Recognizing chromatographic peaks with pattern recognition methods : Part 1. Development of a k-nearest-neighbour technique

Problems in automated peak recognition in chromatography are discussed. An algorithm based on the k-nearest neighbour technique is proposed. Recognition of a peak is done by comparing it with a predefined profile function (normally a Gaussian peak profile). The profile and a part of the chromatogram are both interpreted as points in a multi-dimensional pattern space. The distance between the two points gives the value of the peak recognition function. The effects of different properties of chromatographic peaks (i.e., peak width, peak height and noise) and of the profile parameter (i.e., dimension of the pattern space, shape and width of the function, and characteristics of the distance measure) are evaluated. The method has excellent properties for recognizing peaks with low signal/noise (S/N) ratios; an example with S/N = 1 is shown. Changing peak widths and drifting baselines have little effect on the recognition ability. Difficulties with changing peak heights can be compensated by range scaling. Problems occur when two peaks are not sufficiently separated.     

Testing the capability of a polynomial‐modified gaussian model in the description and simulation of chromatographic peaks of amlodipine and its impurity in ion‐interaction chromatography

In this paper, the capability of a polynomial‐modified Gaussian model to relate the peak shape of basic analytes, amlodipine, and its impurity A, with the change of chromatographic conditions was tested. For the accurate simulation of real chromatographic peaks the authors proposed the three‐step procedure based on indirect modeling of peak width at 10% of peak height (W_0.1), individual values of left‐half width (A) and right‐half width (B), number of theoretical plates (N), and tailing factor (Tf). The values of retention factors corresponding to the peak beginning (k_B), peak apex (k_A), peak ending (k_E), and peak heights (H₀) of the analytes were directly modeled. Then, the investigated experimental domain was divided to acquire a grid of appropriate density, which allowed the subsequent calculation of W_0.1, A, B, N, and Tf. On the basis of the predicted results for Tf and N, as well as the defined criteria for the simulation the following conditions were selected: 33% acetonitrile/67% aqueous phase (55 mM perchloric acid, pH 2.2) at 40°C column temperature. Perfect agreement between predicted and experimental values was obtained confirming the ability of polynomial modified Gaussian model and three‐step procedure to successfully simulate the real chromatograms in ion‐interaction chromatography.     

Chemometrics‐assisted chromatographic fingerprinting: An illicit methamphetamine case study

The volatile chemical constituents in complex mixtures can be analyzed using gas chromatography with mass spectrometry. This analysis allows the tentative identification of diverse impurities of an illicit methamphetamine sample. The acquired two‐dimensional data of liquid–liquid extraction was resolved by multivariate curve resolution alternating curve resolution to elucidate the embedded peaks effectively. This is the first report on the application of a curve resolution approach for chromatogram fingerprinting to identify particularly the embedded impurities of a drug of abuse. Indeed, the strong and broad peak of methamphetamine makes identifying the underlying peaks problematic and even impossible. Mathematical separation instead of conventional chromatographic approaches was performed in a way that trace components embedded in methamphetamine peak were successfully resolved. Comprehensive analysis of the chromatogram, using multivariate curve resolution, resulted in elution profiles and mass spectra for each pure compound. Impurities such as benzaldehyde, benzyl alcohol, benzene, propenyl methyl ketone, benzyl methyl ketone, amphetamine, N‐benzyl‐2‐methylaziridine, phenethylamine, N ,N ,α‐trimethylamine, phenethylamine, N ,α,α‐trimethylmethamphetamine, N‐acetylmethamphetamine, N‐formylmethamphetamine, and other chemicals were identified. A route‐specific impurity, N‐benzyl‐2‐methylaziridine, indicating a synthesis route based on ephedrine/pseudoephedrine was identified. Moreover, this is the first report on the detection of impurities such as phenethylamine, N ,α,α‐trimethylamine (a structurally related impurity), and clonitazene (as an adulterant) in an illicit methamphetamine sample.     

Characterization and complete separation of major cyclolinopeptides in flaxseed oil by reversed‐phase chromatography

Organoleptic properties of flaxseed oil deteriorate during storage due to methionine oxidation in its major cyclolinopeptides. Cyclolinopeptide E was previously identified as being responsible for the manifestation of bitter taste with flaxseed oil ageing. We developed a chromatographic procedure to monitor the oxidation of major cyclic peptides in flaxseed oil. We also used liquid chromatography with mass spectrometry and high‐efficiency core–shell reversed‐phase sorbents to study the separation of cyclolinopeptides in detail. The Kinetex^TM family of stationary phases (C₈, C₁₈, phenyl‐hexyl) was tested, along with the standard porous Luna^TM C₁₈(2) media. We found that only the phenyl‐hexyl stationary phase allows for complete resolution of major cyclolinopeptides, thus permitting direct UV monitoring of degree of conversion for cyclolinopeptide B into C and L into E. We also report, for the first time, a significant effect of peak splitting for some methionine S‐oxide (Mso) containing cyclolinopeptides, which most likely appear due to diastereomerization. This results in poor separation efficiency for cyclolinopeptides F, G, and E, and gives baseline resolution of diastereomeric pairs for cyclolinopeptides I and P. Thus, a single oxidation of cyclolinopeptide N yields three distinct chromatographic peaks corresponding to cyclolinopeptide T (cyclo‐MsoLMPFFWV, reported for the first time) and pair of cyclolinopeptide I (cyclo‐MLMsoPFFWV) diastereomers.     

Preserving the chromatographic integrity of high-speed supercritical fluid chromatography separations using time-of-flight mass spectrometry

The advantage of high-speed time-of-flight-mass spectrometry (TOF-MS) detection for ultrafast qualitative supercritical fluid chromatography/mass spectrometry (SFC/MS) applications allows the superior resolving power of SFC to be exploited in high-throughput analysis. A chromatographic comparison of quadrupole MS and TOF-MS shows high-speed TOF total ion current data point sampling to be more indicative of fast SFC separations and corresponding short (1-2 s) baseline peak widths. Results shown for analysis of a six-compound mixture with two peaks eluting at 0.86 and 0.89 min exhibit >50% resolution by high-speed TOF data sampling, whereas the same peaks appear to coelute using quadrupole MS data sampling. Additionally, a marked improvement in the peak baseline widths is afforded by fast TOF data acquisition of 0.1 s/spectrum, resulting in a reduction in the baseline width, 1.6 s, of sulfanilamide in a four-compound mixture that is more than 2-fold greater than that achieved at the slower data acquisition of 0.5 s/spectrum. The resulting increase in resolution and improved peak shapes allow automatic integration routines to perform more effectively. For most classes of compounds amenable to high performance liquid chromatography, including druglike species, steroids, and polymers, the union of SFC with TOF-MS provides the maximum density of chemical information per unit time available with any high-speed chromatographic/mass spectrometric method.     

Recursive Wavelet Peak Detection of Analytical Signals

A novel algorithm, entitled recursive wavelet peak detection (RWPD), is proposed to detect both normal and overlapped peaks in analytical signals. Recursive peak detection is based on continuous wavelet transforms (CWTs), which can be used to obtain initial peak positions even for overlapped peaks. Genetic algorithm (GA) and Gaussian fitting are used to refine peak parameters (peak positions, widths, and heights). Finally, area of peaks can be calculated by numeric integration. Simulated and ultrahigh performance liquid chromatographic ion trap time-of-flight mass spectrometry (UPLC-IT-TOF-MS) data sets have been analyzed by RWPD, MassSpecWavelet, and peakfit package by Tom O’Haver. Results show that RWPD can obtain more accurate positions and smaller relative fitting errors than MassSpecWavelet and peakfit, especially in overlapped peaks. RWPD is a convenient tool for peak detection and deconvolution of overlapped peaks, and it has been developed in R programming language and is available at https://github.com/zmzhang/RWPD.

     

Stereoselective separation of (1S, 4S)‐sertraline from medicinal reaction mixtures by countercurrent chromatography with hydroxypropyl‐β‐cyclodextrin as stereoselective selector

Four stereoisomeric components were produced during the synthesis of the antidepressant drug (1S, 4S)‐sertraline hydrochloride due to the two chiral carbon centers in its chemical structure, including (1S, 4S), (1R, 4R), (1S, 4R), and (1R, 4S)‐isomer. Stereoselective separation of the target isomer (1S, 4S)‐sertraline from the medicinal reaction mixtures by countercurrent chromatography using hydroxypropyl‐β‐cyclodextrin as the stereoselective selector was investigated. A biphasic solvent system composed of n‐hexane/0.20 mol/L phosphate buffer solution with pH 7.6 containing 0.10 mol/L of hydroxypropyl‐β‐cyclodextrin (1:1, v/v) was selected for separation of cis‐sertraline and trans‐sertraline using reverse phase elution mode and (1S, 4S)‐sertraline was separated with (1R, 4R)‐sertraline using recycling elution mode. A fabricated in‐house analytical countercurrent chromatographic apparatus was used for optimization of the separation conditions. Stationary phase retention and peak resolution were investigated for separation of cis‐sertraline and trans‐sertraline by the analytical apparatus.     

Scope of partial least‐squares regression applied to the enantiomeric composition determination of ketoprofen from strongly overlapped chromatographic profiles

Valuable quantitative information could be obtained from strongly overlapped chromatographic profiles of two enantiomers by using proper chemometric methods. Complete separation profiles where the peaks are fully resolved are difficult to achieve in chiral separation methods, and this becomes a particularly severe problem in case that the analyst needs to measure the chiral purity, i.e., when one of the enantiomers is present in the sample in very low concentrations. In this report, we explore the scope of a multivariate chemometric technique based on unfolded partial least‐squares regression, as a mathematical tool to solve this quite frequent difficulty. This technique was applied to obtain quantitative results from partially overlapped chromatographic profiles of R‐ and S‐ketoprofen, with different values of enantioresolution factors (from 0.81 down to less than 0.2 resolution units), and also at several different S:R enantiomeric ratios. Enantiomeric purity below 1% was determined with excellent precision even from almost completely overlapped signals. All these assays were tested on the most demanding condition, i.e., when the minor peak elutes immediately after the main peak. The results were validated using univariate calibration of completely resolved profiles and the method applied to the determination of enantiomeric purity of commercial pharmaceuticals.     

Chiral separation of N-methyl-dl-aspartic acid in rat brain tissue as N-ethoxycarbonylated (S)-(+)-2-octyl ester derivatives by GC-MS

A selective and sensitive analytical method was developed for enantiomeric separation and determination of N‐methyl‐DL‐aspartic acid (NMA). The method involved the conversion of each enantiomer into N‐ethoxycarbonylated (S)‐(+)‐2‐octyl ester derivative for the direct separation by gas chromatography–mass spectrometry (GC‐MS). The diastereomeric derivatives showed characteristic mass spectral properties for analysis by selected ion monitoring mode (SIM) and enabling enantioseparation on an achiral capillary column. Two enantiomers were baseline separated, and the detection limits for N‐methyl‐L‐aspartic acid (NMLA) and N‐methyl‐D‐aspartic acid (NMDA) were 0.07 and 0.03 ng/g, respectively. When applied to rat brain tissues for absolute configuration of NMA, only NMDA was determined, while NMLA was monitored as lower than the limit of detection. Copyright © 2012 John Wiley & Sons, Ltd.     

Analysis of Morpholinium Ionic Liquid Cations by Hydrophilic Interaction Columns Coupled with Indirect UV Detection

A method was developed for the separation and detection of morpholinium ionic liquid cations by hydrophilic interaction column combined with indirect ultraviolet detection using imidazolium ionic liquids as ultraviolet absorbents in high‐performance liquid chromatography. The effects of the ultraviolet absorbents, organic solvents, and the pH value of the aqueous solution in the mobile phase for the determination of morpholinium cations were investigated by using a hydrophilic column with carbamoyl group as the analytical column. The retention and detection behavior of morpholinium cations was discussed. The suitable chromatographic conditions were 0.8 mmol/L 1‐ethyl‐3‐methyl‐imidazolium tetrafluoroborate aqueous solution (pH 3.5 adjusted with acetic acid)/acetonitrile (45/55, v/v) as the mobile phase and a detection wavelength of 210 nm. Under these conditions, the baseline separation of N‐methyl‐N‐ethyl‐morpholinium cation ([MEMo]⁺) and N‐methyl‐N‐propyl‐morpholinium cation ([MPMo]⁺) was successfully achieved in 15 min. The detection limits of [MEMo]⁺ and [MPMo]⁺ were 0.595 and 0.531 mg/L, respectively. Relative standard deviations were less than 0.2%. This method has been successfully applied to the analysis of morpholinium ionic liquid samples synthesized in chemical laboratories, which is simple, reliable, and practical.     

Peak recognition imitating human judgement

Summary Peak integration is still a major source of error in analytical techniques such as chromatography (LC and GC), aapillary electrophoresis (CE), spectrosocpy, and electrochemistry. If the baseline is complex, e.g. because of matrix effects, or if the peak shape is irregular, e.g. because of peak tailing, the results are often not satisfactory when classical procedures are used. These shortcomings arise because of the stepwise appearance of the chromatogram. An algorithm that copies the human method of considering baseline and peaks as a whole has already been introduced. Here the use of a straight line as a baseline model led to an improvement in several instances. The baseline is, however, usually not exactly straight and rigid. A baseline model with flexible properties is more advantageous. Thus the smoothing cubic spline function is applied in this work. Here the rigidity can be controlled by use of a parameterp _k. The prediction interval of the spline is used for iterative distinction between baseline and peak regions. Afterwards straightforward optimization of the peak boundaries is applied. More than 50 series of consecutive injections of the same sample (n=40 on average) were used to test the performance of this procedure. The same raw data have been integrated by means of the algorithm described here and by use of commercially available software. The reproducibility of the main component peak are within the series was taken as a measure of integration quality. Typically the new procedure reducesRSD % by approximately 33% (e.g. from 1.5% to 1.0%). The improvement is even more impressive for difficult samples with complex matrices, e.g. blood plasma or polymer excipients. for such samples improvements of up to a factor of 6 are obtained.     

Factors influencing chromatographic data in fast gas chromatography with on‐column injection

The use of larger volume injection with on‐column injection and fast GC commercial instrumentation was evaluated with the model mixture of n‐alkanes of a broad range of volatility (C₁₀–C₂₈). The presented configuration allows introduction of 40–80‐fold larger sample volumes without any distortion of peak shapes compared to “usual” fast GC set‐ups using narrow‐bore columns. A normal‐bore retention gap (1–5 m×0.32 mm ID) was coupled to a narrow‐bore (5 m×0.1 mm ID×0.4 μm film thickness) analytical column using a standard press‐fit connector. The connection was tight and reliable, and hence suitable for hydrogen as carrier gas. The effect of pre‐column and analytical column connector, injection volume, pre‐column length, column inlet pressure, and analyte volatility on peak shape, peak broadening, and focusing are discussed. The precision of chromatographic data measurements and peak capacity under optimised temperature programmed conditions for fast separations with large volume injection were found to be very good. The presented fast GC set‐up with on‐column injection extends the applicability of the technique to trace analysis.     

Enantioselective determination of ketamine in dog plasma by chiral liquid chromatography–tandem mass spectrometry

Ketamine is an N‐methyl‐d ‐aspartate receptor antagonist that is usually used clinically as a racemic mixture. Its two enantiomers exhibit different pharmacological activities. To determine whether the enantiomers have different pharmacokinetic profiles, a chiral liquid chromatography–tandem mass spectrometry method was developed and validated for the determination of ketamine enantiomers in dog plasma. The enantiomers of ketamine were extracted from 50 μL of plasma by methyl tert‐butyl ether. Adequate chromatographic retention and baseline resolution of the enantiomers were achieved within a runtime of 5 min on a chiral column coated with polysaccharide derivatives, using a gradient mobile phase of acetonitrile and 10 mm ammonium bicarbonate aqueous solution. Ketamine enantiomers were detected by mass spectrometry with multiple reaction monitoring mode using the transitions of m/z 238.3 → 125.9 for the analytes and m/z 237.1 → 194.1 for carbamazepine (internal standard). The method was linear over the concentration range from 0.5 to 500 ng/mL for each enantiomer. The lower limit of quantification (LLOQ) for each enantiomer was 0.5 ng/mL. The intra‐ and inter‐day precision was <7.3% and 8.5% for R‐ and S‐ketamine, respectively. The accuracy was 92.9–110.4% for R‐ketamine and 99.8–102.4% for S‐ketamine. The method was successfully applied to characterize the stereoselective pharmacokinetic profiles of ketamine in beagle dogs.     

Theoretical Study of Origin of Triple Fluorescences of a Squaraine Dye, Bis[4-(N,N-dimethylamino)phenyl]-squaraine

A new scheme of photo‐fluorescent emission origin, described as S₀ (relaxed state)→S_n (Frank‐Condon state)→ S_m (relaxed state)→S₀ (Frank‐Condon state), is presented to explain the multiple fluorescent emissions of squaraine dyes observed experimentally according to the configuration interaction singles calculations of relaxed excited states of a model compound, bis[4‐(N,N‐dimethylamino)phenyl]squaraine (SQ). It is exhibited that all triple fluorescent emissions of SQ have their significant origin in vertical electron transitions of different relaxed excited states. In addition, some important absorption peaks appearing in higher energy region are most likely to be responsible for the higher energy band observed in solid states of many squaraine dyes.     

One-Stage Determination of the Number of Hydroxyl Groups in Polyatomic Phenols by Gas Chromatography Using Mixed Reagents1

The use of mixtures of compounds of similar chemical nature is recommended for derivatization in chromatographic analysis. It is assumed that every initial compound containing Nreaction centers in the molecule yields (N + 1) derivatives as a set of approximately equidistant retention peaks in the index scale with the intensity ratios close to binomial coefficients. A one-stage procedure for obtaining volatile derivatives of polyatomic phenols by the action of mixed perfluoroacylating reagents is proposed as the simplest way to determine the number of hydroxyl groups in their molecules. The possibility to extend this method to other classes of organic compounds with active hydrogen atoms and mixtures of other reagents is considered.     

Generalized Gaussian reference curve measurement model for high‐performance liquid chromatography with diode array detector separation and its solution by multi‐target intermittent particle swarm optimization

In order to separate a high‐performance liquid chromatography with diode array detector (HPLC‐DAD) data set to chromatogram peaks and spectra for all compounds, a separation method based on the model of generalized Gaussian reference curve measurement (GGRCM) and the algorithm of multi‐target intermittent particle swarm optimization (MIPSO) is proposed in this paper. A parameter θ is constructed to generate a reference curve r(θ) for a chromatogram peak based on its physical principle. The GGRCM model is proposed to calculate the fitness ε(θ) for every θ, which indicates the possibility for the HPLC‐DAD data set to contain a chromatogram peak similar to the r(θ). The smaller the fitness is, the higher the possibility. The algorithm of MIPSO is then introduced to calculate the optimal parameters by minimizing the fitness mentioned earlier. Finally, chromatogram peaks are constructed based on these optimal parameters, and the spectra are calculated by an estimator. Through the simulations and experiments, the following conclusions are drawn: (i) the GGRCM‐MIPSO method can extract chromatogram peaks from simulation data set without knowing the number of the compounds in advance even when a severe overlap and white noise exist and (ii) the GGRCM‐MIPSO method can be applied to the real HPLC‐DAD data set. Copyright © 2014 John Wiley & Sons, Ltd.     

Nucleophile Promiscuity of Engineered Class II Pyruvate Aldolase YfaU from E. Coli

Pyruvate‐dependent aldolases exhibit a stringent selectivity for pyruvate, limiting application of their synthetic potential, which is a drawback shared with other existing aldolases. Structure‐guided rational protein engineering rendered a 2‐keto‐3‐deoxy‐l ‐rhamnonate aldolase variant, fused with a maltose‐binding protein (MBP‐YfaU W23V/L216A), capable of efficiently converting larger pyruvate analogues, for example, those with linear and branched aliphatic chains, in aldol addition reactions. Combination of these nucleophiles with N‐Cbz‐alaninal (Cbz=benzyloxycarbonyl) and N‐Cbz‐prolinal electrophiles gave access to chiral building blocks, for example, derivatives of (2S,3S,4R)‐4‐amino‐3‐hydroxy‐2‐methylpentanoic acid (68 %, d.r. 90:10) and the enantiomer of dolaproine (33 %, d.r. 94:6) as well as a collection of unprecedented α‐amino acid derivatives of the proline and pyrrolizidine type. Conversions varied between 6–93 % and diastereomeric ratios from 50:50 to 95:5 depending on the nucleophilic and electrophilic components.     

A Simple Design to Realize Micro-column Separation by Conventional Analytical HPLC

The conventional analytical HPLC was successfully developed for micro‐column separation by using a simple eluate splitting system, self‐preparation of packing column and on‐capillary column detector in our laboratory. Porous inlet frit in fused silica capillary was rapidly prepared by sintering stainless steel powders under 500 meshes for about 20 s. The use of such frits or metal meshes in capillary to retain C₁₈ particles of chromatographic packing was demonstrated to be stable and specially robust with continuous packing and long chromatographic runs. Furthermore, the chromatographic behavior was detailedly evaluated by changing the flow rate and the percentage of mobile phase using the prepared capillary column. Under the optimal experimental conditions, baseline separation of the model analytes including thiourea, benzene, toluene, ethylbenzene was obtained with a high column efficiency near 70000N (plates/m) by the developed capillary‐HPLC.     

Simultaneous Determination of Trace Zinc and Cadmium by Anodic Stripping Voltammetry Using a Polymeric Film Nanoparticle Self‐Assembled Electrode

A promising modified electrode was fabricated by polymerization a conductive polymer film of dipicolinic acid (DPA) onto gold nanoparticle (AuNP)‐cysteine‐gold electrode (Au). The morphology of poly(DPA)‐AuNP‐Au electrode was investigated by scanning electron microscopy (SEM). This chemically modified electrode was used for electrochemical determination of cadmium and zinc in aqueous media using differential pulse anodic stripping voltammetry. The result showed that the modified electrode could clearly resolve the anodic stripping peaks of zinc and cadmium. The linear analytical curves were obtained in the ranges of 0.020–25.0 and 0.045–17.0 µM for zinc and cadmium respectively. The limit of detections (S/N=3) were 0.008 µM for zinc and 0.015 µM for cadmium.     

Collision‐induced dissociation mass spectra of glucosinolate anions

Collision‐induced dissociation (CID) mass spectra of differently substituted glucosinolates were investigated under negative‐ion mode. Data obtained from several glucosinolates and their isotopologues (³⁴S and ²H) revealed that many peaks observed are independent of the nature of the substituent group. For example, all investigated glucosinolate anions fragment to produce a product ion observed at m/z 195 for the thioglucose anion, which further dissociates via an ion/neutral complex to give two peaks at m/z 75 and 119. The other product ions observed at m/z 80, 96 and 97 are characteristic for the sulfate moiety. The peaks at m/z 259 and 275 have been attributed previously to glucose 1‐sulfate anion and 1‐thioglucose 2‐sulfate anion, respectively. However, based on our tandem mass spectrometric experiments, we propose that the peak at m/z 275 represents the glucose 1‐thiosulfate anion. In addition to the common peaks, the spectrum of phenyl glucosinolate (β‐D ‐Glucopyranose, 1‐thio‐, 1‐[N‐(sulfooxy)benzenecarboximidate] shows a substituent‐group‐specific peak at m/z 152 for C₆H₅‐C(?NOH)S^?, the CID spectrum of which was indistinguishable from that of the anion of synthetic benzothiohydroxamic acid. Similarly, the m/z 201 peak in the spectrum of phenyl glucosinolate was attributed to C₆H₅‐C(?S)OSO₂^?. Copyright © 2009 John Wiley & Sons, Ltd.