Reference materials to evaluate measurement systems for the nutrient composition of foods: results from USDA’s National Food and Nutrient Analysis Program (NFNAP)

Over a 6.5-year period a total of 2554 values were reported by nine laboratories for 259 certified or reference nutrient concentrations in 26 certified reference materials (CRM) submitted to contract laboratories, blinded, as part of the qualifying process for analytical contracts and in the routine sample stream as part of the National Food and Nutrient Analysis Program. Each value was converted to a Z′-score, reflecting the difference from the assigned value related to the combined expected analytical uncertainty plus the uncertainty in the CRM value. Z′-scores >|3.0| were considered unacceptable. For some nutrients (Na, folate, dietary fiber, pantothenic acid, thiamin, tocopherols, carotenoids, monounsaturated, and polyunsaturated fatty acids), >20% of Z′-scores were >|3.0|. For total fat, vitamin C, and niacin >25% of Z′-scores were >|2.0|. Components for which CRM data were best (more than 90% of Z′-scores <|2.0|) were Mg, P, Mn, Se, and vitamin B12. In some cases deviations from assigned values were not uniform across laboratories and materials. For Na almost all high Z′-scores were for low-Na matrices, suggesting analytical problems related to concentration. Figure Z′-scores for vitamins in certified reference materials Disclaimer: Certain commercial equipment, instruments, or materials are identified in this paper in order to specify the experimental procedure adequately. Such identification is not intended to imply recommendation or endorsement by the National Institute of Standards and Technology, or the United States Department of Agriculture, nor is it intended to imply that the materials or equipment identified are necessarily the best available for the purpose.     

Development of a highly precise ID-ICP-SFMS method for analysis of low concentrations of lead in rice flour reference materials

Microwave digestion and isotope dilution inductively coupled plasma mass spectrometry (ID-ICP-SFMS) has been applied to the determination of Pb in rice flour. In order to achieve highly precise determination of low concentrations of Pb, the digestion blank for Pb was reduced to 0.21 ng g⁻¹ after optimization of the digestion conditions, in which 20 mL analysis solution was obtained after digestion of 0.5 g rice flour. The observed value of Pb in a non-fat milk powder certified reference material (CRM), NIST SRM 1549, was 16.8 ± 0.8 ng g⁻¹ (mean ± expanded uncertainty, k = 2; n = 5), which agreed with the certified value of 19 ± 3 ng g⁻¹ and indicated the effectiveness of the method. Analytical results for Pb in three brown rice flour CRMs, NIST SRM 1568a, NIES CRM 10-a, and NIES CRM 10-b, were 7.32 ± 0.24 ng g⁻¹ (n = 5), 1010 ± 10 ng g⁻¹ (n = 5), and 1250 ± 20 ng g⁻¹ (n = 5), respectively. The concentration of Pb in a candidate white rice flour reference material (RM) sample prepared by the National Metrology Institute of Japan (NMIJ) was observed to be 4.36 ± 0.28 ng g⁻¹ (n = 10 bottles). Figure Digestion blank of Pb was carefully reduced to approximately 0.2 ng g^-1 which permitted the highly precise determination of Pb at low ng g^-1 level in foodstuff samples by ID-SFMS     

Certification of methylmercury in cod fish tissue certified reference material by species-specific isotope dilution mass spectrometric analysis

A new cod fish tissue certified reference material, NMIJ CRM 7402-a, for methylmercury analysis was certified by the National Metrological Institute of Japan in the National Institute of Advanced Industrial Science and Technology (NMIJ/AIST). Cod fish was collected from the sea close to Japan. The cod muscle was powdered by freeze-pulverization and was placed into 600 glass bottles (10 g each), which were sterilized with γ-ray irradiation. The certification was carried out using species-specific isotope dilution gas chromatography inductively coupled plasma mass spectrometry (SSID–GC–ICPMS), where ²⁰²Hg-enriched methylmercury (MeHg) was used as the spike compound. In order to avoid any possible analytical biases caused by nonquantitative extraction, degradation and/or formation of MeHg in sample preparations, two different extraction methods (KOH/methanol and HCl/methanol extractions) were performed, and one of these extraction methods utilized two different derivatization methods (ethylation and phenylation). A double ID method was adopted to minimize the uncertainty arising from the analyses. In order to ensure not only the reliability of the analytical results but also traceability to SI units, the standard solution of MeHg used for the reverse-ID was prepared from high-purity MeHg chloride and was carefully assayed as follows: the total mercury was determined by ID–ICPMS following aqua regia digestion, and the ratio of Hg as MeHg to the total Hg content was estimated by GC–ICPMS. The certified value given for MeHg is 0.58 ± 0.02 mg kg⁻¹ as Hg. Figure NMIJ CRM 7402-a: cod fish tissue for MeHg analysis     

Calibration possibilities and modifier use in ETV ICP OES determination of trace and minor elements in plant materials

The possibilities for universal calibration based on multi-element aqueous standard solutions and graphite laboratory reference materials (graphite standards) for the electrothermal vaporization inductively coupled plasma optical emission spectrometric (ETV ICP OES) determination of Al, B, Ba, Cd, Co, Cr, Cu, Fe, Mn, Ni, P, Pb, S, Sr, Ti, V, and Zn in plant materials were investigated. A commercially available state-of-the-art ETV device was coupled with an Echelle ICP spectrometer equipped with a charge-injection-device (CID) camera for spectral detection. The transition area between transport tube and ETV graphite tube and the gas streams for inner gas, bypass gas, and modifier gas were optimized to achieve best transport efficiencies. The influence of four gaseous modifiers (CCl₄, CHCl₃, CCl₂F₂, and C₃H₈) added to the inner gas was studied. Five reference materials (RM P-Alfalfa, Lucerne; NIES CRM No.9 “Sargasso”; CTA-VTL-2 Virginia Tobacco Leaves; NIST SRM 1515 Apple Leaves; IAEA-V-10 Hay Powder) were used for method validation. If certified reference materials are not available, calibration against graphite standards or dried aqueous standard solutions is possible. Three carbonization procedures as sample pretreatment for the plant materials were investigated. Figure Picture of the ETV system (sample changer and graphite-tube furbace) used in this work Presented at the European Symposium on Atomic Spectrometry (ESAS) September 28-October 1, 2008, Weimar, Germany.     

A theoretical framework for evaluating analytical digestion methods for poorly soluble particulate beryllium

Complete digestion of all chemical forms and sizes of particulate analytes in environmental samples is usually necessary to obtain accurate results with atomic spectroscopy. In the current study, we investigate the physicochemical properties of beryllium particles likely to be encountered in samples collected from different occupational environments and present a hypothesis that a dissolution theory can be used as a conceptual framework to guide development of strategies for digestion procedures. For monodisperse single-chemical constituent primary particles, such as those encountered when handling some types of beryllium oxide (BeO) powder, theory predicts that a digestion procedure is sufficient when it completely dissolves all primary particles, independent of cluster size. For polydisperse single-chemical constituent particles, such as those encountered during the handling of some types of beryllium metal powder, theory predicts that a digestion procedure is sufficient only when it completely dissolves the largest particle in the sample. For samples with unknown or multi-chemical constituent particles and with particles having undefined sizes, e.g., fume emissions from a copper–beryllium alloy furnace operation or dust from a beryl ore crushing operation, a surface area-limited and single-constituent-dependent dissolution theory may not predict complete dissolution, thereby requiring non-routine robust treatment procedures with post-digestion filtration, followed by examination of residual particulate material. Additionally, for beryllium, and likely other poorly soluble materials, particulate reference materials of various chemical forms and size distributions are needed to better evaluate and harmonize analytical digestion procedures. Figure Generation of aerosol particles during machining of beryllium oxide The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health.     

Production of malachite green oxalate and leucomalachite green reference materials certified for purity

Malachite green oxalate (MG oxalate) and leucomalachite green (LMG) have been prepared and certified as pure reference materials. The purities of MG oxalate and LMG were assessed by high-performance liquid chromatography–diode array detection (HPLC–DAD), nuclear magnetic resonance (NMR) spectroscopy, differential scanning calorimetry (DSC), Karl Fischer titration, ashing and thermogravimetric analysis (TGA). MG oxalate was purified by supercritical fluid extraction (SFE). Prior to purification, commercial MG oxalate purity was estimated to be about 90%. The main impurities present in SFE-purified MG oxalate were identified and quantified using HPLC–DAD. The main impurities were found to be monode-MG (monodemethylated MG oxalate synthesis impurity), 4-(dimethylamino)benzophenone (4-DMABP), MG-carbinol and LMG. The homogeneity of both reference materials was also determined. Issues associated with the stability of LMG and MG oxalate in solution forced an extensive study investigating different parameters i.e. solvent, acid, analyte concentration and temperature. MG oxalate (100 μg/mL) was found to be stable in acetonitrile containing 1% v/v glacial acetic acid for at least 155 days and LMG (100 μg/mL) was stable in acetonitrile for at least 133 days. The final purity value for MG oxalate was 94.3 ± 1.4% m/m at the 95% confidence interval (or 67% m/m if MG cation is reported). For LMG, the certified purity was found to be 98.8 ± 0.8% m/m at the 95% confidence interval. Figure Calibration reference materials for malachite green and leucomalachite green, certified for purity, are essential in characterising these key analytes in a fish matrix reference material     

Mineral oil certified reference materials for the determination of polychlorinated biphenyls from the National Metrology Institute of Japan (NMIJ)

Four mineral oil certified reference materials (CRMs), NMIJ CRM 7902-a, CRM 7903-a, CRM 7904-a, and CRM 7905-a, have been issued by the National Metrology Institute of Japan, which is part of the National Institute of Advanced Industrial Science and Technology (NMIJ/AIST), for the determination of polychlorinated biphenyls (PCBs). The raw materials for the CRMs were an insulation oil (CRM 7902-a and CRM 7903-a) and a fuel oil (CRM7904-a and CRM 7905-a). A solution of PCB3, PCB8, and technical PCB products, comprising four types of Kaneclor, was added to the oil matrices. The total PCB concentrations in the PCB-fortified oils (CRM 7902-a and CRM 7904-a) are approximately 6 mg kg⁻¹. In addition, the mineral oils which were not fortified with PCBs were also distributed as CRMs (CRM 7903-a and CRM 7905-a). Characterization of these CRMs was conducted by the NMIJ/AIST, where the mineral oils and the PCB solution were analyzed using multiple analytical methods such as dimethylsulfoxide extraction, normal-phase liquid chromatography, gel permeation chromatography, reversed-phase liquid chromatography, and chromatography using sulfoxide-bonded silica; and/or various capillary columns for gas chromatography, and two ionization modes for mass spectrometry. The target compounds in the mineral oils and those in the PCB solution were determined by one of the primary methods of measurement, isotope dilution–mass spectrometry (ID-MS). Certified values have been provided for 11 PCB congeners (PCB3, 8, 28, 52, 101, 118, 138, 153, 180, 194, and 206) in the CRMs. These CRMs have information values for PCB homologue concentrations determined by using a Japanese official method for determination of PCBs in wastes and densities determined with an oscillational density meter. Because oil samples having arbitrary PCB concentrations between respective property values of the PCB-fortified and nonfortified CRMs can be prepared by gravimetric mixing of the CRM pairs, these CRMs can be used for validation of PCB analyses using various instruments which have different sensitivities. Figure Preparation and certification processes of the mineral oil CRMs (example shown is polychlorinated biphenyls in insulation oil, high/low concentrations) Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Evaluation of a multiresidue method for measuring fourteen chemical groups of pesticides in water by use of LC-MS-MS

European Council Directive 98/83/EC on the quality of water intended for human consumption brought a new challenge for water-quality control routine laboratories, mainly on pesticides analysis. Under the guidelines of ISO/IEC 17025:2005, a multiresidue method was developed, validated, implemented in routine, and studied with real samples during a one-year period. The proposed method enables routine laboratories to handle a large number of samples, since 28 pesticides of 14 different chemical groups can be quantitated in a single procedure. The method comprises a solid-phase extraction step and subsequent analysis by liquid chromatography-mass spectrometry (LC-MS-MS). The accuracy was established on the basis of participation in interlaboratory proficiency tests, with encouraging results (majority |z-score| <2), and the precision was consistently analysed over one year. The limits of quantitation (below 0.050 μg L⁻¹) are in agreement with the enforced threshold value for pesticides of 0.10 μg L⁻¹. Overall method performance is suitable for routine use according to accreditation rules, taking into account the data collected over one year. Figure Simultaneous SPE extraction system for high thoughput analysis Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users. Notes: IAREN is an accredited laboratory under the ISO/IEC 17025:2005 for 165 analytical determinations and part of the NORMAN Network of Reference laboratories for monitoring of emerging environmental pollutants in the European Union. This work was presented (podium presentation) at the 12th Symposium on Sampling and Handling of the International Association of Environmental Analytical Chemistry (IAEAC), Zaragoza, 2006.     

Evaluation of different calibration strategies for the analysis of pure copper and zinc samples using femtosecond laser ablation ICP-MS

Solution-doped metal powder pellets as well as aspirated liquids were used as calibration samples to analyze pure copper and zinc certified reference materials (CRMs) by femtosecond laser ablation ICP-MS. It was demonstrated that calibration by copper pellets resulted in relative deviations up to 20%, whereas fs-LA-ICP-MS among copper-based CRMs led to inaccuracies in the same range unless nominal mass fractions were chosen to be <3 mg/kg. Calibration by zinc pellets generally provided better accuracy. Depending on the analyte considered, deviations below 10% were obtained even for mass fractions close to the limit of quantification. Our data, therefore, indicate solution-doped metal powder pellets to be suitable as calibration samples for fs-LA-ICP-MS of metals. Furthermore, the utilization of liquid standards for calibration was found to result in stronger deviations of up to 50% for both copper and zinc samples which, in addition, turned out to be dependent on the plasma conditions.      

Simple and accurate candidate reference measurement procedure for total testosterone in human serum by one-step liquid-liquid extraction coupled with isotope dilution mass spectrometry

Testosterone in human serum is commonly tested in clinical laboratories using immunoassay methods as well as liquid chromatography-tandem mass spectrometry (LC-MS/MS) methods. To standardize and ensure the accuracy of the measurement results, reference procedures with higher metrological order are required. A simple measurement procedure based on one-step liquid-liquid extraction (LLE) and liquid chromatography-isotope dilution tandem mass spectrometry (LC-IDMS/MS) was developed for total testosterone in human serum. The procedure involved serum spiked with ¹³C₃-testosterone, equilibration for 2 h, and extraction with an organic solvent. Testosterone certified reference material (CRM) was used as the calibration standard to ensure the traceability to the International System of Units (SI). Testosterone in serum CRMs from the National Institute for Standards and Technology (NIST) and LGC were used to validate the accuracy of the newly developed method. The deviations of the obtained values from the NIST and LGC certified values ranged from −0.55% to 0.45%. Similarly, the coefficient of variations (CVs) of the replicate measurements were in the range of 0.55% and 0.78%, respectively. The relative expanded uncertainties were comparable with those of the certified materials. The newly developed LC-IDMS/MS procedure demonstrated adequate trueness and precision, and was simple to perform. The method can be used for value assignment of testosterone in external quality assessment (EQA) materials as well as certification of CRMs in the future.

Graphical abstract

     

Characterization of NIES CRM No. 23 Tea Leaves II for the determination of multielements

A candidate environmental certified reference material (CRM) for the determination of multielements in tea leaves and materials of similar matrix, NIES CRM No. 23 Tea Leaves II, has been developed and characterized by the National Institute for Environmental Studies (NIES), Japan. The origin of the material was tea leaves, which were ground, sieved through a 106-μm mesh, homogenized, and then subdivided into amber glass bottles. The results of homogeneity and stability tests indicated that the material was sufficiently homogeneous and stable for use as a reference material. The property values of the material were statistically determined based on chemical analyses by a network of laboratories using a wide range of methods. Sixteen laboratories participated in the characterization, and nine certified values and five reference values were obtained. These property values of the candidate CRM, which are expressed as mass fractions, were close to the median and/or mean values of the mass fractions of elements in various tea products. The candidate CRM is appropriate for use in analytical quality control and in the evaluation of methods used in the analysis of tea and materials of similar matrix.     

Comparison of three strategies to evaluate uncertainty from in-house validation data. A case study: mercury determination in sediments

In the present paper, three approaches are compared for the evaluation of the combined uncertainty in the determination of mercury in aquatic sediments by an aqua regia extraction procedure. For this, the data obtained in validation studies from five certified reference materials (CRMs), covering a range of concentrations from 0.8 to 130 mg kg⁻¹ of mercury and analysed by three atomic spectroscopic techniques (cold vapour generation atomic fluorescence spectrometry, CV-AFS, cold vapour generation atomic absorption spectroscopy, and inductively coupled plasma mass spectroscopy), were considered. The combined uncertainty was firstly assessed by considering separately the data obtained for each CRM analysed (approach A). Moreover, this assessment was also performed with two other calculation approaches (B and C) based on the pooled data obtained from the validation step. The comparison of the results obtained for the different techniques showed a clear bias effect when using CV-AFS with nitric acid as a diluent. In relation to the strategies tested for the combined uncertainty assessment, approach C proved to be the easiest and friendliest method for uncertainty assessment.      

A headspace solid-phase microextraction procedure coupled with gas chromatography–mass spectrometry for the analysis of volatile polycyclic aromatic hydrocarbons in milk samples

A sensitive and solvent-free method for the determination of ten polycyclic aromatic hydrocarbons, namely, naphthalene, acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benzo[a]anthracene and chrysene, with up to four aromatic rings, in milk samples using headspace solid-phase microextraction and gas chromatography–mass spectrometry detection has been developed. A polydimethylsiloxane–divinylbenzene fiber was chosen and used at 75°C for 60 min. Detection limits ranging from 0.2 to 5 ng L⁻¹ were attained at a signal-to-noise ratio of 3, depending on the compound and the milk sample under analysis. The proposed method was applied to ten different milk samples and the presence of six of the analytes studied in a skimmed milk with vegetal fiber sample was confirmed. The reliability of the procedure was verified by analyzing two different certified reference materials and by recovery studies. Figure Milk is safe, healthy food     

Preparation of reference materials for the determination of RoHS-relevant flame retardants in styrenic polymers

Reference materials for the analysis of polybrominated diphenyl ethers, polybrominated biphenyls and other common brominated flame retardants (FR) in styrenic polymers were prepared to suit the demands of actual restriction of the use of certain hazardous substances in electrical and electronic equipment analytics. Three methods of preparation were employed, viz. pellet forming, dissolution/vaporisation and extrusion, whereby extrusion proved to be the most suitable method. For extrusion, three procedures of pre-mixing were investigated: the polymers were either mixed with FR powder, FR solutions or FR concentrates that were taken from waste industrial polymers. The latter procedure proved to be most appropriate in terms of analyte concentration, predictability and recovery. The homogeneity of the samples, as well as the chemical and thermal long-term stabilities, was investigated. The result was an optimised method to prepare a suitable reference material for laboratory use. Figure   Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Determination of 43 polycyclic aromatic hydrocarbons in air particulate matter by use of direct elution and isotope dilution gas chromatography/mass spectrometry

We are reporting a method for measuring 43 polycyclic aromatic hydrocarbons (PAH) and their methylated derivatives (Me-PAHs) in air particulate matter (PM) samples using isotope dilution gas chromatography/high-resolution mass spectrometry (GC/HRMS). In this method, PM samples were spiked with internal standards, loaded into solid phase extraction cartridges, and eluted by dichloromethane. The extracts were concentrated, spiked with a recovery standard, and analyzed by GC/HRMS at 10,000 resolution. Sixteen ¹³C-labeled PAHs and two deuterated Me-PAHs were used as internal standards to account for instrument variability and losses during sample preparation. Recovery of labeled internal standards was in the range of 86–115%. The proposed method is less time-consuming than commonly used extraction methods, such as sonication and accelerated solvent extraction (ASE), and it eliminates the need for a filtration step required after the sonication extraction method. Limits of detection ranged from 41 to 332 pg/sample for the 43 analytes. This method was used to analyze reference materials from the National Institute of Standards and Technology. The results were consistent with those from ASE and sonication extraction, and these results were also in good agreement with the certified or reference concentrations. The proposed method was then used to measure PAHs on PM_2.5 samples collected at three sites (urban, suburban, and rural) in Atlanta, GA. The results showed distinct seasonal and spatial variation and were consistent with an earlier study measuring PM_2.5 samples using an ASE method, further demonstrating the compatibility of this method and the commonly used ASE method.      

Certification and uncertainty evaluation of the certified reference materials of poly(ethylene glycol) for molecular mass fractions by using supercritical fluid chromatography

Poly(ethylene glycol) (PEG) is a useful water-soluble polymer that has attracted considerable interest in medical and biological science applications as well as in polymer physics. Through the use of a well-calibrated evaporative light-scattering detector coupled with high performance supercritical fluid chromatography, we are able to determine exactly not only the average mass but also all of the molecular mass fractions of PEG samples needed for certified reference materials issued by the National Metrology Institute of Japan. In addition, experimental uncertainty was determined in accordance with the Guide to the expression of uncertainty in measurement (GUM). This reference material can be used to calibrate measuring instruments, to control measurement precision, and to confirm the validity of measurement methods when determining molecular mass distributions and average molecular masses. Especially, it is suitable for calibration against both masses and intensities for matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Figure Comparison between the molecular mass fractions of PEG 1000 before calibration (si) (○) and after calibration (wi) (⧫). The error bar shows the expanded uncertainty of k = 2 of each mass fraction     

Evaluation of solid-phase microextraction conditions for the determination of polycyclic aromatic hydrocarbons in aquatic species using gas chromatography

This paper describes a headspace solid-phase microextraction (HS-SPME) procedure coupled to gas chromatography with mass spectrometric detection (GC–MS) for the determination of eight PAHs in aquatic species. The influence of various parameters on the PAH extraction efficiency was carefully examined. At 75 °C and for an extraction time of 60 min, a polydimethylsiloxane–divinylbenzene (PDMS/DVB) fiber coating was found to be most suitable. Under the optimized conditions, detection limits ranged from 8 to 450 pg g⁻¹, depending on the compound and the sample matrix. The repeatability varied between 7 and 15% (RSD). Accuracy was tested using the NIST SRM 1974b reference material. The method was successfully applied to different samples, and the studied PAHs were detected in several of the samples. Figure Headspace SPME sampling followed by GC–MS facilitates routine monitoring of PAHs in aquatic species     

Results of an interlaboratory comparison on the determination of polybrominated flame retardants in poly(ethyleneterephthalate)

An intercomparison to establish the performance of routine laboratories in the determination of polybrominated flame retardants in polymers was organised. Commercial poly(ethyleneterephthalate) was fortified with technical pentabromodiphenyl ether, octabromodiphenyl ether and decabromodiphenyl ether mixtures and with a decabromobiphenyl technical mixture at 0.4–0.8 g/kg. Homogeneity and stability of the total Br content in the material was confirmed. Thirty-seven laboratories from Europe, Asia and the Americas submitted results. Relative repeatability standard deviations for individual congeners ranged from 7 to 17%. Relative between-laboratory standard deviations ranged from 22 to 61%. No significant influence of a common standard, application of a standard method or method parameters could be identified. The quality and uncertainty of the results of this study are significantly worse than those reported in the environmental field and indicate a clear need for a learning process among the laboratories involved. Figure Mandel’s h (between labs): critical level: 1.91     

Integration of an electrochemical-based biolithography technique into an AFM system

An ordinary atomic force microscopy (AFM) was functionalized and applied to electrochemically draw micropatterns of biomolecules. To fabricate an electrochemical AFM probe having an electrode at the tip, a metal-coated AFM probe was first insulated with Parylene C, and then the apex of the tip was ground mechanically to expose the electrode. The effective electrode diameter was estimated to be ca. 500 nm. The electrode probe was positioned close to a heparin-coated antibiofouling substrate and used to locally generate hypobromous acid from a dilute Br⁻ solution to render the substrate surface protein-adhesive. In situ topographical imaging after the electrochemical treatment suggested the heparin layer became detached to allow the adsorption of proteins, in this case fibronectin. The diameter of the drawn fibronectin pattern was 2 μm, which is one order of magnitude smaller than we achieved previously using a microdisk electrode (tip diameter 10 μm). Figure AFM configuration integrated with the electrochemical-based surface modification and resultant micropatterns of fluorescence-labeled fibronectin Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Pollen discrimination and classification by Fourier transform infrared (FT-IR) microspectroscopy and machine learning

The discrimination and classification of allergy-relevant pollen was studied for the first time by mid-infrared Fourier transform infrared (FT-IR) microspectroscopy together with unsupervised and supervised multivariate statistical methods. Pollen samples of 11 different taxa were collected, whose outdoor air concentration during the flowering time is typically measured by aerobiological monitoring networks. Unsupervised hierarchical cluster analysis provided valuable information about the reproducibility of FT-IR spectra of the same taxon acquired either from one pollen grain in a 25 × 25 μm² area or from a group of grains inside a 100 × 100 μm² area. As regards the supervised learning method, best results were achieved using a K nearest neighbors classifier and the leave-one-out cross-validation procedure on the dataset composed of single pollen grain spectra (overall accuracy 84%). FT-IR microspectroscopy is therefore a reliable method for discrimination and classification of allergenic pollen. The limits of its practical application to the monitoring performed in the aerobiological stations were also discussed. Figure Traditional and innovative methods for the identification of airborne pollen grains