihydroxybenzoic, sinapic, gentisic and gallic acids) in honey, iced tea and canned coffee drink samples has been developed. The separation was achieved using a Poroshell 120-EC-C18 column under a gradient elution at a flow rate of 0.6mLmin-1 and mobile phase composed of methanol and acetic acid (1%, v/v). Under the optimum chromatographic conditions, the fourteen phenolic acids were separated in less than 32min. The extraction was performed using a small volume (400µL) of ternary organic solvents (1-pentanol, propyl acetate and 1-hexanol) dispersed into the aqueous sample (10mL) and assisted by vortex agitation (2500rpm for 45s), the analytes were next back-extracted from the organic solvent using 0.02M KOH (40µL) with vortex speed and time of 2500rpm and 60s, respectively. Under these conditions, enrichment factors of 30-193-fold were achieved. The limits of detection (LODs) were 0.05-0.68µgL-1. Recoveries in honey, iced tea and canned coffee drinks were in the range 72.2-112%. The method was successfully applied for the determination of the phenolic acids in honey, iced tea and canned coffee drinks.
  2. Makahleh A, Yap HF, Saad B
    Talanta, 2015 Oct 01;143:394-401.
    PMID: 26078176 DOI: 10.1016/j.talanta.2015.05.011
    A new, rapid and sensitive microextraction technique named vortex-assisted liquid-liquid-liquid microextraction (VALLLME) is proposed. The complete extraction process involves two steps. First, a vortex-assisted liquid-liquid microextraction (VALLME) procedure was used to extract the analytes from a relatively large volume of sample (donor phase) to a small volume of organic solvent (intermediate phase). Next, a micro-vortex-assisted liquid-liquid extraction (µ-VALLE) was used to extract the target analytes from the intermediate phase to a smaller volume of aqueous solution (acceptor phase). The final extract (acceptor phase) can be directly injected into the high performance liquid chromatography or capillary electrophoresis units without any further treatments. The selection of the intermediate phase and the manipulation of pH are key parameters that ensure good extraction efficiency of the technique. The proposed technique has been successfully applied for the determination of carvedilol (used as model analyte) in biological fluid samples. The optimum extraction conditions were: toluene as intermediate phase (150 μL); pH of the donor phase, 9.5; vortex time of the VALLME, 45 s (maximum speed, 2500 rpm); 0.1M HCl (15 μL) as acceptor phase; vortexing time of the µ-VALLME, 75 s (maximum stirring speed, 2500 rpm) and salt concentration in the donor phase, 5% (w/v). Under these conditions, enrichment factors of 51- and 418-fold for VALLME step and VALLLME procedure, respectively, were achieved.
  3. Abu-Bakar NB, Makahleh A, Saad B
    Talanta, 2014 Mar;120:47-54.
    PMID: 24468341 DOI: 10.1016/j.talanta.2013.11.081
    A fast and simple solvent microextraction technique using salting out-vortex-assisted liquid-liquid microextraction (salting out-VALLME) was developed for the extraction of furfurals (2-furfural (2-F), 3-furfural (3-F), 5-methylfurfural (5-MF) and 5-hydroxymethylfurfural (5-HMF)) and patulin (PAT) in fruit juice samples. The optimum extraction conditions for 5 mL sample were: extraction solvent, 1-hexanol; volume of extractant, 200 µL; vortex time, 45 s; salt addition, 20%. The simultaneous determination of the furfurals and PAT were investigated using high performance liquid chromatography coupled with diode array detector (HPLC-DAD). The separation was performed using ODS Hypersil C18 column (4.6 mm i.d × 250 mm, 5 μm) under gradient elution. The detection wavelengths used for all compounds were 280 nm except for 3-F (210 nm). The furfurals and PAT were successfully separated in less than 9 min. Good linearities (r(2)>0.99) were obtained within the range 1-5000 μg L(-1) for all compounds except for 3-F (10-5000 µg L(-1)) and PAT (0.5-100 μg L(-1)). The limits of detection (0.28-3.2 µg L(-1)) were estimated at S/N ratio of 3. The validated salting out-VALLME-HPLC method was applied for the analysis of furfurals and PAT in fruit juice samples (apple, mango and grape).
  4. Safian MT, Sekeri SH, Yaqoob AA, Serrà A, Jamudin MD, Mohamad Ibrahim MN
    Talanta, 2022 Mar 01;239:123109.
    PMID: 34864531 DOI: 10.1016/j.talanta.2021.123109
    With each passing year, the agriculture and wood processing industries generate increasingly high tonnages of biomass waste, which instead of being burned or left to accumulate should be utilized more sustainably. In parallel, advances in green technology have encouraged large companies and nations to begin using eco-friendly materials, including eco-friendly emulsifiers, which are used in various industries and in bio-based materials. The emulsion-conducive properties of lignocellulosic materials such as cellulose, hemicellulose, and lignin, the building blocks of plant and wood structures, have demonstrated a particular ability to alter the landscape of emulsion technology. Beyond that, the further modification of their structure may improve emulsion stability, which often determines the performance of emulsions. Considering those trends, this review examines the performance of lignocellulosic materials after modification according to their stability, droplet size, and distribution by size, all of which suggest their outstanding potential as materials for emulsifying agents.
  5. Citartan M, Gopinath SC, Tominaga J, Chen Y, Tang TH
    Talanta, 2014 Aug;126:103-9.
    PMID: 24881539 DOI: 10.1016/j.talanta.2014.03.043
    Label-free-based detection is pivotal for real-time monitoring of biomolecular interactions and to eliminate the need for labeling with tags that can occupy important binding sites of biomolecules. One simplest form of label-free-based detection is ultraviolet-visible-near-infrared (UV-vis-NIR) spectroscopy, which measure changes in reflectivity as a means to monitor immobilization and interaction of biomolecules with their corresponding partners. In biosensor development, the platform used for the biomolecular interaction should be suitable for different molecular recognition elements. In this study, gold (Au)-coated polycarbonate was used as a platform and as a proof-of-concept, erythropoietin (EPO), a doping substance widely abused by the athletes was used as the target. The interaction of EPO with its corresponding molecular recognition elements (anti-EPO monoclonal antibody and anti-EPO DNA aptamer) is monitored by UV-vis-NIR spectroscopy. Prior to this, to show that UV-vis-NIR spectroscopy is a suitable method for measuring biomolecular interaction, the interaction between biotin and streptavidin was demonstrated via this strategy and reflectivity of this interaction decreased by 25%. Subsequent to this, interaction of the EPO with anti-EPO monoclonal antibody and anti-EPO DNA aptamer resulted in the decrease of reflectivity by 5% and 10%, respectively. The results indicated that Au-coated polycarbonate could be an ideal biosensor platform for monitoring biomolecular interactions using UV-vis-NIR spectroscopy. A smaller version of the Au-coated polycarbonate substrates can be derived from the recent set-up, to be applied towards detecting EPO abuse among atheletes.
  6. Wong YF, Makahleh A, Saad B, Ibrahim MN, Rahim AA, Brosse N
    Talanta, 2014 Dec;130:299-306.
    PMID: 25159413 DOI: 10.1016/j.talanta.2014.07.021
    A sensitive and rapid reversed-phase ultra performance liquid chromatographic (UPLC) method for the simultaneous determination of tocopherols (α-, β-, γ-, δ-), tocotrienols (α-, β-, γ-, δ-), α-tocopherol acetate and α-tocopherol nicotinate is described. The separation was achieved using a Kinetex pentafluorophenyl (PFP) column (150 × 2.1mm, 2.6 µm) with both photodiode array (PDA) and fluorescence (FL) detectors that were connected in series. Column was thermostated at 42°C. Under a gradient system consisting of methanol and water at a constant flow rate of 0.38 mL min(-1), all the ten analytes were well separated in less than 9.5 min. The method was validated in terms of linearity, limits of detection and quantitation, precision and recoveries. Calibration curves of the ten compounds were well correlated (r(2)>0.999) within the range of 100 to 25,000 μg L(-1) for α-tocopherol acetate and α-tocopherol nicotinate, 10 to 25,000 μg L(-1) for α-tocotrienol and 5 to 25,000 μg L(-1) for the other components. The method is simple and sensitive with detection limits (S/N, 3) of 1.0 to 3.0 μg L(-1) (FL detection) and 30 to 74 μg L(-1) (PDA detection). Relative standard deviations for intra- and inter-day retention times (<1%) and peak areas (≤ 4%) were obtained. The method was successfully applied to the determination of vitamin E in vegetable oils (extra virgin olive, virgin olive, pomace olive, blended virgin and refined olive, sunflower, soybean, palm olein, carotino, crude palm, walnut, rice bran and grape seed), margarines and supplements.
  7. Voon CH, Yusop NM, Khor SM
    Talanta, 2022 May 01;241:123271.
    PMID: 35121541 DOI: 10.1016/j.talanta.2022.123271
    Oil and grease content in wastewater is used as an environmental monitoring parameter in the oil and gas industry to prevent serious pollution. Conventional oil and grease laboratory testing is time-consuming and necessitates the use of a hazardous chemical solvent, resulting in non-real-time test data and unnecessary chemical waste. On-site or real-time analysis can enable monitoring of oil and grease in wastewater before discharge to the environment from an operating plant, allowing immediate action to be taken to mitigate environmental impact before contamination spirals out of control. Bioluminescent whole-cell biosensors have been reported to have high sensitivity and selectivity in environmental samples, but only for a few traces of organic compounds such as polycyclic aromatics and naphthalene, allowing for faster analysis times. However, no evaluation of biosensor application for oil and grease (a mixture of hydrocarbons) detection in wastewater, which is critical in the oil and gas industry, has been published to date. Herein, the advantages, disadvantages, challenges, and limitations of using a whole-cell bioluminescent biosensor technology to measure oil and grease content in wastewater are carefully reviewed. This review attempts to bridge the knowledge gap between conventional laboratory methods and biosensor technology in terms of analytical challenges, identifying areas for improvement as well as real-world applications for oil and grease content detection in wastewater.
  8. Ovenden SPB, Webster RL, Micich E, McDowall LJ, McGill NW, Williams J, et al.
    Talanta, 2020 May 01;211:120753.
    PMID: 32070627 DOI: 10.1016/j.talanta.2020.120753
    The organophosphorous nerve agent VX is classified by the Chemical Warfare Convention (CWC) as a Schedule 1 chemical; namely a substance that is highly toxic with no use that is of benefit to society. Even with this classification, the nefarious use of the Schedule 1 chemical VX has been observed, as demonstrated in 2017 in Malaysia. Therefore, undertaking chemical analysis on samples of VX to identify chemical attribution signatures (CAS) for chemical forensics is required. To further understand the chemical profile of VX, and to aid in the identification of potential CAS, three in house synthesised stocks of VX were investigated. The three VX stocks analysed were synthesised in 2014, 2017 and 2018 using the same method, allowing for a comparison of data between each of the stocks at different stages of storage. As opposed to a majority of literature reports, these agent stocks were not stabilised, nor were they subjected to forced degradation. Using NMR, high resolution (HR) LC-HRMS, GC-(EI)MS and GC-(CI)MS to gain a full insight into the CAS profile, a total of 44 compounds were identified. Of these compounds, 30 were readily identified through accurate mass measurement and NIST library matches. A further seven were identified through extensive LC-HRMS/MS studies, with seven remaining unresolved. Several compounds, identified in minor amounts, were able to be traced back to impurities in the precursor compounds used in the synthesis of VX, and hence may be useful as CAS for source attribution.
  9. Wee SS, Ng YH, Ng SM
    Talanta, 2013 Nov 15;116:71-6.
    PMID: 24148375 DOI: 10.1016/j.talanta.2013.04.081
    Carbon dots have great potential to be utilised as an optical sensing probe due to its unique photoluminescence and less toxic properties. This work reports a simple and novel synthesis method of carbon dots via direct acid hydrolysis of bovine serum albumin protein in a one-pot approach. Optimisation of the important synthetic parameters has been performed which consists of temperature effect, acid to protein ratio and kinetics of reaction. Higher temperature has promoted better yield with shorter reaction time. The carbon dots obtained shows a strong emission at the wavelength of 400 nm with an optimum excitation of 305 nm. The potential of the carbon dots as optical sensing probe has been investigated on with different cations that are of environmental and health concern. The fluorescence of the carbon dots was significantly quenched particularly by lead (II) ions in a selective manner. Further analytical study has been performed to leverage the performance of the carbon dots for lead (II) ions sensing using the standard Stern-Volmer relationship. The sensing probe has a dynamic linear range up to 6.0 mM with a Stern-Volmer constant of 605.99 M(-1) and a limit of detection (LOD) of 5.05 μM. The probe performance was highly repeatable with a standard deviation below 3.0%. The probe suggested in this study demonstrates the potential of a more economical and greener approach that uses protein based carbon dots for sensing of heavy metal ions.
  10. Haron MJ, Wan Md ZW, Desa MZ, Kassim A
    Talanta, 1994 May;41(5):805-7.
    PMID: 18966002
    Poly(hydroxamic acid) chelating ion-exchange resin was prepared from crosslinked poly(methacrylate) beads. The starting polymer was prepared by a suspension polymerization of methacrylate and divinyl benzene. Conversion of the ester groups into the hydroxamic acid was carried out by treatment with hydroxylamine in an alkaline solution. Hydroxamic acid capacity of the product was 2.71 mmol/g. The resin exhibited high affinity towards Fe(III) and Pb ions and its capacities for Fe(III), Pb, Cu, Ni and Co ions were pH dependent. The ability of the resin to carry out the separation of Fe(III)CuCo/Ni and PbNi ions is also reported.
  11. Leong CL
    Talanta, 1971 Aug;18(8):845-8.
    PMID: 18960953
    A ternary complex between germanium, Catechol Violet (CV) and cetyltrimethylanunoniuni bromide is proposed for the determination of germanium. The stoichiometric ratio Ge:CV is 1:2. Beer's law is obeyed from 0.1 to 1.0 ppm of Ge. The method is highly selective. Interference from Sn(IV), Fe(III), Bi(III), Cr(VI), Mo(VI), V(V) and Sb(III) in mg amounts is eliminated by extracting the germanium into carbon tetrachloride from 9M HC1 and then stripping into water before the photometric determination.
  12. Saleh MI, Ahmad M, Darus H
    Talanta, 1990 Jul;37(7):757-9.
    PMID: 18965016
    A series of chelating reagents, 1-phenyl-3-methyl-4-(2-fluorobenzoyl)-5-pyrazolone, 1-phenyl-3-methyl-4-(3-fluorobenzoyl)-5-pyrazolone and 1-phenyl-3-methyl-4-(4-fluorobenzoyl)-5-pyrazolone, has been synthesized. The extraction of Ln(III), (Ln = La, Eu and Lu) into chloroform with these reagents at 30 +/- 1 degrees has been studied. The composition of the complexes extracted has been determined by the slope method, and the extraction constants K(ex), were measured. The presence of the fluorine atom in the reagents does not make the K(ex), values much different from those obtained with the parent pyrazolone.
  13. Lim HJ, Saha T, Ooi CW
    Talanta, 2024 Feb 01;268(Pt 2):125376.
    PMID: 37951180 DOI: 10.1016/j.talanta.2023.125376
    Serum levels of dengue virus (DENV) non-structural 1 (NS1) antigen can serve as a valuable prognostic indicator of severe dengue infections. A quartz crystal microbalance (QCM)-based biosensor with a biomimetic recognition element was designed to quantitatively detect DENV NS1 as an early disease biomarker. To mitigate the reliance on costly viral antigens during the molecular imprinting process, a synthetic peptide mimicking a DENV NS1 epitope was used as a surrogate template for the synthesis of an epitope-imprinted polydopamine (EMIPDA) sensing film on the biosensor surface. The maximal frequency shift for DENV NS1 was obtained with an EMIPDA film synthesised using 5 mg mL-1 of dopamine monomer and 0.5 mg mL-1 of peptide template. The EMIPDA-QCM biosensor achieved low detection and quantitation limits of 0.091 μg mL-1 and 0.436 μg mL-1, respectively, allowing acute-phase detection of dengue and prognosis of the disease progression. The EMIPDA-QCM biosensor exhibited remarkable selectivity with up to 68-fold larger frequency responses towards DENV NS1 compared to a major serum protein. The site-specific imprinting approach not only enhanced the biosensing performance but also enabled a 26-fold cost reduction for biosensor functionalisation, providing a cost-effective strategy for label-free biosensing of the dengue biomarker via the biopolymer film.
  14. Semail NF, Abdul Keyon AS, Saad B, Kamaruzaman S, Mohamad Zain NN, Lim V, et al.
    Talanta, 2022 Jan 01;236:122833.
    PMID: 34635223 DOI: 10.1016/j.talanta.2021.122833
    A dynamic pH junction was used in capillary electrophoresis (CE-DAD) to on-line preconcentrate, separate, and determine trace amounts of sulfonamide antibiotics (SAs) in milk and yoghurt samples in this study. A sample matrix with 0.15% acetic acid and 10% methanol (MeOH) at a pH of 4.0, and a background electrolyte (BGE) that contained 35 mM sodium citrate with 10% MeOH at a pH of 8.5, and an acidic barrage of 0.4% acetic acid with 10% MeOH at a pH of 2.5 were utilised to achieve a stacking effect for SAs through a dynamic pH junction. Under optimised conditions, the proposed preconcentration method showed good linearity (30-500 ng/mL, r2 ≥ 0.9940), low limits of detection (LODs) of 4.1-6.3 ng/mL, and acceptable analytes recovery (81.2-106.9%) with relative standard deviations (RSDs) within 5.3-13.7 (n = 9). The limits of quantification (LOQs) were below the maximum residue limit approved by the European Union (EU) in this type of matrices. Sensitivity enhancement factors of up to 129 were reached with the optimised dynamic pH junction using CE with a diode array detector (DAD). The method was used to determine SAs in fresh milk, low-fat milk, full-cream milk, and yoghurt samples.
  15. Ben-Hander GM, Makahleh A, Saad B, Saleh MI, Cheng KW
    Talanta, 2015 Jan;131:590-6.
    PMID: 25281145 DOI: 10.1016/j.talanta.2014.08.037
    A new analytical method for the simultaneous determination of the antidiabetic drugs rosiglitazone (ROS) and metformin hydrochloride (MH) with marked differences in their affinity towards organic solvents (log P of 2.4 and -1.43, respectively) was developed. Prior to the HPLC separation, the drugs were subjected to a sequential hollow fiber liquid phase microextraction (HF-LPME) procedure. Two sequential HF-LPME approaches were considered, the preferred one involves the use of two vials containing solution mixtures for the extraction of ROS (vial 1) and MH (vial 2), respectively, but using the same fiber and acceptor phase. Important parameters that affect the extraction efficiency such as extracting solvent, donor phase conditions, HCl concentration, agitation, extraction time, addition of salt, etc. were studied. Under the optimum conditions, good enrichment factors (EF, 471 and 86.6 for ROS and MH, respectively) were achieved. Calibration curves were linear over the range 1-500 (r(2)=0.998) and 5-2500 ng mL(-1) (r(2)=0.999) for ROS and MH, respectively. The relative standard deviation values (RSD%) for six replicates were below 8.4%. Detection and quantitation limits based on S/N ratio of 3 and 10 were 0.12, 1.0 and 0.36, 3.0 ng mL(-1) for ROS and MH, respectively. The proposed method is simple, sensitive and opens up new opportunities for the microextraction of analytes with contrasting properties.
  16. Taib M, Tan LL, Abd Karim NH, Ta GC, Heng LY, Khalid B
    Talanta, 2020 Jan 15;207:120321.
    PMID: 31594568 DOI: 10.1016/j.talanta.2019.120321
    An optical aptasensor-based sensing platform for rapid insulin detection was fabricated. Aminated porous silica microparticles (PSiMPs) were synthesized via a facile mini-emulsion method to provide large surface area for covalent immobilization of insulin-binding DNA aptamer (IGA3) by glutaraldehyde cross-linking protocol. A Nickel-salphen type complex with piperidine side chain [Ni(II)-SP] was synthesized with a simple one-pot reaction, and functionalized as an optical label due to strong π-π interaction between aromatic carbons of G-quadruplex DNA aptamer and planar aromatic groups of Ni(II)-SP to form the immobilized IGA3-Ni(II)-SP complex, i.e. the dye-labeled aptamer, thereby bringing yellow colouration to the immobilized G-quartet plane. Optical characterization of aptasensor towards insulin binding was carried out with a fiber optic reflectance spectrophotometer. The maximum reflectance intensity of the immobilized IGA3-Ni(II)-SP complex at 656 nm decreased upon binding with insulin as aptasensor changed to brownish orange colouration in the background. This allows optical detection of insulin as the colour change of aptasensor is dependent on the insulin concentration. The linear detection range of the aptasensor is obtained from 10 to 50 μIU mL-1 (R2 = 0.9757), which conformed to the normal fasting insulin levels in human with a limit of detection (LOD) at 3.71 μIU mL-1. The aptasensor showed fast response time of 40 min and long shelf life stability of >3 weeks. Insulin detection using healthy human serums with informed consent provided by participants suggests the DNA aptamer biosensor was in good agreement with ELISA standard method using BIOMATIK Human INS (Insulin) ELISA Kit.
  17. Citartan M, Tang TH
    Talanta, 2019 Jul 01;199:556-566.
    PMID: 30952298 DOI: 10.1016/j.talanta.2019.02.066
    Aptamers are nucleic acid-based molecular recognition elements that are specific and have high binding affinity against their respective targets. On account of their target recognition capacity, aptamers are widely utilized in a number of applications including diagnostics. This review aims to highlight the recent developments of aptasensors expedient for point-of-care (POC) diagnostics. Significant focus is given on the primary assay formats of aptamers such as fluorescence, electrochemical, surface plasmon resonance (SPR) and colorimetric assays. A potpourri of platforms such as paper-based device, lateral flow assay, portable electrodes, portable SPR and smart phones expedient for point-of-care (POC) diagnostics are discussed. Emphasis is also given on the technicalities and assay configurations associated with the sensors.
  18. Poh BL, Khairuddean M
    Talanta, 1996 Oct;43(10):1727-31.
    PMID: 18966659
    A non-cyclic tetrameric structure has been suggested for calcichrome (calcion). This structure is consistent with its mass spectrum, proton NMR spectrum, elemental composition and complexing ability with polyaromatic hydrocarbons in water. The stability constants of the 1:1 complexes formed between calcichrome and seven polyaromatic hydrocarbons in water at room temperature have been measured.
  19. Rozaini MNH, Yahaya N, Saad B, Kamaruzaman S, Hanapi NSM
    Talanta, 2017 Aug 15;171:242-249.
    PMID: 28551135 DOI: 10.1016/j.talanta.2017.05.006
    Molecularly imprinted polymer (MIP) was employed as sorbent in ultrasound assisted emulsification molecularly imprinted polymer micro-solid phase extraction (USAE-MIP-µ-SPE) of bisphenol A (BPA) in water, beverages and the aqueous liquid in canned foods prior to high performance liquid chromatography-diode array detector (HPLC-DAD) analysis. Several effective variables, such as types of emulsification solvent and its volume, types of desorption solvent and its volume, salting out effect, pH of sample solution, mass of sorbent, extraction and desorption time, and sample volume, were optimized comprehensively. Under the optimized USAE-MIP-µ-SPE and HPLC-DAD conditions, the method demonstrated good linearity over the range of 0.5-700μgL-1with a coefficient determination of R2=0.9973, low limit of detection (0.07μgL-1), good analyte recoveries (82.2-118.9%) and acceptable RSDs (0.7-14.2%, n=3) with enrichment factor of 49. The method was applied to thirty samples of drinking water, mineral water, river water, lake water, as well as beverages and canned foods, the presence of BPA was identified in four samples. The proposed method showed good selectivity and reusability for extraction of BPA, and hence the USAE-MIP-µ-SPE is rapid, simple, cost effective and environmentally friendly.
  20. Goh KM, Maulidiani M, Rudiyanto R, Wong YH, Ang MY, Yew WM, et al.
    Talanta, 2019 Jun 01;198:215-223.
    PMID: 30876552 DOI: 10.1016/j.talanta.2019.01.111
    The technique of Fourier transform infrared spectroscopy is widely used to generate spectral data for use in the detection of food contaminants. Monochloropropanediol (MCPD) is a refining process-induced contaminant that is found in palm-based fats and oils. In this study, a chemometric approach was used to evaluate the relationship between the FTIR spectra and the total MCPD content of a palm-based cooking oil. A total of 156 samples were used to develop partial least squares regression (PLSR), artificial neural network (nnet), average artificial neural network (avNNET), random forest (RF) and cubist models. In addition, a consensus approach was used to generate fusion result consisted from all the model mentioned above. All the models were evaluated based on validation performed using training and testing datasets. In addition, the box plot of coefficient of determination (R2), root mean square error (RMSE), slopes and intercepts by 100 times randomization was also compared. Evaluation of performance based on the testing R2 and RMSE suggested that the cubist model predicted total MCPD content with the highest accuracy, followed by the RF, avNNET, nnet and PLSR models. The overfitting tendency was assessed based on differences in R2 and RMSE in the training and testing calibrations. The observations showed that the cubist and avNNET models possessed a certain degree of overfitting. However, the accuracy of these models in predicting the total MCPD content was high. Results of the consensus model showed that it slightly improved the accuracy of prediction as well as significantly reduced its uncertainty. The important variables derived from the cubist and RF models suggested that the wavenumbers corresponding to the MCPDs originated from the -CH=CH2 or CH=CH (990-900 cm-1) and C-Cl stretch (800-700 cm-1) regions of the FTIR spectrum data. In short, chemometrics in combination with FTIR analysis especially for the consensus model represent a potential and flexible technique for estimating the total MCPD content of refined vegetable oils.
Filters
Contact Us

Please provide feedback to Administrator (afdal@afpm.org.my)

External Links