Guillaume L. Erny

Direct photodegradation of carbamazepine followed by micellar electrokinetic chromatography and mass spectrometry

Carbamazepine, a widely consumed psychotropic pharmaceutical, is one of the most commonly detected drugs in the environment. To better assess the environmental persistence of carbamazepine in aqueous matrices, the effect of pH and dissolved oxygen on the direct photodegradation rate of this pharmaceutical was evaluated in this study, using simulated solar irradiation. In order to follow the degradation and the emergence of photoproducts, a micellar electrokinetic chromatography based method was developed, consisting on the use of a dynamically coated capillary column. The developed methodology showed good repeatability and efficiency in the separation of carbamazepine and photoirradiation products. Also, seven photodegradation products were identified by electrospray mass spectrometry (ESI-MS), including the known carcinogenic acridine that was produced under all the pH and oxygenation levels studied and one newly identified photoproduct. This paper gives new insights into the role of dissolved oxygen on the photodegradation rate of carbamazepine. The results indicate that acidic pH, combined with the absence of dissolved oxygen in the aqueous matrix, results in very high direct photodegradation rates. At basic pH, dissolved oxygen does not interfere with the process and very low rates were observed. At environmentally relevant conditions, carbamazepine was shown to persist in the environment from 4.5 to 25 days.

Applications of molecularly imprinted polymers to the analysis and removal of personal care products: A review

Personal-care products (PCPs) involve a variety of chemicals whose persistency along with their constant release into the environment raised concern to their potential impact on wildlife and humans health. Regarded as emergent contaminants, PCPs demonstrated estrogenic activity leading to the need of new methodologies to detect and remove those compounds from the environment. Molecular imprinting starts with a complex between a template molecule and a functional monomer, which is then polymerized in the presence of a cross-linker. After template removal, the polymer will contain specific cavities. Based on a good selectivity towards the template, molecularly imprinted polymers (MIPs) have been investigated as efficient materials for the analysis and extraction of the so called emergent pollutants contaminants. Rather than lowering the limit of detections, the key theoretical advantage of MIP over existing methodologies is the potential to target specific chemicals. This unique feature, sometime named specificity (as synonym to very high selectivity) allows to use cheap, simple and/or rapid quantitative techniques such as fast separation with ultra-violet (UV) detection, sensors or even spectrometric techniques. When a high degree of selectivity is achieved, samples extracted with MIPs can be directly analyzed without the need of a separation step. However, while some papers clearly demonstrated the specificity of their MIP toward the targeted PCP, such prove is often lacking, especially with real matrices, making it difficult to assess the success of the different approaches. This review paper focusses on the latest development of MIPs for the analysis of personal care products in the environment, with particular emphasis on design, preparation and practical applications of MIPs.

Quantification of organic acids in beer by nuclear magnetic resonance (NMR)-based methods

The organic acids present in beer provide important information on the products quality and history, determining organoleptic properties and being useful indicators of fermentation performance. NMR spectroscopy may be used for rapid quantification of organic acids in beer and different NMR-based methodologies are hereby compared for the six main acids found in beer (acetic, citric, lactic, malic, pyruvic and succinic). The use of partial least squares (PLS) regression enables faster quantification, compared to traditional integration methods, and the performance of PLS models built using different reference methods (capillary electrophoresis (CE), both with direct and indirect UV detection, and enzymatic essays) was investigated. The best multivariate models were obtained using CE/indirect detection and enzymatic essays as reference and their response was compared with NMR integration, either using an internal reference or an electrical reference signal (Electronic REference To access In vivo Concentrations, ERETIC). NMR integration results generally agree with those obtained by PLS, with some overestimation for malic and pyruvic acids, probably due to peak overlap and subsequent integral errors, and an apparent relative underestimation for citric acid. Overall, these results make the PLS-NMR method an interesting choice for organic acid quantification in beer.

Time of flight versus ion trap MS coupled to CE to analyse intact proteins.

In this work, two different CE-MS instruments, namely, CE-ESI-IT-MS and CE-ESI-TOF-MS, applied to analyse intact proteins from complex samples are investigated. The aim of this work was to compare both instruments in terms of LOD, number of proteins detected, and precision and repeatability in the determination of the protein relative molecular mass. Results show that although CE-ESI-IT-MS provides cleaner MS spectra of intact proteins, CE-ESI-TOF-MS allows the identification of a higher number of proteins from complex matrices in an easier way. Performance in terms of peak area reproducibility, LOD and precision in the determination of the molecular mass were similar for both instruments. The usefulness of the optimised CE-ESI-IT-MS and CE-ESI-TOF-MS conditions was demonstrated by studying the zein-proteins composition of three natural maize lines and their corresponding transgenic lines, showing no significant differences.

Electromigration dispersion in capillary zone electrophoresis. Experimental validation of use of the Haarhoff-Van der Linde function.

This paper provides experimental validation of the use of the Haarhoff-Van der Linde (HVL) peak fitting function to fit experimental capillary zone electrophoresis (CZE) electropherograms. The test mixtures were composed of paraquat over a five order of magnitude concentration range (1.2 microM to 120 mM) and 4-aminopyridine at constant concentration (0.53 mM) as internal standard. Peak descriptors and electrophoresis parameters were extracted reliably by a Gaussian function from 4 to 40 microM; by the HVL function from 120 microM to 4 mM; and by a triangular function from 4 to 120 mM. The HVL function can be used where there is significant peak asymmetry due to electromigration distortion (EMD) and the Gaussian contribution toward the peak variance is greater than 25%. The peak centre (a1) and the Gaussian variance (a2) of the paraquat peak are shown to be independent of concentration. Diffusion coefficients obtained from a2 for both analytes were found to be in good agreement with their theoretical values. For all peaks where the distortion coefficient (a3) can be extracted, this parameter is shown to be directly proportional to the sample loading, as predicted by EMD theory. For the 4-aminopyridinium ion, mobilities calculated from a3 and measured independently are in excellent agreement. These results show that the HVL function accurately describes the two major processes, diffusion and EMD. contributing to the variance during a CZE separation.

Comparison between MEKC and UV spectral deconvolution to follow sorption experiment in soil.

In this study a simple and rapid procedure for monitoring sorption of atrazine onto soil samples was developed. This method is based on a multiwavelength UV spectral deconvolution (UVSD) where the UV spectrum of a sample is considered as a linear combination of absorption spectra, named reference spectra. The combination of the reference spectra allows the restitution of the shape of the UV spectrum of any unknown sample. In order to evaluate the accuracy of this method, a sorption study of atrazine to three different soil samples was monitored using both UVSD and micellar electrokinetic chromatography (MEKC) methods. The Freundlich parameters (K(F) and N) obtained for all soil samples tested were similar for both methods and UVSD has proven to be an accurate methodology, since the results did not present statistical significant differences at 95% confidence level. Moreover, the LOD obtained using UVSD presented lower values (0.066-0.12mgL(-1)) than the one obtained using MEKC (0.26mgL(-1)). The proposed UVSD method has been proven useful as a valid alternative, to the more common procedures to follow sorption experiment in soil matrix samples with no need of sample pre-treatment or column conditioning. This method is much faster and requires less sample manipulation than traditional analytical separation methods. Moreover, most modern spreadsheet-applications, such as Excel, now include statistical packages that allow performing multi-linear regression, which make this approach particularly inexpensive since the only requirement is a UV spectrometer. The regression output was made using the Microsoft Excel, a very easy and fast program, allowing anyone, with less knowledge about complex statistical programs, to perform the UV spectral.

Introducing the concept of centergram. A new tool to squeeze data from separation techniques-mass spectrometry couplings

In separation techniques hyphenated to mass spectrometry (MS) the bulk from the separation step is continuously flowing into the mass spectrometer where the compounds, arriving at each separation time, are ionized and further separated based on their m/z ratio. An MS detector is recognized as being a universal detector, although it can also be a very selective instrument. In spite of these advantages, classical two dimensional representations from these hyphenated systems, such as those based on the base peak of electropherogram/chromatogram or on the total ion of electropherogram/chromatogram, usually hide a large number of features that if correctly assessed will show the presence of co-migrating species and/or the low abundant ones. The uses of peak picking algorithms to detect and measure as many peaks as possible from a dataset allow extracting much more information. However, a single migrating compound usually produces a multiplicity of ions, making difficult to differentiate peaks generated by the same compound from other peaks due e.g., to closely co-migrating/eluting species. In this work, a new representation is proposed and its usefulness demonstrated with experimental data from capillary electrophoresis-hyphenated to a time of flight mass spectrometer via an electrospray interface. This representation, called centergram, is obtained after using a peak picking methodology that detects electrophoretic peaks of single ions and measure their positions. The centergram is the histogram (i.e. the count of the number of observations that fall into each one of the intervals, known as bins, as determined by the user) of the measured positions. The intensity of the bars in this histogram will indicate the amount of peaks in the whole dataset whose centers are within each interval. As a compound that has been separated and has entered the MS instrument will produce multiple images at the same position along the m/z dimension, the centergram will exhibit a series of intense bars around the migration time. Those bars will allow defining a centergram peak whose area will be proportional to the number of different types of ions that have been generated in the ionization chamber, the position will be equal to the migration/retention time of the parent compounds and the width will depend on the precision in the measurement of the peak positions. The efficiency of this peak is determined to be up to thirty times higher than the equivalent peak in the classical base peak electropherogram allowing detecting easily co-migrating peaks or the presence of compounds at very low abundance. The number of peaks detected by using this new tool called centergram was increased by more than a factor of 3 compared to the standard representations.

Liquid chromatography with diode array detection combined with spectral deconvolution for the analysis of some diterpene esters in Arabica coffee brew

In this manuscript, the separation of kahweol and cafestol esters from Arabica coffee brews was investigated using liquid chromatography with a diode array detector. When detected in conjunction, cafestol, and kahweol esters were eluted together, but, after optimization, the kahweol esters could be selectively detected by setting the wavelength at 290 nm to allow their quantification. Such an approach was not possible for the cafestol esters, and spectral deconvolution was used to obtain deconvoluted chromatograms. In each of those chromatograms, the four esters were baseline separated allowing for the quantification of the eight targeted compounds. Because kahweol esters could be quantified either using the chromatogram obtained by setting the wavelength at 290 nm or using the deconvoluted chromatogram, those compounds were used to compare the analytical performances. Slightly better limits of detection were obtained using the deconvoluted chromatogram. Identical concentrations were found in a real sample with both approaches. The peak areas in the deconvoluted chromatograms were repeatable (intraday repeatability of 0.8%, interday repeatability of 1.0%). This work demonstrates the accuracy of spectral deconvolution when using liquid chromatography to mathematically separate coeluting compounds using the full spectra recorded by a diode array detector.

Application of MEKC to the monitoring of atrazine sorption behaviour on soils.

Measuring interactions between soils and xenobiotics is critical to assess the threat to water resources posed by potentially harmful chemicals. Such studies are done via batch experiments where the chemical of interest is incubated with the soil for a long period of time. However, the analysis of the remaining concentration in the aqueous phase is often troublesome, due to dissolved organic matter that can interfere with the analysis. Thus, for such application a lengthy sample clean-up is often needed prior to analysis. The aim of this work was to develop an alternative approach to monitor atrazine using MEKC. A buffer of 10 mM NaH(2)PO(4) and 50 mM SDS was used to separate atrazine and soil organic matter in less than 7 min. Samples were injected without any pre-treatment leading to a significantly lower total analysis time (sample clean-up, column conditioning, separation) than HPLC. After full optimisation, good repeatability of retention time (RSD<1%) and peak areas (RSD<3%) was achieved as well as a good linearity (r(2)=0.9994) and relatively low LOD (0.26 mg/L). The method was applied to study the sorption of atrazine to a soil sample using batch equilibrium technique.

Noise normalisation in capillary electrophoresis using a diode array detector

When using capillary electrophoresis with a diode array detector, the wavelength at maximum absorbance is often chosen to quantify a given analyte. However, the background noise for every wavelength should be taken into account as it is by maximising the signal to noise ratio that the lowest limit of detection will be obtained. Here, we proposed an algorithm allowing to correct an electropherogram from its background absorption and to estimate the background noise. Applying it to all the electropherograms obtained in each wavelength channel allows obtaining the background noise as a function of the wavelength, which can be used to calculate the signal to noise ratio. This not only allows selecting the best wavelength to maximise the limit of detection of a given analyte, but also to generate a noise normalised base peak electropherogram (nn-BPE). It is shown that the noise normalised base peak electropherograms substantially improve the peaks visualisation. The algorithm is part of a graphic user interface that runs under MatLab environment; it does not require any programming knowledge and is freely available.