Investigation and application of liquid chromatography mass spectrometry in the analysis of polar, less volatile and thermal unstable organic pollutants in environmental and biological samples 1

More importantly, with the development of atmospheric pressure ionization API interface techniques, LC-API-MS has assumed great importance in the analysis of polar, low volatile, and the

1 Chapter One

Introduction & Literature Review

1.1 GENERAL INTRODUCTION TO LIQUID CHROMATOGRAPHY-MASS SPECTROMETRY (LC-MS)

In recent decades, LC-MS has experienced impressive progress, both in terms of technological development and application The combination of chromatography (for separation) and mass spectrometry (as a sensitive detector that provides structural information) has proved to be a primary technology for the detection and characterization

of various molecules, providing the analytical chemist with one of the most powerful analytical tools of modern times The key advantages of LC-MS are [1]:

¾ Selectivity: LC-MS is not constrained by chromatographic resolution coeluting peaks can be isolated by “mass selectivity.”

¾ Peak assignment: LC-MS creates a unique “chemical fingerprint” for the compound of interest, ensuring correct peak assignment even in the presence

¾ Quantitation: quantitative and qualitative data can be obtained easily with limited instrument optimization

More importantly, with the development of atmospheric pressure ionization (API) interface techniques, LC-API-MS has assumed great importance in the analysis of polar, low volatile, and thermally instable organic compounds, for which direct analysis by GC-

MS is not possible [2]

1.2 INTRODUCTION AND BASIC PRINCIPLES

1.2.1 LC and MS

Liquid chromatography (LC) is a physical separation method by which the components to

be separated are selectively distributed between two immiscible phases: a liquid mobile phase and a stationary phase bed through which the liquid mobile phase passes The chromatographic process occurs as a result of repeated sorption/desorption steps during the movement of the analytes along the stationary phase The separation is due to differences in distribution coefficients of the individual analytes within the sample [3-6] Because a high-pressure pump is required to move the mobile phase and analytes through the column, LC is also known as high performance liquid chromatography (HPLC)

Mass spectrometry (MS), as one of LC detection techniques, has attracted growing interest because of the potential to yield information on the relative molecular mass (Mr)

such as ultraviolet (UV), refractive index (RI) and fluorescence At present, MS is the most sensitive method of molecular analysis MS is based on the detection of emitted ions that have been separated or filtered according to their mass-to-charge (m/z) ratio The resulting mass spectrum is a plot of the relative abundance of the generated ions as a function of the m/z ratio Because extreme selectivity can be obtained, mass detection combined with chromatographic separation is invaluable in quantitative trace analyses [7-9]

1.2.2 LC-MS Coupling

1.2.2.1 General problems

When searching for a suitable technique for the analysis of mixtures that, often contain unknowns and/or analytes in low concentration, the combination of chromatography (for separation) with mass spectrometry (as detector that provides sensitive and structural information) appears to be an obvious choice However, liquid chromatography coupled

to mass spectrometry is an odd combination In HPLC, the sample is in solution and at atmospheric pressure, while the mass analysis of ions takes place within a vacuum [10]

In developing on-line LC-MS, three fundamental compatibility problems had to be solved Firstly, LC is preferred over gas chromatography (GC) in analysis of samples with high polarity and low volatility In contrast, one of the prerequisites for mass spectrometric analysis is the formation of volatilized ions Second difficult problem to be solved was the necessity to eliminate the mobile phase Water flowing at a rate of 1

ml/min from a conventional 4.6-mm-ID LC column is converted into vapor (at atmospheric pressure) at 1244 ml/min, which is far too much for standard MS vacuum systems to handle Third, salts and other additives within the mobile phase are often involatile (e.g., phosphates, NaCl, etc.) [10]

1.2.2.2 General introduction of interface techniques

The aforementioned problems have been solved by the development of interface techniques that: 1) transfer the analyte from the LC column to the MS ion source and prepare the analyte for ionization, 2) cope with the solvent from the LC, 3) bridge the pressure difference between column outlet and mass analyzer [11-16]

The first experiments that couple LC with MS date back to the late 1960s The actual breakthrough in the development of coupling techniques was the introduction of the moving-belt interface (MB) in 1974 that, became the first commercial LC-MS interface [11] Subsequently, many interface techniques have been rapidly developed, such as particle beam (PB) [12], continuous flow fast atom bombardment (cf-FAB) [13,14], direct liquid introduction (DLI) [15], thermospray (TSP) [16], and more recently, the API technique (APCI and ESI) [16]

Both MB and PB interfaces rely on the removal of solvent prior to entering the MS [17]

MB separates the condensed liquid-phase side of the LC from the high vacuum of the MS and uses a belt to transport the analytes from one to the other The mobile phase of the

LC is deposited on a band and evaporated [18, 19] Most moving-belt analyses deal with volatile analytes, that consequently limit the application of MB LC-MS The PB interface provides the opportunity to use EI/CI without the mechanical transportation portion used

in MB [20] The LC elutent is forced through a small nebulizer with the aid of following

He gas to form a stream of uniform droplets These droplets move through a desolvation chamber and evaporate, leaving a solid particle These particles are separated from the gas and transported into the MS source using a differentially pumped momentum separator.The PB interface can be considered as the successor of the MB, giving better stability and more robustness Furthermore, this is of growing importance in LC-MS interfacing because relatively volatile analytes can be examined Many pesticides are not sufficiently volatile to be amenable to GC/MS, but can be analyzed by PB LC-MS DLI

or cf-FAB interfaces reduce the flow entering the MS using a splitting device A serious drawback of this approach is the reduction in sensitivity caused by the split factor The flow rate magnitude, used with a classic 4.6 mm i.d column (1 ml/min) is only tolerated

by techniques such as TSP and API [21]

The TSP interface was developed by Vestal et al [22-24] As the name thermospray

implies, heating of liquid flow leaving an LC system creates a spray of superheated mist containing small liquid droplets A major advantage of TSP over other LC-MS interfaces

is its ability to handle the high flow-rates delivered by LC (up to 2 mL/min) The TSP interface was developed to solve two problems: 1) to reduce the pressure of the solvents, and 2) generate ions of the analyte, that are frequently nonvolatile This particular interface was developed for a system that, for a number of years, was considered to be the

easiest, most versatile, and powerful LC-MS interface technique [23, 24] However, with the advent of new, more adaptable and robust LC-MS interfaces based on atmospheric-pressure ionization (e.g., APCI and ESI), the use of TSP diminished rapidly [25]

1.2.2.3 API technique in LC-MS

The earliest LC-MS techniques (DLI, TSP, MB, and PB), though commercialized, were often difficult to use, had limited sensitivity, and were not robust However, they were very useful for specific applications The overwhelming increase in LC-MS applications

is primarily the result of the sensitivity and ruggedness of ESI and APCI, which are both API techniques The advantages of API techniques were summarized by Voyksner with four key points [26]:

¾ API approaches can handle volumes of liquid typically used in LC,

¾ API is suitable for the analysis of non-volatile, polar, and thermally unstable compounds typically analyzed by LC,

¾ API-MS systems are sensitive, offering comparable or better detection limits than achieved by GC-MS,

¾ API systems are very rugged and relatively easy to use

1.2.2.3.1 ESI

The first attempts to use electrospray (ES) as an MS interface date back to the late 1960s

and early 1970s At that time, Dole et al investigated the possibility of producing

gas-phase ions from macromolecules in solution, utilizing an atmospheric pressure electrostatic sprayer for analysis via ion mobility measurements [27-29] The actual origins of the LC–ES-MS coupling were reported in 1984, almost simultaneously by

Yamashita and Fenn and Aleksandrov et al [30-32] Yet, the real breakthrough of ESI in

the early 1990’s relied on extremely low flow-rates that, introduced protein solutions at the concentration of pmol/µl into the mass spectrometer [33, 34] This breakthrough has revolutionized the applications of MS, especially in biological and environmental analyses [35, 36]

The ES process can be divided into three stages: droplet formation, droplet shrinkage and gaseous ion formation [37] The ES process is shown in Fig 1-1 [38], here the liquid flows from the HPLC column, enters a small caliber stainless steel capillary (maintained

at a voltage of 3000±4000 V), and are then dispersed into a very fine spray of charged droplets with the same polarity The solvent then evaporates, shrinking the droplet size and increasing the charge concentration at the droplet's surface Eventually, at the Rayleigh limit, Coulombic repulsion overcomes the droplet's surface tension and the droplet explodes This “Coulombic explosion” forms a series of smaller, less charged droplets The process of shrinking immediately followed by explosion is repeated, leading to very small (3–10 nm) charged droplets that are capable of producing gas-phase ions, which gives a very soft ionization technique The ions are sampled through a set of skimmer electrodes and finally analyzed in the MS analyzer

Fig 1-1 Droplets and ion production under ES conditions [39]

In contrast with all other LC-MS combinations, ESI appears to be a sensitive device; that is, the response is directly proportional to the concentration of the analyte entering the source, irrespective of the flow-rate at which it is delivered This allows miniaturization of the technique without a loss in sensitivity

concentration-Another interesting characteristic of ESI is its “softness.” Specifically, very labile structures can be carried as ions into the gas phase without disrupting their structures For the same reason, ESI spectra contain little to no structural information because of the absence of fragmentation Molecular weight information is obtained in the first instance

If more structural information is needed (e.g., sequence information of peptides), fragmentation must be induced This is most conveniently done by applying tandem MS

There are many reasons for the predominant use of ESI: ease of operation, sensitivity, reliability, robustness, and expanded areas of application ESI is a highly efficient ionization technique that has greatly extended the analytical potential of MS [35], and its interface is currently the most widely applied method for liquid introduction into an MS

Additionally, ESI is especially suited for the analysis of compounds ionized in the liquid phase, due to the fact that, the observed ions, generally formed by protonation or deprotonation of the molecule or by adduct formation with solvent ions, directly reflect acid-base equilibrium in solution

ESI is a mild ionization technique that efficiently produce protonated (or deprotonated) molecular ions of polar, non-volatile, high molecular mass, and themolabile compounds Therefore, LC-ESI-MS is now widely used for the analysis of polar and small ionic molecules, high molecular-weight proteins, and other biomacromolecules [38]

1.2.2.3.2 APCI

The exploration of APCI for LC-MS started in the early 1970s with the research work of

Hornning et al on the use of a modified plasma chromatograph-MS combination [40]

Further research by the same group led to an APCI source, equipped with either a 63Ni foil or corona discharge needle as the primary source of electrons used to generate reactant ions [41] The vaporizer for sample and solvent was a heated glass tube filled with a plug of glass wool, directly attached to a small APCI source APCI interface

techniques were proven to be fully developed only after Fenn et al [42], demonstrated

rapid and accurate molecular-mass determination for large proteins by means of ESI At present, APCI is commercially available and a growing interest in applications involving biological and environmental analysis

In an APCI source, the column effluent is nebulized in a heated vaporizer tube (350-500

°C), where solvent evaporation is nearly complete The gas-vapor mixture enters an atmospheric-pressure ion source, wherein analyte ionization is initiated within a corona discharge needle The solvent vapor acts as reagent gas At atmospheric pressure, the ions are extracted and moved into the mass spectrometer by the exactly same set of skimmers used for electrospray (Fig 1-2) [43,44]

Fig 1-2 Schematic diagram of the typical layout of an APCI source [45]

Ionization occurs through a corona discharge (Fig 1-3), creating reagent ions from the solvent vapor Chemical ionization of the sample molecules is very efficient when at atmospheric pressure, due to the high collision frequency The moderate influences of

solvent clusters and high-pressure gas exposure on the reagent ions, reduced fragmentation during ionization and results in primarily molecular ions

Fig 1-3 Schematic diagram of ionization in APCI-MS [46]

In APCI, positive and negative chemical ionization modes are often used Proton transfer (protonation [M+H]+ reactions) occurs in the positive mode, and either electron transfer

or proton transfer (proton loss, [M-H]-) occurs in the negative mode [M+H] + or [M-H] - are usually formed to give molecular weight information and structural information is obtained through fragmentation induced in the source by increasing the cone voltage [47, 48]

In the positive ionization mode, an electron beam passes across methane-filled ion source, methane being the reagent gas (see Fig 1-4) Interaction of methane (CH4) with electrons (e-) yields methane molecular ions (CH4 •+) Newly formed ions collide several times with neutral molecules (CH4) to give carbonium ions (CH5+) The substance (M) to

be investigated is vaporized in the ion source where it collides with these ions (CH5+) Proton exchange usually occurs because the organic substance (M) is a stronger base than

CH4; hence, protonated molecular ions (MH+ or better [M+H]+) are formed

Fig 1-4 Formation of reactive ions (CH5+) from methane (CH 4 ) reagent gas and their reaction with sample molecules (M) to form protonated molecular ions MH+ [49]

In the negative ionization mode, negative ions are formed by deprotonation (Fig 1-5) Ions of reagent gas, such as OH- or O•-, are strongly basic and capable of extracting a proton from sample molecules M to give [M-H]- ions, one mass unit less than the true molecular mass Negative ion CI is a useful and sensitive technique for substances having

a high electron affinity, such as halogenated compounds or polycyclic aromatic hydrocarbons

Fig 1-5 Negative reactant gas ions, O•- and OH- may be produced easily from (a) NO and (b)

NO/CH4, respectively [49]

In comparison to ESI, APCI allows for the use of solvents that are unfavourable for ion formation because the solvent-evaporation and ion-formation processes are separated in APCI These low-polarity solvents are commonly used in normal-phase LC in conjunction with low polarity samples that can generally be evaporated for APCI ionization Another major difference between APCI and ESI can be found in the LC flow-rates that are used APCI is a technique with optimal performance at high flow-rates

(up to 2 ml/min), while ESI flow rates must be less than 1 ml/min When flow-rates are too low, however, controlling the stability of the corona discharge may become problematic Whereas ESI is ideally suited to miniaturization, reducing the flows and LC dimensions using APCI is much more laborious

APCI is widely used to analyze relatively non-polar, semi-volatile samples of less than

1200 Daltons and it is an especially good ionization source for LC The analytes should have some degree of volatility and should not be too thermolabile Typical molecules are pesticides, drugs, steroids, etc [50-54] Currently, LC-APCI-MS is often used in various stages of drug analysis, environmental analysis of pesticides and herbicides (in water, air, and soil), and metabolite studies

1.3 SAMPLE PREPARATION TECHNIQUES

1.3.1 General Introduction

The development of a complete analytical method includes sample storage, sample preparation, analyte separation, identification and quantitation Among these, sample preparation is the most tedious and time-consuming step, and usually the source of imprecision and inaccuracy of the overall analysis

Currently, chromatography combined with various existing detection systems is the most common analytical technique used for the separation and determination of pesticide

residues in environmental and biological samples In order to obtain qualitative and quantitative information on the presence of trace levels of pesticides from various complicated solid matrices such as soils, food, animals and plants, optimized sample preparation is necessary prior to chromatography analysis

The conventional extraction techniques used for solid matrices are mainly Soxhlet extraction, and sonication Soxhlet extraction has some advantages: it allows for the use

of large sample amount, and no filtration is required after the extraction Nonetheless, significant amounts of solvents, which are often toxic and flammable, are needed and the procedure itself is tedious and time-consuming Sonication is faster than Soxhlet extraction and enables one to extract large amounts of sample at a relatively low cost [55] Still, it uses about as much solvent as Soxhlet extraction, is labor intensive, and filtration is required after extraction In recent years, some novel extraction techniques for solid matrices such as supercritical fluid extraction (SFE), microwave-assisted extraction (MAE) and accelerated-solvent extraction (ASE), have attracted growing interest as they are less time-consuming, environmentally friendlier, and smaller volumes of solvents The principles, advantages and disadvantages of the above mentioned extraction techniques for solid matrices are summarized in Table 1-1 [56,57]

Table 1-1 Preparation techniques for solid matrices-principles, advantages and disadvantages [56,57]

Extraction

Soxhlet Sample is placed in an extraction thimble and

leached with hot solvent in a Soxhlet extractor for 12-24h Solvent evaporation/concentration

is done separately

Standard method Large amount of sample (10-30g) Filtration not required

Not matrix dependent Low cost

Long extraction time, up to 24-48 hrs High consumption of hazardous solvents (300-500 ml)

Large amount of sample (10-30g) Not matrix dependent

Low cost

High consumption of hazardous solvents (300-500 ml)

Labor intensive Filtration required Exposure to solvent vapor Supercritical

Fast (30-60min) CO2 is nontoxic, nonflammable, and environmentally friendly

Small amount of solvent (5-10 ml) Filtration not required

1) Fast (4-30min) 2) Small amount of solvent (10-30ml) 3) Full control of extraction parameters (time, power, temperature)

4) Stirring possible 5) Higher temperatures 6) 12-14 samples can be processed simultaneously

Extracts must be filtered Polar solvent needed

1.3.2 Microwave-assisted Extraction (MAE)

1.3.2.1 Basic principle of MAE

The principle of MAE is based on heating the extractants (mostly liquid organic solvents)

in contact with the sample with microwave energy [58] Partitioning of the analytes of interest from the sample matrix to the extractant depends on the efficiency of microwave heating and the nature of the extractant

Microwave energy is a non-ionizing radiation (frequency 300 - 3×105 MHz) that causes molecular motion by migration of ions and rotation of dipoles Dipole rotation refers to alignment, owing to the electric field, of molecules in the solvent and samples that have dipole moments (either permanent or induced by the electric field) As the field decreases, thermal disorder is restored which results in thermal energy being released At

2450 MHz (the frequency used in commercial systems), alignment of the molecules followed by their return to disorder occurs 4.9 x 109 times per second, resulting in rapid heating [59]

In order to absorb microwave energy, the solvent must exhibit dielectric polarization The greater the dielectric constant, the more thermal energy is released and the more rapid is the heating at a given frequency Besides the absorption of microwave energy, the efficiency of the solvent converting it into heat is also important, which is expressed

by the dielectric loss factor In fact, the dielectric constant and the dielectric loss factor of

a certain solvent can differ greatly Therefore, the overall efficiency of heating using microwave energy is always given by the dissipation factor (Tan δ = ε″ / ε′), which is the ratio of the sample’s dielectric loss (i.e the loss factor ε″) to its dielectric constant (ε′) [60, 61] The dielectric constant is a measure of the sample’s ability to absorb microwave energy while the loss factor is the ability to dissipate the absorbed energy Molecules and ionic solutions (usually acids) will absorb microwave energy easily because they have a permanent dipole moment that is affected by the microwaves On the other hand, non-polar solvents such as hexane will not heat up when exposed to microwaves radiation In Table 1-2, selected physical parameters, including dielectric constants and dissipation factors, are shown for solvents that are used in more than 90% of the applications [62, 63]

Unlike classical heating, microwave radiation heats the entire sample simultaneously without heating the vessel Thus, the solution reaches its boiling point very rapidly, especially in a closed system, and the solvent can be heated to above its boiling point at atmospheric pressure This leads to very short extraction times and high extraction efficiency Table 1-3 shows the comparison of the boiling points of solvents in an open system and closed vessel From Table 1-3, we also can see that there is no heating under microwave irradiation in some non-polar compounds such as hexane, cyclohexane, and petroleum ether[64]

Table 1-2 Dielectric constant, dielectric loss factors and dissipation factors of different materials

at a frequency 3 GHz and a temperature 25°C [63]

Table 1-3 Comparison of the boiling points of solvents in open and closed systems [64]

1.3.2.2 Parameter’s influence on the extraction process

The optimization of MAE conditions for extraction purposes has been investigated by many researchers The most commonly studied parameters are solvent composition, solvent volume, extraction temperature, extraction time and matrix characteristics (including water content) [65, 66]

A correct choice of solvent is paramount to optimization of the extraction process When selecting a solvent, consideration should be given to the microwave-absorbing properties

of the solvent, the interaction of the solvent with the matrix, and the analyte solubility in the solvent Preferably, the solvent should have a high selectivity towards the analyte of interest, excluding unwanted matrix components Another important aspect is the compatibility of the extraction solvent with the analytical method used for the final analysis step Optimal extraction solvents cannot be deduced directly from those used in conventional procedures If the solvent molecule is unable to absorb microwave energy there will be no heating, and hence no effective extraction [67, 68]

The amount of solvent needed for a single sample is often between 10–30 ml [69] In some cases, solvent volume may be an important parameter for efficient extractions The solvent volume must be sufficient to ensure that the entire sample is immersed, especially when utilizing a matrix that will absorb solvent and swell during the extraction process [70]

The most investigated parameter in MAE is extraction temperature, not surprisingly, since the temperature is an important factor contributing to increased recoveries, not only for MAE but for all extraction techniques When MAE is conducted in closed vessels, the temperature may reach well above the boiling point of the solvent These elevated temperatures result in improved extraction efficiencies, since desorption of analytes from active matrix sites will increase Additionally, solvents have a higher capacity to solubilize analytes at higher temperatures Likewise, surface tension and solvent viscosity decrease with temperature, thus improving sample wetting and matrix penetration, respectively

Extraction times in MAE are short compared to conventional techniques Normally, 10 min is sufficient, for the case of extraction of organic pollutants [71, 72], but even 3 min has been demonstrated to give full recovery for pesticides from soils and sediments [73, 74] In the extraction of sulfonylurea herbicides from soils, it was demonstrated that increasing the extraction time from 5 to 30 min did not adversely affect the recovery [75] With thermolabile compounds, however, long extraction times may result in degradation, which was reported for the extraction of pesticides [76] This was observed in the present work (see below): the extraction of carbamates pesticides from soil, wherein the recoveries of some carbamates decrease with increasing extraction times, while other analytes were unaffected

The nature of the matrix, in which the analytes of interest are bound, can have a profound effect on the recoveries of compounds This has been illustrated by spiking experiments,